diff options
Diffstat (limited to 'lib')
237 files changed, 54978 insertions, 0 deletions
diff --git a/lib/Makefile.am b/lib/Makefile.am new file mode 100644 index 0000000..b9fc8a4 --- /dev/null +++ b/lib/Makefile.am @@ -0,0 +1,41 @@ +# Makefile for GNU tar library. -*- Makefile -*- + +# Copyright (C) 1994, 1995, 1996, 1997, 1999, 2000, 2001, 2003, 2004, +# 2005, 2006 Free Software Foundation, Inc. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License along +# with this program; if not, write to the Free Software Foundation, Inc., +# 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + +include gnulib.mk + +rmt-command.h : Makefile + rm -f $@-t $@ + echo "#ifndef DEFAULT_RMT_COMMAND" >> $@-t + echo "# define DEFAULT_RMT_COMMAND \"$(DEFAULT_RMT_DIR)/`echo rmt | sed '$(transform)'`$(EXEEXT)\"" >> $@-t + echo "#endif" >> $@-t + mv $@-t $@ +BUILT_SOURCES += rmt-command.h +CLEANFILES += rmt-command.h rmt-command.h-t + +noinst_HEADERS += system.h system-ioctl.h rmt.h paxlib.h stdopen.h +libtar_a_SOURCES += \ + paxerror.c paxexit.c paxlib.h paxnames.c \ + prepargs.c prepargs.h \ + rtapelib.c \ + rmt.h \ + stdopen.c stdopen.h \ + system.h system-ioctl.h + +libtar_a_LIBADD += $(LIBOBJS) +libtar_a_DEPENDENCIES += $(LIBOBJS) diff --git a/lib/Makefile.in b/lib/Makefile.in new file mode 100644 index 0000000..28589ae --- /dev/null +++ b/lib/Makefile.in @@ -0,0 +1,1459 @@ +# Makefile.in generated by automake 1.10a from Makefile.am. +# @configure_input@ + +# Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, +# 2003, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. +# This Makefile.in is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY, to the extent permitted by law; without +# even the implied warranty of MERCHANTABILITY or FITNESS FOR A +# PARTICULAR PURPOSE. + +@SET_MAKE@ + +# Makefile for GNU tar library. -*- Makefile -*- + +# Copyright (C) 1994, 1995, 1996, 1997, 1999, 2000, 2001, 2003, 2004, +# 2005, 2006 Free Software Foundation, Inc. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License along +# with this program; if not, write to the Free Software Foundation, Inc., +# 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + +# Copyright (C) 2004-2007 Free Software Foundation, Inc. +# +# This file is free software, distributed under the terms of the GNU +# General Public License. As a special exception to the GNU General +# Public License, this file may be distributed as part of a program +# that contains a configuration script generated by Autoconf, under +# the same distribution terms as the rest of that program. +# +# Generated by gnulib-tool. +# Reproduce by: gnulib-tool --import --dir=. --local-dir=gl --lib=libtar --source-base=.#bootmp/lib --m4-base=.#bootmp/m4 --doc-base=.#bootmp/doc --aux-dir=.#bootmp/build-aux --avoid=lock --avoid=size_max --avoid=xsize --no-libtool --macro-prefix=gl alloca argmatch argp backupfile closeout configmake dirname error exclude exitfail fileblocks fnmatch-gnu ftruncate full-write getdate getline getopt getpagesize gettext gettime hash human inttostr inttypes lchown localcharset memset mkdtemp modechange obstack quote quotearg rmdir rpmatch safe-read save-cwd savedir setenv stat-time stdbool stdint stpcpy strdup strerror strtol strtoul timespec unlinkdir unlocked-io utime utimens version-etc-fsf xalloc xalloc-die xgetcwd xstrtoumax + + + +VPATH = @srcdir@ +pkgdatadir = $(datadir)/@PACKAGE@ +pkgincludedir = $(includedir)/@PACKAGE@ +pkglibdir = $(libdir)/@PACKAGE@ +pkglibexecdir = $(libexecdir)/@PACKAGE@ +am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd +install_sh_DATA = $(install_sh) -c -m 644 +install_sh_PROGRAM = $(install_sh) -c +install_sh_SCRIPT = $(install_sh) -c +INSTALL_HEADER = $(INSTALL_DATA) +transform = $(program_transform_name) +NORMAL_INSTALL = : +PRE_INSTALL = : +POST_INSTALL = : +NORMAL_UNINSTALL = : +PRE_UNINSTALL = : +POST_UNINSTALL = : +build_triplet = @build@ +host_triplet = @host@ +DIST_COMMON = $(noinst_HEADERS) $(srcdir)/Makefile.am \ + $(srcdir)/Makefile.in $(srcdir)/gnulib.mk alloca.c getdate.c \ + waitpid.c +subdir = lib +ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 +am__aclocal_m4_deps = $(top_srcdir)/m4/absolute-header.m4 \ + $(top_srcdir)/m4/alloca.m4 $(top_srcdir)/m4/allocsa.m4 \ + $(top_srcdir)/m4/argmatch.m4 $(top_srcdir)/m4/argp.m4 \ + $(top_srcdir)/m4/backupfile.m4 $(top_srcdir)/m4/bison.m4 \ + $(top_srcdir)/m4/canonicalize-lgpl.m4 \ + $(top_srcdir)/m4/chdir-long.m4 $(top_srcdir)/m4/chown.m4 \ + $(top_srcdir)/m4/clock_time.m4 \ + $(top_srcdir)/m4/close-stream.m4 $(top_srcdir)/m4/closeout.m4 \ + $(top_srcdir)/m4/codeset.m4 $(top_srcdir)/m4/d-ino.m4 \ + $(top_srcdir)/m4/dirfd.m4 $(top_srcdir)/m4/dirname.m4 \ + $(top_srcdir)/m4/dos.m4 $(top_srcdir)/m4/double-slash-root.m4 \ + $(top_srcdir)/m4/dup2.m4 $(top_srcdir)/m4/eealloc.m4 \ + $(top_srcdir)/m4/eoverflow.m4 $(top_srcdir)/m4/error.m4 \ + $(top_srcdir)/m4/exclude.m4 $(top_srcdir)/m4/exitfail.m4 \ + $(top_srcdir)/m4/extensions.m4 $(top_srcdir)/m4/fchdir.m4 \ + $(top_srcdir)/m4/fcntl-safer.m4 $(top_srcdir)/m4/fcntl_h.m4 \ + $(top_srcdir)/m4/fileblocks.m4 $(top_srcdir)/m4/float_h.m4 \ + $(top_srcdir)/m4/fnmatch.m4 $(top_srcdir)/m4/fpending.m4 \ + $(top_srcdir)/m4/ftruncate.m4 \ + $(top_srcdir)/m4/getcwd-abort-bug.m4 \ + $(top_srcdir)/m4/getcwd-path-max.m4 $(top_srcdir)/m4/getcwd.m4 \ + $(top_srcdir)/m4/getdate.m4 $(top_srcdir)/m4/getdelim.m4 \ + $(top_srcdir)/m4/getline.m4 $(top_srcdir)/m4/getopt.m4 \ + $(top_srcdir)/m4/getpagesize.m4 $(top_srcdir)/m4/gettext.m4 \ + $(top_srcdir)/m4/gettime.m4 $(top_srcdir)/m4/gettimeofday.m4 \ + $(top_srcdir)/m4/glibc21.m4 $(top_srcdir)/m4/gnulib-common.m4 \ + $(top_srcdir)/m4/gnulib-comp.m4 $(top_srcdir)/m4/hash.m4 \ + $(top_srcdir)/m4/human.m4 $(top_srcdir)/m4/iconv.m4 \ + $(top_srcdir)/m4/inline.m4 $(top_srcdir)/m4/intmax_t.m4 \ + $(top_srcdir)/m4/inttostr.m4 $(top_srcdir)/m4/inttypes-pri.m4 \ + $(top_srcdir)/m4/inttypes.m4 $(top_srcdir)/m4/inttypes_h.m4 \ + $(top_srcdir)/m4/lchown.m4 $(top_srcdir)/m4/lib-ld.m4 \ + $(top_srcdir)/m4/lib-link.m4 $(top_srcdir)/m4/lib-prefix.m4 \ + $(top_srcdir)/m4/localcharset.m4 $(top_srcdir)/m4/longlong.m4 \ + $(top_srcdir)/m4/lstat.m4 $(top_srcdir)/m4/mbchar.m4 \ + $(top_srcdir)/m4/mbiter.m4 $(top_srcdir)/m4/mbrtowc.m4 \ + $(top_srcdir)/m4/mbscasecmp.m4 $(top_srcdir)/m4/mbstate_t.m4 \ + $(top_srcdir)/m4/memchr.m4 $(top_srcdir)/m4/mempcpy.m4 \ + $(top_srcdir)/m4/memrchr.m4 $(top_srcdir)/m4/memset.m4 \ + $(top_srcdir)/m4/mkdtemp.m4 $(top_srcdir)/m4/mktime.m4 \ + $(top_srcdir)/m4/modechange.m4 $(top_srcdir)/m4/nls.m4 \ + $(top_srcdir)/m4/openat.m4 $(top_srcdir)/m4/pathmax.m4 \ + $(top_srcdir)/m4/paxutils.m4 $(top_srcdir)/m4/po.m4 \ + $(top_srcdir)/m4/progtest.m4 $(top_srcdir)/m4/quote.m4 \ + $(top_srcdir)/m4/quotearg.m4 $(top_srcdir)/m4/readlink.m4 \ + $(top_srcdir)/m4/regex.m4 $(top_srcdir)/m4/rmdir.m4 \ + $(top_srcdir)/m4/rmt.m4 $(top_srcdir)/m4/rpmatch.m4 \ + $(top_srcdir)/m4/rtapelib.m4 $(top_srcdir)/m4/safe-read.m4 \ + $(top_srcdir)/m4/safe-write.m4 $(top_srcdir)/m4/save-cwd.m4 \ + $(top_srcdir)/m4/savedir.m4 $(top_srcdir)/m4/setenv.m4 \ + $(top_srcdir)/m4/sleep.m4 $(top_srcdir)/m4/ssize_t.m4 \ + $(top_srcdir)/m4/stat-time.m4 $(top_srcdir)/m4/stdarg.m4 \ + $(top_srcdir)/m4/stdbool.m4 $(top_srcdir)/m4/stdint.m4 \ + $(top_srcdir)/m4/stdint_h.m4 $(top_srcdir)/m4/stdio_h.m4 \ + $(top_srcdir)/m4/stdlib_h.m4 $(top_srcdir)/m4/stpcpy.m4 \ + $(top_srcdir)/m4/strcase.m4 $(top_srcdir)/m4/strchrnul.m4 \ + $(top_srcdir)/m4/strdup.m4 $(top_srcdir)/m4/strerror.m4 \ + $(top_srcdir)/m4/string_h.m4 $(top_srcdir)/m4/strndup.m4 \ + $(top_srcdir)/m4/strnlen.m4 $(top_srcdir)/m4/strtoimax.m4 \ + $(top_srcdir)/m4/strtol.m4 $(top_srcdir)/m4/strtoll.m4 \ + $(top_srcdir)/m4/strtoul.m4 $(top_srcdir)/m4/strtoull.m4 \ + $(top_srcdir)/m4/strtoumax.m4 $(top_srcdir)/m4/sys_stat_h.m4 \ + $(top_srcdir)/m4/sys_time_h.m4 $(top_srcdir)/m4/sysexits.m4 \ + $(top_srcdir)/m4/system.m4 $(top_srcdir)/m4/tempname.m4 \ + $(top_srcdir)/m4/time_h.m4 $(top_srcdir)/m4/time_r.m4 \ + $(top_srcdir)/m4/timespec.m4 $(top_srcdir)/m4/tm_gmtoff.m4 \ + $(top_srcdir)/m4/unistd-safer.m4 $(top_srcdir)/m4/unistd_h.m4 \ + $(top_srcdir)/m4/unlinkdir.m4 $(top_srcdir)/m4/unlocked-io.m4 \ + $(top_srcdir)/m4/utimbuf.m4 $(top_srcdir)/m4/utime.m4 \ + $(top_srcdir)/m4/utimens.m4 $(top_srcdir)/m4/utimes-null.m4 \ + $(top_srcdir)/m4/utimes.m4 $(top_srcdir)/m4/vasnprintf.m4 \ + $(top_srcdir)/m4/vsnprintf.m4 $(top_srcdir)/m4/wchar.m4 \ + $(top_srcdir)/m4/wchar_t.m4 $(top_srcdir)/m4/wctype.m4 \ + $(top_srcdir)/m4/wcwidth.m4 $(top_srcdir)/m4/wint_t.m4 \ + $(top_srcdir)/m4/xalloc.m4 $(top_srcdir)/m4/xgetcwd.m4 \ + $(top_srcdir)/m4/xstrndup.m4 $(top_srcdir)/m4/xstrtol.m4 \ + $(top_srcdir)/configure.ac +am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ + $(ACLOCAL_M4) +mkinstalldirs = $(SHELL) $(top_srcdir)/build-aux/mkinstalldirs +CONFIG_HEADER = $(top_builddir)/config.h +CONFIG_CLEAN_FILES = +LIBRARIES = $(noinst_LIBRARIES) +AR = ar +ARFLAGS = cru +libtar_a_AR = $(AR) $(ARFLAGS) +am__DEPENDENCIES_1 = +am_libtar_a_OBJECTS = allocsa.$(OBJEXT) argp-ba.$(OBJEXT) \ + argp-eexst.$(OBJEXT) argp-fmtstream.$(OBJEXT) \ + argp-fs-xinl.$(OBJEXT) argp-help.$(OBJEXT) \ + argp-parse.$(OBJEXT) argp-pin.$(OBJEXT) argp-pv.$(OBJEXT) \ + argp-pvh.$(OBJEXT) argp-xinl.$(OBJEXT) full-write.$(OBJEXT) \ + getdate.$(OBJEXT) localcharset.$(OBJEXT) mbchar.$(OBJEXT) \ + mbscasecmp.$(OBJEXT) openat-die.$(OBJEXT) strnlen1.$(OBJEXT) \ + version-etc.$(OBJEXT) version-etc-fsf.$(OBJEXT) \ + xalloc-die.$(OBJEXT) xstrndup.$(OBJEXT) xstrtoumax.$(OBJEXT) \ + paxerror.$(OBJEXT) paxexit.$(OBJEXT) paxnames.$(OBJEXT) \ + prepargs.$(OBJEXT) rtapelib.$(OBJEXT) stdopen.$(OBJEXT) +libtar_a_OBJECTS = $(am_libtar_a_OBJECTS) +LTLIBRARIES = $(noinst_LTLIBRARIES) +DEFAULT_INCLUDES = -I. -I$(top_builddir)@am__isrc@ +depcomp = $(SHELL) $(top_srcdir)/build-aux/depcomp +am__depfiles_maybe = depfiles +COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ + $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) +CCLD = $(CC) +LINK = $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) $(LDFLAGS) -o $@ +YACCCOMPILE = $(YACC) $(YFLAGS) $(AM_YFLAGS) +YLWRAP = $(top_srcdir)/build-aux/ylwrap +SOURCES = $(libtar_a_SOURCES) $(EXTRA_libtar_a_SOURCES) +DIST_SOURCES = $(libtar_a_SOURCES) $(EXTRA_libtar_a_SOURCES) +HEADERS = $(noinst_HEADERS) +ETAGS = etags +CTAGS = ctags +DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) +ABSOLUTE_DIRENT_H = @ABSOLUTE_DIRENT_H@ +ABSOLUTE_FCNTL_H = @ABSOLUTE_FCNTL_H@ +ABSOLUTE_FLOAT_H = @ABSOLUTE_FLOAT_H@ +ABSOLUTE_INTTYPES_H = @ABSOLUTE_INTTYPES_H@ +ABSOLUTE_STDINT_H = @ABSOLUTE_STDINT_H@ +ABSOLUTE_STDIO_H = @ABSOLUTE_STDIO_H@ +ABSOLUTE_STDLIB_H = @ABSOLUTE_STDLIB_H@ +ABSOLUTE_STRING_H = @ABSOLUTE_STRING_H@ +ABSOLUTE_SYSEXITS_H = @ABSOLUTE_SYSEXITS_H@ +ABSOLUTE_SYS_STAT_H = @ABSOLUTE_SYS_STAT_H@ +ABSOLUTE_SYS_TIME_H = @ABSOLUTE_SYS_TIME_H@ +ABSOLUTE_TIME_H = @ABSOLUTE_TIME_H@ +ABSOLUTE_UNISTD_H = @ABSOLUTE_UNISTD_H@ +ABSOLUTE_WCHAR_H = @ABSOLUTE_WCHAR_H@ +ABSOLUTE_WCTYPE_H = @ABSOLUTE_WCTYPE_H@ +ACLOCAL = @ACLOCAL@ +ALLOCA = @ALLOCA@ +ALLOCA_H = @ALLOCA_H@ +AMTAR = @AMTAR@ +AUTOCONF = @AUTOCONF@ +AUTOHEADER = @AUTOHEADER@ +AUTOM4TE = @AUTOM4TE@ +AUTOMAKE = @AUTOMAKE@ +AWK = @AWK@ +BACKUP_LIBEXEC_SCRIPTS = @BACKUP_LIBEXEC_SCRIPTS@ +BACKUP_SBIN_SCRIPTS = @BACKUP_SBIN_SCRIPTS@ +BACKUP_SED_COND = @BACKUP_SED_COND@ +BITSIZEOF_PTRDIFF_T = @BITSIZEOF_PTRDIFF_T@ +BITSIZEOF_SIG_ATOMIC_T = @BITSIZEOF_SIG_ATOMIC_T@ +BITSIZEOF_SIZE_T = @BITSIZEOF_SIZE_T@ +BITSIZEOF_WCHAR_T = @BITSIZEOF_WCHAR_T@ +BITSIZEOF_WINT_T = @BITSIZEOF_WINT_T@ +CC = @CC@ +CCDEPMODE = @CCDEPMODE@ +CFLAGS = @CFLAGS@ +CPP = @CPP@ +CPPFLAGS = @CPPFLAGS@ +CYGPATH_W = @CYGPATH_W@ +DEFAULT_ARCHIVE = @DEFAULT_ARCHIVE@ +DEFAULT_ARCHIVE_FORMAT = @DEFAULT_ARCHIVE_FORMAT@ +DEFAULT_BLOCKING = @DEFAULT_BLOCKING@ +DEFAULT_QUOTING_STYLE = @DEFAULT_QUOTING_STYLE@ +DEFAULT_RMT_COMMAND = @DEFAULT_RMT_COMMAND@ +DEFAULT_RMT_DIR = @DEFAULT_RMT_DIR@ +DEFS = @DEFS@ +DEPDIR = @DEPDIR@ +DIRENT_H = @DIRENT_H@ +ECHO_C = @ECHO_C@ +ECHO_N = @ECHO_N@ +ECHO_T = @ECHO_T@ +EGREP = @EGREP@ +EOVERFLOW = @EOVERFLOW@ +EXEEXT = @EXEEXT@ +FCNTL_H = @FCNTL_H@ +FLOAT_H = @FLOAT_H@ +FNMATCH_H = @FNMATCH_H@ +GETOPT_H = @GETOPT_H@ +GLIBC21 = @GLIBC21@ +GMSGFMT = @GMSGFMT@ +GMSGFMT_015 = @GMSGFMT_015@ +GNULIB_CHOWN = @GNULIB_CHOWN@ +GNULIB_DUP2 = @GNULIB_DUP2@ +GNULIB_FCHDIR = @GNULIB_FCHDIR@ +GNULIB_FFLUSH = @GNULIB_FFLUSH@ +GNULIB_FPRINTF_POSIX = @GNULIB_FPRINTF_POSIX@ +GNULIB_FSEEK = @GNULIB_FSEEK@ +GNULIB_FSEEKO = @GNULIB_FSEEKO@ +GNULIB_FTELL = @GNULIB_FTELL@ +GNULIB_FTELLO = @GNULIB_FTELLO@ +GNULIB_FTRUNCATE = @GNULIB_FTRUNCATE@ +GNULIB_GETCWD = @GNULIB_GETCWD@ +GNULIB_GETLOGIN_R = @GNULIB_GETLOGIN_R@ +GNULIB_GETSUBOPT = @GNULIB_GETSUBOPT@ +GNULIB_IMAXABS = @GNULIB_IMAXABS@ +GNULIB_IMAXDIV = @GNULIB_IMAXDIV@ +GNULIB_LSEEK = @GNULIB_LSEEK@ +GNULIB_MBSCASECMP = @GNULIB_MBSCASECMP@ +GNULIB_MBSCASESTR = @GNULIB_MBSCASESTR@ +GNULIB_MBSCHR = @GNULIB_MBSCHR@ +GNULIB_MBSCSPN = @GNULIB_MBSCSPN@ +GNULIB_MBSLEN = @GNULIB_MBSLEN@ +GNULIB_MBSNCASECMP = @GNULIB_MBSNCASECMP@ +GNULIB_MBSPBRK = @GNULIB_MBSPBRK@ +GNULIB_MBSPCASECMP = @GNULIB_MBSPCASECMP@ +GNULIB_MBSRCHR = @GNULIB_MBSRCHR@ +GNULIB_MBSSEP = @GNULIB_MBSSEP@ +GNULIB_MBSSPN = @GNULIB_MBSSPN@ +GNULIB_MBSSTR = @GNULIB_MBSSTR@ +GNULIB_MBSTOK_R = @GNULIB_MBSTOK_R@ +GNULIB_MEMMEM = @GNULIB_MEMMEM@ +GNULIB_MEMPCPY = @GNULIB_MEMPCPY@ +GNULIB_MEMRCHR = @GNULIB_MEMRCHR@ +GNULIB_MKDTEMP = @GNULIB_MKDTEMP@ +GNULIB_MKSTEMP = @GNULIB_MKSTEMP@ +GNULIB_PRINTF_POSIX = @GNULIB_PRINTF_POSIX@ +GNULIB_READLINK = @GNULIB_READLINK@ +GNULIB_SLEEP = @GNULIB_SLEEP@ +GNULIB_SNPRINTF = @GNULIB_SNPRINTF@ +GNULIB_SPRINTF_POSIX = @GNULIB_SPRINTF_POSIX@ +GNULIB_STPCPY = @GNULIB_STPCPY@ +GNULIB_STPNCPY = @GNULIB_STPNCPY@ +GNULIB_STRCASESTR = @GNULIB_STRCASESTR@ +GNULIB_STRCHRNUL = @GNULIB_STRCHRNUL@ +GNULIB_STRDUP = @GNULIB_STRDUP@ +GNULIB_STRNDUP = @GNULIB_STRNDUP@ +GNULIB_STRNLEN = @GNULIB_STRNLEN@ +GNULIB_STRPBRK = @GNULIB_STRPBRK@ +GNULIB_STRSEP = @GNULIB_STRSEP@ +GNULIB_STRTOIMAX = @GNULIB_STRTOIMAX@ +GNULIB_STRTOK_R = @GNULIB_STRTOK_R@ +GNULIB_STRTOUMAX = @GNULIB_STRTOUMAX@ +GNULIB_VASPRINTF = @GNULIB_VASPRINTF@ +GNULIB_VFPRINTF_POSIX = @GNULIB_VFPRINTF_POSIX@ +GNULIB_VPRINTF_POSIX = @GNULIB_VPRINTF_POSIX@ +GNULIB_VSNPRINTF = @GNULIB_VSNPRINTF@ +GNULIB_VSPRINTF_POSIX = @GNULIB_VSPRINTF_POSIX@ +GREP = @GREP@ +HAVE_DECL_GETLOGIN_R = @HAVE_DECL_GETLOGIN_R@ +HAVE_DECL_IMAXABS = @HAVE_DECL_IMAXABS@ +HAVE_DECL_IMAXDIV = @HAVE_DECL_IMAXDIV@ +HAVE_DECL_MEMMEM = @HAVE_DECL_MEMMEM@ +HAVE_DECL_MEMRCHR = @HAVE_DECL_MEMRCHR@ +HAVE_DECL_MKDIR = @HAVE_DECL_MKDIR@ +HAVE_DECL_SNPRINTF = @HAVE_DECL_SNPRINTF@ +HAVE_DECL_STRDUP = @HAVE_DECL_STRDUP@ +HAVE_DECL_STRNCASECMP = @HAVE_DECL_STRNCASECMP@ +HAVE_DECL_STRNDUP = @HAVE_DECL_STRNDUP@ +HAVE_DECL_STRNLEN = @HAVE_DECL_STRNLEN@ +HAVE_DECL_STRTOIMAX = @HAVE_DECL_STRTOIMAX@ +HAVE_DECL_STRTOK_R = @HAVE_DECL_STRTOK_R@ +HAVE_DECL_STRTOUMAX = @HAVE_DECL_STRTOUMAX@ +HAVE_DECL_VSNPRINTF = @HAVE_DECL_VSNPRINTF@ +HAVE_DUP2 = @HAVE_DUP2@ +HAVE_FSEEKO = @HAVE_FSEEKO@ +HAVE_FTELLO = @HAVE_FTELLO@ +HAVE_FTRUNCATE = @HAVE_FTRUNCATE@ +HAVE_GETSUBOPT = @HAVE_GETSUBOPT@ +HAVE_INTTYPES_H = @HAVE_INTTYPES_H@ +HAVE_IO_H = @HAVE_IO_H@ +HAVE_ISWCNTRL = @HAVE_ISWCNTRL@ +HAVE_LONG_LONG_INT = @HAVE_LONG_LONG_INT@ +HAVE_LSTAT = @HAVE_LSTAT@ +HAVE_MEMPCPY = @HAVE_MEMPCPY@ +HAVE_MKDTEMP = @HAVE_MKDTEMP@ +HAVE_READLINK = @HAVE_READLINK@ +HAVE_SIGNED_SIG_ATOMIC_T = @HAVE_SIGNED_SIG_ATOMIC_T@ +HAVE_SIGNED_WCHAR_T = @HAVE_SIGNED_WCHAR_T@ +HAVE_SIGNED_WINT_T = @HAVE_SIGNED_WINT_T@ +HAVE_SLEEP = @HAVE_SLEEP@ +HAVE_STDINT_H = @HAVE_STDINT_H@ +HAVE_STPCPY = @HAVE_STPCPY@ +HAVE_STPNCPY = @HAVE_STPNCPY@ +HAVE_STRCASECMP = @HAVE_STRCASECMP@ +HAVE_STRCASESTR = @HAVE_STRCASESTR@ +HAVE_STRCHRNUL = @HAVE_STRCHRNUL@ +HAVE_STRNDUP = @HAVE_STRNDUP@ +HAVE_STRPBRK = @HAVE_STRPBRK@ +HAVE_STRSEP = @HAVE_STRSEP@ +HAVE_STRUCT_TIMEVAL = @HAVE_STRUCT_TIMEVAL@ +HAVE_SYSEXITS_H = @HAVE_SYSEXITS_H@ +HAVE_SYS_BITYPES_H = @HAVE_SYS_BITYPES_H@ +HAVE_SYS_INTTYPES_H = @HAVE_SYS_INTTYPES_H@ +HAVE_SYS_TIME_H = @HAVE_SYS_TIME_H@ +HAVE_SYS_TYPES_H = @HAVE_SYS_TYPES_H@ +HAVE_UNISTD_H = @HAVE_UNISTD_H@ +HAVE_UNSIGNED_LONG_LONG_INT = @HAVE_UNSIGNED_LONG_LONG_INT@ +HAVE_VASPRINTF = @HAVE_VASPRINTF@ +HAVE_WCTYPE_H = @HAVE_WCTYPE_H@ +HAVE_WINT_T = @HAVE_WINT_T@ +HAVE__BOOL = @HAVE__BOOL@ +INSTALL = @INSTALL@ +INSTALL_DATA = @INSTALL_DATA@ +INSTALL_PROGRAM = @INSTALL_PROGRAM@ +INSTALL_SCRIPT = @INSTALL_SCRIPT@ +INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ +INTLLIBS = @INTLLIBS@ +INTL_MACOSX_LIBS = @INTL_MACOSX_LIBS@ +INTTYPES_H = @INTTYPES_H@ +LDFLAGS = @LDFLAGS@ +LIBICONV = @LIBICONV@ +LIBINTL = @LIBINTL@ +LIBOBJS = @LIBOBJS@ +LIBS = @LIBS@ +LIBTAR_LIBDEPS = @LIBTAR_LIBDEPS@ +LIBTAR_LTLIBDEPS = @LIBTAR_LTLIBDEPS@ +LIB_CLOCK_GETTIME = @LIB_CLOCK_GETTIME@ +LIB_SETSOCKOPT = @LIB_SETSOCKOPT@ +LTLIBICONV = @LTLIBICONV@ +LTLIBINTL = @LTLIBINTL@ +LTLIBOBJS = @LTLIBOBJS@ +MAKEINFO = @MAKEINFO@ +MKDIR_P = @MKDIR_P@ +MSGFMT = @MSGFMT@ +MSGFMT_015 = @MSGFMT_015@ +MSGMERGE = @MSGMERGE@ +OBJEXT = @OBJEXT@ +PACKAGE = @PACKAGE@ +PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ +PACKAGE_NAME = @PACKAGE_NAME@ +PACKAGE_STRING = @PACKAGE_STRING@ +PACKAGE_TARNAME = @PACKAGE_TARNAME@ +PACKAGE_VERSION = @PACKAGE_VERSION@ +PATH_SEPARATOR = @PATH_SEPARATOR@ +POSUB = @POSUB@ +PRIPTR_PREFIX = @PRIPTR_PREFIX@ +PRI_MACROS_BROKEN = @PRI_MACROS_BROKEN@ +PTRDIFF_T_SUFFIX = @PTRDIFF_T_SUFFIX@ +PU_RMT_PROG = @PU_RMT_PROG@ +RANLIB = @RANLIB@ +REPLACE_CHOWN = @REPLACE_CHOWN@ +REPLACE_FCHDIR = @REPLACE_FCHDIR@ +REPLACE_FFLUSH = @REPLACE_FFLUSH@ +REPLACE_FPRINTF = @REPLACE_FPRINTF@ +REPLACE_FSEEK = @REPLACE_FSEEK@ +REPLACE_FSEEKO = @REPLACE_FSEEKO@ +REPLACE_FTELL = @REPLACE_FTELL@ +REPLACE_FTELLO = @REPLACE_FTELLO@ +REPLACE_GETCWD = @REPLACE_GETCWD@ +REPLACE_GETTIMEOFDAY = @REPLACE_GETTIMEOFDAY@ +REPLACE_LOCALTIME_R = @REPLACE_LOCALTIME_R@ +REPLACE_LSEEK = @REPLACE_LSEEK@ +REPLACE_MKSTEMP = @REPLACE_MKSTEMP@ +REPLACE_NANOSLEEP = @REPLACE_NANOSLEEP@ +REPLACE_PRINTF = @REPLACE_PRINTF@ +REPLACE_SNPRINTF = @REPLACE_SNPRINTF@ +REPLACE_SPRINTF = @REPLACE_SPRINTF@ +REPLACE_STRPTIME = @REPLACE_STRPTIME@ +REPLACE_TIMEGM = @REPLACE_TIMEGM@ +REPLACE_VASPRINTF = @REPLACE_VASPRINTF@ +REPLACE_VFPRINTF = @REPLACE_VFPRINTF@ +REPLACE_VPRINTF = @REPLACE_VPRINTF@ +REPLACE_VSNPRINTF = @REPLACE_VSNPRINTF@ +REPLACE_VSPRINTF = @REPLACE_VSPRINTF@ +SET_MAKE = @SET_MAKE@ +SHELL = @SHELL@ +SIG_ATOMIC_T_SUFFIX = @SIG_ATOMIC_T_SUFFIX@ +SIZE_T_SUFFIX = @SIZE_T_SUFFIX@ +STDBOOL_H = @STDBOOL_H@ +STDINT_H = @STDINT_H@ +STRIP = @STRIP@ +SYSEXITS_H = @SYSEXITS_H@ +SYS_STAT_H = @SYS_STAT_H@ +SYS_TIME_H = @SYS_TIME_H@ +SYS_TIME_H_DEFINES_STRUCT_TIMESPEC = @SYS_TIME_H_DEFINES_STRUCT_TIMESPEC@ +TIME_H_DEFINES_STRUCT_TIMESPEC = @TIME_H_DEFINES_STRUCT_TIMESPEC@ +USE_NLS = @USE_NLS@ +VERSION = @VERSION@ +WCHAR_H = @WCHAR_H@ +WCHAR_T_SUFFIX = @WCHAR_T_SUFFIX@ +WCTYPE_H = @WCTYPE_H@ +WINT_T_SUFFIX = @WINT_T_SUFFIX@ +XGETTEXT = @XGETTEXT@ +XGETTEXT_015 = @XGETTEXT_015@ +YACC = @YACC@ +YFLAGS = @YFLAGS@ +abs_builddir = @abs_builddir@ +abs_srcdir = @abs_srcdir@ +abs_top_builddir = @abs_top_builddir@ +abs_top_srcdir = @abs_top_srcdir@ +ac_ct_CC = @ac_ct_CC@ +am__include = @am__include@ +am__leading_dot = @am__leading_dot@ +am__quote = @am__quote@ +am__tar = @am__tar@ +am__untar = @am__untar@ +bindir = @bindir@ +build = @build@ +build_alias = @build_alias@ +build_cpu = @build_cpu@ +build_os = @build_os@ +build_vendor = @build_vendor@ +builddir = @builddir@ +datadir = @datadir@ +datarootdir = @datarootdir@ +docdir = @docdir@ +dvidir = @dvidir@ +exec_prefix = @exec_prefix@ +gl_LIBOBJS = @gl_LIBOBJS@ +gl_LTLIBOBJS = @gl_LTLIBOBJS@ +host = @host@ +host_alias = @host_alias@ +host_cpu = @host_cpu@ +host_os = @host_os@ +host_vendor = @host_vendor@ +htmldir = @htmldir@ +includedir = @includedir@ +infodir = @infodir@ +install_sh = @install_sh@ +libdir = @libdir@ +libexecdir = @libexecdir@ +localedir = @localedir@ +localstatedir = @localstatedir@ +mandir = @mandir@ +mkdir_p = @mkdir_p@ +oldincludedir = @oldincludedir@ +pdfdir = @pdfdir@ +prefix = @prefix@ +program_transform_name = @program_transform_name@ +psdir = @psdir@ +sbindir = @sbindir@ +sharedstatedir = @sharedstatedir@ +srcdir = @srcdir@ +sysconfdir = @sysconfdir@ +target_alias = @target_alias@ +top_builddir = @top_builddir@ +top_srcdir = @top_srcdir@ +AUTOMAKE_OPTIONS = 1.5 gnits +noinst_HEADERS = system.h system-ioctl.h rmt.h paxlib.h stdopen.h +noinst_LIBRARIES = libtar.a +noinst_LTLIBRARIES = +EXTRA_DIST = alloca.c alloca_.h allocsa.valgrind argmatch.c argmatch.h \ + backupfile.c backupfile.h canonicalize-lgpl.c canonicalize.h \ + chdir-long.c chdir-long.h chown.c fchown-stub.c close-stream.c \ + close-stream.h closeout.c closeout.h dirfd.c dirfd.h \ + basename.c dirname.c dirname.h stripslash.c dup2.c error.c \ + error.h exclude.c exclude.h exitfail.c exitfail.h dirent_.h \ + fchdir.c fcntl_.h creat-safer.c fcntl--.h fcntl-safer.h \ + open-safer.c fileblocks.c float_.h fnmatch.c fnmatch_.h \ + fnmatch_loop.c __fpending.c __fpending.h ftruncate.c getcwd.c \ + getdate.c getdate.h getdelim.c getdelim.h getline.c getline.h \ + getopt.c getopt1.c getopt_.h getopt_int.h getpagesize.h \ + gettime.c gettimeofday.c hash.c hash.h human.c human.h \ + intprops.h imaxtostr.c inttostr.c inttostr.h offtostr.c \ + uinttostr.c umaxtostr.c inttypes_.h lchown.c lchown.h \ + config.charset ref-add.sin ref-del.sin lstat.c lstat.h \ + malloc.c mbchar.h memchr.c mempcpy.c memrchr.c memset.c \ + mkdtemp.c mktime.c modechange.c modechange.h obstack.c \ + obstack.h at-func.c fchmodat.c fchownat.c fstatat.c mkdirat.c \ + openat-priv.h openat-proc.c openat.c openat.h pathmax.h \ + quote.c quote.h quotearg.c quotearg.h readlink.c regcomp.c \ + regex.c regex.h regex_internal.c regex_internal.h regexec.c \ + rmdir.c rpmatch.c safe-read.c safe-read.h safe-write.c \ + safe-write.h same-inode.h save-cwd.c save-cwd.h savedir.c \ + savedir.h setenv.c setenv.h unsetenv.c sleep.c stat-macros.h \ + stat-time.h stdbool_.h stdint_.h stdio_.h stdlib_.h stpcpy.c \ + strcasecmp.c strncasecmp.c strchrnul.c strdup.c strerror.c \ + string_.h strndup.c strnlen.c strtoimax.c strtol.c strtoll.c \ + strtoul.c strtoull.c strtoumax.c sys_stat_.h sys_time_.h \ + sysexits_.h tempname.c tempname.h time_.h time_r.c timespec.h \ + unistd_.h dup-safer.c fd-safer.c pipe-safer.c unistd--.h \ + unistd-safer.h unlinkdir.c unlinkdir.h unlocked-io.h utime.c \ + utimens.c utimens.h asnprintf.c float+.h printf-args.c \ + printf-args.h printf-parse.c printf-parse.h vasnprintf.c \ + vasnprintf.h vsnprintf.c wchar_.h wctype_.h xalloc.h xmalloc.c \ + xgetcwd.c xgetcwd.h xstrtol.c xstrtol.h xstrtoul.c +BUILT_SOURCES = $(ALLOCA_H) configmake.h $(DIRENT_H) $(FCNTL_H) \ + $(FLOAT_H) $(FNMATCH_H) getdate.c $(GETOPT_H) $(INTTYPES_H) \ + $(STDBOOL_H) $(STDINT_H) stdio.h stdlib.h string.h \ + $(SYS_STAT_H) $(SYS_TIME_H) $(SYSEXITS_H) time.h unistd.h \ + $(WCHAR_H) $(WCTYPE_H) rmt-command.h +SUFFIXES = .sed .sin +MOSTLYCLEANFILES = core *.stackdump alloca.h alloca.h-t dirent.h \ + dirent.h-t fcntl.h fcntl.h-t float.h float.h-t fnmatch.h \ + fnmatch.h-t getopt.h getopt.h-t inttypes.h inttypes.h-t \ + stdbool.h stdbool.h-t stdint.h stdint.h-t stdio.h stdio.h-t \ + stdlib.h stdlib.h-t string.h string.h-t sys/stat.h \ + sys/stat.h-t sys/time.h sys/time.h-t sysexits.h sysexits.h-t \ + time.h time.h-t unistd.h unistd.h-t wchar.h wchar.h-t wctype.h \ + wctype.h-t +MOSTLYCLEANDIRS = sys +CLEANFILES = configmake.h configmake.h-t charset.alias ref-add.sed \ + ref-del.sed rmt-command.h rmt-command.h-t +DISTCLEANFILES = +MAINTAINERCLEANFILES = getdate.c +AM_CPPFLAGS = + +# This is for those projects which use "gettextize --intl" to put a source-code +# copy of libintl into their package. In such projects, every Makefile.am needs +# -I$(top_builddir)/intl, so that <libintl.h> can be found in this directory. +# For the Makefile.ams in other directories it is the maintainer's +# responsibility; for the one from gnulib we do it here. +# This option has no effect when the user disables NLS (because then the intl +# directory contains no libintl.h file) or when the project does not use +# "gettextize --intl". +#AM_CPPFLAGS += -I$(top_builddir)/intl +libtar_a_SOURCES = allocsa.h allocsa.c argp.h argp-ba.c argp-eexst.c \ + argp-fmtstream.c argp-fmtstream.h argp-fs-xinl.c argp-help.c \ + argp-namefrob.h argp-parse.c argp-pin.c argp-pv.c argp-pvh.c \ + argp-xinl.c full-write.h full-write.c getdate.y gettext.h \ + localcharset.h localcharset.c mbchar.c mbscasecmp.c mbuiter.h \ + openat-die.c strnlen1.h strnlen1.c verify.h version-etc.h \ + version-etc.c version-etc-fsf.c wcwidth.h xalloc-die.c \ + xstrndup.h xstrndup.c xstrtoumax.c paxerror.c paxexit.c \ + paxlib.h paxnames.c prepargs.c prepargs.h rtapelib.c rmt.h \ + stdopen.c stdopen.h system.h system-ioctl.h +libtar_a_LIBADD = $(gl_LIBOBJS) @ALLOCA@ $(LIBOBJS) +libtar_a_DEPENDENCIES = $(gl_LIBOBJS) @ALLOCA@ $(LIBOBJS) +EXTRA_libtar_a_SOURCES = alloca.c argmatch.c backupfile.c \ + canonicalize-lgpl.c chdir-long.c chown.c fchown-stub.c \ + close-stream.c closeout.c dirfd.c basename.c dirname.c \ + stripslash.c dup2.c error.c exclude.c exitfail.c fchdir.c \ + creat-safer.c open-safer.c fileblocks.c fnmatch.c \ + fnmatch_loop.c __fpending.c ftruncate.c getcwd.c getdelim.c \ + getline.c getopt.c getopt1.c gettime.c gettimeofday.c hash.c \ + human.c imaxtostr.c inttostr.c offtostr.c uinttostr.c \ + umaxtostr.c lchown.c lstat.c malloc.c memchr.c mempcpy.c \ + memrchr.c memset.c mkdtemp.c mktime.c modechange.c obstack.c \ + at-func.c fchmodat.c fchownat.c fstatat.c mkdirat.c \ + openat-proc.c openat.c quote.c quotearg.c readlink.c regcomp.c \ + regex.c regex_internal.c regexec.c rmdir.c rpmatch.c \ + safe-read.c safe-write.c save-cwd.c savedir.c setenv.c \ + unsetenv.c sleep.c stpcpy.c strcasecmp.c strncasecmp.c \ + strchrnul.c strdup.c strerror.c strndup.c strnlen.c \ + strtoimax.c strtol.c strtoll.c strtoul.c strtoull.c \ + strtoumax.c tempname.c time_r.c dup-safer.c fd-safer.c \ + pipe-safer.c unlinkdir.c utime.c utimens.c asnprintf.c \ + printf-args.c printf-parse.c vasnprintf.c vsnprintf.c \ + xmalloc.c xgetcwd.c xstrtol.c xstrtoul.c + +# Use this preprocessor expression to decide whether #include_next works. +# Do not rely on a 'configure'-time test for this, since the expression +# might appear in an installed header, which is used by some other compiler. +HAVE_INCLUDE_NEXT = (__GNUC__ || 60000000 <= __DECC_VER) +LINK_WARNING_H = $(top_srcdir)/build-aux/link-warning.h +charset_alias = $(DESTDIR)$(libdir)/charset.alias +charset_tmp = $(DESTDIR)$(libdir)/charset.tmp +all: $(BUILT_SOURCES) + $(MAKE) $(AM_MAKEFLAGS) all-am + +.SUFFIXES: +.SUFFIXES: .sed .sin .c .o .obj .y +$(srcdir)/Makefile.in: $(srcdir)/Makefile.am $(srcdir)/gnulib.mk $(am__configure_deps) + @for dep in $?; do \ + case '$(am__configure_deps)' in \ + *$$dep*) \ + cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh \ + && exit 0; \ + exit 1;; \ + esac; \ + done; \ + echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnits lib/Makefile'; \ + cd $(top_srcdir) && \ + $(AUTOMAKE) --gnits lib/Makefile +.PRECIOUS: Makefile +Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status + @case '$?' in \ + *config.status*) \ + cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ + *) \ + echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ + cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ + esac; + +$(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) + cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh + +$(top_srcdir)/configure: $(am__configure_deps) + cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh +$(ACLOCAL_M4): $(am__aclocal_m4_deps) + cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh + +clean-noinstLIBRARIES: + -test -z "$(noinst_LIBRARIES)" || rm -f $(noinst_LIBRARIES) +libtar.a: $(libtar_a_OBJECTS) $(libtar_a_DEPENDENCIES) + -rm -f libtar.a + $(libtar_a_AR) libtar.a $(libtar_a_OBJECTS) $(libtar_a_LIBADD) + $(RANLIB) libtar.a + +clean-noinstLTLIBRARIES: + -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) + @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ + dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ + test "$$dir" != "$$p" || dir=.; \ + echo "rm -f \"$${dir}/so_locations\""; \ + rm -f "$${dir}/so_locations"; \ + done + +mostlyclean-compile: + -rm -f *.$(OBJEXT) + +distclean-compile: + -rm -f *.tab.c + +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/alloca.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/waitpid.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/__fpending.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/alloca.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/allocsa.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argmatch.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-ba.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-eexst.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-fmtstream.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-fs-xinl.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-help.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-parse.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-pin.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-pv.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-pvh.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/argp-xinl.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/asnprintf.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/at-func.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/backupfile.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/basename.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/canonicalize-lgpl.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/chdir-long.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/chown.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/close-stream.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/closeout.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/creat-safer.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dirfd.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dirname.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dup-safer.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dup2.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/error.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/exclude.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/exitfail.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fchdir.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fchmodat.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fchown-stub.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fchownat.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fd-safer.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fileblocks.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fnmatch.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fnmatch_loop.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fstatat.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ftruncate.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/full-write.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/getcwd.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/getdate.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/getdelim.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/getline.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/getopt.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/getopt1.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/gettime.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/gettimeofday.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hash.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/human.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/imaxtostr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/inttostr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/lchown.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/localcharset.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/lstat.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/malloc.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mbchar.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mbscasecmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/memchr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mempcpy.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/memrchr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/memset.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mkdirat.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mkdtemp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mktime.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/modechange.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/obstack.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/offtostr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/open-safer.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/openat-die.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/openat-proc.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/openat.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/paxerror.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/paxexit.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/paxnames.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/pipe-safer.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/prepargs.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/printf-args.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/printf-parse.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/quote.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/quotearg.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/readlink.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/regcomp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/regex.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/regex_internal.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/regexec.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rmdir.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rpmatch.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rtapelib.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/safe-read.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/safe-write.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/save-cwd.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/savedir.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/setenv.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/sleep.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/stdopen.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/stpcpy.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strcasecmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strchrnul.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strdup.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strerror.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/stripslash.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strncasecmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strndup.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strnlen.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strnlen1.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strtoimax.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strtol.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strtoll.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strtoul.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strtoull.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strtoumax.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tempname.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/time_r.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/uinttostr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/umaxtostr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/unlinkdir.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/unsetenv.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/utime.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/utimens.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/vasnprintf.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/version-etc-fsf.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/version-etc.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/vsnprintf.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xalloc-die.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xgetcwd.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xmalloc.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xstrndup.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xstrtol.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xstrtoul.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xstrtoumax.Po@am__quote@ + +.c.o: +@am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< +@am__fastdepCC_TRUE@ mv -f $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po +@AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ +@AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ +@am__fastdepCC_FALSE@ $(COMPILE) -c $< + +.c.obj: +@am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` +@am__fastdepCC_TRUE@ mv -f $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po +@AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ +@AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ +@am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` + +.y.c: + $(am__skipyacc) $(SHELL) $(YLWRAP) $< y.tab.c $@ y.tab.h $*.h y.output $*.output -- $(YACCCOMPILE) + +ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) + list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ + unique=`for i in $$list; do \ + if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ + done | \ + $(AWK) ' { files[$$0] = 1; } \ + END { for (i in files) print i; }'`; \ + mkid -fID $$unique +tags: TAGS + +TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ + $(TAGS_FILES) $(LISP) + tags=; \ + here=`pwd`; \ + list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ + unique=`for i in $$list; do \ + if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ + done | \ + $(AWK) ' { files[$$0] = 1; } \ + END { for (i in files) print i; }'`; \ + if test -z "$(ETAGS_ARGS)$$tags$$unique"; then :; else \ + test -n "$$unique" || unique=$$empty_fix; \ + $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ + $$tags $$unique; \ + fi +ctags: CTAGS +CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ + $(TAGS_FILES) $(LISP) + tags=; \ + here=`pwd`; \ + list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ + unique=`for i in $$list; do \ + if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ + done | \ + $(AWK) ' { files[$$0] = 1; } \ + END { for (i in files) print i; }'`; \ + test -z "$(CTAGS_ARGS)$$tags$$unique" \ + || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ + $$tags $$unique + +GTAGS: + here=`$(am__cd) $(top_builddir) && pwd` \ + && cd $(top_srcdir) \ + && gtags -i $(GTAGS_ARGS) $$here + +distclean-tags: + -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags + +distdir: $(DISTFILES) + @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ + topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ + list='$(DISTFILES)'; \ + dist_files=`for file in $$list; do echo $$file; done | \ + sed -e "s|^$$srcdirstrip/||;t" \ + -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ + case $$dist_files in \ + */*) $(MKDIR_P) `echo "$$dist_files" | \ + sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ + sort -u` ;; \ + esac; \ + for file in $$dist_files; do \ + if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ + if test -d $$d/$$file; then \ + dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ + if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ + cp -pR $(srcdir)/$$file $(distdir)$$dir || exit 1; \ + fi; \ + cp -pR $$d/$$file $(distdir)$$dir || exit 1; \ + else \ + test -f $(distdir)/$$file \ + || cp -p $$d/$$file $(distdir)/$$file \ + || exit 1; \ + fi; \ + done +check-am: all-am +check: $(BUILT_SOURCES) + $(MAKE) $(AM_MAKEFLAGS) check-am +all-am: Makefile $(LIBRARIES) $(LTLIBRARIES) $(HEADERS) all-local +installdirs: +install: $(BUILT_SOURCES) + $(MAKE) $(AM_MAKEFLAGS) install-am +install-exec: install-exec-am +install-data: install-data-am +uninstall: uninstall-am + +install-am: all-am + @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am + +installcheck: installcheck-am +install-strip: + $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ + install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ + `test -z '$(STRIP)' || \ + echo "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'"` install +mostlyclean-generic: + -test -z "$(MOSTLYCLEANFILES)" || rm -f $(MOSTLYCLEANFILES) + +clean-generic: + -test -z "$(CLEANFILES)" || rm -f $(CLEANFILES) + +distclean-generic: + -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) + -test -z "$(DISTCLEANFILES)" || rm -f $(DISTCLEANFILES) + +maintainer-clean-generic: + @echo "This command is intended for maintainers to use" + @echo "it deletes files that may require special tools to rebuild." + -rm -f getdate.c + -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) + -test -z "$(MAINTAINERCLEANFILES)" || rm -f $(MAINTAINERCLEANFILES) +clean: clean-am + +clean-am: clean-generic clean-noinstLIBRARIES clean-noinstLTLIBRARIES \ + mostlyclean-am + +distclean: distclean-am + -rm -rf $(DEPDIR) ./$(DEPDIR) + -rm -f Makefile +distclean-am: clean-am distclean-compile distclean-generic \ + distclean-tags + +dvi: dvi-am + +dvi-am: + +html: html-am + +info: info-am + +info-am: + +install-data-am: + +install-dvi: install-dvi-am + +install-exec-am: install-exec-local + +install-html: install-html-am + +install-info: install-info-am + +install-man: + +install-pdf: install-pdf-am + +install-ps: install-ps-am + +installcheck-am: + +maintainer-clean: maintainer-clean-am + -rm -rf $(DEPDIR) ./$(DEPDIR) + -rm -f Makefile +maintainer-clean-am: distclean-am maintainer-clean-generic + +mostlyclean: mostlyclean-am + +mostlyclean-am: mostlyclean-compile mostlyclean-generic \ + mostlyclean-local + +pdf: pdf-am + +pdf-am: + +ps: ps-am + +ps-am: + +uninstall-am: uninstall-local + +.MAKE: install-am install-strip + +.PHONY: CTAGS GTAGS all all-am all-local check check-am clean \ + clean-generic clean-noinstLIBRARIES clean-noinstLTLIBRARIES \ + ctags distclean distclean-compile distclean-generic \ + distclean-tags distdir dvi dvi-am html html-am info info-am \ + install install-am install-data install-data-am install-dvi \ + install-dvi-am install-exec install-exec-am install-exec-local \ + install-html install-html-am install-info install-info-am \ + install-man install-pdf install-pdf-am install-ps \ + install-ps-am install-strip installcheck installcheck-am \ + installdirs maintainer-clean maintainer-clean-generic \ + mostlyclean mostlyclean-compile mostlyclean-generic \ + mostlyclean-local pdf pdf-am ps ps-am tags uninstall \ + uninstall-am uninstall-local + + +# We need the following in order to create <alloca.h> when the system +# doesn't have one that works with the given compiler. +alloca.h: alloca_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/alloca_.h; \ + } > $@-t + mv -f $@-t $@ + +# Retrieve values of the variables through 'configure' followed by +# 'make', not directly through 'configure', so that a user who +# sets some of these variables consistently on the 'make' command +# line gets correct results. +# +# One advantage of this approach, compared to the classical +# approach of adding -DLIBDIR=\"$(libdir)\" etc. to AM_CPPFLAGS, +# is that it protects against the use of undefined variables. +# If, say, $(libdir) is not set in the Makefile, LIBDIR is not +# defined by this module, and code using LIBDIR gives a +# compilation error. +# +# Another advantage is that 'make' output is shorter. +# +# Listed in the same order as the GNU makefile conventions. +# The Automake-defined pkg* macros are appended, in the order +# listed in the Automake 1.10a+ documentation. +configmake.h: Makefile + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + echo '#define PREFIX "$(prefix)"'; \ + echo '#define EXEC_PREFIX "$(exec_prefix)"'; \ + echo '#define BINDIR "$(bindir)"'; \ + echo '#define SBINDIR "$(sbindir)"'; \ + echo '#define LIBEXECDIR "$(libexecdir)"'; \ + echo '#define DATAROOTDIR "$(datarootdir)"'; \ + echo '#define DATADIR "$(datadir)"'; \ + echo '#define SYSCONFDIR "$(sysconfdir)"'; \ + echo '#define SHAREDSTATEDIR "$(sharedstatedir)"'; \ + echo '#define LOCALSTATEDIR "$(localstatedir)"'; \ + echo '#define INCLUDEDIR "$(includedir)"'; \ + echo '#define OLDINCLUDEDIR "$(oldincludedir)"'; \ + echo '#define DOCDIR "$(docdir)"'; \ + echo '#define INFODIR "$(infodir)"'; \ + echo '#define HTMLDIR "$(htmldir)"'; \ + echo '#define DVIDIR "$(dvidir)"'; \ + echo '#define PDFDIR "$(pdfdir)"'; \ + echo '#define PSDIR "$(psdir)"'; \ + echo '#define LIBDIR "$(libdir)"'; \ + echo '#define LISPDIR "$(lispdir)"'; \ + echo '#define LOCALEDIR "$(localedir)"'; \ + echo '#define MANDIR "$(mandir)"'; \ + echo '#define MANEXT "$(manext)"'; \ + echo '#define PKGDATADIR "$(pkgdatadir)"'; \ + echo '#define PKGINCLUDEDIR "$(pkgincludedir)"'; \ + echo '#define PKGLIBDIR "$(pkglibdir)"'; \ + echo '#define PKGLIBEXECDIR "$(pkglibexecdir)"'; \ + } | sed '/""/d' > $@-t + mv $@-t $@ + +# We need the following in order to create <dirent.h> when the system +# doesn't have one that works with the given compiler. +dirent.h: dirent_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_DIRENT_H''@|$(ABSOLUTE_DIRENT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''REPLACE_FCHDIR''@|$(REPLACE_FCHDIR)|g' \ + < $(srcdir)/dirent_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <fcntl.h> when the system +# doesn't have one that works with the given compiler. +fcntl.h: fcntl_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_FCNTL_H''@|$(ABSOLUTE_FCNTL_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/fcntl_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <float.h> when the system +# doesn't have one that works with the given compiler. +float.h: float_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_FLOAT_H''@|$(ABSOLUTE_FLOAT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/float_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <fnmatch.h> when the system +# doesn't have one that supports the required API. +fnmatch.h: fnmatch_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/fnmatch_.h; \ + } > $@-t + mv -f $@-t $@ + +# We need the following in order to create <getopt.h> when the system +# doesn't have one that works with the given compiler. +getopt.h: getopt_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/getopt_.h; \ + } > $@-t + mv -f $@-t $@ + +# We need the following in order to create <inttypes.h> when the system +# doesn't have one that works with the given compiler. +inttypes.h: inttypes_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_INTTYPES_H''@/$(HAVE_INTTYPES_H)/g' \ + -e 's|@''ABSOLUTE_INTTYPES_H''@|$(ABSOLUTE_INTTYPES_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''PRI_MACROS_BROKEN''@/$(PRI_MACROS_BROKEN)/g' \ + -e 's/@''HAVE_LONG_LONG_INT''@/$(HAVE_LONG_LONG_INT)/g' \ + -e 's/@''HAVE_UNSIGNED_LONG_LONG_INT''@/$(HAVE_UNSIGNED_LONG_LONG_INT)/g' \ + -e 's/@''PRIPTR_PREFIX''@/$(PRIPTR_PREFIX)/g' \ + -e 's/@''GNULIB_IMAXABS''@/$(GNULIB_IMAXABS)/g' \ + -e 's/@''GNULIB_IMAXDIV''@/$(GNULIB_IMAXDIV)/g' \ + -e 's/@''GNULIB_STRTOIMAX''@/$(GNULIB_STRTOIMAX)/g' \ + -e 's/@''GNULIB_STRTOUMAX''@/$(GNULIB_STRTOUMAX)/g' \ + -e 's/@''HAVE_DECL_IMAXABS''@/$(HAVE_DECL_IMAXABS)/g' \ + -e 's/@''HAVE_DECL_IMAXDIV''@/$(HAVE_DECL_IMAXDIV)/g' \ + -e 's/@''HAVE_DECL_STRTOIMAX''@/$(HAVE_DECL_STRTOIMAX)/g' \ + -e 's/@''HAVE_DECL_STRTOUMAX''@/$(HAVE_DECL_STRTOUMAX)/g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/inttypes_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to install a simple file in $(libdir) +# which is shared with other installed packages. We use a list of referencing +# packages so that "make uninstall" will remove the file if and only if it +# is not used by another installed package. +# On systems with glibc-2.1 or newer, the file is redundant, therefore we +# avoid installing it. + +all-local: charset.alias ref-add.sed ref-del.sed +install-exec-local: all-local + test $(GLIBC21) != no || $(mkinstalldirs) $(DESTDIR)$(libdir) + if test -f $(charset_alias); then \ + sed -f ref-add.sed $(charset_alias) > $(charset_tmp) ; \ + $(INSTALL_DATA) $(charset_tmp) $(charset_alias) ; \ + rm -f $(charset_tmp) ; \ + else \ + if test $(GLIBC21) = no; then \ + sed -f ref-add.sed charset.alias > $(charset_tmp) ; \ + $(INSTALL_DATA) $(charset_tmp) $(charset_alias) ; \ + rm -f $(charset_tmp) ; \ + fi ; \ + fi + +uninstall-local: all-local + if test -f $(charset_alias); then \ + sed -f ref-del.sed $(charset_alias) > $(charset_tmp); \ + if grep '^# Packages using this file: $$' $(charset_tmp) \ + > /dev/null; then \ + rm -f $(charset_alias); \ + else \ + $(INSTALL_DATA) $(charset_tmp) $(charset_alias); \ + fi; \ + rm -f $(charset_tmp); \ + fi + +charset.alias: config.charset + rm -f t-$@ $@ + $(SHELL) $(srcdir)/config.charset '$(host)' > t-$@ + mv t-$@ $@ +.sin.sed: + rm -f t-$@ $@ + sed -e '/^#/d' -e 's/@''PACKAGE''@/$(PACKAGE)/g' $< > t-$@ + mv t-$@ $@ + +# We need the following in order to create <stdbool.h> when the system +# doesn't have one that works. +stdbool.h: stdbool_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE__BOOL''@/$(HAVE__BOOL)/g' < $(srcdir)/stdbool_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <stdint.h> when the system +# doesn't have one that works with the given compiler. +stdint.h: stdint_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_STDINT_H''@/$(HAVE_STDINT_H)/g' \ + -e 's|@''ABSOLUTE_STDINT_H''@|$(ABSOLUTE_STDINT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''HAVE_SYS_TYPES_H''@/$(HAVE_SYS_TYPES_H)/g' \ + -e 's/@''HAVE_INTTYPES_H''@/$(HAVE_INTTYPES_H)/g' \ + -e 's/@''HAVE_SYS_INTTYPES_H''@/$(HAVE_SYS_INTTYPES_H)/g' \ + -e 's/@''HAVE_SYS_BITYPES_H''@/$(HAVE_SYS_BITYPES_H)/g' \ + -e 's/@''HAVE_LONG_LONG_INT''@/$(HAVE_LONG_LONG_INT)/g' \ + -e 's/@''HAVE_UNSIGNED_LONG_LONG_INT''@/$(HAVE_UNSIGNED_LONG_LONG_INT)/g' \ + -e 's/@''BITSIZEOF_PTRDIFF_T''@/$(BITSIZEOF_PTRDIFF_T)/g' \ + -e 's/@''PTRDIFF_T_SUFFIX''@/$(PTRDIFF_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_SIG_ATOMIC_T''@/$(BITSIZEOF_SIG_ATOMIC_T)/g' \ + -e 's/@''HAVE_SIGNED_SIG_ATOMIC_T''@/$(HAVE_SIGNED_SIG_ATOMIC_T)/g' \ + -e 's/@''SIG_ATOMIC_T_SUFFIX''@/$(SIG_ATOMIC_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_SIZE_T''@/$(BITSIZEOF_SIZE_T)/g' \ + -e 's/@''SIZE_T_SUFFIX''@/$(SIZE_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_WCHAR_T''@/$(BITSIZEOF_WCHAR_T)/g' \ + -e 's/@''HAVE_SIGNED_WCHAR_T''@/$(HAVE_SIGNED_WCHAR_T)/g' \ + -e 's/@''WCHAR_T_SUFFIX''@/$(WCHAR_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_WINT_T''@/$(BITSIZEOF_WINT_T)/g' \ + -e 's/@''HAVE_SIGNED_WINT_T''@/$(HAVE_SIGNED_WINT_T)/g' \ + -e 's/@''WINT_T_SUFFIX''@/$(WINT_T_SUFFIX)/g' \ + < $(srcdir)/stdint_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <stdio.h> when the system +# doesn't have one that works with the given compiler. +stdio.h: stdio_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_STDIO_H''@|$(ABSOLUTE_STDIO_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_FPRINTF_POSIX''@|$(GNULIB_FPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_PRINTF_POSIX''@|$(GNULIB_PRINTF_POSIX)|g' \ + -e 's|@''GNULIB_SNPRINTF''@|$(GNULIB_SNPRINTF)|g' \ + -e 's|@''GNULIB_SPRINTF_POSIX''@|$(GNULIB_SPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VFPRINTF_POSIX''@|$(GNULIB_VFPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VPRINTF_POSIX''@|$(GNULIB_VPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VSNPRINTF''@|$(GNULIB_VSNPRINTF)|g' \ + -e 's|@''GNULIB_VSPRINTF_POSIX''@|$(GNULIB_VSPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VASPRINTF''@|$(GNULIB_VASPRINTF)|g' \ + -e 's|@''GNULIB_FSEEK''@|$(GNULIB_FSEEK)|g' \ + -e 's|@''GNULIB_FSEEKO''@|$(GNULIB_FSEEKO)|g' \ + -e 's|@''GNULIB_FTELL''@|$(GNULIB_FTELL)|g' \ + -e 's|@''GNULIB_FTELLO''@|$(GNULIB_FTELLO)|g' \ + -e 's|@''GNULIB_FFLUSH''@|$(GNULIB_FFLUSH)|g' \ + -e 's|@''REPLACE_FPRINTF''@|$(REPLACE_FPRINTF)|g' \ + -e 's|@''REPLACE_VFPRINTF''@|$(REPLACE_VFPRINTF)|g' \ + -e 's|@''REPLACE_PRINTF''@|$(REPLACE_PRINTF)|g' \ + -e 's|@''REPLACE_VPRINTF''@|$(REPLACE_VPRINTF)|g' \ + -e 's|@''REPLACE_SNPRINTF''@|$(REPLACE_SNPRINTF)|g' \ + -e 's|@''HAVE_DECL_SNPRINTF''@|$(HAVE_DECL_SNPRINTF)|g' \ + -e 's|@''REPLACE_VSNPRINTF''@|$(REPLACE_VSNPRINTF)|g' \ + -e 's|@''HAVE_DECL_VSNPRINTF''@|$(HAVE_DECL_VSNPRINTF)|g' \ + -e 's|@''REPLACE_SPRINTF''@|$(REPLACE_SPRINTF)|g' \ + -e 's|@''REPLACE_VSPRINTF''@|$(REPLACE_VSPRINTF)|g' \ + -e 's|@''HAVE_VASPRINTF''@|$(HAVE_VASPRINTF)|g' \ + -e 's|@''REPLACE_VASPRINTF''@|$(REPLACE_VASPRINTF)|g' \ + -e 's|@''REPLACE_FSEEKO''@|$(REPLACE_FSEEKO)|g' \ + -e 's|@''REPLACE_FSEEK''@|$(REPLACE_FSEEK)|g' \ + -e 's|@''REPLACE_FTELLO''@|$(REPLACE_FTELLO)|g' \ + -e 's|@''REPLACE_FTELL''@|$(REPLACE_FTELL)|g' \ + -e 's|@''REPLACE_FFLUSH''@|$(REPLACE_FFLUSH)|g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/stdio_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <stdlib.h> when the system +# doesn't have one that works with the given compiler. +stdlib.h: stdlib_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_STDLIB_H''@|$(ABSOLUTE_STDLIB_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_GETSUBOPT''@|$(GNULIB_GETSUBOPT)|g' \ + -e 's|@''GNULIB_MKDTEMP''@|$(GNULIB_MKDTEMP)|g' \ + -e 's|@''GNULIB_MKSTEMP''@|$(GNULIB_MKSTEMP)|g' \ + -e 's|@''HAVE_GETSUBOPT''@|$(HAVE_GETSUBOPT)|g' \ + -e 's|@''HAVE_MKDTEMP''@|$(HAVE_MKDTEMP)|g' \ + -e 's|@''REPLACE_MKSTEMP''@|$(REPLACE_MKSTEMP)|g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/stdlib_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <string.h> when the system +# doesn't have one that works with the given compiler. +string.h: string_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_STRING_H''@|$(ABSOLUTE_STRING_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_MBSLEN''@|$(GNULIB_MBSLEN)|g' \ + -e 's|@''GNULIB_MBSCHR''@|$(GNULIB_MBSCHR)|g' \ + -e 's|@''GNULIB_MBSRCHR''@|$(GNULIB_MBSRCHR)|g' \ + -e 's|@''GNULIB_MBSSTR''@|$(GNULIB_MBSSTR)|g' \ + -e 's|@''GNULIB_MBSCASECMP''@|$(GNULIB_MBSCASECMP)|g' \ + -e 's|@''GNULIB_MBSNCASECMP''@|$(GNULIB_MBSNCASECMP)|g' \ + -e 's|@''GNULIB_MBSPCASECMP''@|$(GNULIB_MBSPCASECMP)|g' \ + -e 's|@''GNULIB_MBSCASESTR''@|$(GNULIB_MBSCASESTR)|g' \ + -e 's|@''GNULIB_MBSCSPN''@|$(GNULIB_MBSCSPN)|g' \ + -e 's|@''GNULIB_MBSPBRK''@|$(GNULIB_MBSPBRK)|g' \ + -e 's|@''GNULIB_MBSSPN''@|$(GNULIB_MBSSPN)|g' \ + -e 's|@''GNULIB_MBSSEP''@|$(GNULIB_MBSSEP)|g' \ + -e 's|@''GNULIB_MBSTOK_R''@|$(GNULIB_MBSTOK_R)|g' \ + -e 's|@''GNULIB_MEMMEM''@|$(GNULIB_MEMMEM)|g' \ + -e 's|@''GNULIB_MEMPCPY''@|$(GNULIB_MEMPCPY)|g' \ + -e 's|@''GNULIB_MEMRCHR''@|$(GNULIB_MEMRCHR)|g' \ + -e 's|@''GNULIB_STPCPY''@|$(GNULIB_STPCPY)|g' \ + -e 's|@''GNULIB_STPNCPY''@|$(GNULIB_STPNCPY)|g' \ + -e 's|@''GNULIB_STRCHRNUL''@|$(GNULIB_STRCHRNUL)|g' \ + -e 's|@''GNULIB_STRDUP''@|$(GNULIB_STRDUP)|g' \ + -e 's|@''GNULIB_STRNDUP''@|$(GNULIB_STRNDUP)|g' \ + -e 's|@''GNULIB_STRNLEN''@|$(GNULIB_STRNLEN)|g' \ + -e 's|@''GNULIB_STRPBRK''@|$(GNULIB_STRPBRK)|g' \ + -e 's|@''GNULIB_STRSEP''@|$(GNULIB_STRSEP)|g' \ + -e 's|@''GNULIB_STRCASESTR''@|$(GNULIB_STRCASESTR)|g' \ + -e 's|@''GNULIB_STRTOK_R''@|$(GNULIB_STRTOK_R)|g' \ + -e 's|@''HAVE_DECL_MEMMEM''@|$(HAVE_DECL_MEMMEM)|g' \ + -e 's|@''HAVE_MEMPCPY''@|$(HAVE_MEMPCPY)|g' \ + -e 's|@''HAVE_DECL_MEMRCHR''@|$(HAVE_DECL_MEMRCHR)|g' \ + -e 's|@''HAVE_STPCPY''@|$(HAVE_STPCPY)|g' \ + -e 's|@''HAVE_STPNCPY''@|$(HAVE_STPNCPY)|g' \ + -e 's|@''HAVE_STRCASECMP''@|$(HAVE_STRCASECMP)|g' \ + -e 's|@''HAVE_DECL_STRNCASECMP''@|$(HAVE_DECL_STRNCASECMP)|g' \ + -e 's|@''HAVE_STRCHRNUL''@|$(HAVE_STRCHRNUL)|g' \ + -e 's|@''HAVE_DECL_STRDUP''@|$(HAVE_DECL_STRDUP)|g' \ + -e 's|@''HAVE_STRNDUP''@|$(HAVE_STRNDUP)|g' \ + -e 's|@''HAVE_DECL_STRNDUP''@|$(HAVE_DECL_STRNDUP)|g' \ + -e 's|@''HAVE_DECL_STRNLEN''@|$(HAVE_DECL_STRNLEN)|g' \ + -e 's|@''HAVE_STRPBRK''@|$(HAVE_STRPBRK)|g' \ + -e 's|@''HAVE_STRSEP''@|$(HAVE_STRSEP)|g' \ + -e 's|@''HAVE_STRCASESTR''@|$(HAVE_STRCASESTR)|g' \ + -e 's|@''HAVE_DECL_STRTOK_R''@|$(HAVE_DECL_STRTOK_R)|g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/string_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <sys/stat.h> when the system +# has one that is incomplete. +sys/stat.h: sys_stat_.h + @MKDIR_P@ sys + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_SYS_STAT_H''@|$(ABSOLUTE_SYS_STAT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''HAVE_IO_H''@|$(HAVE_IO_H)|g' \ + -e 's|@''HAVE_LSTAT''@|$(HAVE_LSTAT)|g' \ + -e 's|@''HAVE_DECL_MKDIR''@|$(HAVE_DECL_MKDIR)|g' \ + < $(srcdir)/sys_stat_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <sys/time.h> when the system +# doesn't have one that works with the given compiler. +sys/time.h: sys_time_.h + @MKDIR_P@ sys + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_SYS_TIME_H''@/$(HAVE_SYS_TIME_H)/g' \ + -e 's|@''ABSOLUTE_SYS_TIME_H''@|$(ABSOLUTE_SYS_TIME_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''REPLACE_GETTIMEOFDAY''@/$(REPLACE_GETTIMEOFDAY)/g' \ + -e 's/@''HAVE_STRUCT_TIMEVAL''@/$(HAVE_STRUCT_TIMEVAL)/g' \ + < $(srcdir)/sys_time_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <sysexits.h> when the system +# doesn't have one that works with the given compiler. +sysexits.h: sysexits_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''HAVE_SYSEXITS_H''@|$(HAVE_SYSEXITS_H)|g' \ + -e 's|@''ABSOLUTE_SYSEXITS_H''@|$(ABSOLUTE_SYSEXITS_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/sysexits_.h; \ + } > $@-t + mv -f $@-t $@ + +# We need the following in order to create <time.h> when the system +# doesn't have one that works with the given compiler. +time.h: time_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@ABSOLUTE_TIME_H''@|$(ABSOLUTE_TIME_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@REPLACE_LOCALTIME_R''@|$(REPLACE_LOCALTIME_R)|g' \ + -e 's|@REPLACE_NANOSLEEP''@|$(REPLACE_NANOSLEEP)|g' \ + -e 's|@REPLACE_STRPTIME''@|$(REPLACE_STRPTIME)|g' \ + -e 's|@REPLACE_TIMEGM''@|$(REPLACE_TIMEGM)|g' \ + -e 's|@SYS_TIME_H_DEFINES_STRUCT_TIMESPEC''@|$(SYS_TIME_H_DEFINES_STRUCT_TIMESPEC)|g' \ + -e 's|@TIME_H_DEFINES_STRUCT_TIMESPEC''@|$(TIME_H_DEFINES_STRUCT_TIMESPEC)|g' \ + < $(srcdir)/time_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create an empty placeholder for +# <unistd.h> when the system doesn't have one. +unistd.h: unistd_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''HAVE_UNISTD_H''@|$(HAVE_UNISTD_H)|g' \ + -e 's|@''ABSOLUTE_UNISTD_H''@|$(ABSOLUTE_UNISTD_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_CHOWN''@|$(GNULIB_CHOWN)|g' \ + -e 's|@''GNULIB_DUP2''@|$(GNULIB_DUP2)|g' \ + -e 's|@''GNULIB_FCHDIR''@|$(GNULIB_FCHDIR)|g' \ + -e 's|@''GNULIB_FTRUNCATE''@|$(GNULIB_FTRUNCATE)|g' \ + -e 's|@''GNULIB_GETCWD''@|$(GNULIB_GETCWD)|g' \ + -e 's|@''GNULIB_GETLOGIN_R''@|$(GNULIB_GETLOGIN_R)|g' \ + -e 's|@''GNULIB_LSEEK''@|$(GNULIB_LSEEK)|g' \ + -e 's|@''GNULIB_READLINK''@|$(GNULIB_READLINK)|g' \ + -e 's|@''GNULIB_SLEEP''@|$(GNULIB_SLEEP)|g' \ + -e 's|@''HAVE_DUP2''@|$(HAVE_DUP2)|g' \ + -e 's|@''HAVE_FTRUNCATE''@|$(HAVE_FTRUNCATE)|g' \ + -e 's|@''HAVE_READLINK''@|$(HAVE_READLINK)|g' \ + -e 's|@''HAVE_SLEEP''@|$(HAVE_SLEEP)|g' \ + -e 's|@''HAVE_DECL_GETLOGIN_R''@|$(HAVE_DECL_GETLOGIN_R)|g' \ + -e 's|@''REPLACE_CHOWN''@|$(REPLACE_CHOWN)|g' \ + -e 's|@''REPLACE_FCHDIR''@|$(REPLACE_FCHDIR)|g' \ + -e 's|@''REPLACE_GETCWD''@|$(REPLACE_GETCWD)|g' \ + -e 's|@''REPLACE_LSEEK''@|$(REPLACE_LSEEK)|g' \ + < $(srcdir)/unistd_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <wchar.h> when the system +# version does not work standalone. +wchar.h: wchar_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_WCHAR_H''@|$(ABSOLUTE_WCHAR_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/wchar_.h; \ + } > $@-t + mv $@-t $@ + +# We need the following in order to create <wctype.h> when the system +# doesn't have one that works with the given compiler. +wctype.h: wctype_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_WCTYPE_H''@/$(HAVE_WCTYPE_H)/g' \ + -e 's|@''ABSOLUTE_WCTYPE_H''@|$(ABSOLUTE_WCTYPE_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''HAVE_ISWCNTRL''@/$(HAVE_ISWCNTRL)/g' \ + -e 's/@''HAVE_WINT_T''@/$(HAVE_WINT_T)/g' \ + < $(srcdir)/wctype_.h; \ + } > $@-t + mv $@-t $@ + +mostlyclean-local: mostlyclean-generic + @for dir in '' $(MOSTLYCLEANDIRS); do \ + if test -n "$$dir" && test -d $$dir; then \ + echo "rmdir $$dir"; rmdir $$dir; \ + fi; \ + done + +rmt-command.h : Makefile + rm -f $@-t $@ + echo "#ifndef DEFAULT_RMT_COMMAND" >> $@-t + echo "# define DEFAULT_RMT_COMMAND \"$(DEFAULT_RMT_DIR)/`echo rmt | sed '$(transform)'`$(EXEEXT)\"" >> $@-t + echo "#endif" >> $@-t + mv $@-t $@ +# Tell versions [3.59,3.63) of GNU make to not export all variables. +# Otherwise a system limit (for SysV at least) may be exceeded. +.NOEXPORT: diff --git a/lib/__fpending.c b/lib/__fpending.c new file mode 100644 index 0000000..221aee6 --- /dev/null +++ b/lib/__fpending.c @@ -0,0 +1,30 @@ +/* __fpending.c -- return the number of pending output bytes on a stream + Copyright (C) 2000, 2004, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "__fpending.h" + +/* Return the number of pending (aka buffered, unflushed) + bytes on the stream, FP, that is open for writing. */ +size_t +__fpending (FILE *fp) +{ + return PENDING_OUTPUT_N_BYTES; +} diff --git a/lib/__fpending.h b/lib/__fpending.h new file mode 100644 index 0000000..8a8aabc --- /dev/null +++ b/lib/__fpending.h @@ -0,0 +1,34 @@ +/* Declare __fpending. + + Copyright (C) 2000, 2003, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + + Written by Jim Meyering. */ + +#include <stddef.h> +#include <stdio.h> + +#ifndef HAVE_DECL___FPENDING +"this configure-time declaration test was not run" +#endif + +#if HAVE_DECL___FPENDING +# if HAVE_STDIO_EXT_H +# include <stdio_ext.h> +# endif +#else +size_t __fpending (FILE *); +#endif diff --git a/lib/alloca.c b/lib/alloca.c new file mode 100644 index 0000000..3a1f4e2 --- /dev/null +++ b/lib/alloca.c @@ -0,0 +1,489 @@ +/* alloca.c -- allocate automatically reclaimed memory + (Mostly) portable public-domain implementation -- D A Gwyn + + This implementation of the PWB library alloca function, + which is used to allocate space off the run-time stack so + that it is automatically reclaimed upon procedure exit, + was inspired by discussions with J. Q. Johnson of Cornell. + J.Otto Tennant <jot@cray.com> contributed the Cray support. + + There are some preprocessor constants that can + be defined when compiling for your specific system, for + improved efficiency; however, the defaults should be okay. + + The general concept of this implementation is to keep + track of all alloca-allocated blocks, and reclaim any + that are found to be deeper in the stack than the current + invocation. This heuristic does not reclaim storage as + soon as it becomes invalid, but it will do so eventually. + + As a special case, alloca(0) reclaims storage without + allocating any. It is a good idea to use alloca(0) in + your main control loop, etc. to force garbage collection. */ + +#include <config.h> + +#include <alloca.h> + +#include <string.h> +#include <stdlib.h> + +#ifdef emacs +# include "lisp.h" +# include "blockinput.h" +# ifdef EMACS_FREE +# undef free +# define free EMACS_FREE +# endif +#else +# define memory_full() abort () +#endif + +/* If compiling with GCC 2, this file's not needed. */ +#if !defined (__GNUC__) || __GNUC__ < 2 + +/* If someone has defined alloca as a macro, + there must be some other way alloca is supposed to work. */ +# ifndef alloca + +# ifdef emacs +# ifdef static +/* actually, only want this if static is defined as "" + -- this is for usg, in which emacs must undefine static + in order to make unexec workable + */ +# ifndef STACK_DIRECTION +you +lose +-- must know STACK_DIRECTION at compile-time +/* Using #error here is not wise since this file should work for + old and obscure compilers. */ +# endif /* STACK_DIRECTION undefined */ +# endif /* static */ +# endif /* emacs */ + +/* If your stack is a linked list of frames, you have to + provide an "address metric" ADDRESS_FUNCTION macro. */ + +# if defined (CRAY) && defined (CRAY_STACKSEG_END) +long i00afunc (); +# define ADDRESS_FUNCTION(arg) (char *) i00afunc (&(arg)) +# else +# define ADDRESS_FUNCTION(arg) &(arg) +# endif + +/* Define STACK_DIRECTION if you know the direction of stack + growth for your system; otherwise it will be automatically + deduced at run-time. + + STACK_DIRECTION > 0 => grows toward higher addresses + STACK_DIRECTION < 0 => grows toward lower addresses + STACK_DIRECTION = 0 => direction of growth unknown */ + +# ifndef STACK_DIRECTION +# define STACK_DIRECTION 0 /* Direction unknown. */ +# endif + +# if STACK_DIRECTION != 0 + +# define STACK_DIR STACK_DIRECTION /* Known at compile-time. */ + +# else /* STACK_DIRECTION == 0; need run-time code. */ + +static int stack_dir; /* 1 or -1 once known. */ +# define STACK_DIR stack_dir + +static void +find_stack_direction (void) +{ + static char *addr = NULL; /* Address of first `dummy', once known. */ + auto char dummy; /* To get stack address. */ + + if (addr == NULL) + { /* Initial entry. */ + addr = ADDRESS_FUNCTION (dummy); + + find_stack_direction (); /* Recurse once. */ + } + else + { + /* Second entry. */ + if (ADDRESS_FUNCTION (dummy) > addr) + stack_dir = 1; /* Stack grew upward. */ + else + stack_dir = -1; /* Stack grew downward. */ + } +} + +# endif /* STACK_DIRECTION == 0 */ + +/* An "alloca header" is used to: + (a) chain together all alloca'ed blocks; + (b) keep track of stack depth. + + It is very important that sizeof(header) agree with malloc + alignment chunk size. The following default should work okay. */ + +# ifndef ALIGN_SIZE +# define ALIGN_SIZE sizeof(double) +# endif + +typedef union hdr +{ + char align[ALIGN_SIZE]; /* To force sizeof(header). */ + struct + { + union hdr *next; /* For chaining headers. */ + char *deep; /* For stack depth measure. */ + } h; +} header; + +static header *last_alloca_header = NULL; /* -> last alloca header. */ + +/* Return a pointer to at least SIZE bytes of storage, + which will be automatically reclaimed upon exit from + the procedure that called alloca. Originally, this space + was supposed to be taken from the current stack frame of the + caller, but that method cannot be made to work for some + implementations of C, for example under Gould's UTX/32. */ + +void * +alloca (size_t size) +{ + auto char probe; /* Probes stack depth: */ + register char *depth = ADDRESS_FUNCTION (probe); + +# if STACK_DIRECTION == 0 + if (STACK_DIR == 0) /* Unknown growth direction. */ + find_stack_direction (); +# endif + + /* Reclaim garbage, defined as all alloca'd storage that + was allocated from deeper in the stack than currently. */ + + { + register header *hp; /* Traverses linked list. */ + +# ifdef emacs + BLOCK_INPUT; +# endif + + for (hp = last_alloca_header; hp != NULL;) + if ((STACK_DIR > 0 && hp->h.deep > depth) + || (STACK_DIR < 0 && hp->h.deep < depth)) + { + register header *np = hp->h.next; + + free (hp); /* Collect garbage. */ + + hp = np; /* -> next header. */ + } + else + break; /* Rest are not deeper. */ + + last_alloca_header = hp; /* -> last valid storage. */ + +# ifdef emacs + UNBLOCK_INPUT; +# endif + } + + if (size == 0) + return NULL; /* No allocation required. */ + + /* Allocate combined header + user data storage. */ + + { + /* Address of header. */ + register header *new; + + size_t combined_size = sizeof (header) + size; + if (combined_size < sizeof (header)) + memory_full (); + + new = malloc (combined_size); + + if (! new) + memory_full (); + + new->h.next = last_alloca_header; + new->h.deep = depth; + + last_alloca_header = new; + + /* User storage begins just after header. */ + + return (void *) (new + 1); + } +} + +# if defined (CRAY) && defined (CRAY_STACKSEG_END) + +# ifdef DEBUG_I00AFUNC +# include <stdio.h> +# endif + +# ifndef CRAY_STACK +# define CRAY_STACK +# ifndef CRAY2 +/* Stack structures for CRAY-1, CRAY X-MP, and CRAY Y-MP */ +struct stack_control_header + { + long shgrow:32; /* Number of times stack has grown. */ + long shaseg:32; /* Size of increments to stack. */ + long shhwm:32; /* High water mark of stack. */ + long shsize:32; /* Current size of stack (all segments). */ + }; + +/* The stack segment linkage control information occurs at + the high-address end of a stack segment. (The stack + grows from low addresses to high addresses.) The initial + part of the stack segment linkage control information is + 0200 (octal) words. This provides for register storage + for the routine which overflows the stack. */ + +struct stack_segment_linkage + { + long ss[0200]; /* 0200 overflow words. */ + long sssize:32; /* Number of words in this segment. */ + long ssbase:32; /* Offset to stack base. */ + long:32; + long sspseg:32; /* Offset to linkage control of previous + segment of stack. */ + long:32; + long sstcpt:32; /* Pointer to task common address block. */ + long sscsnm; /* Private control structure number for + microtasking. */ + long ssusr1; /* Reserved for user. */ + long ssusr2; /* Reserved for user. */ + long sstpid; /* Process ID for pid based multi-tasking. */ + long ssgvup; /* Pointer to multitasking thread giveup. */ + long sscray[7]; /* Reserved for Cray Research. */ + long ssa0; + long ssa1; + long ssa2; + long ssa3; + long ssa4; + long ssa5; + long ssa6; + long ssa7; + long sss0; + long sss1; + long sss2; + long sss3; + long sss4; + long sss5; + long sss6; + long sss7; + }; + +# else /* CRAY2 */ +/* The following structure defines the vector of words + returned by the STKSTAT library routine. */ +struct stk_stat + { + long now; /* Current total stack size. */ + long maxc; /* Amount of contiguous space which would + be required to satisfy the maximum + stack demand to date. */ + long high_water; /* Stack high-water mark. */ + long overflows; /* Number of stack overflow ($STKOFEN) calls. */ + long hits; /* Number of internal buffer hits. */ + long extends; /* Number of block extensions. */ + long stko_mallocs; /* Block allocations by $STKOFEN. */ + long underflows; /* Number of stack underflow calls ($STKRETN). */ + long stko_free; /* Number of deallocations by $STKRETN. */ + long stkm_free; /* Number of deallocations by $STKMRET. */ + long segments; /* Current number of stack segments. */ + long maxs; /* Maximum number of stack segments so far. */ + long pad_size; /* Stack pad size. */ + long current_address; /* Current stack segment address. */ + long current_size; /* Current stack segment size. This + number is actually corrupted by STKSTAT to + include the fifteen word trailer area. */ + long initial_address; /* Address of initial segment. */ + long initial_size; /* Size of initial segment. */ + }; + +/* The following structure describes the data structure which trails + any stack segment. I think that the description in 'asdef' is + out of date. I only describe the parts that I am sure about. */ + +struct stk_trailer + { + long this_address; /* Address of this block. */ + long this_size; /* Size of this block (does not include + this trailer). */ + long unknown2; + long unknown3; + long link; /* Address of trailer block of previous + segment. */ + long unknown5; + long unknown6; + long unknown7; + long unknown8; + long unknown9; + long unknown10; + long unknown11; + long unknown12; + long unknown13; + long unknown14; + }; + +# endif /* CRAY2 */ +# endif /* not CRAY_STACK */ + +# ifdef CRAY2 +/* Determine a "stack measure" for an arbitrary ADDRESS. + I doubt that "lint" will like this much. */ + +static long +i00afunc (long *address) +{ + struct stk_stat status; + struct stk_trailer *trailer; + long *block, size; + long result = 0; + + /* We want to iterate through all of the segments. The first + step is to get the stack status structure. We could do this + more quickly and more directly, perhaps, by referencing the + $LM00 common block, but I know that this works. */ + + STKSTAT (&status); + + /* Set up the iteration. */ + + trailer = (struct stk_trailer *) (status.current_address + + status.current_size + - 15); + + /* There must be at least one stack segment. Therefore it is + a fatal error if "trailer" is null. */ + + if (trailer == 0) + abort (); + + /* Discard segments that do not contain our argument address. */ + + while (trailer != 0) + { + block = (long *) trailer->this_address; + size = trailer->this_size; + if (block == 0 || size == 0) + abort (); + trailer = (struct stk_trailer *) trailer->link; + if ((block <= address) && (address < (block + size))) + break; + } + + /* Set the result to the offset in this segment and add the sizes + of all predecessor segments. */ + + result = address - block; + + if (trailer == 0) + { + return result; + } + + do + { + if (trailer->this_size <= 0) + abort (); + result += trailer->this_size; + trailer = (struct stk_trailer *) trailer->link; + } + while (trailer != 0); + + /* We are done. Note that if you present a bogus address (one + not in any segment), you will get a different number back, formed + from subtracting the address of the first block. This is probably + not what you want. */ + + return (result); +} + +# else /* not CRAY2 */ +/* Stack address function for a CRAY-1, CRAY X-MP, or CRAY Y-MP. + Determine the number of the cell within the stack, + given the address of the cell. The purpose of this + routine is to linearize, in some sense, stack addresses + for alloca. */ + +static long +i00afunc (long address) +{ + long stkl = 0; + + long size, pseg, this_segment, stack; + long result = 0; + + struct stack_segment_linkage *ssptr; + + /* Register B67 contains the address of the end of the + current stack segment. If you (as a subprogram) store + your registers on the stack and find that you are past + the contents of B67, you have overflowed the segment. + + B67 also points to the stack segment linkage control + area, which is what we are really interested in. */ + + stkl = CRAY_STACKSEG_END (); + ssptr = (struct stack_segment_linkage *) stkl; + + /* If one subtracts 'size' from the end of the segment, + one has the address of the first word of the segment. + + If this is not the first segment, 'pseg' will be + nonzero. */ + + pseg = ssptr->sspseg; + size = ssptr->sssize; + + this_segment = stkl - size; + + /* It is possible that calling this routine itself caused + a stack overflow. Discard stack segments which do not + contain the target address. */ + + while (!(this_segment <= address && address <= stkl)) + { +# ifdef DEBUG_I00AFUNC + fprintf (stderr, "%011o %011o %011o\n", this_segment, address, stkl); +# endif + if (pseg == 0) + break; + stkl = stkl - pseg; + ssptr = (struct stack_segment_linkage *) stkl; + size = ssptr->sssize; + pseg = ssptr->sspseg; + this_segment = stkl - size; + } + + result = address - this_segment; + + /* If you subtract pseg from the current end of the stack, + you get the address of the previous stack segment's end. + This seems a little convoluted to me, but I'll bet you save + a cycle somewhere. */ + + while (pseg != 0) + { +# ifdef DEBUG_I00AFUNC + fprintf (stderr, "%011o %011o\n", pseg, size); +# endif + stkl = stkl - pseg; + ssptr = (struct stack_segment_linkage *) stkl; + size = ssptr->sssize; + pseg = ssptr->sspseg; + result += size; + } + return (result); +} + +# endif /* not CRAY2 */ +# endif /* CRAY */ + +# endif /* no alloca */ +#endif /* not GCC version 2 */ diff --git a/lib/alloca_.h b/lib/alloca_.h new file mode 100644 index 0000000..af274b9 --- /dev/null +++ b/lib/alloca_.h @@ -0,0 +1,54 @@ +/* Memory allocation on the stack. + + Copyright (C) 1995, 1999, 2001-2004, 2006-2007 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published + by the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public + License along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, + USA. */ + +/* Avoid using the symbol _ALLOCA_H here, as Bison assumes _ALLOCA_H + means there is a real alloca function. */ +#ifndef _GL_ALLOCA_H +#define _GL_ALLOCA_H + +/* alloca (N) returns a pointer to N bytes of memory + allocated on the stack, which will last until the function returns. + Use of alloca should be avoided: + - inside arguments of function calls - undefined behaviour, + - in inline functions - the allocation may actually last until the + calling function returns, + - for huge N (say, N >= 65536) - you never know how large (or small) + the stack is, and when the stack cannot fulfill the memory allocation + request, the program just crashes. + */ + +#ifndef alloca +# ifdef __GNUC__ +# define alloca __builtin_alloca +# elif defined _AIX +# define alloca __alloca +# elif defined _MSC_VER +# include <malloc.h> +# define alloca _alloca +# else +# include <stddef.h> +# ifdef __cplusplus +extern "C" +# endif +void *alloca (size_t); +# endif +#endif + +#endif /* _GL_ALLOCA_H */ diff --git a/lib/allocsa.c b/lib/allocsa.c new file mode 100644 index 0000000..97652e6 --- /dev/null +++ b/lib/allocsa.c @@ -0,0 +1,137 @@ +/* Safe automatic memory allocation. + Copyright (C) 2003, 2006 Free Software Foundation, Inc. + Written by Bruno Haible <bruno@clisp.org>, 2003. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "allocsa.h" + +/* The speed critical point in this file is freesa() applied to an alloca() + result: it must be fast, to match the speed of alloca(). The speed of + mallocsa() and freesa() in the other case are not critical, because they + are only invoked for big memory sizes. */ + +#if HAVE_ALLOCA + +/* Store the mallocsa() results in a hash table. This is needed to reliably + distinguish a mallocsa() result and an alloca() result. + + Although it is possible that the same pointer is returned by alloca() and + by mallocsa() at different times in the same application, it does not lead + to a bug in freesa(), because: + - Before a pointer returned by alloca() can point into malloc()ed memory, + the function must return, and once this has happened the programmer must + not call freesa() on it anyway. + - Before a pointer returned by mallocsa() can point into the stack, it + must be freed. The only function that can free it is freesa(), and + when freesa() frees it, it also removes it from the hash table. */ + +#define MAGIC_NUMBER 0x1415fb4a +#define MAGIC_SIZE sizeof (int) +/* This is how the header info would look like without any alignment + considerations. */ +struct preliminary_header { void *next; char room[MAGIC_SIZE]; }; +/* But the header's size must be a multiple of sa_alignment_max. */ +#define HEADER_SIZE \ + (((sizeof (struct preliminary_header) + sa_alignment_max - 1) / sa_alignment_max) * sa_alignment_max) +struct header { void *next; char room[HEADER_SIZE - sizeof (struct preliminary_header) + MAGIC_SIZE]; }; +/* Verify that HEADER_SIZE == sizeof (struct header). */ +typedef int verify1[2 * (HEADER_SIZE == sizeof (struct header)) - 1]; +/* We make the hash table quite big, so that during lookups the probability + of empty hash buckets is quite high. There is no need to make the hash + table resizable, because when the hash table gets filled so much that the + lookup becomes slow, it means that the application has memory leaks. */ +#define HASH_TABLE_SIZE 257 +static void * mallocsa_results[HASH_TABLE_SIZE]; + +#endif + +void * +mallocsa (size_t n) +{ +#if HAVE_ALLOCA + /* Allocate one more word, that serves as an indicator for malloc()ed + memory, so that freesa() of an alloca() result is fast. */ + size_t nplus = n + HEADER_SIZE; + + if (nplus >= n) + { + char *p = (char *) malloc (nplus); + + if (p != NULL) + { + size_t slot; + + p += HEADER_SIZE; + + /* Put a magic number into the indicator word. */ + ((int *) p)[-1] = MAGIC_NUMBER; + + /* Enter p into the hash table. */ + slot = (unsigned long) p % HASH_TABLE_SIZE; + ((struct header *) (p - HEADER_SIZE))->next = mallocsa_results[slot]; + mallocsa_results[slot] = p; + + return p; + } + } + /* Out of memory. */ + return NULL; +#else +# if !MALLOC_0_IS_NONNULL + if (n == 0) + n = 1; +# endif + return malloc (n); +#endif +} + +#if HAVE_ALLOCA +void +freesa (void *p) +{ + /* mallocsa() may have returned NULL. */ + if (p != NULL) + { + /* Attempt to quickly distinguish the mallocsa() result - which has + a magic indicator word - and the alloca() result - which has an + uninitialized indicator word. It is for this test that sa_increment + additional bytes are allocated in the alloca() case. */ + if (((int *) p)[-1] == MAGIC_NUMBER) + { + /* Looks like a mallocsa() result. To see whether it really is one, + perform a lookup in the hash table. */ + size_t slot = (unsigned long) p % HASH_TABLE_SIZE; + void **chain = &mallocsa_results[slot]; + for (; *chain != NULL;) + { + if (*chain == p) + { + /* Found it. Remove it from the hash table and free it. */ + char *p_begin = (char *) p - HEADER_SIZE; + *chain = ((struct header *) p_begin)->next; + free (p_begin); + return; + } + chain = &((struct header *) ((char *) *chain - HEADER_SIZE))->next; + } + } + /* At this point, we know it was not a mallocsa() result. */ + } +} +#endif diff --git a/lib/allocsa.h b/lib/allocsa.h new file mode 100644 index 0000000..ffee917 --- /dev/null +++ b/lib/allocsa.h @@ -0,0 +1,124 @@ +/* Safe automatic memory allocation. + Copyright (C) 2003-2007 Free Software Foundation, Inc. + Written by Bruno Haible <bruno@clisp.org>, 2003. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _ALLOCSA_H +#define _ALLOCSA_H + +#include <alloca.h> +#include <stddef.h> +#include <stdlib.h> + + +#ifdef __cplusplus +extern "C" { +#endif + + +/* safe_alloca(N) is equivalent to alloca(N) when it is safe to call + alloca(N); otherwise it returns NULL. It either returns N bytes of + memory allocated on the stack, that lasts until the function returns, + or NULL. + Use of safe_alloca should be avoided: + - inside arguments of function calls - undefined behaviour, + - in inline functions - the allocation may actually last until the + calling function returns. +*/ +#if HAVE_ALLOCA +/* The OS usually guarantees only one guard page at the bottom of the stack, + and a page size can be as small as 4096 bytes. So we cannot safely + allocate anything larger than 4096 bytes. Also care for the possibility + of a few compiler-allocated temporary stack slots. + This must be a macro, not an inline function. */ +# define safe_alloca(N) ((N) < 4032 ? alloca (N) : NULL) +#else +# define safe_alloca(N) ((void) (N), NULL) +#endif + +/* allocsa(N) is a safe variant of alloca(N). It allocates N bytes of + memory allocated on the stack, that must be freed using freesa() before + the function returns. Upon failure, it returns NULL. */ +#if HAVE_ALLOCA +# define allocsa(N) \ + ((N) < 4032 - sa_increment \ + ? (void *) ((char *) alloca ((N) + sa_increment) + sa_increment) \ + : mallocsa (N)) +#else +# define allocsa(N) \ + mallocsa (N) +#endif +extern void * mallocsa (size_t n); + +/* Free a block of memory allocated through allocsa(). */ +#if HAVE_ALLOCA +extern void freesa (void *p); +#else +# define freesa free +#endif + +/* Maybe we should also define a variant + nallocsa (size_t n, size_t s) - behaves like allocsa (n * s) + If this would be useful in your application. please speak up. */ + + +#ifdef __cplusplus +} +#endif + + +/* ------------------- Auxiliary, non-public definitions ------------------- */ + +/* Determine the alignment of a type at compile time. */ +#if defined __GNUC__ +# define sa_alignof __alignof__ +#elif defined __cplusplus + template <class type> struct sa_alignof_helper { char __slot1; type __slot2; }; +# define sa_alignof(type) offsetof (sa_alignof_helper<type>, __slot2) +#elif defined __hpux + /* Work around a HP-UX 10.20 cc bug with enums constants defined as offsetof + values. */ +# define sa_alignof(type) (sizeof (type) <= 4 ? 4 : 8) +#elif defined _AIX + /* Work around an AIX 3.2.5 xlc bug with enums constants defined as offsetof + values. */ +# define sa_alignof(type) (sizeof (type) <= 4 ? 4 : 8) +#else +# define sa_alignof(type) offsetof (struct { char __slot1; type __slot2; }, __slot2) +#endif + +enum +{ +/* The desired alignment of memory allocations is the maximum alignment + among all elementary types. */ + sa_alignment_long = sa_alignof (long), + sa_alignment_double = sa_alignof (double), +#if HAVE_LONG_LONG_INT + sa_alignment_longlong = sa_alignof (long long), +#endif + sa_alignment_longdouble = sa_alignof (long double), + sa_alignment_max = ((sa_alignment_long - 1) | (sa_alignment_double - 1) +#if HAVE_LONG_LONG_INT + | (sa_alignment_longlong - 1) +#endif + | (sa_alignment_longdouble - 1) + ) + 1, +/* The increment that guarantees room for a magic word must be >= sizeof (int) + and a multiple of sa_alignment_max. */ + sa_increment = ((sizeof (int) + sa_alignment_max - 1) / sa_alignment_max) * sa_alignment_max +}; + +#endif /* _ALLOCSA_H */ diff --git a/lib/allocsa.valgrind b/lib/allocsa.valgrind new file mode 100644 index 0000000..f4c77d6 --- /dev/null +++ b/lib/allocsa.valgrind @@ -0,0 +1,7 @@ +# Suppress a valgrind message about use of uninitialized memory in freesa(). +# This use is OK because it provides only a speedup. +{ + freesa + Memcheck:Cond + fun:freesa +} diff --git a/lib/argmatch.c b/lib/argmatch.c new file mode 100644 index 0000000..72d9248 --- /dev/null +++ b/lib/argmatch.c @@ -0,0 +1,278 @@ +/* argmatch.c -- find a match for a string in an array + + Copyright (C) 1990, 1998, 1999, 2001, 2002, 2003, 2004, 2005, 2006, 2007 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@ai.mit.edu> + Modified by Akim Demaille <demaille@inf.enst.fr> */ + +#include <config.h> + +/* Specification. */ +#include "argmatch.h" + +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +#include "error.h" +#include "quotearg.h" +#include "quote.h" + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +/* When reporting an invalid argument, show nonprinting characters + by using the quoting style ARGMATCH_QUOTING_STYLE. Do not use + literal_quoting_style. */ +#ifndef ARGMATCH_QUOTING_STYLE +# define ARGMATCH_QUOTING_STYLE locale_quoting_style +#endif + +/* Non failing version of argmatch call this function after failing. */ +#ifndef ARGMATCH_DIE +# include "exitfail.h" +# define ARGMATCH_DIE exit (exit_failure) +#endif + +#ifdef ARGMATCH_DIE_DECL +ARGMATCH_DIE_DECL; +#endif + +static void +__argmatch_die (void) +{ + ARGMATCH_DIE; +} + +/* Used by XARGMATCH and XARGCASEMATCH. See description in argmatch.h. + Default to __argmatch_die, but allow caller to change this at run-time. */ +argmatch_exit_fn argmatch_die = __argmatch_die; + + +/* If ARG is an unambiguous match for an element of the + NULL-terminated array ARGLIST, return the index in ARGLIST + of the matched element, else -1 if it does not match any element + or -2 if it is ambiguous (is a prefix of more than one element). + + If VALLIST is none null, use it to resolve ambiguities limited to + synonyms, i.e., for + "yes", "yop" -> 0 + "no", "nope" -> 1 + "y" is a valid argument, for `0', and "n" for `1'. */ + +ptrdiff_t +argmatch (const char *arg, const char *const *arglist, + const char *vallist, size_t valsize) +{ + size_t i; /* Temporary index in ARGLIST. */ + size_t arglen; /* Length of ARG. */ + ptrdiff_t matchind = -1; /* Index of first nonexact match. */ + bool ambiguous = false; /* If true, multiple nonexact match(es). */ + + arglen = strlen (arg); + + /* Test all elements for either exact match or abbreviated matches. */ + for (i = 0; arglist[i]; i++) + { + if (!strncmp (arglist[i], arg, arglen)) + { + if (strlen (arglist[i]) == arglen) + /* Exact match found. */ + return i; + else if (matchind == -1) + /* First nonexact match found. */ + matchind = i; + else + { + /* Second nonexact match found. */ + if (vallist == NULL + || memcmp (vallist + valsize * matchind, + vallist + valsize * i, valsize)) + { + /* There is a real ambiguity, or we could not + disambiguate. */ + ambiguous = true; + } + } + } + } + if (ambiguous) + return -2; + else + return matchind; +} + +/* Error reporting for argmatch. + CONTEXT is a description of the type of entity that was being matched. + VALUE is the invalid value that was given. + PROBLEM is the return value from argmatch. */ + +void +argmatch_invalid (const char *context, const char *value, ptrdiff_t problem) +{ + char const *format = (problem == -1 + ? _("invalid argument %s for %s") + : _("ambiguous argument %s for %s")); + + error (0, 0, format, quotearg_n_style (0, ARGMATCH_QUOTING_STYLE, value), + quote_n (1, context)); +} + +/* List the valid arguments for argmatch. + ARGLIST is the same as in argmatch. + VALLIST is a pointer to an array of values. + VALSIZE is the size of the elements of VALLIST */ +void +argmatch_valid (const char *const *arglist, + const char *vallist, size_t valsize) +{ + size_t i; + const char *last_val = NULL; + + /* We try to put synonyms on the same line. The assumption is that + synonyms follow each other */ + fprintf (stderr, _("Valid arguments are:")); + for (i = 0; arglist[i]; i++) + if ((i == 0) + || memcmp (last_val, vallist + valsize * i, valsize)) + { + fprintf (stderr, "\n - `%s'", arglist[i]); + last_val = vallist + valsize * i; + } + else + { + fprintf (stderr, ", `%s'", arglist[i]); + } + putc ('\n', stderr); +} + +/* Never failing versions of the previous functions. + + CONTEXT is the context for which argmatch is called (e.g., + "--version-control", or "$VERSION_CONTROL" etc.). Upon failure, + calls the (supposed never to return) function EXIT_FN. */ + +ptrdiff_t +__xargmatch_internal (const char *context, + const char *arg, const char *const *arglist, + const char *vallist, size_t valsize, + argmatch_exit_fn exit_fn) +{ + ptrdiff_t res = argmatch (arg, arglist, vallist, valsize); + if (res >= 0) + /* Success. */ + return res; + + /* We failed. Explain why. */ + argmatch_invalid (context, arg, res); + argmatch_valid (arglist, vallist, valsize); + (*exit_fn) (); + + return -1; /* To please the compilers. */ +} + +/* Look for VALUE in VALLIST, an array of objects of size VALSIZE and + return the first corresponding argument in ARGLIST */ +const char * +argmatch_to_argument (const char *value, + const char *const *arglist, + const char *vallist, size_t valsize) +{ + size_t i; + + for (i = 0; arglist[i]; i++) + if (!memcmp (value, vallist + valsize * i, valsize)) + return arglist[i]; + return NULL; +} + +#ifdef TEST +/* + * Based on "getversion.c" by David MacKenzie <djm@gnu.ai.mit.edu> + */ +char *program_name; + +/* When to make backup files. */ +enum backup_type +{ + /* Never make backups. */ + no_backups, + + /* Make simple backups of every file. */ + simple_backups, + + /* Make numbered backups of files that already have numbered backups, + and simple backups of the others. */ + numbered_existing_backups, + + /* Make numbered backups of every file. */ + numbered_backups +}; + +/* Two tables describing arguments (keys) and their corresponding + values */ +static const char *const backup_args[] = +{ + "no", "none", "off", + "simple", "never", + "existing", "nil", + "numbered", "t", + 0 +}; + +static const enum backup_type backup_vals[] = +{ + no_backups, no_backups, no_backups, + simple_backups, simple_backups, + numbered_existing_backups, numbered_existing_backups, + numbered_backups, numbered_backups +}; + +int +main (int argc, const char *const *argv) +{ + const char *cp; + enum backup_type backup_type = no_backups; + + program_name = (char *) argv[0]; + + if (argc > 2) + { + fprintf (stderr, "Usage: %s [VERSION_CONTROL]\n", program_name); + exit (1); + } + + if ((cp = getenv ("VERSION_CONTROL"))) + backup_type = XARGMATCH ("$VERSION_CONTROL", cp, + backup_args, backup_vals); + + if (argc == 2) + backup_type = XARGMATCH (program_name, argv[1], + backup_args, backup_vals); + + printf ("The version control is `%s'\n", + ARGMATCH_TO_ARGUMENT (backup_type, backup_args, backup_vals)); + + return 0; +} +#endif diff --git a/lib/argmatch.h b/lib/argmatch.h new file mode 100644 index 0000000..f2dfe59 --- /dev/null +++ b/lib/argmatch.h @@ -0,0 +1,103 @@ +/* argmatch.h -- definitions and prototypes for argmatch.c + + Copyright (C) 1990, 1998, 1999, 2001, 2002, 2004, 2005 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@ai.mit.edu> + Modified by Akim Demaille <demaille@inf.enst.fr> */ + +#ifndef ARGMATCH_H_ +# define ARGMATCH_H_ 1 + +# include <stddef.h> + +# include "verify.h" + +# define ARRAY_CARDINALITY(Array) (sizeof (Array) / sizeof *(Array)) + +/* Assert there are as many real arguments as there are values + (argument list ends with a NULL guard). */ + +# define ARGMATCH_VERIFY(Arglist, Vallist) \ + verify (ARRAY_CARDINALITY (Arglist) == ARRAY_CARDINALITY (Vallist) + 1) + +/* Return the index of the element of ARGLIST (NULL terminated) that + matches with ARG. If VALLIST is not NULL, then use it to resolve + false ambiguities (i.e., different matches of ARG but corresponding + to the same values in VALLIST). */ + +ptrdiff_t argmatch (char const *arg, char const *const *arglist, + char const *vallist, size_t valsize); + +# define ARGMATCH(Arg, Arglist, Vallist) \ + argmatch (Arg, Arglist, (char const *) (Vallist), sizeof *(Vallist)) + +/* xargmatch calls this function when it fails. This function should not + return. By default, this is a function that calls ARGMATCH_DIE which + in turn defaults to `exit (exit_failure)'. */ +typedef void (*argmatch_exit_fn) (void); +extern argmatch_exit_fn argmatch_die; + +/* Report on stderr why argmatch failed. Report correct values. */ + +void argmatch_invalid (char const *context, char const *value, + ptrdiff_t problem); + +/* Left for compatibility with the old name invalid_arg */ + +# define invalid_arg(Context, Value, Problem) \ + argmatch_invalid (Context, Value, Problem) + + + +/* Report on stderr the list of possible arguments. */ + +void argmatch_valid (char const *const *arglist, + char const *vallist, size_t valsize); + +# define ARGMATCH_VALID(Arglist, Vallist) \ + argmatch_valid (Arglist, (char const *) (Vallist), sizeof *(Vallist)) + + + +/* Same as argmatch, but upon failure, reports a explanation on the + failure, and exits using the function EXIT_FN. */ + +ptrdiff_t __xargmatch_internal (char const *context, + char const *arg, char const *const *arglist, + char const *vallist, size_t valsize, + argmatch_exit_fn exit_fn); + +/* Programmer friendly interface to __xargmatch_internal. */ + +# define XARGMATCH(Context, Arg, Arglist, Vallist) \ + ((Vallist) [__xargmatch_internal (Context, Arg, Arglist, \ + (char const *) (Vallist), \ + sizeof *(Vallist), \ + argmatch_die)]) + +/* Convert a value into a corresponding argument. */ + +char const *argmatch_to_argument (char const *value, + char const *const *arglist, + char const *vallist, size_t valsize); + +# define ARGMATCH_TO_ARGUMENT(Value, Arglist, Vallist) \ + argmatch_to_argument (Value, Arglist, \ + (char const *) (Vallist), sizeof *(Vallist)) + +#endif /* ARGMATCH_H_ */ diff --git a/lib/argp-ba.c b/lib/argp-ba.c new file mode 100644 index 0000000..8bb7309 --- /dev/null +++ b/lib/argp-ba.c @@ -0,0 +1,25 @@ +/* Default definition for ARGP_PROGRAM_BUG_ADDRESS. + Copyright (C) 1996, 1997, 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* If set by the user program, it should point to string that is the + bug-reporting address for the program. It will be printed by argp_help if + the ARGP_HELP_BUG_ADDR flag is set (as it is by various standard help + messages), embedded in a sentence that says something like `Report bugs to + ADDR.'. */ +const char *argp_program_bug_address; diff --git a/lib/argp-eexst.c b/lib/argp-eexst.c new file mode 100644 index 0000000..bcab1c0 --- /dev/null +++ b/lib/argp-eexst.c @@ -0,0 +1,31 @@ +/* Default definition for ARGP_ERR_EXIT_STATUS + Copyright (C) 1997 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include <sysexits.h> + +#include "argp.h" + +/* The exit status that argp will use when exiting due to a parsing error. + If not defined or set by the user program, this defaults to EX_USAGE from + <sysexits.h>. */ +error_t argp_err_exit_status = EX_USAGE; diff --git a/lib/argp-fmtstream.c b/lib/argp-fmtstream.c new file mode 100644 index 0000000..0dd9256 --- /dev/null +++ b/lib/argp-fmtstream.c @@ -0,0 +1,435 @@ +/* Word-wrapping and line-truncating streams + Copyright (C) 1997-1999,2001,2002,2003,2005 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* This package emulates glibc `line_wrap_stream' semantics for systems that + don't have that. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include <stdlib.h> +#include <string.h> +#include <errno.h> +#include <stdarg.h> +#include <ctype.h> + +#include "argp-fmtstream.h" +#include "argp-namefrob.h" + +#ifndef ARGP_FMTSTREAM_USE_LINEWRAP + +#ifndef isblank +#define isblank(ch) ((ch)==' ' || (ch)=='\t') +#endif + +#if defined _LIBC && defined USE_IN_LIBIO +# include <wchar.h> +# include <libio/libioP.h> +# define __vsnprintf(s, l, f, a) _IO_vsnprintf (s, l, f, a) +#endif + +#define INIT_BUF_SIZE 200 +#define PRINTF_SIZE_GUESS 150 + +/* Return an argp_fmtstream that outputs to STREAM, and which prefixes lines + written on it with LMARGIN spaces and limits them to RMARGIN columns + total. If WMARGIN >= 0, words that extend past RMARGIN are wrapped by + replacing the whitespace before them with a newline and WMARGIN spaces. + Otherwise, chars beyond RMARGIN are simply dropped until a newline. + Returns NULL if there was an error. */ +argp_fmtstream_t +__argp_make_fmtstream (FILE *stream, + size_t lmargin, size_t rmargin, ssize_t wmargin) +{ + argp_fmtstream_t fs; + + fs = (struct argp_fmtstream *) malloc (sizeof (struct argp_fmtstream)); + if (fs != NULL) + { + fs->stream = stream; + + fs->lmargin = lmargin; + fs->rmargin = rmargin; + fs->wmargin = wmargin; + fs->point_col = 0; + fs->point_offs = 0; + + fs->buf = (char *) malloc (INIT_BUF_SIZE); + if (! fs->buf) + { + free (fs); + fs = 0; + } + else + { + fs->p = fs->buf; + fs->end = fs->buf + INIT_BUF_SIZE; + } + } + + return fs; +} +#if 0 +/* Not exported. */ +#ifdef weak_alias +weak_alias (__argp_make_fmtstream, argp_make_fmtstream) +#endif +#endif + +/* Flush FS to its stream, and free it (but don't close the stream). */ +void +__argp_fmtstream_free (argp_fmtstream_t fs) +{ + __argp_fmtstream_update (fs); + if (fs->p > fs->buf) + { +#ifdef USE_IN_LIBIO + __fxprintf (fs->stream, "%.*s", (int) (fs->p - fs->buf), fs->buf); +#else + fwrite_unlocked (fs->buf, 1, fs->p - fs->buf, fs->stream); +#endif + } + free (fs->buf); + free (fs); +} +#if 0 +/* Not exported. */ +#ifdef weak_alias +weak_alias (__argp_fmtstream_free, argp_fmtstream_free) +#endif +#endif + +/* Process FS's buffer so that line wrapping is done from POINT_OFFS to the + end of its buffer. This code is mostly from glibc stdio/linewrap.c. */ +void +__argp_fmtstream_update (argp_fmtstream_t fs) +{ + char *buf, *nl; + size_t len; + + /* Scan the buffer for newlines. */ + buf = fs->buf + fs->point_offs; + while (buf < fs->p) + { + size_t r; + + if (fs->point_col == 0 && fs->lmargin != 0) + { + /* We are starting a new line. Print spaces to the left margin. */ + const size_t pad = fs->lmargin; + if (fs->p + pad < fs->end) + { + /* We can fit in them in the buffer by moving the + buffer text up and filling in the beginning. */ + memmove (buf + pad, buf, fs->p - buf); + fs->p += pad; /* Compensate for bigger buffer. */ + memset (buf, ' ', pad); /* Fill in the spaces. */ + buf += pad; /* Don't bother searching them. */ + } + else + { + /* No buffer space for spaces. Must flush. */ + size_t i; + for (i = 0; i < pad; i++) + { +#ifdef USE_IN_LIBIO + if (_IO_fwide (fs->stream, 0) > 0) + putwc_unlocked (L' ', fs->stream); + else +#endif + putc_unlocked (' ', fs->stream); + } + } + fs->point_col = pad; + } + + len = fs->p - buf; + nl = memchr (buf, '\n', len); + + if (fs->point_col < 0) + fs->point_col = 0; + + if (!nl) + { + /* The buffer ends in a partial line. */ + + if (fs->point_col + len < fs->rmargin) + { + /* The remaining buffer text is a partial line and fits + within the maximum line width. Advance point for the + characters to be written and stop scanning. */ + fs->point_col += len; + break; + } + else + /* Set the end-of-line pointer for the code below to + the end of the buffer. */ + nl = fs->p; + } + else if (fs->point_col + (nl - buf) < (ssize_t) fs->rmargin) + { + /* The buffer contains a full line that fits within the maximum + line width. Reset point and scan the next line. */ + fs->point_col = 0; + buf = nl + 1; + continue; + } + + /* This line is too long. */ + r = fs->rmargin - 1; + + if (fs->wmargin < 0) + { + /* Truncate the line by overwriting the excess with the + newline and anything after it in the buffer. */ + if (nl < fs->p) + { + memmove (buf + (r - fs->point_col), nl, fs->p - nl); + fs->p -= buf + (r - fs->point_col) - nl; + /* Reset point for the next line and start scanning it. */ + fs->point_col = 0; + buf += r + 1; /* Skip full line plus \n. */ + } + else + { + /* The buffer ends with a partial line that is beyond the + maximum line width. Advance point for the characters + written, and discard those past the max from the buffer. */ + fs->point_col += len; + fs->p -= fs->point_col - r; + break; + } + } + else + { + /* Do word wrap. Go to the column just past the maximum line + width and scan back for the beginning of the word there. + Then insert a line break. */ + + char *p, *nextline; + int i; + + p = buf + (r + 1 - fs->point_col); + while (p >= buf && !isblank (*p)) + --p; + nextline = p + 1; /* This will begin the next line. */ + + if (nextline > buf) + { + /* Swallow separating blanks. */ + if (p >= buf) + do + --p; + while (p >= buf && isblank (*p)); + nl = p + 1; /* The newline will replace the first blank. */ + } + else + { + /* A single word that is greater than the maximum line width. + Oh well. Put it on an overlong line by itself. */ + p = buf + (r + 1 - fs->point_col); + /* Find the end of the long word. */ + if (p < nl) + do + ++p; + while (p < nl && !isblank (*p)); + if (p == nl) + { + /* It already ends a line. No fussing required. */ + fs->point_col = 0; + buf = nl + 1; + continue; + } + /* We will move the newline to replace the first blank. */ + nl = p; + /* Swallow separating blanks. */ + do + ++p; + while (isblank (*p)); + /* The next line will start here. */ + nextline = p; + } + + /* Note: There are a bunch of tests below for + NEXTLINE == BUF + LEN + 1; this case is where NL happens to fall + at the end of the buffer, and NEXTLINE is in fact empty (and so + we need not be careful to maintain its contents). */ + + if ((nextline == buf + len + 1 + ? fs->end - nl < fs->wmargin + 1 + : nextline - (nl + 1) < fs->wmargin) + && fs->p > nextline) + { + /* The margin needs more blanks than we removed. */ + if (fs->end - fs->p > fs->wmargin + 1) + /* Make some space for them. */ + { + size_t mv = fs->p - nextline; + memmove (nl + 1 + fs->wmargin, nextline, mv); + nextline = nl + 1 + fs->wmargin; + len = nextline + mv - buf; + *nl++ = '\n'; + } + else + /* Output the first line so we can use the space. */ + { +#ifdef _LIBC + __fxprintf (fs->stream, "%.*s\n", + (int) (nl - fs->buf), fs->buf); +#else + if (nl > fs->buf) + fwrite_unlocked (fs->buf, 1, nl - fs->buf, fs->stream); + putc_unlocked ('\n', fs->stream); +#endif + + len += buf - fs->buf; + nl = buf = fs->buf; + } + } + else + /* We can fit the newline and blanks in before + the next word. */ + *nl++ = '\n'; + + if (nextline - nl >= fs->wmargin + || (nextline == buf + len + 1 && fs->end - nextline >= fs->wmargin)) + /* Add blanks up to the wrap margin column. */ + for (i = 0; i < fs->wmargin; ++i) + *nl++ = ' '; + else + for (i = 0; i < fs->wmargin; ++i) +#ifdef USE_IN_LIBIO + if (_IO_fwide (fs->stream, 0) > 0) + putwc_unlocked (L' ', fs->stream); + else +#endif + putc_unlocked (' ', fs->stream); + + /* Copy the tail of the original buffer into the current buffer + position. */ + if (nl < nextline) + memmove (nl, nextline, buf + len - nextline); + len -= nextline - buf; + + /* Continue the scan on the remaining lines in the buffer. */ + buf = nl; + + /* Restore bufp to include all the remaining text. */ + fs->p = nl + len; + + /* Reset the counter of what has been output this line. If wmargin + is 0, we want to avoid the lmargin getting added, so we set + point_col to a magic value of -1 in that case. */ + fs->point_col = fs->wmargin ? fs->wmargin : -1; + } + } + + /* Remember that we've scanned as far as the end of the buffer. */ + fs->point_offs = fs->p - fs->buf; +} + +/* Ensure that FS has space for AMOUNT more bytes in its buffer, either by + growing the buffer, or by flushing it. True is returned iff we succeed. */ +int +__argp_fmtstream_ensure (struct argp_fmtstream *fs, size_t amount) +{ + if ((size_t) (fs->end - fs->p) < amount) + { + ssize_t wrote; + + /* Flush FS's buffer. */ + __argp_fmtstream_update (fs); + +#ifdef _LIBC + __fxprintf (fs->stream, "%.*s", (int) (fs->p - fs->buf), fs->buf); + wrote = fs->p - fs->buf; +#else + wrote = fwrite_unlocked (fs->buf, 1, fs->p - fs->buf, fs->stream); +#endif + if (wrote == fs->p - fs->buf) + { + fs->p = fs->buf; + fs->point_offs = 0; + } + else + { + fs->p -= wrote; + fs->point_offs -= wrote; + memmove (fs->buf, fs->buf + wrote, fs->p - fs->buf); + return 0; + } + + if ((size_t) (fs->end - fs->buf) < amount) + /* Gotta grow the buffer. */ + { + size_t old_size = fs->end - fs->buf; + size_t new_size = old_size + amount; + char *new_buf; + + if (new_size < old_size || ! (new_buf = realloc (fs->buf, new_size))) + { + __set_errno (ENOMEM); + return 0; + } + + fs->buf = new_buf; + fs->end = new_buf + new_size; + fs->p = fs->buf; + } + } + + return 1; +} + +ssize_t +__argp_fmtstream_printf (struct argp_fmtstream *fs, const char *fmt, ...) +{ + int out; + size_t avail; + size_t size_guess = PRINTF_SIZE_GUESS; /* How much space to reserve. */ + + do + { + va_list args; + + if (! __argp_fmtstream_ensure (fs, size_guess)) + return -1; + + va_start (args, fmt); + avail = fs->end - fs->p; + out = __vsnprintf (fs->p, avail, fmt, args); + va_end (args); + if ((size_t) out >= avail) + size_guess = out + 1; + } + while ((size_t) out >= avail); + + fs->p += out; + + return out; +} +#if 0 +/* Not exported. */ +#ifdef weak_alias +weak_alias (__argp_fmtstream_printf, argp_fmtstream_printf) +#endif +#endif + +#endif /* !ARGP_FMTSTREAM_USE_LINEWRAP */ diff --git a/lib/argp-fmtstream.h b/lib/argp-fmtstream.h new file mode 100644 index 0000000..e045a72 --- /dev/null +++ b/lib/argp-fmtstream.h @@ -0,0 +1,301 @@ +/* Word-wrapping and line-truncating streams. + Copyright (C) 1997, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* This package emulates glibc `line_wrap_stream' semantics for systems that + don't have that. If the system does have it, it is just a wrapper for + that. This header file is only used internally while compiling argp, and + shouldn't be installed. */ + +#ifndef _ARGP_FMTSTREAM_H +#define _ARGP_FMTSTREAM_H + +#include <stdio.h> +#include <string.h> +#include <unistd.h> + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) || __STRICT_ANSI__ +# define __format__ format +# define __printf__ printf +# endif +#endif + +#if (_LIBC - 0 && !defined (USE_IN_LIBIO)) \ + || (defined (__GNU_LIBRARY__) && defined (HAVE_LINEWRAP_H)) +/* line_wrap_stream is available, so use that. */ +#define ARGP_FMTSTREAM_USE_LINEWRAP +#endif + +#ifdef ARGP_FMTSTREAM_USE_LINEWRAP +/* Just be a simple wrapper for line_wrap_stream; the semantics are + *slightly* different, as line_wrap_stream doesn't actually make a new + object, it just modifies the given stream (reversibly) to do + line-wrapping. Since we control who uses this code, it doesn't matter. */ + +#include <linewrap.h> + +typedef FILE *argp_fmtstream_t; + +#define argp_make_fmtstream line_wrap_stream +#define __argp_make_fmtstream line_wrap_stream +#define argp_fmtstream_free line_unwrap_stream +#define __argp_fmtstream_free line_unwrap_stream + +#define __argp_fmtstream_putc(fs,ch) putc(ch,fs) +#define argp_fmtstream_putc(fs,ch) putc(ch,fs) +#define __argp_fmtstream_puts(fs,str) fputs(str,fs) +#define argp_fmtstream_puts(fs,str) fputs(str,fs) +#define __argp_fmtstream_write(fs,str,len) fwrite(str,1,len,fs) +#define argp_fmtstream_write(fs,str,len) fwrite(str,1,len,fs) +#define __argp_fmtstream_printf fprintf +#define argp_fmtstream_printf fprintf + +#define __argp_fmtstream_lmargin line_wrap_lmargin +#define argp_fmtstream_lmargin line_wrap_lmargin +#define __argp_fmtstream_set_lmargin line_wrap_set_lmargin +#define argp_fmtstream_set_lmargin line_wrap_set_lmargin +#define __argp_fmtstream_rmargin line_wrap_rmargin +#define argp_fmtstream_rmargin line_wrap_rmargin +#define __argp_fmtstream_set_rmargin line_wrap_set_rmargin +#define argp_fmtstream_set_rmargin line_wrap_set_rmargin +#define __argp_fmtstream_wmargin line_wrap_wmargin +#define argp_fmtstream_wmargin line_wrap_wmargin +#define __argp_fmtstream_set_wmargin line_wrap_set_wmargin +#define argp_fmtstream_set_wmargin line_wrap_set_wmargin +#define __argp_fmtstream_point line_wrap_point +#define argp_fmtstream_point line_wrap_point + +#else /* !ARGP_FMTSTREAM_USE_LINEWRAP */ +/* Guess we have to define our own version. */ + +struct argp_fmtstream +{ + FILE *stream; /* The stream we're outputting to. */ + + size_t lmargin, rmargin; /* Left and right margins. */ + ssize_t wmargin; /* Margin to wrap to, or -1 to truncate. */ + + /* Point in buffer to which we've processed for wrapping, but not output. */ + size_t point_offs; + /* Output column at POINT_OFFS, or -1 meaning 0 but don't add lmargin. */ + ssize_t point_col; + + char *buf; /* Output buffer. */ + char *p; /* Current end of text in BUF. */ + char *end; /* Absolute end of BUF. */ +}; + +typedef struct argp_fmtstream *argp_fmtstream_t; + +/* Return an argp_fmtstream that outputs to STREAM, and which prefixes lines + written on it with LMARGIN spaces and limits them to RMARGIN columns + total. If WMARGIN >= 0, words that extend past RMARGIN are wrapped by + replacing the whitespace before them with a newline and WMARGIN spaces. + Otherwise, chars beyond RMARGIN are simply dropped until a newline. + Returns NULL if there was an error. */ +extern argp_fmtstream_t __argp_make_fmtstream (FILE *__stream, + size_t __lmargin, + size_t __rmargin, + ssize_t __wmargin); +extern argp_fmtstream_t argp_make_fmtstream (FILE *__stream, + size_t __lmargin, + size_t __rmargin, + ssize_t __wmargin); + +/* Flush __FS to its stream, and free it (but don't close the stream). */ +extern void __argp_fmtstream_free (argp_fmtstream_t __fs); +extern void argp_fmtstream_free (argp_fmtstream_t __fs); + +extern ssize_t __argp_fmtstream_printf (argp_fmtstream_t __fs, + const char *__fmt, ...) + __attribute__ ((__format__ (printf, 2, 3))); +extern ssize_t argp_fmtstream_printf (argp_fmtstream_t __fs, + const char *__fmt, ...) + __attribute__ ((__format__ (printf, 2, 3))); + +extern int __argp_fmtstream_putc (argp_fmtstream_t __fs, int __ch); +extern int argp_fmtstream_putc (argp_fmtstream_t __fs, int __ch); + +extern int __argp_fmtstream_puts (argp_fmtstream_t __fs, const char *__str); +extern int argp_fmtstream_puts (argp_fmtstream_t __fs, const char *__str); + +extern size_t __argp_fmtstream_write (argp_fmtstream_t __fs, + const char *__str, size_t __len); +extern size_t argp_fmtstream_write (argp_fmtstream_t __fs, + const char *__str, size_t __len); + +/* Access macros for various bits of state. */ +#define argp_fmtstream_lmargin(__fs) ((__fs)->lmargin) +#define argp_fmtstream_rmargin(__fs) ((__fs)->rmargin) +#define argp_fmtstream_wmargin(__fs) ((__fs)->wmargin) +#define __argp_fmtstream_lmargin argp_fmtstream_lmargin +#define __argp_fmtstream_rmargin argp_fmtstream_rmargin +#define __argp_fmtstream_wmargin argp_fmtstream_wmargin + +/* Set __FS's left margin to LMARGIN and return the old value. */ +extern size_t argp_fmtstream_set_lmargin (argp_fmtstream_t __fs, + size_t __lmargin); +extern size_t __argp_fmtstream_set_lmargin (argp_fmtstream_t __fs, + size_t __lmargin); + +/* Set __FS's right margin to __RMARGIN and return the old value. */ +extern size_t argp_fmtstream_set_rmargin (argp_fmtstream_t __fs, + size_t __rmargin); +extern size_t __argp_fmtstream_set_rmargin (argp_fmtstream_t __fs, + size_t __rmargin); + +/* Set __FS's wrap margin to __WMARGIN and return the old value. */ +extern size_t argp_fmtstream_set_wmargin (argp_fmtstream_t __fs, + size_t __wmargin); +extern size_t __argp_fmtstream_set_wmargin (argp_fmtstream_t __fs, + size_t __wmargin); + +/* Return the column number of the current output point in __FS. */ +extern size_t argp_fmtstream_point (argp_fmtstream_t __fs); +extern size_t __argp_fmtstream_point (argp_fmtstream_t __fs); + +/* Internal routines. */ +extern void _argp_fmtstream_update (argp_fmtstream_t __fs); +extern void __argp_fmtstream_update (argp_fmtstream_t __fs); +extern int _argp_fmtstream_ensure (argp_fmtstream_t __fs, size_t __amount); +extern int __argp_fmtstream_ensure (argp_fmtstream_t __fs, size_t __amount); + +#ifdef __OPTIMIZE__ +/* Inline versions of above routines. */ + +#if !_LIBC +#define __argp_fmtstream_putc argp_fmtstream_putc +#define __argp_fmtstream_puts argp_fmtstream_puts +#define __argp_fmtstream_write argp_fmtstream_write +#define __argp_fmtstream_set_lmargin argp_fmtstream_set_lmargin +#define __argp_fmtstream_set_rmargin argp_fmtstream_set_rmargin +#define __argp_fmtstream_set_wmargin argp_fmtstream_set_wmargin +#define __argp_fmtstream_point argp_fmtstream_point +#define __argp_fmtstream_update _argp_fmtstream_update +#define __argp_fmtstream_ensure _argp_fmtstream_ensure +#endif + +#ifndef ARGP_FS_EI +#define ARGP_FS_EI extern inline +#endif + +ARGP_FS_EI size_t +__argp_fmtstream_write (argp_fmtstream_t __fs, + const char *__str, size_t __len) +{ + if (__fs->p + __len <= __fs->end || __argp_fmtstream_ensure (__fs, __len)) + { + memcpy (__fs->p, __str, __len); + __fs->p += __len; + return __len; + } + else + return 0; +} + +ARGP_FS_EI int +__argp_fmtstream_puts (argp_fmtstream_t __fs, const char *__str) +{ + size_t __len = strlen (__str); + if (__len) + { + size_t __wrote = __argp_fmtstream_write (__fs, __str, __len); + return __wrote == __len ? 0 : -1; + } + else + return 0; +} + +ARGP_FS_EI int +__argp_fmtstream_putc (argp_fmtstream_t __fs, int __ch) +{ + if (__fs->p < __fs->end || __argp_fmtstream_ensure (__fs, 1)) + return *__fs->p++ = __ch; + else + return EOF; +} + +/* Set __FS's left margin to __LMARGIN and return the old value. */ +ARGP_FS_EI size_t +__argp_fmtstream_set_lmargin (argp_fmtstream_t __fs, size_t __lmargin) +{ + size_t __old; + if ((size_t) (__fs->p - __fs->buf) > __fs->point_offs) + __argp_fmtstream_update (__fs); + __old = __fs->lmargin; + __fs->lmargin = __lmargin; + return __old; +} + +/* Set __FS's right margin to __RMARGIN and return the old value. */ +ARGP_FS_EI size_t +__argp_fmtstream_set_rmargin (argp_fmtstream_t __fs, size_t __rmargin) +{ + size_t __old; + if ((size_t) (__fs->p - __fs->buf) > __fs->point_offs) + __argp_fmtstream_update (__fs); + __old = __fs->rmargin; + __fs->rmargin = __rmargin; + return __old; +} + +/* Set FS's wrap margin to __WMARGIN and return the old value. */ +ARGP_FS_EI size_t +__argp_fmtstream_set_wmargin (argp_fmtstream_t __fs, size_t __wmargin) +{ + size_t __old; + if ((size_t) (__fs->p - __fs->buf) > __fs->point_offs) + __argp_fmtstream_update (__fs); + __old = __fs->wmargin; + __fs->wmargin = __wmargin; + return __old; +} + +/* Return the column number of the current output point in __FS. */ +ARGP_FS_EI size_t +__argp_fmtstream_point (argp_fmtstream_t __fs) +{ + if ((size_t) (__fs->p - __fs->buf) > __fs->point_offs) + __argp_fmtstream_update (__fs); + return __fs->point_col >= 0 ? __fs->point_col : 0; +} + +#if !_LIBC +#undef __argp_fmtstream_putc +#undef __argp_fmtstream_puts +#undef __argp_fmtstream_write +#undef __argp_fmtstream_set_lmargin +#undef __argp_fmtstream_set_rmargin +#undef __argp_fmtstream_set_wmargin +#undef __argp_fmtstream_point +#undef __argp_fmtstream_update +#undef __argp_fmtstream_ensure +#endif + +#endif /* __OPTIMIZE__ */ + +#endif /* ARGP_FMTSTREAM_USE_LINEWRAP */ + +#endif /* argp-fmtstream.h */ diff --git a/lib/argp-fs-xinl.c b/lib/argp-fs-xinl.c new file mode 100644 index 0000000..3b4c917 --- /dev/null +++ b/lib/argp-fs-xinl.c @@ -0,0 +1,43 @@ +/* Real definitions for extern inline functions in argp-fmtstream.h + Copyright (C) 1997, 2003, 2004 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#define ARGP_FS_EI +#undef __OPTIMIZE__ +#define __OPTIMIZE__ 1 +#include "argp-fmtstream.h" + +#if 0 +/* Not exported. */ +/* Add weak aliases. */ +#if _LIBC - 0 && !defined (ARGP_FMTSTREAM_USE_LINEWRAP) && defined (weak_alias) + +weak_alias (__argp_fmtstream_putc, argp_fmtstream_putc) +weak_alias (__argp_fmtstream_puts, argp_fmtstream_puts) +weak_alias (__argp_fmtstream_write, argp_fmtstream_write) +weak_alias (__argp_fmtstream_set_lmargin, argp_fmtstream_set_lmargin) +weak_alias (__argp_fmtstream_set_rmargin, argp_fmtstream_set_rmargin) +weak_alias (__argp_fmtstream_set_wmargin, argp_fmtstream_set_wmargin) +weak_alias (__argp_fmtstream_point, argp_fmtstream_point) + +#endif +#endif diff --git a/lib/argp-help.c b/lib/argp-help.c new file mode 100644 index 0000000..396e733 --- /dev/null +++ b/lib/argp-help.c @@ -0,0 +1,1954 @@ +/* Hierarchial argument parsing help output + Copyright (C) 1995-2005, 2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GNU_SOURCE +# define _GNU_SOURCE 1 +#endif + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include <alloca.h> +#include <errno.h> +#include <stddef.h> +#include <stdlib.h> +#include <string.h> +#include <assert.h> +#include <stdarg.h> +#include <ctype.h> +#include <limits.h> +#ifdef USE_IN_LIBIO +# include <wchar.h> +#endif + +#ifdef _LIBC +# include <libintl.h> +# undef dgettext +# define dgettext(domain, msgid) \ + INTUSE(__dcgettext) (domain, msgid, LC_MESSAGES) +#else +# include "gettext.h" +#endif + +#include "argp.h" +#include "argp-fmtstream.h" +#include "argp-namefrob.h" + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* User-selectable (using an environment variable) formatting parameters. + + These may be specified in an environment variable called `ARGP_HELP_FMT', + with a contents like: VAR1=VAL1,VAR2=VAL2,BOOLVAR2,no-BOOLVAR2 + Where VALn must be a positive integer. The list of variables is in the + UPARAM_NAMES vector, below. */ + +/* Default parameters. */ +#define DUP_ARGS 0 /* True if option argument can be duplicated. */ +#define DUP_ARGS_NOTE 1 /* True to print a note about duplicate args. */ +#define SHORT_OPT_COL 2 /* column in which short options start */ +#define LONG_OPT_COL 6 /* column in which long options start */ +#define DOC_OPT_COL 2 /* column in which doc options start */ +#define OPT_DOC_COL 29 /* column in which option text starts */ +#define HEADER_COL 1 /* column in which group headers are printed */ +#define USAGE_INDENT 12 /* indentation of wrapped usage lines */ +#define RMARGIN 79 /* right margin used for wrapping */ + +/* User-selectable (using an environment variable) formatting parameters. + They must all be of type `int' for the parsing code to work. */ +struct uparams +{ + /* If true, arguments for an option are shown with both short and long + options, even when a given option has both, e.g. `-x ARG, --longx=ARG'. + If false, then if an option has both, the argument is only shown with + the long one, e.g., `-x, --longx=ARG', and a message indicating that + this really means both is printed below the options. */ + int dup_args; + + /* This is true if when DUP_ARGS is false, and some duplicate arguments have + been suppressed, an explanatory message should be printed. */ + int dup_args_note; + + /* Various output columns. */ + int short_opt_col; /* column in which short options start */ + int long_opt_col; /* column in which long options start */ + int doc_opt_col; /* column in which doc options start */ + int opt_doc_col; /* column in which option text starts */ + int header_col; /* column in which group headers are printed */ + int usage_indent; /* indentation of wrapped usage lines */ + int rmargin; /* right margin used for wrapping */ + + int valid; /* True when the values in here are valid. */ +}; + +/* This is a global variable, as user options are only ever read once. */ +static struct uparams uparams = { + DUP_ARGS, DUP_ARGS_NOTE, + SHORT_OPT_COL, LONG_OPT_COL, DOC_OPT_COL, OPT_DOC_COL, HEADER_COL, + USAGE_INDENT, RMARGIN, + 0 +}; + +/* A particular uparam, and what the user name is. */ +struct uparam_name +{ + const char *name; /* User name. */ + int is_bool; /* Whether it's `boolean'. */ + size_t uparams_offs; /* Location of the (int) field in UPARAMS. */ +}; + +/* The name-field mappings we know about. */ +static const struct uparam_name uparam_names[] = +{ + { "dup-args", 1, offsetof (struct uparams, dup_args) }, + { "dup-args-note", 1, offsetof (struct uparams, dup_args_note) }, + { "short-opt-col", 0, offsetof (struct uparams, short_opt_col) }, + { "long-opt-col", 0, offsetof (struct uparams, long_opt_col) }, + { "doc-opt-col", 0, offsetof (struct uparams, doc_opt_col) }, + { "opt-doc-col", 0, offsetof (struct uparams, opt_doc_col) }, + { "header-col", 0, offsetof (struct uparams, header_col) }, + { "usage-indent", 0, offsetof (struct uparams, usage_indent) }, + { "rmargin", 0, offsetof (struct uparams, rmargin) }, + { 0 } +}; + +static void +validate_uparams (const struct argp_state *state, struct uparams *upptr) +{ + const struct uparam_name *up; + + for (up = uparam_names; up->name; up++) + { + if (up->is_bool + || up->uparams_offs == offsetof (struct uparams, rmargin)) + continue; + if (*(int *)((char *)upptr + up->uparams_offs) >= upptr->rmargin) + { + __argp_failure (state, 0, 0, + dgettext (state->root_argp->argp_domain, + "\ +ARGP_HELP_FMT: %s value is less than or equal to %s"), + "rmargin", up->name); + return; + } + } + uparams = *upptr; + uparams.valid = 1; +} + +/* Read user options from the environment, and fill in UPARAMS appropiately. */ +static void +fill_in_uparams (const struct argp_state *state) +{ + const char *var = getenv ("ARGP_HELP_FMT"); + struct uparams new_params = uparams; + +#define SKIPWS(p) do { while (isspace ((unsigned char) *p)) p++; } while (0); + + if (var) + { + /* Parse var. */ + while (*var) + { + SKIPWS (var); + + if (isalpha ((unsigned char) *var)) + { + size_t var_len; + const struct uparam_name *un; + int unspec = 0, val = 0; + const char *arg = var; + + while (isalnum ((unsigned char) *arg) || *arg == '-' || *arg == '_') + arg++; + var_len = arg - var; + + SKIPWS (arg); + + if (*arg == '\0' || *arg == ',') + unspec = 1; + else if (*arg == '=') + { + arg++; + SKIPWS (arg); + } + + if (unspec) + { + if (var[0] == 'n' && var[1] == 'o' && var[2] == '-') + { + val = 0; + var += 3; + var_len -= 3; + } + else + val = 1; + } + else if (isdigit ((unsigned char) *arg)) + { + val = atoi (arg); + while (isdigit ((unsigned char) *arg)) + arg++; + SKIPWS (arg); + } + + for (un = uparam_names; un->name; un++) + if (strlen (un->name) == var_len + && strncmp (var, un->name, var_len) == 0) + { + if (unspec && !un->is_bool) + __argp_failure (state, 0, 0, + dgettext (state->root_argp->argp_domain, + "\ +%.*s: ARGP_HELP_FMT parameter requires a value"), + (int) var_len, var); + else if (val < 0) + __argp_failure (state, 0, 0, + dgettext (state->root_argp->argp_domain, + "\ +%.*s: ARGP_HELP_FMT parameter must be positive"), + (int) var_len, var); + else + *(int *)((char *)&new_params + un->uparams_offs) = val; + break; + } + if (! un->name) + __argp_failure (state, 0, 0, + dgettext (state->root_argp->argp_domain, "\ +%.*s: Unknown ARGP_HELP_FMT parameter"), + (int) var_len, var); + + var = arg; + if (*var == ',') + var++; + } + else if (*var) + { + __argp_failure (state, 0, 0, + dgettext (state->root_argp->argp_domain, + "Garbage in ARGP_HELP_FMT: %s"), var); + break; + } + } + validate_uparams (state, &new_params); + } +} + +/* Returns true if OPT hasn't been marked invisible. Visibility only affects + whether OPT is displayed or used in sorting, not option shadowing. */ +#define ovisible(opt) (! ((opt)->flags & OPTION_HIDDEN)) + +/* Returns true if OPT is an alias for an earlier option. */ +#define oalias(opt) ((opt)->flags & OPTION_ALIAS) + +/* Returns true if OPT is an documentation-only entry. */ +#define odoc(opt) ((opt)->flags & OPTION_DOC) + +/* Returns true if OPT should not be translated */ +#define onotrans(opt) ((opt)->flags & OPTION_NO_TRANS) + +/* Returns true if OPT is the end-of-list marker for a list of options. */ +#define oend(opt) __option_is_end (opt) + +/* Returns true if OPT has a short option. */ +#define oshort(opt) __option_is_short (opt) + +/* + The help format for a particular option is like: + + -xARG, -yARG, --long1=ARG, --long2=ARG Documentation... + + Where ARG will be omitted if there's no argument, for this option, or + will be surrounded by "[" and "]" appropiately if the argument is + optional. The documentation string is word-wrapped appropiately, and if + the list of options is long enough, it will be started on a separate line. + If there are no short options for a given option, the first long option is + indented slighly in a way that's supposed to make most long options appear + to be in a separate column. + + For example, the following output (from ps): + + -p PID, --pid=PID List the process PID + --pgrp=PGRP List processes in the process group PGRP + -P, -x, --no-parent Include processes without parents + -Q, --all-fields Don't elide unusable fields (normally if there's + some reason ps can't print a field for any + process, it's removed from the output entirely) + -r, --reverse, --gratuitously-long-reverse-option + Reverse the order of any sort + --session[=SID] Add the processes from the session SID (which + defaults to the sid of the current process) + + Here are some more options: + -f ZOT, --foonly=ZOT Glork a foonly + -z, --zaza Snit a zar + + -?, --help Give this help list + --usage Give a short usage message + -V, --version Print program version + + The struct argp_option array for the above could look like: + + { + {"pid", 'p', "PID", 0, "List the process PID"}, + {"pgrp", OPT_PGRP, "PGRP", 0, "List processes in the process group PGRP"}, + {"no-parent", 'P', 0, 0, "Include processes without parents"}, + {0, 'x', 0, OPTION_ALIAS}, + {"all-fields",'Q', 0, 0, "Don't elide unusable fields (normally" + " if there's some reason ps can't" + " print a field for any process, it's" + " removed from the output entirely)" }, + {"reverse", 'r', 0, 0, "Reverse the order of any sort"}, + {"gratuitously-long-reverse-option", 0, 0, OPTION_ALIAS}, + {"session", OPT_SESS, "SID", OPTION_ARG_OPTIONAL, + "Add the processes from the session" + " SID (which defaults to the sid of" + " the current process)" }, + + {0,0,0,0, "Here are some more options:"}, + {"foonly", 'f', "ZOT", 0, "Glork a foonly"}, + {"zaza", 'z', 0, 0, "Snit a zar"}, + + {0} + } + + Note that the last three options are automatically supplied by argp_parse, + unless you tell it not to with ARGP_NO_HELP. + +*/ + +/* Returns true if CH occurs between BEG and END. */ +static int +find_char (char ch, char *beg, char *end) +{ + while (beg < end) + if (*beg == ch) + return 1; + else + beg++; + return 0; +} + +struct hol_cluster; /* fwd decl */ + +struct hol_entry +{ + /* First option. */ + const struct argp_option *opt; + /* Number of options (including aliases). */ + unsigned num; + + /* A pointers into the HOL's short_options field, to the first short option + letter for this entry. The order of the characters following this point + corresponds to the order of options pointed to by OPT, and there are at + most NUM. A short option recorded in a option following OPT is only + valid if it occurs in the right place in SHORT_OPTIONS (otherwise it's + probably been shadowed by some other entry). */ + char *short_options; + + /* Entries are sorted by their group first, in the order: + 1, 2, ..., n, 0, -m, ..., -2, -1 + and then alphabetically within each group. The default is 0. */ + int group; + + /* The cluster of options this entry belongs to, or 0 if none. */ + struct hol_cluster *cluster; + + /* The argp from which this option came. */ + const struct argp *argp; + + /* Position in the array */ + unsigned ord; +}; + +/* A cluster of entries to reflect the argp tree structure. */ +struct hol_cluster +{ + /* A descriptive header printed before options in this cluster. */ + const char *header; + + /* Used to order clusters within the same group with the same parent, + according to the order in which they occurred in the parent argp's child + list. */ + int index; + + /* How to sort this cluster with respect to options and other clusters at the + same depth (clusters always follow options in the same group). */ + int group; + + /* The cluster to which this cluster belongs, or 0 if it's at the base + level. */ + struct hol_cluster *parent; + + /* The argp from which this cluster is (eventually) derived. */ + const struct argp *argp; + + /* The distance this cluster is from the root. */ + int depth; + + /* Clusters in a given hol are kept in a linked list, to make freeing them + possible. */ + struct hol_cluster *next; +}; + +/* A list of options for help. */ +struct hol +{ + /* An array of hol_entry's. */ + struct hol_entry *entries; + /* The number of entries in this hol. If this field is zero, the others + are undefined. */ + unsigned num_entries; + + /* A string containing all short options in this HOL. Each entry contains + pointers into this string, so the order can't be messed with blindly. */ + char *short_options; + + /* Clusters of entries in this hol. */ + struct hol_cluster *clusters; +}; + +/* Create a struct hol from the options in ARGP. CLUSTER is the + hol_cluster in which these entries occur, or 0, if at the root. */ +static struct hol * +make_hol (const struct argp *argp, struct hol_cluster *cluster) +{ + char *so; + const struct argp_option *o; + const struct argp_option *opts = argp->options; + struct hol_entry *entry; + unsigned num_short_options = 0; + struct hol *hol = malloc (sizeof (struct hol)); + + assert (hol); + + hol->num_entries = 0; + hol->clusters = 0; + + if (opts) + { + int cur_group = 0; + + /* The first option must not be an alias. */ + assert (! oalias (opts)); + + /* Calculate the space needed. */ + for (o = opts; ! oend (o); o++) + { + if (! oalias (o)) + hol->num_entries++; + if (oshort (o)) + num_short_options++; /* This is an upper bound. */ + } + + hol->entries = malloc (sizeof (struct hol_entry) * hol->num_entries); + hol->short_options = malloc (num_short_options + 1); + + assert (hol->entries && hol->short_options); + if (SIZE_MAX <= UINT_MAX) + assert (hol->num_entries <= SIZE_MAX / sizeof (struct hol_entry)); + + /* Fill in the entries. */ + so = hol->short_options; + for (o = opts, entry = hol->entries; ! oend (o); entry++) + { + entry->opt = o; + entry->num = 0; + entry->short_options = so; + entry->group = cur_group = + o->group + ? o->group + : ((!o->name && !o->key) + ? cur_group + 1 + : cur_group); + entry->cluster = cluster; + entry->argp = argp; + + do + { + entry->num++; + if (oshort (o) && ! find_char (o->key, hol->short_options, so)) + /* O has a valid short option which hasn't already been used.*/ + *so++ = o->key; + o++; + } + while (! oend (o) && oalias (o)); + } + *so = '\0'; /* null terminated so we can find the length */ + } + + return hol; +} + +/* Add a new cluster to HOL, with the given GROUP and HEADER (taken from the + associated argp child list entry), INDEX, and PARENT, and return a pointer + to it. ARGP is the argp that this cluster results from. */ +static struct hol_cluster * +hol_add_cluster (struct hol *hol, int group, const char *header, int index, + struct hol_cluster *parent, const struct argp *argp) +{ + struct hol_cluster *cl = malloc (sizeof (struct hol_cluster)); + if (cl) + { + cl->group = group; + cl->header = header; + + cl->index = index; + cl->parent = parent; + cl->argp = argp; + cl->depth = parent ? parent->depth + 1 : 0; + + cl->next = hol->clusters; + hol->clusters = cl; + } + return cl; +} + +/* Free HOL and any resources it uses. */ +static void +hol_free (struct hol *hol) +{ + struct hol_cluster *cl = hol->clusters; + + while (cl) + { + struct hol_cluster *next = cl->next; + free (cl); + cl = next; + } + + if (hol->num_entries > 0) + { + free (hol->entries); + free (hol->short_options); + } + + free (hol); +} + +static int +hol_entry_short_iterate (const struct hol_entry *entry, + int (*func)(const struct argp_option *opt, + const struct argp_option *real, + const char *domain, void *cookie), + const char *domain, void *cookie) +{ + unsigned nopts; + int val = 0; + const struct argp_option *opt, *real = entry->opt; + char *so = entry->short_options; + + for (opt = real, nopts = entry->num; nopts > 0 && !val; opt++, nopts--) + if (oshort (opt) && *so == opt->key) + { + if (!oalias (opt)) + real = opt; + if (ovisible (opt)) + val = (*func)(opt, real, domain, cookie); + so++; + } + + return val; +} + +static inline int +__attribute__ ((always_inline)) +hol_entry_long_iterate (const struct hol_entry *entry, + int (*func)(const struct argp_option *opt, + const struct argp_option *real, + const char *domain, void *cookie), + const char *domain, void *cookie) +{ + unsigned nopts; + int val = 0; + const struct argp_option *opt, *real = entry->opt; + + for (opt = real, nopts = entry->num; nopts > 0 && !val; opt++, nopts--) + if (opt->name) + { + if (!oalias (opt)) + real = opt; + if (ovisible (opt)) + val = (*func)(opt, real, domain, cookie); + } + + return val; +} + +/* Iterator that returns true for the first short option. */ +static inline int +until_short (const struct argp_option *opt, const struct argp_option *real, + const char *domain, void *cookie) +{ + return oshort (opt) ? opt->key : 0; +} + +/* Returns the first valid short option in ENTRY, or 0 if there is none. */ +static char +hol_entry_first_short (const struct hol_entry *entry) +{ + return hol_entry_short_iterate (entry, until_short, + entry->argp->argp_domain, 0); +} + +/* Returns the first valid long option in ENTRY, or 0 if there is none. */ +static const char * +hol_entry_first_long (const struct hol_entry *entry) +{ + const struct argp_option *opt; + unsigned num; + for (opt = entry->opt, num = entry->num; num > 0; opt++, num--) + if (opt->name && ovisible (opt)) + return opt->name; + return 0; +} + +/* Returns the entry in HOL with the long option name NAME, or 0 if there is + none. */ +static struct hol_entry * +hol_find_entry (struct hol *hol, const char *name) +{ + struct hol_entry *entry = hol->entries; + unsigned num_entries = hol->num_entries; + + while (num_entries-- > 0) + { + const struct argp_option *opt = entry->opt; + unsigned num_opts = entry->num; + + while (num_opts-- > 0) + if (opt->name && ovisible (opt) && strcmp (opt->name, name) == 0) + return entry; + else + opt++; + + entry++; + } + + return 0; +} + +/* If an entry with the long option NAME occurs in HOL, set it's special + sort position to GROUP. */ +static void +hol_set_group (struct hol *hol, const char *name, int group) +{ + struct hol_entry *entry = hol_find_entry (hol, name); + if (entry) + entry->group = group; +} + +/* Order by group: 0, 1, 2, ..., n, -m, ..., -2, -1. + EQ is what to return if GROUP1 and GROUP2 are the same. */ +static int +group_cmp (int group1, int group2, int eq) +{ + if (group1 == group2) + return eq; + else if ((group1 < 0 && group2 < 0) || (group1 >= 0 && group2 >= 0)) + return group1 - group2; + else + return group2 - group1; +} + +/* Compare clusters CL1 & CL2 by the order that they should appear in + output. */ +static int +hol_cluster_cmp (const struct hol_cluster *cl1, const struct hol_cluster *cl2) +{ + /* If one cluster is deeper than the other, use its ancestor at the same + level, so that finding the common ancestor is straightforward. + + clN->depth > 0 means that clN->parent != NULL (see hol_add_cluster) */ + while (cl1->depth > cl2->depth) + cl1 = cl1->parent; + while (cl2->depth > cl1->depth) + cl2 = cl2->parent; + + /* Now reduce both clusters to their ancestors at the point where both have + a common parent; these can be directly compared. */ + while (cl1->parent != cl2->parent) + cl1 = cl1->parent, cl2 = cl2->parent; + + return group_cmp (cl1->group, cl2->group, cl2->index - cl1->index); +} + +/* Return the ancestor of CL that's just below the root (i.e., has a parent + of 0). */ +static struct hol_cluster * +hol_cluster_base (struct hol_cluster *cl) +{ + while (cl->parent) + cl = cl->parent; + return cl; +} + +/* Return true if CL1 is a child of CL2. */ +static int +hol_cluster_is_child (const struct hol_cluster *cl1, + const struct hol_cluster *cl2) +{ + while (cl1 && cl1 != cl2) + cl1 = cl1->parent; + return cl1 == cl2; +} + +/* Given the name of a OPTION_DOC option, modifies NAME to start at the tail + that should be used for comparisons, and returns true iff it should be + treated as a non-option. */ +static int +canon_doc_option (const char **name) +{ + int non_opt; + + if (!*name) + non_opt = 1; + else + { + /* Skip initial whitespace. */ + while (isspace ((unsigned char) **name)) + (*name)++; + /* Decide whether this looks like an option (leading `-') or not. */ + non_opt = (**name != '-'); + /* Skip until part of name used for sorting. */ + while (**name && !isalnum ((unsigned char) **name)) + (*name)++; + } + return non_opt; +} + +#define HOL_ENTRY_PTRCMP(a,b) ((a)->ord < (b)->ord ? -1 : 1) + +/* Order ENTRY1 & ENTRY2 by the order which they should appear in a help + listing. */ +static int +hol_entry_cmp (const struct hol_entry *entry1, + const struct hol_entry *entry2) +{ + /* The group numbers by which the entries should be ordered; if either is + in a cluster, then this is just the group within the cluster. */ + int group1 = entry1->group, group2 = entry2->group; + int rc; + + if (entry1->cluster != entry2->cluster) + { + /* The entries are not within the same cluster, so we can't compare them + directly, we have to use the appropiate clustering level too. */ + if (! entry1->cluster) + /* ENTRY1 is at the `base level', not in a cluster, so we have to + compare it's group number with that of the base cluster in which + ENTRY2 resides. Note that if they're in the same group, the + clustered option always comes laster. */ + return group_cmp (group1, hol_cluster_base (entry2->cluster)->group, -1); + else if (! entry2->cluster) + /* Likewise, but ENTRY2's not in a cluster. */ + return group_cmp (hol_cluster_base (entry1->cluster)->group, group2, 1); + else + /* Both entries are in clusters, we can just compare the clusters. */ + return (rc = hol_cluster_cmp (entry1->cluster, entry2->cluster)) ? + rc : HOL_ENTRY_PTRCMP(entry1, entry2); + } + else if (group1 == group2) + /* The entries are both in the same cluster and group, so compare them + alphabetically. */ + { + int short1 = hol_entry_first_short (entry1); + int short2 = hol_entry_first_short (entry2); + int doc1 = odoc (entry1->opt); + int doc2 = odoc (entry2->opt); + const char *long1 = hol_entry_first_long (entry1); + const char *long2 = hol_entry_first_long (entry2); + + if (doc1) + doc1 = canon_doc_option (&long1); + if (doc2) + doc2 = canon_doc_option (&long2); + + if (doc1 != doc2) + /* `documentation' options always follow normal options (or + documentation options that *look* like normal options). */ + return doc1 - doc2; + else if (!short1 && !short2 && long1 && long2) + /* Only long options. */ + return (rc = __strcasecmp (long1, long2)) ? + rc : HOL_ENTRY_PTRCMP(entry1, entry2); + else + /* Compare short/short, long/short, short/long, using the first + character of long options. Entries without *any* valid + options (such as options with OPTION_HIDDEN set) will be put + first, but as they're not displayed, it doesn't matter where + they are. */ + { + char first1 = short1 ? short1 : long1 ? *long1 : 0; + char first2 = short2 ? short2 : long2 ? *long2 : 0; +#ifdef _tolower + int lower_cmp = _tolower (first1) - _tolower (first2); +#else + int lower_cmp = tolower (first1) - tolower (first2); +#endif + /* Compare ignoring case, except when the options are both the + same letter, in which case lower-case always comes first. */ + return lower_cmp ? lower_cmp : + (rc = first2 - first1) ? + rc : HOL_ENTRY_PTRCMP(entry1, entry2); + } + } + else + /* Within the same cluster, but not the same group, so just compare + groups. */ + return group_cmp (group1, group2, HOL_ENTRY_PTRCMP(entry1, entry2)); +} + +/* Version of hol_entry_cmp with correct signature for qsort. */ +static int +hol_entry_qcmp (const void *entry1_v, const void *entry2_v) +{ + return hol_entry_cmp (entry1_v, entry2_v); +} + +/* Sort HOL by group and alphabetically by option name (with short options + taking precedence over long). Since the sorting is for display purposes + only, the shadowing of options isn't effected. */ +static void +hol_sort (struct hol *hol) +{ + if (hol->num_entries > 0) + { + unsigned i; + struct hol_entry *e; + for (i = 0, e = hol->entries; i < hol->num_entries; i++, e++) + e->ord = i; + qsort (hol->entries, hol->num_entries, sizeof (struct hol_entry), + hol_entry_qcmp); + } +} + +/* Append MORE to HOL, destroying MORE in the process. Options in HOL shadow + any in MORE with the same name. */ +static void +hol_append (struct hol *hol, struct hol *more) +{ + struct hol_cluster **cl_end = &hol->clusters; + + /* Steal MORE's cluster list, and add it to the end of HOL's. */ + while (*cl_end) + cl_end = &(*cl_end)->next; + *cl_end = more->clusters; + more->clusters = 0; + + /* Merge entries. */ + if (more->num_entries > 0) + { + if (hol->num_entries == 0) + { + hol->num_entries = more->num_entries; + hol->entries = more->entries; + hol->short_options = more->short_options; + more->num_entries = 0; /* Mark MORE's fields as invalid. */ + } + else + /* Append the entries in MORE to those in HOL, taking care to only add + non-shadowed SHORT_OPTIONS values. */ + { + unsigned left; + char *so, *more_so; + struct hol_entry *e; + unsigned num_entries = hol->num_entries + more->num_entries; + struct hol_entry *entries = + malloc (num_entries * sizeof (struct hol_entry)); + unsigned hol_so_len = strlen (hol->short_options); + char *short_options = + malloc (hol_so_len + strlen (more->short_options) + 1); + + assert (entries && short_options); + if (SIZE_MAX <= UINT_MAX) + assert (num_entries <= SIZE_MAX / sizeof (struct hol_entry)); + + __mempcpy (__mempcpy (entries, hol->entries, + hol->num_entries * sizeof (struct hol_entry)), + more->entries, + more->num_entries * sizeof (struct hol_entry)); + + __mempcpy (short_options, hol->short_options, hol_so_len); + + /* Fix up the short options pointers from HOL. */ + for (e = entries, left = hol->num_entries; left > 0; e++, left--) + e->short_options += (short_options - hol->short_options); + + /* Now add the short options from MORE, fixing up its entries + too. */ + so = short_options + hol_so_len; + more_so = more->short_options; + for (left = more->num_entries; left > 0; e++, left--) + { + int opts_left; + const struct argp_option *opt; + + e->short_options = so; + + for (opts_left = e->num, opt = e->opt; opts_left; opt++, opts_left--) + { + int ch = *more_so; + if (oshort (opt) && ch == opt->key) + /* The next short option in MORE_SO, CH, is from OPT. */ + { + if (! find_char (ch, short_options, + short_options + hol_so_len)) + /* The short option CH isn't shadowed by HOL's options, + so add it to the sum. */ + *so++ = ch; + more_so++; + } + } + } + + *so = '\0'; + + free (hol->entries); + free (hol->short_options); + + hol->entries = entries; + hol->num_entries = num_entries; + hol->short_options = short_options; + } + } + + hol_free (more); +} + +/* Inserts enough spaces to make sure STREAM is at column COL. */ +static void +indent_to (argp_fmtstream_t stream, unsigned col) +{ + int needed = col - __argp_fmtstream_point (stream); + while (needed-- > 0) + __argp_fmtstream_putc (stream, ' '); +} + +/* Output to STREAM either a space, or a newline if there isn't room for at + least ENSURE characters before the right margin. */ +static void +space (argp_fmtstream_t stream, size_t ensure) +{ + if (__argp_fmtstream_point (stream) + ensure + >= __argp_fmtstream_rmargin (stream)) + __argp_fmtstream_putc (stream, '\n'); + else + __argp_fmtstream_putc (stream, ' '); +} + +/* If the option REAL has an argument, we print it in using the printf + format REQ_FMT or OPT_FMT depending on whether it's a required or + optional argument. */ +static void +arg (const struct argp_option *real, const char *req_fmt, const char *opt_fmt, + const char *domain, argp_fmtstream_t stream) +{ + if (real->arg) + { + if (real->flags & OPTION_ARG_OPTIONAL) + __argp_fmtstream_printf (stream, opt_fmt, + dgettext (domain, real->arg)); + else + __argp_fmtstream_printf (stream, req_fmt, + dgettext (domain, real->arg)); + } +} + +/* Helper functions for hol_entry_help. */ + +/* State used during the execution of hol_help. */ +struct hol_help_state +{ + /* PREV_ENTRY should contain the previous entry printed, or 0. */ + struct hol_entry *prev_entry; + + /* If an entry is in a different group from the previous one, and SEP_GROUPS + is true, then a blank line will be printed before any output. */ + int sep_groups; + + /* True if a duplicate option argument was suppressed (only ever set if + UPARAMS.dup_args is false). */ + int suppressed_dup_arg; +}; + +/* Some state used while printing a help entry (used to communicate with + helper functions). See the doc for hol_entry_help for more info, as most + of the fields are copied from its arguments. */ +struct pentry_state +{ + const struct hol_entry *entry; + argp_fmtstream_t stream; + struct hol_help_state *hhstate; + + /* True if nothing's been printed so far. */ + int first; + + /* If non-zero, the state that was used to print this help. */ + const struct argp_state *state; +}; + +/* If a user doc filter should be applied to DOC, do so. */ +static const char * +filter_doc (const char *doc, int key, const struct argp *argp, + const struct argp_state *state) +{ + if (argp->help_filter) + /* We must apply a user filter to this output. */ + { + void *input = __argp_input (argp, state); + return (*argp->help_filter) (key, doc, input); + } + else + /* No filter. */ + return doc; +} + +/* Prints STR as a header line, with the margin lines set appropiately, and + notes the fact that groups should be separated with a blank line. ARGP is + the argp that should dictate any user doc filtering to take place. Note + that the previous wrap margin isn't restored, but the left margin is reset + to 0. */ +static void +print_header (const char *str, const struct argp *argp, + struct pentry_state *pest) +{ + const char *tstr = dgettext (argp->argp_domain, str); + const char *fstr = filter_doc (tstr, ARGP_KEY_HELP_HEADER, argp, pest->state); + + if (fstr) + { + if (*fstr) + { + if (pest->hhstate->prev_entry) + /* Precede with a blank line. */ + __argp_fmtstream_putc (pest->stream, '\n'); + indent_to (pest->stream, uparams.header_col); + __argp_fmtstream_set_lmargin (pest->stream, uparams.header_col); + __argp_fmtstream_set_wmargin (pest->stream, uparams.header_col); + __argp_fmtstream_puts (pest->stream, fstr); + __argp_fmtstream_set_lmargin (pest->stream, 0); + __argp_fmtstream_putc (pest->stream, '\n'); + } + + pest->hhstate->sep_groups = 1; /* Separate subsequent groups. */ + } + + if (fstr != tstr) + free ((char *) fstr); +} + +/* Inserts a comma if this isn't the first item on the line, and then makes + sure we're at least to column COL. If this *is* the first item on a line, + prints any pending whitespace/headers that should precede this line. Also + clears FIRST. */ +static void +comma (unsigned col, struct pentry_state *pest) +{ + if (pest->first) + { + const struct hol_entry *pe = pest->hhstate->prev_entry; + const struct hol_cluster *cl = pest->entry->cluster; + + if (pest->hhstate->sep_groups && pe && pest->entry->group != pe->group) + __argp_fmtstream_putc (pest->stream, '\n'); + + if (cl && cl->header && *cl->header + && (!pe + || (pe->cluster != cl + && !hol_cluster_is_child (pe->cluster, cl)))) + /* If we're changing clusters, then this must be the start of the + ENTRY's cluster unless that is an ancestor of the previous one + (in which case we had just popped into a sub-cluster for a bit). + If so, then print the cluster's header line. */ + { + int old_wm = __argp_fmtstream_wmargin (pest->stream); + print_header (cl->header, cl->argp, pest); + __argp_fmtstream_set_wmargin (pest->stream, old_wm); + } + + pest->first = 0; + } + else + __argp_fmtstream_puts (pest->stream, ", "); + + indent_to (pest->stream, col); +} + +/* Print help for ENTRY to STREAM. */ +static void +hol_entry_help (struct hol_entry *entry, const struct argp_state *state, + argp_fmtstream_t stream, struct hol_help_state *hhstate) +{ + unsigned num; + const struct argp_option *real = entry->opt, *opt; + char *so = entry->short_options; + int have_long_opt = 0; /* We have any long options. */ + /* Saved margins. */ + int old_lm = __argp_fmtstream_set_lmargin (stream, 0); + int old_wm = __argp_fmtstream_wmargin (stream); + /* PEST is a state block holding some of our variables that we'd like to + share with helper functions. */ + struct pentry_state pest; + + pest.entry = entry; + pest.stream = stream; + pest.hhstate = hhstate; + pest.first = 1; + pest.state = state; + + if (! odoc (real)) + for (opt = real, num = entry->num; num > 0; opt++, num--) + if (opt->name && ovisible (opt)) + { + have_long_opt = 1; + break; + } + + /* First emit short options. */ + __argp_fmtstream_set_wmargin (stream, uparams.short_opt_col); /* For truly bizarre cases. */ + for (opt = real, num = entry->num; num > 0; opt++, num--) + if (oshort (opt) && opt->key == *so) + /* OPT has a valid (non shadowed) short option. */ + { + if (ovisible (opt)) + { + comma (uparams.short_opt_col, &pest); + __argp_fmtstream_putc (stream, '-'); + __argp_fmtstream_putc (stream, *so); + if (!have_long_opt || uparams.dup_args) + arg (real, " %s", "[%s]", state->root_argp->argp_domain, stream); + else if (real->arg) + hhstate->suppressed_dup_arg = 1; + } + so++; + } + + /* Now, long options. */ + if (odoc (real)) + /* A `documentation' option. */ + { + __argp_fmtstream_set_wmargin (stream, uparams.doc_opt_col); + for (opt = real, num = entry->num; num > 0; opt++, num--) + if (opt->name && *opt->name && ovisible (opt)) + { + comma (uparams.doc_opt_col, &pest); + /* Calling dgettext here isn't quite right, since sorting will + have been done on the original; but documentation options + should be pretty rare anyway... */ + __argp_fmtstream_puts (stream, + onotrans (opt) ? + opt->name : + dgettext (state->root_argp->argp_domain, + opt->name)); + } + } + else + /* A real long option. */ + { + int first_long_opt = 1; + + __argp_fmtstream_set_wmargin (stream, uparams.long_opt_col); + for (opt = real, num = entry->num; num > 0; opt++, num--) + if (opt->name && ovisible (opt)) + { + comma (uparams.long_opt_col, &pest); + __argp_fmtstream_printf (stream, "--%s", opt->name); + if (first_long_opt || uparams.dup_args) + arg (real, "=%s", "[=%s]", state->root_argp->argp_domain, + stream); + else if (real->arg) + hhstate->suppressed_dup_arg = 1; + } + } + + /* Next, documentation strings. */ + __argp_fmtstream_set_lmargin (stream, 0); + + if (pest.first) + { + /* Didn't print any switches, what's up? */ + if (!oshort (real) && !real->name) + /* This is a group header, print it nicely. */ + print_header (real->doc, entry->argp, &pest); + else + /* Just a totally shadowed option or null header; print nothing. */ + goto cleanup; /* Just return, after cleaning up. */ + } + else + { + const char *tstr = real->doc ? dgettext (state->root_argp->argp_domain, + real->doc) : 0; + const char *fstr = filter_doc (tstr, real->key, entry->argp, state); + if (fstr && *fstr) + { + unsigned int col = __argp_fmtstream_point (stream); + + __argp_fmtstream_set_lmargin (stream, uparams.opt_doc_col); + __argp_fmtstream_set_wmargin (stream, uparams.opt_doc_col); + + if (col > (unsigned int) (uparams.opt_doc_col + 3)) + __argp_fmtstream_putc (stream, '\n'); + else if (col >= (unsigned int) uparams.opt_doc_col) + __argp_fmtstream_puts (stream, " "); + else + indent_to (stream, uparams.opt_doc_col); + + __argp_fmtstream_puts (stream, fstr); + } + if (fstr && fstr != tstr) + free ((char *) fstr); + + /* Reset the left margin. */ + __argp_fmtstream_set_lmargin (stream, 0); + __argp_fmtstream_putc (stream, '\n'); + } + + hhstate->prev_entry = entry; + +cleanup: + __argp_fmtstream_set_lmargin (stream, old_lm); + __argp_fmtstream_set_wmargin (stream, old_wm); +} + +/* Output a long help message about the options in HOL to STREAM. */ +static void +hol_help (struct hol *hol, const struct argp_state *state, + argp_fmtstream_t stream) +{ + unsigned num; + struct hol_entry *entry; + struct hol_help_state hhstate = { 0, 0, 0 }; + + for (entry = hol->entries, num = hol->num_entries; num > 0; entry++, num--) + hol_entry_help (entry, state, stream, &hhstate); + + if (hhstate.suppressed_dup_arg && uparams.dup_args_note) + { + const char *tstr = dgettext (state->root_argp->argp_domain, "\ +Mandatory or optional arguments to long options are also mandatory or \ +optional for any corresponding short options."); + const char *fstr = filter_doc (tstr, ARGP_KEY_HELP_DUP_ARGS_NOTE, + state ? state->root_argp : 0, state); + if (fstr && *fstr) + { + __argp_fmtstream_putc (stream, '\n'); + __argp_fmtstream_puts (stream, fstr); + __argp_fmtstream_putc (stream, '\n'); + } + if (fstr && fstr != tstr) + free ((char *) fstr); + } +} + +/* Helper functions for hol_usage. */ + +/* If OPT is a short option without an arg, append its key to the string + pointer pointer to by COOKIE, and advance the pointer. */ +static int +add_argless_short_opt (const struct argp_option *opt, + const struct argp_option *real, + const char *domain, void *cookie) +{ + char **snao_end = cookie; + if (!(opt->arg || real->arg) + && !((opt->flags | real->flags) & OPTION_NO_USAGE)) + *(*snao_end)++ = opt->key; + return 0; +} + +/* If OPT is a short option with an arg, output a usage entry for it to the + stream pointed at by COOKIE. */ +static int +usage_argful_short_opt (const struct argp_option *opt, + const struct argp_option *real, + const char *domain, void *cookie) +{ + argp_fmtstream_t stream = cookie; + const char *arg = opt->arg; + int flags = opt->flags | real->flags; + + if (! arg) + arg = real->arg; + + if (arg && !(flags & OPTION_NO_USAGE)) + { + arg = dgettext (domain, arg); + + if (flags & OPTION_ARG_OPTIONAL) + __argp_fmtstream_printf (stream, " [-%c[%s]]", opt->key, arg); + else + { + /* Manually do line wrapping so that it (probably) won't + get wrapped at the embedded space. */ + space (stream, 6 + strlen (arg)); + __argp_fmtstream_printf (stream, "[-%c %s]", opt->key, arg); + } + } + + return 0; +} + +/* Output a usage entry for the long option opt to the stream pointed at by + COOKIE. */ +static int +usage_long_opt (const struct argp_option *opt, + const struct argp_option *real, + const char *domain, void *cookie) +{ + argp_fmtstream_t stream = cookie; + const char *arg = opt->arg; + int flags = opt->flags | real->flags; + + if (! arg) + arg = real->arg; + + if (! (flags & OPTION_NO_USAGE) && !odoc (opt)) + { + if (arg) + { + arg = dgettext (domain, arg); + if (flags & OPTION_ARG_OPTIONAL) + __argp_fmtstream_printf (stream, " [--%s[=%s]]", opt->name, arg); + else + __argp_fmtstream_printf (stream, " [--%s=%s]", opt->name, arg); + } + else + __argp_fmtstream_printf (stream, " [--%s]", opt->name); + } + + return 0; +} + +/* Print a short usage description for the arguments in HOL to STREAM. */ +static void +hol_usage (struct hol *hol, argp_fmtstream_t stream) +{ + if (hol->num_entries > 0) + { + unsigned nentries; + struct hol_entry *entry; + char *short_no_arg_opts = alloca (strlen (hol->short_options) + 1); + char *snao_end = short_no_arg_opts; + + /* First we put a list of short options without arguments. */ + for (entry = hol->entries, nentries = hol->num_entries + ; nentries > 0 + ; entry++, nentries--) + hol_entry_short_iterate (entry, add_argless_short_opt, + entry->argp->argp_domain, &snao_end); + if (snao_end > short_no_arg_opts) + { + *snao_end++ = 0; + __argp_fmtstream_printf (stream, " [-%s]", short_no_arg_opts); + } + + /* Now a list of short options *with* arguments. */ + for (entry = hol->entries, nentries = hol->num_entries + ; nentries > 0 + ; entry++, nentries--) + hol_entry_short_iterate (entry, usage_argful_short_opt, + entry->argp->argp_domain, stream); + + /* Finally, a list of long options (whew!). */ + for (entry = hol->entries, nentries = hol->num_entries + ; nentries > 0 + ; entry++, nentries--) + hol_entry_long_iterate (entry, usage_long_opt, + entry->argp->argp_domain, stream); + } +} + +/* Make a HOL containing all levels of options in ARGP. CLUSTER is the + cluster in which ARGP's entries should be clustered, or 0. */ +static struct hol * +argp_hol (const struct argp *argp, struct hol_cluster *cluster) +{ + const struct argp_child *child = argp->children; + struct hol *hol = make_hol (argp, cluster); + if (child) + while (child->argp) + { + struct hol_cluster *child_cluster = + ((child->group || child->header) + /* Put CHILD->argp within its own cluster. */ + ? hol_add_cluster (hol, child->group, child->header, + child - argp->children, cluster, argp) + /* Just merge it into the parent's cluster. */ + : cluster); + hol_append (hol, argp_hol (child->argp, child_cluster)) ; + child++; + } + return hol; +} + +/* Calculate how many different levels with alternative args strings exist in + ARGP. */ +static size_t +argp_args_levels (const struct argp *argp) +{ + size_t levels = 0; + const struct argp_child *child = argp->children; + + if (argp->args_doc && strchr (argp->args_doc, '\n')) + levels++; + + if (child) + while (child->argp) + levels += argp_args_levels ((child++)->argp); + + return levels; +} + +/* Print all the non-option args documented in ARGP to STREAM. Any output is + preceded by a space. LEVELS is a pointer to a byte vector the length + returned by argp_args_levels; it should be initialized to zero, and + updated by this routine for the next call if ADVANCE is true. True is + returned as long as there are more patterns to output. */ +static int +argp_args_usage (const struct argp *argp, const struct argp_state *state, + char **levels, int advance, argp_fmtstream_t stream) +{ + char *our_level = *levels; + int multiple = 0; + const struct argp_child *child = argp->children; + const char *tdoc = dgettext (argp->argp_domain, argp->args_doc), *nl = 0; + const char *fdoc = filter_doc (tdoc, ARGP_KEY_HELP_ARGS_DOC, argp, state); + + if (fdoc) + { + const char *cp = fdoc; + nl = __strchrnul (cp, '\n'); + if (*nl != '\0') + /* This is a `multi-level' args doc; advance to the correct position + as determined by our state in LEVELS, and update LEVELS. */ + { + int i; + multiple = 1; + for (i = 0; i < *our_level; i++) + cp = nl + 1, nl = __strchrnul (cp, '\n'); + (*levels)++; + } + + /* Manually do line wrapping so that it (probably) won't get wrapped at + any embedded spaces. */ + space (stream, 1 + nl - cp); + + __argp_fmtstream_write (stream, cp, nl - cp); + } + if (fdoc && fdoc != tdoc) + free ((char *)fdoc); /* Free user's modified doc string. */ + + if (child) + while (child->argp) + advance = !argp_args_usage ((child++)->argp, state, levels, advance, stream); + + if (advance && multiple) + { + /* Need to increment our level. */ + if (*nl) + /* There's more we can do here. */ + { + (*our_level)++; + advance = 0; /* Our parent shouldn't advance also. */ + } + else if (*our_level > 0) + /* We had multiple levels, but used them up; reset to zero. */ + *our_level = 0; + } + + return !advance; +} + +/* Print the documentation for ARGP to STREAM; if POST is false, then + everything preceeding a `\v' character in the documentation strings (or + the whole string, for those with none) is printed, otherwise, everything + following the `\v' character (nothing for strings without). Each separate + bit of documentation is separated a blank line, and if PRE_BLANK is true, + then the first is as well. If FIRST_ONLY is true, only the first + occurrence is output. Returns true if anything was output. */ +static int +argp_doc (const struct argp *argp, const struct argp_state *state, + int post, int pre_blank, int first_only, + argp_fmtstream_t stream) +{ + const char *text; + const char *inp_text; + size_t inp_text_len = 0; + const char *trans_text; + void *input = 0; + int anything = 0; + const struct argp_child *child = argp->children; + + if (argp->doc) + { + char *vt = strchr (argp->doc, '\v'); + if (vt) + { + if (post) + inp_text = vt + 1; + else + { + inp_text_len = vt - argp->doc; + inp_text = __strndup (argp->doc, inp_text_len); + } + } + else + inp_text = post ? 0 : argp->doc; + trans_text = inp_text ? dgettext (argp->argp_domain, inp_text) : NULL; + } + else + trans_text = inp_text = 0; + + if (argp->help_filter) + /* We have to filter the doc strings. */ + { + input = __argp_input (argp, state); + text = + (*argp->help_filter) (post + ? ARGP_KEY_HELP_POST_DOC + : ARGP_KEY_HELP_PRE_DOC, + trans_text, input); + } + else + text = (const char *) trans_text; + + if (text) + { + if (pre_blank) + __argp_fmtstream_putc (stream, '\n'); + + __argp_fmtstream_puts (stream, text); + + if (__argp_fmtstream_point (stream) > __argp_fmtstream_lmargin (stream)) + __argp_fmtstream_putc (stream, '\n'); + + anything = 1; + } + + if (text && text != trans_text) + free ((char *) text); /* Free TEXT returned from the help filter. */ + + if (inp_text && inp_text_len) + free ((char *) inp_text); /* We copied INP_TEXT, so free it now. */ + + if (post && argp->help_filter) + /* Now see if we have to output a ARGP_KEY_HELP_EXTRA text. */ + { + text = (*argp->help_filter) (ARGP_KEY_HELP_EXTRA, 0, input); + if (text) + { + if (anything || pre_blank) + __argp_fmtstream_putc (stream, '\n'); + __argp_fmtstream_puts (stream, text); + free ((char *) text); + if (__argp_fmtstream_point (stream) + > __argp_fmtstream_lmargin (stream)) + __argp_fmtstream_putc (stream, '\n'); + anything = 1; + } + } + + if (child) + while (child->argp && !(first_only && anything)) + anything |= + argp_doc ((child++)->argp, state, + post, anything || pre_blank, first_only, + stream); + + return anything; +} + +/* Output a usage message for ARGP to STREAM. If called from + argp_state_help, STATE is the relevent parsing state. FLAGS are from the + set ARGP_HELP_*. NAME is what to use wherever a `program name' is + needed. */ +static void +_help (const struct argp *argp, const struct argp_state *state, FILE *stream, + unsigned flags, char *name) +{ + int anything = 0; /* Whether we've output anything. */ + struct hol *hol = 0; + argp_fmtstream_t fs; + + if (! stream) + return; + +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __flockfile (stream); +#endif + + if (! uparams.valid) + fill_in_uparams (state); + + fs = __argp_make_fmtstream (stream, 0, uparams.rmargin, 0); + if (! fs) + { +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __funlockfile (stream); +#endif + return; + } + + if (flags & (ARGP_HELP_USAGE | ARGP_HELP_SHORT_USAGE | ARGP_HELP_LONG)) + { + hol = argp_hol (argp, 0); + + /* If present, these options always come last. */ + hol_set_group (hol, "help", -1); + hol_set_group (hol, "version", -1); + + hol_sort (hol); + } + + if (flags & (ARGP_HELP_USAGE | ARGP_HELP_SHORT_USAGE)) + /* Print a short `Usage:' message. */ + { + int first_pattern = 1, more_patterns; + size_t num_pattern_levels = argp_args_levels (argp); + char *pattern_levels = alloca (num_pattern_levels); + + memset (pattern_levels, 0, num_pattern_levels); + + do + { + int old_lm; + int old_wm = __argp_fmtstream_set_wmargin (fs, uparams.usage_indent); + char *levels = pattern_levels; + + if (first_pattern) + __argp_fmtstream_printf (fs, "%s %s", + dgettext (argp->argp_domain, "Usage:"), + name); + else + __argp_fmtstream_printf (fs, "%s %s", + dgettext (argp->argp_domain, " or: "), + name); + + /* We set the lmargin as well as the wmargin, because hol_usage + manually wraps options with newline to avoid annoying breaks. */ + old_lm = __argp_fmtstream_set_lmargin (fs, uparams.usage_indent); + + if (flags & ARGP_HELP_SHORT_USAGE) + /* Just show where the options go. */ + { + if (hol->num_entries > 0) + __argp_fmtstream_puts (fs, dgettext (argp->argp_domain, + " [OPTION...]")); + } + else + /* Actually print the options. */ + { + hol_usage (hol, fs); + flags |= ARGP_HELP_SHORT_USAGE; /* But only do so once. */ + } + + more_patterns = argp_args_usage (argp, state, &levels, 1, fs); + + __argp_fmtstream_set_wmargin (fs, old_wm); + __argp_fmtstream_set_lmargin (fs, old_lm); + + __argp_fmtstream_putc (fs, '\n'); + anything = 1; + + first_pattern = 0; + } + while (more_patterns); + } + + if (flags & ARGP_HELP_PRE_DOC) + anything |= argp_doc (argp, state, 0, 0, 1, fs); + + if (flags & ARGP_HELP_SEE) + { + __argp_fmtstream_printf (fs, dgettext (argp->argp_domain, "\ +Try `%s --help' or `%s --usage' for more information.\n"), + name, name); + anything = 1; + } + + if (flags & ARGP_HELP_LONG) + /* Print a long, detailed help message. */ + { + /* Print info about all the options. */ + if (hol->num_entries > 0) + { + if (anything) + __argp_fmtstream_putc (fs, '\n'); + hol_help (hol, state, fs); + anything = 1; + } + } + + if (flags & ARGP_HELP_POST_DOC) + /* Print any documentation strings at the end. */ + anything |= argp_doc (argp, state, 1, anything, 0, fs); + + if ((flags & ARGP_HELP_BUG_ADDR) && argp_program_bug_address) + { + if (anything) + __argp_fmtstream_putc (fs, '\n'); + __argp_fmtstream_printf (fs, dgettext (argp->argp_domain, + "Report bugs to %s.\n"), + argp_program_bug_address); + anything = 1; + } + +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __funlockfile (stream); +#endif + + if (hol) + hol_free (hol); + + __argp_fmtstream_free (fs); +} + +/* Output a usage message for ARGP to STREAM. FLAGS are from the set + ARGP_HELP_*. NAME is what to use wherever a `program name' is needed. */ +void __argp_help (const struct argp *argp, FILE *stream, + unsigned flags, char *name) +{ + struct argp_state state; + memset (&state, 0, sizeof state); + state.root_argp = argp; + _help (argp, &state, stream, flags, name); +} +#ifdef weak_alias +weak_alias (__argp_help, argp_help) +#endif + +#if ! (defined _LIBC || HAVE_DECL_PROGRAM_INVOCATION_SHORT_NAME) +char * +__argp_short_program_name (void) +{ +# if HAVE_DECL_PROGRAM_INVOCATION_NAME + return __argp_base_name (program_invocation_name); +# else + /* FIXME: What now? Miles suggests that it is better to use NULL, + but currently the value is passed on directly to fputs_unlocked, + so that requires more changes. */ +# if __GNUC__ +# warning No reasonable value to return +# endif /* __GNUC__ */ + return ""; +# endif +} +#endif + +/* Output, if appropriate, a usage message for STATE to STREAM. FLAGS are + from the set ARGP_HELP_*. */ +void +__argp_state_help (const struct argp_state *state, FILE *stream, unsigned flags) +{ + if ((!state || ! (state->flags & ARGP_NO_ERRS)) && stream) + { + if (state && (state->flags & ARGP_LONG_ONLY)) + flags |= ARGP_HELP_LONG_ONLY; + + _help (state ? state->root_argp : 0, state, stream, flags, + state ? state->name : __argp_short_program_name ()); + + if (!state || ! (state->flags & ARGP_NO_EXIT)) + { + if (flags & ARGP_HELP_EXIT_ERR) + exit (argp_err_exit_status); + if (flags & ARGP_HELP_EXIT_OK) + exit (0); + } + } +} +#ifdef weak_alias +weak_alias (__argp_state_help, argp_state_help) +#endif + +/* If appropriate, print the printf string FMT and following args, preceded + by the program name and `:', to stderr, and followed by a `Try ... --help' + message, then exit (1). */ +void +__argp_error (const struct argp_state *state, const char *fmt, ...) +{ + if (!state || !(state->flags & ARGP_NO_ERRS)) + { + FILE *stream = state ? state->err_stream : stderr; + + if (stream) + { + va_list ap; + +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __flockfile (stream); +#endif + + va_start (ap, fmt); + +#ifdef USE_IN_LIBIO + if (_IO_fwide (stream, 0) > 0) + { + char *buf; + + if (__asprintf (&buf, fmt, ap) < 0) + buf = NULL; + + __fwprintf (stream, L"%s: %s\n", + state ? state->name : __argp_short_program_name (), + buf); + + free (buf); + } + else +#endif + { + fputs_unlocked (state + ? state->name : __argp_short_program_name (), + stream); + putc_unlocked (':', stream); + putc_unlocked (' ', stream); + + vfprintf (stream, fmt, ap); + + putc_unlocked ('\n', stream); + } + + __argp_state_help (state, stream, ARGP_HELP_STD_ERR); + + va_end (ap); + +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __funlockfile (stream); +#endif + } + } +} +#ifdef weak_alias +weak_alias (__argp_error, argp_error) +#endif + +/* Similar to the standard gnu error-reporting function error(), but will + respect the ARGP_NO_EXIT and ARGP_NO_ERRS flags in STATE, and will print + to STATE->err_stream. This is useful for argument parsing code that is + shared between program startup (when exiting is desired) and runtime + option parsing (when typically an error code is returned instead). The + difference between this function and argp_error is that the latter is for + *parsing errors*, and the former is for other problems that occur during + parsing but don't reflect a (syntactic) problem with the input. */ +void +__argp_failure (const struct argp_state *state, int status, int errnum, + const char *fmt, ...) +{ + if (!state || !(state->flags & ARGP_NO_ERRS)) + { + FILE *stream = state ? state->err_stream : stderr; + + if (stream) + { +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __flockfile (stream); +#endif + +#ifdef USE_IN_LIBIO + if (_IO_fwide (stream, 0) > 0) + __fwprintf (stream, L"%s", + state ? state->name : __argp_short_program_name ()); + else +#endif + fputs_unlocked (state + ? state->name : __argp_short_program_name (), + stream); + + if (fmt) + { + va_list ap; + + va_start (ap, fmt); +#ifdef USE_IN_LIBIO + if (_IO_fwide (stream, 0) > 0) + { + char *buf; + + if (__asprintf (&buf, fmt, ap) < 0) + buf = NULL; + + __fwprintf (stream, L": %s", buf); + + free (buf); + } + else +#endif + { + putc_unlocked (':', stream); + putc_unlocked (' ', stream); + + vfprintf (stream, fmt, ap); + } + + va_end (ap); + } + + if (errnum) + { + char buf[200]; + +#ifdef USE_IN_LIBIO + if (_IO_fwide (stream, 0) > 0) + __fwprintf (stream, L": %s", + __strerror_r (errnum, buf, sizeof (buf))); + else +#endif + { + char const *s = NULL; + putc_unlocked (':', stream); + putc_unlocked (' ', stream); +#if _LIBC || (HAVE_DECL_STRERROR_R && STRERROR_R_CHAR_P) + s = __strerror_r (errnum, buf, sizeof buf); +#elif HAVE_DECL_STRERROR_R + if (__strerror_r (errnum, buf, sizeof buf) == 0) + s = buf; +#endif +#if !_LIBC + if (! s && ! (s = strerror (errnum))) + s = dgettext (state->root_argp->argp_domain, + "Unknown system error"); +#endif + fputs (s, stream); + } + } + +#ifdef USE_IN_LIBIO + if (_IO_fwide (stream, 0) > 0) + putwc_unlocked (L'\n', stream); + else +#endif + putc_unlocked ('\n', stream); + +#if _LIBC || (HAVE_FLOCKFILE && HAVE_FUNLOCKFILE) + __funlockfile (stream); +#endif + + if (status && (!state || !(state->flags & ARGP_NO_EXIT))) + exit (status); + } + } +} +#ifdef weak_alias +weak_alias (__argp_failure, argp_failure) +#endif diff --git a/lib/argp-namefrob.h b/lib/argp-namefrob.h new file mode 100644 index 0000000..6fe99cd --- /dev/null +++ b/lib/argp-namefrob.h @@ -0,0 +1,158 @@ +/* Name frobnication for compiling argp outside of glibc + Copyright (C) 1997, 2003, 2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if !_LIBC +/* This code is written for inclusion in gnu-libc, and uses names in the + namespace reserved for libc. If we're not compiling in libc, define those + names to be the normal ones instead. */ + +/* argp-parse functions */ +#undef __argp_parse +#define __argp_parse argp_parse +#undef __option_is_end +#define __option_is_end _option_is_end +#undef __option_is_short +#define __option_is_short _option_is_short +#undef __argp_input +#define __argp_input _argp_input + +/* argp-help functions */ +#undef __argp_help +#define __argp_help argp_help +#undef __argp_error +#define __argp_error argp_error +#undef __argp_failure +#define __argp_failure argp_failure +#undef __argp_state_help +#define __argp_state_help argp_state_help +#undef __argp_usage +#define __argp_usage argp_usage + +/* argp-fmtstream functions */ +#undef __argp_make_fmtstream +#define __argp_make_fmtstream argp_make_fmtstream +#undef __argp_fmtstream_free +#define __argp_fmtstream_free argp_fmtstream_free +#undef __argp_fmtstream_putc +#define __argp_fmtstream_putc argp_fmtstream_putc +#undef __argp_fmtstream_puts +#define __argp_fmtstream_puts argp_fmtstream_puts +#undef __argp_fmtstream_write +#define __argp_fmtstream_write argp_fmtstream_write +#undef __argp_fmtstream_printf +#define __argp_fmtstream_printf argp_fmtstream_printf +#undef __argp_fmtstream_set_lmargin +#define __argp_fmtstream_set_lmargin argp_fmtstream_set_lmargin +#undef __argp_fmtstream_set_rmargin +#define __argp_fmtstream_set_rmargin argp_fmtstream_set_rmargin +#undef __argp_fmtstream_set_wmargin +#define __argp_fmtstream_set_wmargin argp_fmtstream_set_wmargin +#undef __argp_fmtstream_point +#define __argp_fmtstream_point argp_fmtstream_point +#undef __argp_fmtstream_update +#define __argp_fmtstream_update _argp_fmtstream_update +#undef __argp_fmtstream_ensure +#define __argp_fmtstream_ensure _argp_fmtstream_ensure +#undef __argp_fmtstream_lmargin +#define __argp_fmtstream_lmargin argp_fmtstream_lmargin +#undef __argp_fmtstream_rmargin +#define __argp_fmtstream_rmargin argp_fmtstream_rmargin +#undef __argp_fmtstream_wmargin +#define __argp_fmtstream_wmargin argp_fmtstream_wmargin + +/* normal libc functions we call */ +#undef __flockfile +#define __flockfile flockfile +#undef __funlockfile +#define __funlockfile funlockfile +#undef __mempcpy +#define __mempcpy mempcpy +#undef __sleep +#define __sleep sleep +#undef __strcasecmp +#define __strcasecmp strcasecmp +#undef __strchrnul +#define __strchrnul strchrnul +#undef __strerror_r +#define __strerror_r strerror_r +#undef __strndup +#define __strndup strndup +#undef __vsnprintf +#define __vsnprintf vsnprintf + +#if defined(HAVE_DECL_CLEARERR_UNLOCKED) && !HAVE_DECL_CLEARERR_UNLOCKED +# define clearerr_unlocked(x) clearerr (x) +#endif +#if defined(HAVE_DECL_FEOF_UNLOCKED) && !HAVE_DECL_FEOF_UNLOCKED +# define feof_unlocked(x) feof (x) +# endif +#if defined(HAVE_DECL_FERROR_UNLOCKED) && !HAVE_DECL_FERROR_UNLOCKED +# define ferror_unlocked(x) ferror (x) +# endif +#if defined(HAVE_DECL_FFLUSH_UNLOCKED) && !HAVE_DECL_FFLUSH_UNLOCKED +# define fflush_unlocked(x) fflush (x) +# endif +#if defined(HAVE_DECL_FGETS_UNLOCKED) && !HAVE_DECL_FGETS_UNLOCKED +# define fgets_unlocked(x,y,z) fgets (x,y,z) +# endif +#if defined(HAVE_DECL_FPUTC_UNLOCKED) && !HAVE_DECL_FPUTC_UNLOCKED +# define fputc_unlocked(x,y) fputc (x,y) +# endif +#if defined(HAVE_DECL_FPUTS_UNLOCKED) && !HAVE_DECL_FPUTS_UNLOCKED +# define fputs_unlocked(x,y) fputs (x,y) +# endif +#if defined(HAVE_DECL_FREAD_UNLOCKED) && !HAVE_DECL_FREAD_UNLOCKED +# define fread_unlocked(w,x,y,z) fread (w,x,y,z) +# endif +#if defined(HAVE_DECL_FWRITE_UNLOCKED) && !HAVE_DECL_FWRITE_UNLOCKED +# define fwrite_unlocked(w,x,y,z) fwrite (w,x,y,z) +# endif +#if defined(HAVE_DECL_GETC_UNLOCKED) && !HAVE_DECL_GETC_UNLOCKED +# define getc_unlocked(x) getc (x) +# endif +#if defined(HAVE_DECL_GETCHAR_UNLOCKED) && !HAVE_DECL_GETCHAR_UNLOCKED +# define getchar_unlocked() getchar () +# endif +#if defined(HAVE_DECL_PUTC_UNLOCKED) && !HAVE_DECL_PUTC_UNLOCKED +# define putc_unlocked(x,y) putc (x,y) +# endif +#if defined(HAVE_DECL_PUTCHAR_UNLOCKED) && !HAVE_DECL_PUTCHAR_UNLOCKED +# define putchar_unlocked(x) putchar (x) +# endif + +#endif /* !_LIBC */ + +#ifndef __set_errno +#define __set_errno(e) (errno = (e)) +#endif + +#if defined GNULIB_ARGP_DISABLE_DIRNAME +# define __argp_base_name(arg) arg +#elif defined GNULIB_ARGP_EXTERN_BASENAME +extern char *__argp_base_name(const char *arg); +#else +# include "dirname.h" +# define __argp_base_name base_name +#endif + +#if defined _LIBC || HAVE_DECL_PROGRAM_INVOCATION_SHORT_NAME +# define __argp_short_program_name() (program_invocation_short_name) +#else +extern char *__argp_short_program_name (void); +#endif diff --git a/lib/argp-parse.c b/lib/argp-parse.c new file mode 100644 index 0000000..a7de729 --- /dev/null +++ b/lib/argp-parse.c @@ -0,0 +1,953 @@ +/* Hierarchial argument parsing, layered over getopt + Copyright (C) 1995-2000, 2002, 2003, 2004 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include <alloca.h> +#include <stddef.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#include <limits.h> +#include <getopt.h> +#include <getopt_int.h> + +#ifdef _LIBC +# include <libintl.h> +# undef dgettext +# define dgettext(domain, msgid) \ + INTUSE(__dcgettext) (domain, msgid, LC_MESSAGES) +#else +# include "gettext.h" +#endif +#define N_(msgid) msgid + +#include "argp.h" +#include "argp-namefrob.h" + +#define alignof(type) offsetof (struct { char c; type x; }, x) +#define alignto(n, d) ((((n) + (d) - 1) / (d)) * (d)) + +/* Getopt return values. */ +#define KEY_END (-1) /* The end of the options. */ +#define KEY_ARG 1 /* A non-option argument. */ +#define KEY_ERR '?' /* An error parsing the options. */ + +/* The meta-argument used to prevent any further arguments being interpreted + as options. */ +#define QUOTE "--" + +/* The number of bits we steal in a long-option value for our own use. */ +#define GROUP_BITS CHAR_BIT + +/* The number of bits available for the user value. */ +#define USER_BITS ((sizeof ((struct option *)0)->val * CHAR_BIT) - GROUP_BITS) +#define USER_MASK ((1 << USER_BITS) - 1) + +/* EZ alias for ARGP_ERR_UNKNOWN. */ +#define EBADKEY ARGP_ERR_UNKNOWN + +/* Default options. */ + +/* When argp is given the --HANG switch, _ARGP_HANG is set and argp will sleep + for one second intervals, decrementing _ARGP_HANG until it's zero. Thus + you can force the program to continue by attaching a debugger and setting + it to 0 yourself. */ +static volatile int _argp_hang; + +#define OPT_PROGNAME -2 +#define OPT_USAGE -3 +#define OPT_HANG -4 + +static const struct argp_option argp_default_options[] = +{ + {"help", '?', 0, 0, N_("give this help list"), -1}, + {"usage", OPT_USAGE, 0, 0, N_("give a short usage message"), 0}, + {"program-name",OPT_PROGNAME,N_("NAME"), OPTION_HIDDEN, N_("set the program name"), 0}, + {"HANG", OPT_HANG, N_("SECS"), OPTION_ARG_OPTIONAL | OPTION_HIDDEN, + N_("hang for SECS seconds (default 3600)"), 0}, + {NULL, 0, 0, 0, NULL, 0} +}; + +static error_t +argp_default_parser (int key, char *arg, struct argp_state *state) +{ + switch (key) + { + case '?': + __argp_state_help (state, state->out_stream, ARGP_HELP_STD_HELP); + break; + case OPT_USAGE: + __argp_state_help (state, state->out_stream, + ARGP_HELP_USAGE | ARGP_HELP_EXIT_OK); + break; + + case OPT_PROGNAME: /* Set the program name. */ +#if defined _LIBC || HAVE_DECL_PROGRAM_INVOCATION_NAME + program_invocation_name = arg; +#endif + /* [Note that some systems only have PROGRAM_INVOCATION_SHORT_NAME (aka + __PROGNAME), in which case, PROGRAM_INVOCATION_NAME is just defined + to be that, so we have to be a bit careful here.] */ + + /* Update what we use for messages. */ + state->name = __argp_base_name (arg); + +#if defined _LIBC || HAVE_DECL_PROGRAM_INVOCATION_SHORT_NAME + program_invocation_short_name = state->name; +#endif + + if ((state->flags & (ARGP_PARSE_ARGV0 | ARGP_NO_ERRS)) + == ARGP_PARSE_ARGV0) + /* Update what getopt uses too. */ + state->argv[0] = arg; + + break; + + case OPT_HANG: + _argp_hang = atoi (arg ? arg : "3600"); + while (_argp_hang-- > 0) + __sleep (1); + break; + + default: + return EBADKEY; + } + return 0; +} + +static const struct argp argp_default_argp = + {argp_default_options, &argp_default_parser, NULL, NULL, NULL, NULL, "libc"}; + + +static const struct argp_option argp_version_options[] = +{ + {"version", 'V', 0, 0, N_("print program version"), -1}, + {NULL, 0, 0, 0, NULL, 0} +}; + +static error_t +argp_version_parser (int key, char *arg, struct argp_state *state) +{ + switch (key) + { + case 'V': + if (argp_program_version_hook) + (*argp_program_version_hook) (state->out_stream, state); + else if (argp_program_version) + fprintf (state->out_stream, "%s\n", argp_program_version); + else + __argp_error (state, dgettext (state->root_argp->argp_domain, + "(PROGRAM ERROR) No version known!?")); + if (! (state->flags & ARGP_NO_EXIT)) + exit (0); + break; + default: + return EBADKEY; + } + return 0; +} + +static const struct argp argp_version_argp = + {argp_version_options, &argp_version_parser, NULL, NULL, NULL, NULL, "libc"}; + +/* Returns the offset into the getopt long options array LONG_OPTIONS of a + long option with called NAME, or -1 if none is found. Passing NULL as + NAME will return the number of options. */ +static int +find_long_option (struct option *long_options, const char *name) +{ + struct option *l = long_options; + while (l->name != NULL) + if (name != NULL && strcmp (l->name, name) == 0) + return l - long_options; + else + l++; + if (name == NULL) + return l - long_options; + else + return -1; +} + + +/* The state of a `group' during parsing. Each group corresponds to a + particular argp structure from the tree of such descending from the top + level argp passed to argp_parse. */ +struct group +{ + /* This group's parsing function. */ + argp_parser_t parser; + + /* Which argp this group is from. */ + const struct argp *argp; + + /* Points to the point in SHORT_OPTS corresponding to the end of the short + options for this group. We use it to determine from which group a + particular short options is from. */ + char *short_end; + + /* The number of non-option args sucessfully handled by this parser. */ + unsigned args_processed; + + /* This group's parser's parent's group. */ + struct group *parent; + unsigned parent_index; /* And the our position in the parent. */ + + /* These fields are swapped into and out of the state structure when + calling this group's parser. */ + void *input, **child_inputs; + void *hook; +}; + +/* Call GROUP's parser with KEY and ARG, swapping any group-specific info + from STATE before calling, and back into state afterwards. If GROUP has + no parser, EBADKEY is returned. */ +static error_t +group_parse (struct group *group, struct argp_state *state, int key, char *arg) +{ + if (group->parser) + { + error_t err; + state->hook = group->hook; + state->input = group->input; + state->child_inputs = group->child_inputs; + state->arg_num = group->args_processed; + err = (*group->parser)(key, arg, state); + group->hook = state->hook; + return err; + } + else + return EBADKEY; +} + +struct parser +{ + const struct argp *argp; + + /* SHORT_OPTS is the getopt short options string for the union of all the + groups of options. */ + char *short_opts; + /* LONG_OPTS is the array of getop long option structures for the union of + all the groups of options. */ + struct option *long_opts; + /* OPT_DATA is the getopt data used for the re-entrant getopt. */ + struct _getopt_data opt_data; + + /* States of the various parsing groups. */ + struct group *groups; + /* The end of the GROUPS array. */ + struct group *egroup; + /* An vector containing storage for the CHILD_INPUTS field in all groups. */ + void **child_inputs; + + /* True if we think using getopt is still useful; if false, then + remaining arguments are just passed verbatim with ARGP_KEY_ARG. This is + cleared whenever getopt returns KEY_END, but may be set again if the user + moves the next argument pointer backwards. */ + int try_getopt; + + /* State block supplied to parsing routines. */ + struct argp_state state; + + /* Memory used by this parser. */ + void *storage; +}; + +/* The next usable entries in the various parser tables being filled in by + convert_options. */ +struct parser_convert_state +{ + struct parser *parser; + char *short_end; + struct option *long_end; + void **child_inputs_end; +}; + +/* Converts all options in ARGP (which is put in GROUP) and ancestors + into getopt options stored in SHORT_OPTS and LONG_OPTS; SHORT_END and + CVT->LONG_END are the points at which new options are added. Returns the + next unused group entry. CVT holds state used during the conversion. */ +static struct group * +convert_options (const struct argp *argp, + struct group *parent, unsigned parent_index, + struct group *group, struct parser_convert_state *cvt) +{ + /* REAL is the most recent non-alias value of OPT. */ + const struct argp_option *real = argp->options; + const struct argp_child *children = argp->children; + + if (real || argp->parser) + { + const struct argp_option *opt; + + if (real) + for (opt = real; !__option_is_end (opt); opt++) + { + if (! (opt->flags & OPTION_ALIAS)) + /* OPT isn't an alias, so we can use values from it. */ + real = opt; + + if (! (real->flags & OPTION_DOC)) + /* A real option (not just documentation). */ + { + if (__option_is_short (opt)) + /* OPT can be used as a short option. */ + { + *cvt->short_end++ = opt->key; + if (real->arg) + { + *cvt->short_end++ = ':'; + if (real->flags & OPTION_ARG_OPTIONAL) + *cvt->short_end++ = ':'; + } + *cvt->short_end = '\0'; /* keep 0 terminated */ + } + + if (opt->name + && find_long_option (cvt->parser->long_opts, opt->name) < 0) + /* OPT can be used as a long option. */ + { + cvt->long_end->name = opt->name; + cvt->long_end->has_arg = + (real->arg + ? (real->flags & OPTION_ARG_OPTIONAL + ? optional_argument + : required_argument) + : no_argument); + cvt->long_end->flag = 0; + /* we add a disambiguating code to all the user's + values (which is removed before we actually call + the function to parse the value); this means that + the user loses use of the high 8 bits in all his + values (the sign of the lower bits is preserved + however)... */ + cvt->long_end->val = + ((opt->key | real->key) & USER_MASK) + + (((group - cvt->parser->groups) + 1) << USER_BITS); + + /* Keep the LONG_OPTS list terminated. */ + (++cvt->long_end)->name = NULL; + } + } + } + + group->parser = argp->parser; + group->argp = argp; + group->short_end = cvt->short_end; + group->args_processed = 0; + group->parent = parent; + group->parent_index = parent_index; + group->input = 0; + group->hook = 0; + group->child_inputs = 0; + + if (children) + /* Assign GROUP's CHILD_INPUTS field some space from + CVT->child_inputs_end.*/ + { + unsigned num_children = 0; + while (children[num_children].argp) + num_children++; + group->child_inputs = cvt->child_inputs_end; + cvt->child_inputs_end += num_children; + } + + parent = group++; + } + else + parent = 0; + + if (children) + { + unsigned index = 0; + while (children->argp) + group = + convert_options (children++->argp, parent, index++, group, cvt); + } + + return group; +} + +/* Find the merged set of getopt options, with keys appropiately prefixed. */ +static void +parser_convert (struct parser *parser, const struct argp *argp, int flags) +{ + struct parser_convert_state cvt; + + cvt.parser = parser; + cvt.short_end = parser->short_opts; + cvt.long_end = parser->long_opts; + cvt.child_inputs_end = parser->child_inputs; + + if (flags & ARGP_IN_ORDER) + *cvt.short_end++ = '-'; + else if (flags & ARGP_NO_ARGS) + *cvt.short_end++ = '+'; + *cvt.short_end = '\0'; + + cvt.long_end->name = NULL; + + parser->argp = argp; + + if (argp) + parser->egroup = convert_options (argp, 0, 0, parser->groups, &cvt); + else + parser->egroup = parser->groups; /* No parsers at all! */ +} + +/* Lengths of various parser fields which we will allocated. */ +struct parser_sizes +{ + size_t short_len; /* Getopt short options string. */ + size_t long_len; /* Getopt long options vector. */ + size_t num_groups; /* Group structures we allocate. */ + size_t num_child_inputs; /* Child input slots. */ +}; + +/* For ARGP, increments the NUM_GROUPS field in SZS by the total number of + argp structures descended from it, and the SHORT_LEN & LONG_LEN fields by + the maximum lengths of the resulting merged getopt short options string and + long-options array, respectively. */ +static void +calc_sizes (const struct argp *argp, struct parser_sizes *szs) +{ + const struct argp_child *child = argp->children; + const struct argp_option *opt = argp->options; + + if (opt || argp->parser) + { + szs->num_groups++; + if (opt) + { + int num_opts = 0; + while (!__option_is_end (opt++)) + num_opts++; + szs->short_len += num_opts * 3; /* opt + up to 2 `:'s */ + szs->long_len += num_opts; + } + } + + if (child) + while (child->argp) + { + calc_sizes ((child++)->argp, szs); + szs->num_child_inputs++; + } +} + +/* Initializes PARSER to parse ARGP in a manner described by FLAGS. */ +static error_t +parser_init (struct parser *parser, const struct argp *argp, + int argc, char **argv, int flags, void *input) +{ + error_t err = 0; + struct group *group; + struct parser_sizes szs; + struct _getopt_data opt_data = _GETOPT_DATA_INITIALIZER; + char *storage; + size_t glen, gsum; + size_t clen, csum; + size_t llen, lsum; + size_t slen, ssum; + + szs.short_len = (flags & ARGP_NO_ARGS) ? 0 : 1; + szs.long_len = 0; + szs.num_groups = 0; + szs.num_child_inputs = 0; + + if (argp) + calc_sizes (argp, &szs); + + /* Lengths of the various bits of storage used by PARSER. */ + glen = (szs.num_groups + 1) * sizeof (struct group); + clen = szs.num_child_inputs * sizeof (void *); + llen = (szs.long_len + 1) * sizeof (struct option); + slen = szs.short_len + 1; + + /* Sums of previous lengths, properly aligned. There's no need to + align gsum, since struct group is aligned at least as strictly as + void * (since it contains a void * member). And there's no need + to align lsum, since struct option is aligned at least as + strictly as char. */ + gsum = glen; + csum = alignto (gsum + clen, alignof (struct option)); + lsum = csum + llen; + ssum = lsum + slen; + + parser->storage = malloc (ssum); + if (! parser->storage) + return ENOMEM; + + storage = parser->storage; + parser->groups = parser->storage; + parser->child_inputs = (void **) (storage + gsum); + parser->long_opts = (struct option *) (storage + csum); + parser->short_opts = storage + lsum; + parser->opt_data = opt_data; + + memset (parser->child_inputs, 0, clen); + parser_convert (parser, argp, flags); + + memset (&parser->state, 0, sizeof (struct argp_state)); + parser->state.root_argp = parser->argp; + parser->state.argc = argc; + parser->state.argv = argv; + parser->state.flags = flags; + parser->state.err_stream = stderr; + parser->state.out_stream = stdout; + parser->state.next = 0; /* Tell getopt to initialize. */ + parser->state.pstate = parser; + + parser->try_getopt = 1; + + /* Call each parser for the first time, giving it a chance to propagate + values to child parsers. */ + if (parser->groups < parser->egroup) + parser->groups->input = input; + for (group = parser->groups; + group < parser->egroup && (!err || err == EBADKEY); + group++) + { + if (group->parent) + /* If a child parser, get the initial input value from the parent. */ + group->input = group->parent->child_inputs[group->parent_index]; + + if (!group->parser + && group->argp->children && group->argp->children->argp) + /* For the special case where no parsing function is supplied for an + argp, propagate its input to its first child, if any (this just + makes very simple wrapper argps more convenient). */ + group->child_inputs[0] = group->input; + + err = group_parse (group, &parser->state, ARGP_KEY_INIT, 0); + } + if (err == EBADKEY) + err = 0; /* Some parser didn't understand. */ + + if (err) + return err; + + if (parser->state.flags & ARGP_NO_ERRS) + { + parser->opt_data.opterr = 0; + if (parser->state.flags & ARGP_PARSE_ARGV0) + /* getopt always skips ARGV[0], so we have to fake it out. As long + as OPTERR is 0, then it shouldn't actually try to access it. */ + parser->state.argv--, parser->state.argc++; + } + else + parser->opt_data.opterr = 1; /* Print error messages. */ + + if (parser->state.argv == argv && argv[0]) + /* There's an argv[0]; use it for messages. */ + parser->state.name = __argp_base_name (argv[0]); + else + parser->state.name = __argp_short_program_name (); + + return 0; +} + +/* Free any storage consumed by PARSER (but not PARSER itself). */ +static error_t +parser_finalize (struct parser *parser, + error_t err, int arg_ebadkey, int *end_index) +{ + struct group *group; + + if (err == EBADKEY && arg_ebadkey) + /* Suppress errors generated by unparsed arguments. */ + err = 0; + + if (! err) + { + if (parser->state.next == parser->state.argc) + /* We successfully parsed all arguments! Call all the parsers again, + just a few more times... */ + { + for (group = parser->groups; + group < parser->egroup && (!err || err==EBADKEY); + group++) + if (group->args_processed == 0) + err = group_parse (group, &parser->state, ARGP_KEY_NO_ARGS, 0); + for (group = parser->egroup - 1; + group >= parser->groups && (!err || err==EBADKEY); + group--) + err = group_parse (group, &parser->state, ARGP_KEY_END, 0); + + if (err == EBADKEY) + err = 0; /* Some parser didn't understand. */ + + /* Tell the user that all arguments are parsed. */ + if (end_index) + *end_index = parser->state.next; + } + else if (end_index) + /* Return any remaining arguments to the user. */ + *end_index = parser->state.next; + else + /* No way to return the remaining arguments, they must be bogus. */ + { + if (!(parser->state.flags & ARGP_NO_ERRS) + && parser->state.err_stream) + fprintf (parser->state.err_stream, + dgettext (parser->argp->argp_domain, + "%s: Too many arguments\n"), + parser->state.name); + err = EBADKEY; + } + } + + /* Okay, we're all done, with either an error or success; call the parsers + to indicate which one. */ + + if (err) + { + /* Maybe print an error message. */ + if (err == EBADKEY) + /* An appropriate message describing what the error was should have + been printed earlier. */ + __argp_state_help (&parser->state, parser->state.err_stream, + ARGP_HELP_STD_ERR); + + /* Since we didn't exit, give each parser an error indication. */ + for (group = parser->groups; group < parser->egroup; group++) + group_parse (group, &parser->state, ARGP_KEY_ERROR, 0); + } + else + /* Notify parsers of success, and propagate back values from parsers. */ + { + /* We pass over the groups in reverse order so that child groups are + given a chance to do there processing before passing back a value to + the parent. */ + for (group = parser->egroup - 1 + ; group >= parser->groups && (!err || err == EBADKEY) + ; group--) + err = group_parse (group, &parser->state, ARGP_KEY_SUCCESS, 0); + if (err == EBADKEY) + err = 0; /* Some parser didn't understand. */ + } + + /* Call parsers once more, to do any final cleanup. Errors are ignored. */ + for (group = parser->egroup - 1; group >= parser->groups; group--) + group_parse (group, &parser->state, ARGP_KEY_FINI, 0); + + if (err == EBADKEY) + err = EINVAL; + + free (parser->storage); + + return err; +} + +/* Call the user parsers to parse the non-option argument VAL, at the current + position, returning any error. The state NEXT pointer is assumed to have + been adjusted (by getopt) to point after this argument; this function will + adjust it correctly to reflect however many args actually end up being + consumed. */ +static error_t +parser_parse_arg (struct parser *parser, char *val) +{ + /* Save the starting value of NEXT, first adjusting it so that the arg + we're parsing is again the front of the arg vector. */ + int index = --parser->state.next; + error_t err = EBADKEY; + struct group *group; + int key = 0; /* Which of ARGP_KEY_ARG[S] we used. */ + + /* Try to parse the argument in each parser. */ + for (group = parser->groups + ; group < parser->egroup && err == EBADKEY + ; group++) + { + parser->state.next++; /* For ARGP_KEY_ARG, consume the arg. */ + key = ARGP_KEY_ARG; + err = group_parse (group, &parser->state, key, val); + + if (err == EBADKEY) + /* This parser doesn't like ARGP_KEY_ARG; try ARGP_KEY_ARGS instead. */ + { + parser->state.next--; /* For ARGP_KEY_ARGS, put back the arg. */ + key = ARGP_KEY_ARGS; + err = group_parse (group, &parser->state, key, 0); + } + } + + if (! err) + { + if (key == ARGP_KEY_ARGS) + /* The default for ARGP_KEY_ARGS is to assume that if NEXT isn't + changed by the user, *all* arguments should be considered + consumed. */ + parser->state.next = parser->state.argc; + + if (parser->state.next > index) + /* Remember that we successfully processed a non-option + argument -- but only if the user hasn't gotten tricky and set + the clock back. */ + (--group)->args_processed += (parser->state.next - index); + else + /* The user wants to reparse some args, give getopt another try. */ + parser->try_getopt = 1; + } + + return err; +} + +/* Call the user parsers to parse the option OPT, with argument VAL, at the + current position, returning any error. */ +static error_t +parser_parse_opt (struct parser *parser, int opt, char *val) +{ + /* The group key encoded in the high bits; 0 for short opts or + group_number + 1 for long opts. */ + int group_key = opt >> USER_BITS; + error_t err = EBADKEY; + + if (group_key == 0) + /* A short option. By comparing OPT's position in SHORT_OPTS to the + various starting positions in each group's SHORT_END field, we can + determine which group OPT came from. */ + { + struct group *group; + char *short_index = strchr (parser->short_opts, opt); + + if (short_index) + for (group = parser->groups; group < parser->egroup; group++) + if (group->short_end > short_index) + { + err = group_parse (group, &parser->state, opt, + parser->opt_data.optarg); + break; + } + } + else + /* A long option. We use shifts instead of masking for extracting + the user value in order to preserve the sign. */ + err = + group_parse (&parser->groups[group_key - 1], &parser->state, + (opt << GROUP_BITS) >> GROUP_BITS, + parser->opt_data.optarg); + + if (err == EBADKEY) + /* At least currently, an option not recognized is an error in the + parser, because we pre-compute which parser is supposed to deal + with each option. */ + { + static const char bad_key_err[] = + N_("(PROGRAM ERROR) Option should have been recognized!?"); + if (group_key == 0) + __argp_error (&parser->state, "-%c: %s", opt, + dgettext (parser->argp->argp_domain, bad_key_err)); + else + { + struct option *long_opt = parser->long_opts; + while (long_opt->val != opt && long_opt->name) + long_opt++; + __argp_error (&parser->state, "--%s: %s", + long_opt->name ? long_opt->name : "???", + dgettext (parser->argp->argp_domain, bad_key_err)); + } + } + + return err; +} + +/* Parse the next argument in PARSER (as indicated by PARSER->state.next). + Any error from the parsers is returned, and *ARGP_EBADKEY indicates + whether a value of EBADKEY is due to an unrecognized argument (which is + generally not fatal). */ +static error_t +parser_parse_next (struct parser *parser, int *arg_ebadkey) +{ + int opt; + error_t err = 0; + + if (parser->state.quoted && parser->state.next < parser->state.quoted) + /* The next argument pointer has been moved to before the quoted + region, so pretend we never saw the quoting `--', and give getopt + another chance. If the user hasn't removed it, getopt will just + process it again. */ + parser->state.quoted = 0; + + if (parser->try_getopt && !parser->state.quoted) + /* Give getopt a chance to parse this. */ + { + /* Put it back in OPTIND for getopt. */ + parser->opt_data.optind = parser->state.next; + /* Distinguish KEY_ERR from a real option. */ + parser->opt_data.optopt = KEY_END; + if (parser->state.flags & ARGP_LONG_ONLY) + opt = _getopt_long_only_r (parser->state.argc, parser->state.argv, + parser->short_opts, parser->long_opts, 0, + &parser->opt_data); + else + opt = _getopt_long_r (parser->state.argc, parser->state.argv, + parser->short_opts, parser->long_opts, 0, + &parser->opt_data); + /* And see what getopt did. */ + parser->state.next = parser->opt_data.optind; + + if (opt == KEY_END) + /* Getopt says there are no more options, so stop using + getopt; we'll continue if necessary on our own. */ + { + parser->try_getopt = 0; + if (parser->state.next > 1 + && strcmp (parser->state.argv[parser->state.next - 1], QUOTE) + == 0) + /* Not only is this the end of the options, but it's a + `quoted' region, which may have args that *look* like + options, so we definitely shouldn't try to use getopt past + here, whatever happens. */ + parser->state.quoted = parser->state.next; + } + else if (opt == KEY_ERR && parser->opt_data.optopt != KEY_END) + /* KEY_ERR can have the same value as a valid user short + option, but in the case of a real error, getopt sets OPTOPT + to the offending character, which can never be KEY_END. */ + { + *arg_ebadkey = 0; + return EBADKEY; + } + } + else + opt = KEY_END; + + if (opt == KEY_END) + { + /* We're past what getopt considers the options. */ + if (parser->state.next >= parser->state.argc + || (parser->state.flags & ARGP_NO_ARGS)) + /* Indicate that we're done. */ + { + *arg_ebadkey = 1; + return EBADKEY; + } + else + /* A non-option arg; simulate what getopt might have done. */ + { + opt = KEY_ARG; + parser->opt_data.optarg = parser->state.argv[parser->state.next++]; + } + } + + if (opt == KEY_ARG) + /* A non-option argument; try each parser in turn. */ + err = parser_parse_arg (parser, parser->opt_data.optarg); + else + err = parser_parse_opt (parser, opt, parser->opt_data.optarg); + + if (err == EBADKEY) + *arg_ebadkey = (opt == KEY_END || opt == KEY_ARG); + + return err; +} + +/* Parse the options strings in ARGC & ARGV according to the argp in ARGP. + FLAGS is one of the ARGP_ flags above. If END_INDEX is non-NULL, the + index in ARGV of the first unparsed option is returned in it. If an + unknown option is present, EINVAL is returned; if some parser routine + returned a non-zero value, it is returned; otherwise 0 is returned. */ +error_t +__argp_parse (const struct argp *argp, int argc, char **argv, unsigned flags, + int *end_index, void *input) +{ + error_t err; + struct parser parser; + + /* If true, then err == EBADKEY is a result of a non-option argument failing + to be parsed (which in some cases isn't actually an error). */ + int arg_ebadkey = 0; + +#ifndef _LIBC + if (!(flags & ARGP_PARSE_ARGV0)) + { +#ifdef HAVE_DECL_PROGRAM_INVOCATION_NAME + if (!program_invocation_name) + program_invocation_name = argv[0]; +#endif +#ifdef HAVE_DECL_PROGRAM_INVOCATION_SHORT_NAME + if (!program_invocation_short_name) + program_invocation_short_name = __argp_base_name (argv[0]); +#endif + } +#endif + + if (! (flags & ARGP_NO_HELP)) + /* Add our own options. */ + { + struct argp_child *child = alloca (4 * sizeof (struct argp_child)); + struct argp *top_argp = alloca (sizeof (struct argp)); + + /* TOP_ARGP has no options, it just serves to group the user & default + argps. */ + memset (top_argp, 0, sizeof (*top_argp)); + top_argp->children = child; + + memset (child, 0, 4 * sizeof (struct argp_child)); + + if (argp) + (child++)->argp = argp; + (child++)->argp = &argp_default_argp; + if (argp_program_version || argp_program_version_hook) + (child++)->argp = &argp_version_argp; + child->argp = 0; + + argp = top_argp; + } + + /* Construct a parser for these arguments. */ + err = parser_init (&parser, argp, argc, argv, flags, input); + + if (! err) + /* Parse! */ + { + while (! err) + err = parser_parse_next (&parser, &arg_ebadkey); + err = parser_finalize (&parser, err, arg_ebadkey, end_index); + } + + return err; +} +#ifdef weak_alias +weak_alias (__argp_parse, argp_parse) +#endif + +/* Return the input field for ARGP in the parser corresponding to STATE; used + by the help routines. */ +void * +__argp_input (const struct argp *argp, const struct argp_state *state) +{ + if (state) + { + struct group *group; + struct parser *parser = state->pstate; + + for (group = parser->groups; group < parser->egroup; group++) + if (group->argp == argp) + return group->input; + } + + return 0; +} +#ifdef weak_alias +weak_alias (__argp_input, _argp_input) +#endif diff --git a/lib/argp-pin.c b/lib/argp-pin.c new file mode 100644 index 0000000..852c6d6 --- /dev/null +++ b/lib/argp-pin.c @@ -0,0 +1,28 @@ +/* Full and short program names for argp module + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#ifndef HAVE_PROGRAM_INVOCATION_SHORT_NAME +char *program_invocation_short_name = 0; +#endif +#ifndef HAVE_PROGRAM_INVOCATION_NAME +char *program_invocation_name = 0; +#endif + diff --git a/lib/argp-pv.c b/lib/argp-pv.c new file mode 100644 index 0000000..a11298b --- /dev/null +++ b/lib/argp-pv.c @@ -0,0 +1,24 @@ +/* Default definition for ARGP_PROGRAM_VERSION. + Copyright (C) 1996, 1997, 1999, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* If set by the user program to a non-zero value, then a default option + --version is added (unless the ARGP_NO_HELP flag is used), which will + print this string followed by a newline and exit (unless the + ARGP_NO_EXIT flag is used). Overridden by ARGP_PROGRAM_VERSION_HOOK. */ +const char *argp_program_version; diff --git a/lib/argp-pvh.c b/lib/argp-pvh.c new file mode 100644 index 0000000..6bf7c49 --- /dev/null +++ b/lib/argp-pvh.c @@ -0,0 +1,31 @@ +/* Default definition for ARGP_PROGRAM_VERSION_HOOK. + Copyright (C) 1996, 1997, 1999, 2004 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include "argp.h" + +/* If set by the user program to a non-zero value, then a default option + --version is added (unless the ARGP_NO_HELP flag is used), which calls + this function with a stream to print the version to and a pointer to the + current parsing state, and then exits (unless the ARGP_NO_EXIT flag is + used). This variable takes precedent over ARGP_PROGRAM_VERSION. */ +void (*argp_program_version_hook) (FILE *stream, struct argp_state *state) = NULL; diff --git a/lib/argp-xinl.c b/lib/argp-xinl.c new file mode 100644 index 0000000..a6afb1f --- /dev/null +++ b/lib/argp-xinl.c @@ -0,0 +1,43 @@ +/* Real definitions for extern inline functions in argp.h + Copyright (C) 1997, 1998, 2004 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#if defined _LIBC || defined HAVE_FEATURES_H +# include <features.h> +#endif + +#ifndef __USE_EXTERN_INLINES +# define __USE_EXTERN_INLINES 1 +#endif +#define ARGP_EI +#undef __OPTIMIZE__ +#define __OPTIMIZE__ 1 +#include "argp.h" + +/* Add weak aliases. */ +#if _LIBC - 0 && defined (weak_alias) + +weak_alias (__argp_usage, argp_usage) +weak_alias (__option_is_short, _option_is_short) +weak_alias (__option_is_end, _option_is_end) + +#endif diff --git a/lib/argp.h b/lib/argp.h new file mode 100644 index 0000000..1c4e06f --- /dev/null +++ b/lib/argp.h @@ -0,0 +1,622 @@ +/* Hierarchial argument parsing, layered over getopt. + Copyright (C) 1995-1999,2003-2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Miles Bader <miles@gnu.ai.mit.edu>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _ARGP_H +#define _ARGP_H + +#include <stdio.h> +#include <ctype.h> +#include <getopt.h> +#include <limits.h> + +#define __need_error_t +#include <errno.h> + +#ifndef __THROW +# define __THROW +#endif +#ifndef __NTH +# define __NTH(fct) fct __THROW +#endif + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) || __STRICT_ANSI__ +# define __format__ format +# define __printf__ printf +# endif +#endif + +/* GCC 2.95 and later have "__restrict"; C99 compilers have + "restrict", and "configure" may have defined "restrict". + Other compilers use __restrict, __restrict__, and _Restrict, and + 'configure' might #define 'restrict' to those words. */ +#ifndef __restrict +# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) +# if 199901L <= __STDC_VERSION__ +# define __restrict restrict +# else +# define __restrict +# endif +# endif +#endif + +#ifndef __error_t_defined +typedef int error_t; +# define __error_t_defined +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* A description of a particular option. A pointer to an array of + these is passed in the OPTIONS field of an argp structure. Each option + entry can correspond to one long option and/or one short option; more + names for the same option can be added by following an entry in an option + array with options having the OPTION_ALIAS flag set. */ +struct argp_option +{ + /* The long option name. For more than one name for the same option, you + can use following options with the OPTION_ALIAS flag set. */ + const char *name; + + /* What key is returned for this option. If > 0 and printable, then it's + also accepted as a short option. */ + int key; + + /* If non-NULL, this is the name of the argument associated with this + option, which is required unless the OPTION_ARG_OPTIONAL flag is set. */ + const char *arg; + + /* OPTION_ flags. */ + int flags; + + /* The doc string for this option. If both NAME and KEY are 0, This string + will be printed outdented from the normal option column, making it + useful as a group header (it will be the first thing printed in its + group); in this usage, it's conventional to end the string with a `:'. + + Write the initial value as N_("TEXT") if you want xgettext to collect + it into a POT file. */ + const char *doc; + + /* The group this option is in. In a long help message, options are sorted + alphabetically within each group, and the groups presented in the order + 0, 1, 2, ..., n, -m, ..., -2, -1. Every entry in an options array with + if this field 0 will inherit the group number of the previous entry, or + zero if it's the first one, unless its a group header (NAME and KEY both + 0), in which case, the previous entry + 1 is the default. Automagic + options such as --help are put into group -1. */ + int group; +}; + +/* The argument associated with this option is optional. */ +#define OPTION_ARG_OPTIONAL 0x1 + +/* This option isn't displayed in any help messages. */ +#define OPTION_HIDDEN 0x2 + +/* This option is an alias for the closest previous non-alias option. This + means that it will be displayed in the same help entry, and will inherit + fields other than NAME and KEY from the aliased option. */ +#define OPTION_ALIAS 0x4 + +/* This option isn't actually an option (and so should be ignored by the + actual option parser), but rather an arbitrary piece of documentation that + should be displayed in much the same manner as the options. If this flag + is set, then the option NAME field is displayed unmodified (e.g., no `--' + prefix is added) at the left-margin (where a *short* option would normally + be displayed), and the documentation string in the normal place. The NAME + field will be translated using gettext, unless OPTION_NO_TRANS is set (see + below). For purposes of sorting, any leading whitespace and punctuation is + ignored, except that if the first non-whitespace character is not `-', this + entry is displayed after all options (and OPTION_DOC entries with a leading + `-') in the same group. */ +#define OPTION_DOC 0x8 + +/* This option shouldn't be included in `long' usage messages (but is still + included in help messages). This is mainly intended for options that are + completely documented in an argp's ARGS_DOC field, in which case including + the option in the generic usage list would be redundant. For instance, + if ARGS_DOC is "FOO BAR\n-x BLAH", and the `-x' option's purpose is to + distinguish these two cases, -x should probably be marked + OPTION_NO_USAGE. */ +#define OPTION_NO_USAGE 0x10 + +/* Valid only in conjunction with OPTION_DOC. This option disables translation + of option name. */ +#define OPTION_NO_TRANS 0x20 + + +struct argp; /* fwd declare this type */ +struct argp_state; /* " */ +struct argp_child; /* " */ + +/* The type of a pointer to an argp parsing function. */ +typedef error_t (*argp_parser_t) (int key, char *arg, + struct argp_state *state); + +/* What to return for unrecognized keys. For special ARGP_KEY_ keys, such + returns will simply be ignored. For user keys, this error will be turned + into EINVAL (if the call to argp_parse is such that errors are propagated + back to the user instead of exiting); returning EINVAL itself would result + in an immediate stop to parsing in *all* cases. */ +#define ARGP_ERR_UNKNOWN E2BIG /* Hurd should never need E2BIG. XXX */ + +/* Special values for the KEY argument to an argument parsing function. + ARGP_ERR_UNKNOWN should be returned if they aren't understood. + + The sequence of keys to a parsing function is either (where each + uppercased word should be prefixed by `ARGP_KEY_' and opt is a user key): + + INIT opt... NO_ARGS END SUCCESS -- No non-option arguments at all + or INIT (opt | ARG)... END SUCCESS -- All non-option args parsed + or INIT (opt | ARG)... SUCCESS -- Some non-option arg unrecognized + + The third case is where every parser returned ARGP_KEY_UNKNOWN for an + argument, in which case parsing stops at that argument (returning the + unparsed arguments to the caller of argp_parse if requested, or stopping + with an error message if not). + + If an error occurs (either detected by argp, or because the parsing + function returned an error value), then the parser is called with + ARGP_KEY_ERROR, and no further calls are made. */ + +/* This is not an option at all, but rather a command line argument. If a + parser receiving this key returns success, the fact is recorded, and the + ARGP_KEY_NO_ARGS case won't be used. HOWEVER, if while processing the + argument, a parser function decrements the NEXT field of the state it's + passed, the option won't be considered processed; this is to allow you to + actually modify the argument (perhaps into an option), and have it + processed again. */ +#define ARGP_KEY_ARG 0 +/* There are remaining arguments not parsed by any parser, which may be found + starting at (STATE->argv + STATE->next). If success is returned, but + STATE->next left untouched, it's assumed that all arguments were consume, + otherwise, the parser should adjust STATE->next to reflect any arguments + consumed. */ +#define ARGP_KEY_ARGS 0x1000006 +/* There are no more command line arguments at all. */ +#define ARGP_KEY_END 0x1000001 +/* Because it's common to want to do some special processing if there aren't + any non-option args, user parsers are called with this key if they didn't + successfully process any non-option arguments. Called just before + ARGP_KEY_END (where more general validity checks on previously parsed + arguments can take place). */ +#define ARGP_KEY_NO_ARGS 0x1000002 +/* Passed in before any parsing is done. Afterwards, the values of each + element of the CHILD_INPUT field, if any, in the state structure is + copied to each child's state to be the initial value of the INPUT field. */ +#define ARGP_KEY_INIT 0x1000003 +/* Use after all other keys, including SUCCESS & END. */ +#define ARGP_KEY_FINI 0x1000007 +/* Passed in when parsing has successfully been completed (even if there are + still arguments remaining). */ +#define ARGP_KEY_SUCCESS 0x1000004 +/* Passed in if an error occurs. */ +#define ARGP_KEY_ERROR 0x1000005 + +/* An argp structure contains a set of options declarations, a function to + deal with parsing one, documentation string, a possible vector of child + argp's, and perhaps a function to filter help output. When actually + parsing options, getopt is called with the union of all the argp + structures chained together through their CHILD pointers, with conflicts + being resolved in favor of the first occurrence in the chain. */ +struct argp +{ + /* An array of argp_option structures, terminated by an entry with both + NAME and KEY having a value of 0. */ + const struct argp_option *options; + + /* What to do with an option from this structure. KEY is the key + associated with the option, and ARG is any associated argument (NULL if + none was supplied). If KEY isn't understood, ARGP_ERR_UNKNOWN should be + returned. If a non-zero, non-ARGP_ERR_UNKNOWN value is returned, then + parsing is stopped immediately, and that value is returned from + argp_parse(). For special (non-user-supplied) values of KEY, see the + ARGP_KEY_ definitions below. */ + argp_parser_t parser; + + /* A string describing what other arguments are wanted by this program. It + is only used by argp_usage to print the `Usage:' message. If it + contains newlines, the strings separated by them are considered + alternative usage patterns, and printed on separate lines (lines after + the first are prefix by ` or: ' instead of `Usage:'). */ + const char *args_doc; + + /* If non-NULL, a string containing extra text to be printed before and + after the options in a long help message (separated by a vertical tab + `\v' character). + Write the initial value as N_("BEFORE-TEXT") "\v" N_("AFTER-TEXT") if + you want xgettext to collect the two pieces of text into a POT file. */ + const char *doc; + + /* A vector of argp_children structures, terminated by a member with a 0 + argp field, pointing to child argps should be parsed with this one. Any + conflicts are resolved in favor of this argp, or early argps in the + CHILDREN list. This field is useful if you use libraries that supply + their own argp structure, which you want to use in conjunction with your + own. */ + const struct argp_child *children; + + /* If non-zero, this should be a function to filter the output of help + messages. KEY is either a key from an option, in which case TEXT is + that option's help text, or a special key from the ARGP_KEY_HELP_ + defines, below, describing which other help text TEXT is. The function + should return either TEXT, if it should be used as-is, a replacement + string, which should be malloced, and will be freed by argp, or NULL, + meaning `print nothing'. The value for TEXT is *after* any translation + has been done, so if any of the replacement text also needs translation, + that should be done by the filter function. INPUT is either the input + supplied to argp_parse, or NULL, if argp_help was called directly. */ + char *(*help_filter) (int __key, const char *__text, void *__input); + + /* If non-zero the strings used in the argp library are translated using + the domain described by this string. Otherwise the currently installed + default domain is used. */ + const char *argp_domain; +}; + +/* Possible KEY arguments to a help filter function. */ +#define ARGP_KEY_HELP_PRE_DOC 0x2000001 /* Help text preceeding options. */ +#define ARGP_KEY_HELP_POST_DOC 0x2000002 /* Help text following options. */ +#define ARGP_KEY_HELP_HEADER 0x2000003 /* Option header string. */ +#define ARGP_KEY_HELP_EXTRA 0x2000004 /* After all other documentation; + TEXT is NULL for this key. */ +/* Explanatory note emitted when duplicate option arguments have been + suppressed. */ +#define ARGP_KEY_HELP_DUP_ARGS_NOTE 0x2000005 +#define ARGP_KEY_HELP_ARGS_DOC 0x2000006 /* Argument doc string. */ + +/* When an argp has a non-zero CHILDREN field, it should point to a vector of + argp_child structures, each of which describes a subsidiary argp. */ +struct argp_child +{ + /* The child parser. */ + const struct argp *argp; + + /* Flags for this child. */ + int flags; + + /* If non-zero, an optional header to be printed in help output before the + child options. As a side-effect, a non-zero value forces the child + options to be grouped together; to achieve this effect without actually + printing a header string, use a value of "". */ + const char *header; + + /* Where to group the child options relative to the other (`consolidated') + options in the parent argp; the values are the same as the GROUP field + in argp_option structs, but all child-groupings follow parent options at + a particular group level. If both this field and HEADER are zero, then + they aren't grouped at all, but rather merged with the parent options + (merging the child's grouping levels with the parents). */ + int group; +}; + +/* Parsing state. This is provided to parsing functions called by argp, + which may examine and, as noted, modify fields. */ +struct argp_state +{ + /* The top level ARGP being parsed. */ + const struct argp *root_argp; + + /* The argument vector being parsed. May be modified. */ + int argc; + char **argv; + + /* The index in ARGV of the next arg that to be parsed. May be modified. */ + int next; + + /* The flags supplied to argp_parse. May be modified. */ + unsigned flags; + + /* While calling a parsing function with a key of ARGP_KEY_ARG, this is the + number of the current arg, starting at zero, and incremented after each + such call returns. At all other times, this is the number of such + arguments that have been processed. */ + unsigned arg_num; + + /* If non-zero, the index in ARGV of the first argument following a special + `--' argument (which prevents anything following being interpreted as an + option). Only set once argument parsing has proceeded past this point. */ + int quoted; + + /* An arbitrary pointer passed in from the user. */ + void *input; + /* Values to pass to child parsers. This vector will be the same length as + the number of children for the current parser. */ + void **child_inputs; + + /* For the parser's use. Initialized to 0. */ + void *hook; + + /* The name used when printing messages. This is initialized to ARGV[0], + or PROGRAM_INVOCATION_NAME if that is unavailable. */ + char *name; + + /* Streams used when argp prints something. */ + FILE *err_stream; /* For errors; initialized to stderr. */ + FILE *out_stream; /* For information; initialized to stdout. */ + + void *pstate; /* Private, for use by argp. */ +}; + +/* Flags for argp_parse (note that the defaults are those that are + convenient for program command line parsing): */ + +/* Don't ignore the first element of ARGV. Normally (and always unless + ARGP_NO_ERRS is set) the first element of the argument vector is + skipped for option parsing purposes, as it corresponds to the program name + in a command line. */ +#define ARGP_PARSE_ARGV0 0x01 + +/* Don't print error messages for unknown options to stderr; unless this flag + is set, ARGP_PARSE_ARGV0 is ignored, as ARGV[0] is used as the program + name in the error messages. This flag implies ARGP_NO_EXIT (on the + assumption that silent exiting upon errors is bad behaviour). */ +#define ARGP_NO_ERRS 0x02 + +/* Don't parse any non-option args. Normally non-option args are parsed by + calling the parse functions with a key of ARGP_KEY_ARG, and the actual arg + as the value. Since it's impossible to know which parse function wants to + handle it, each one is called in turn, until one returns 0 or an error + other than ARGP_ERR_UNKNOWN; if an argument is handled by no one, the + argp_parse returns prematurely (but with a return value of 0). If all + args have been parsed without error, all parsing functions are called one + last time with a key of ARGP_KEY_END. This flag needn't normally be set, + as the normal behavior is to stop parsing as soon as some argument can't + be handled. */ +#define ARGP_NO_ARGS 0x04 + +/* Parse options and arguments in the same order they occur on the command + line -- normally they're rearranged so that all options come first. */ +#define ARGP_IN_ORDER 0x08 + +/* Don't provide the standard long option --help, which causes usage and + option help information to be output to stdout, and exit (0) called. */ +#define ARGP_NO_HELP 0x10 + +/* Don't exit on errors (they may still result in error messages). */ +#define ARGP_NO_EXIT 0x20 + +/* Use the gnu getopt `long-only' rules for parsing arguments. */ +#define ARGP_LONG_ONLY 0x40 + +/* Turns off any message-printing/exiting options. */ +#define ARGP_SILENT (ARGP_NO_EXIT | ARGP_NO_ERRS | ARGP_NO_HELP) + +/* Parse the options strings in ARGC & ARGV according to the options in ARGP. + FLAGS is one of the ARGP_ flags above. If ARG_INDEX is non-NULL, the + index in ARGV of the first unparsed option is returned in it. If an + unknown option is present, ARGP_ERR_UNKNOWN is returned; if some parser + routine returned a non-zero value, it is returned; otherwise 0 is + returned. This function may also call exit unless the ARGP_NO_HELP flag + is set. INPUT is a pointer to a value to be passed in to the parser. */ +extern error_t argp_parse (const struct argp *__restrict __argp, + int /*argc*/, char **__restrict /*argv*/, + unsigned __flags, int *__restrict __arg_index, + void *__restrict __input); +extern error_t __argp_parse (const struct argp *__restrict __argp, + int /*argc*/, char **__restrict /*argv*/, + unsigned __flags, int *__restrict __arg_index, + void *__restrict __input); + +/* Global variables. */ + +/* GNULIB makes sure both program_invocation_name and + program_invocation_short_name are available */ +#ifdef GNULIB_PROGRAM_INVOCATION_NAME +extern char *program_invocation_name; +# undef HAVE_DECL_PROGRAM_INVOCATION_NAME +# define HAVE_DECL_PROGRAM_INVOCATION_NAME 1 +#endif + +#ifdef GNULIB_PROGRAM_INVOCATION_SHORT_NAME +extern char *program_invocation_short_name; +# undef HAVE_DECL_PROGRAM_INVOCATION_SHORT_NAME +# define HAVE_DECL_PROGRAM_INVOCATION_SHORT_NAME 1 +#endif + +/* If defined or set by the user program to a non-zero value, then a default + option --version is added (unless the ARGP_NO_HELP flag is used), which + will print this string followed by a newline and exit (unless the + ARGP_NO_EXIT flag is used). Overridden by ARGP_PROGRAM_VERSION_HOOK. */ +extern const char *argp_program_version; + +/* If defined or set by the user program to a non-zero value, then a default + option --version is added (unless the ARGP_NO_HELP flag is used), which + calls this function with a stream to print the version to and a pointer to + the current parsing state, and then exits (unless the ARGP_NO_EXIT flag is + used). This variable takes precedent over ARGP_PROGRAM_VERSION. */ +extern void (*argp_program_version_hook) (FILE *__restrict __stream, + struct argp_state *__restrict + __state); + +/* If defined or set by the user program, it should point to string that is + the bug-reporting address for the program. It will be printed by + argp_help if the ARGP_HELP_BUG_ADDR flag is set (as it is by various + standard help messages), embedded in a sentence that says something like + `Report bugs to ADDR.'. */ +extern const char *argp_program_bug_address; + +/* The exit status that argp will use when exiting due to a parsing error. + If not defined or set by the user program, this defaults to EX_USAGE from + <sysexits.h>. */ +extern error_t argp_err_exit_status; + +/* Flags for argp_help. */ +#define ARGP_HELP_USAGE 0x01 /* a Usage: message. */ +#define ARGP_HELP_SHORT_USAGE 0x02 /* " but don't actually print options. */ +#define ARGP_HELP_SEE 0x04 /* a `Try ... for more help' message. */ +#define ARGP_HELP_LONG 0x08 /* a long help message. */ +#define ARGP_HELP_PRE_DOC 0x10 /* doc string preceding long help. */ +#define ARGP_HELP_POST_DOC 0x20 /* doc string following long help. */ +#define ARGP_HELP_DOC (ARGP_HELP_PRE_DOC | ARGP_HELP_POST_DOC) +#define ARGP_HELP_BUG_ADDR 0x40 /* bug report address */ +#define ARGP_HELP_LONG_ONLY 0x80 /* modify output appropriately to + reflect ARGP_LONG_ONLY mode. */ + +/* These ARGP_HELP flags are only understood by argp_state_help. */ +#define ARGP_HELP_EXIT_ERR 0x100 /* Call exit(1) instead of returning. */ +#define ARGP_HELP_EXIT_OK 0x200 /* Call exit(0) instead of returning. */ + +/* The standard thing to do after a program command line parsing error, if an + error message has already been printed. */ +#define ARGP_HELP_STD_ERR \ + (ARGP_HELP_SEE | ARGP_HELP_EXIT_ERR) +/* The standard thing to do after a program command line parsing error, if no + more specific error message has been printed. */ +#define ARGP_HELP_STD_USAGE \ + (ARGP_HELP_SHORT_USAGE | ARGP_HELP_SEE | ARGP_HELP_EXIT_ERR) +/* The standard thing to do in response to a --help option. */ +#define ARGP_HELP_STD_HELP \ + (ARGP_HELP_SHORT_USAGE | ARGP_HELP_LONG | ARGP_HELP_EXIT_OK \ + | ARGP_HELP_DOC | ARGP_HELP_BUG_ADDR) + +/* Output a usage message for ARGP to STREAM. FLAGS are from the set + ARGP_HELP_*. */ +extern void argp_help (const struct argp *__restrict __argp, + FILE *__restrict __stream, + unsigned __flags, char *__restrict __name); +extern void __argp_help (const struct argp *__restrict __argp, + FILE *__restrict __stream, unsigned __flags, + char *__name); + +/* The following routines are intended to be called from within an argp + parsing routine (thus taking an argp_state structure as the first + argument). They may or may not print an error message and exit, depending + on the flags in STATE -- in any case, the caller should be prepared for + them *not* to exit, and should return an appropiate error after calling + them. [argp_usage & argp_error should probably be called argp_state_..., + but they're used often enough that they should be short] */ + +/* Output, if appropriate, a usage message for STATE to STREAM. FLAGS are + from the set ARGP_HELP_*. */ +extern void argp_state_help (const struct argp_state *__restrict __state, + FILE *__restrict __stream, + unsigned int __flags); +extern void __argp_state_help (const struct argp_state *__restrict __state, + FILE *__restrict __stream, + unsigned int __flags); + +/* Possibly output the standard usage message for ARGP to stderr and exit. */ +extern void argp_usage (const struct argp_state *__state); +extern void __argp_usage (const struct argp_state *__state); + +/* If appropriate, print the printf string FMT and following args, preceded + by the program name and `:', to stderr, and followed by a `Try ... --help' + message, then exit (1). */ +extern void argp_error (const struct argp_state *__restrict __state, + const char *__restrict __fmt, ...) + __attribute__ ((__format__ (__printf__, 2, 3))); +extern void __argp_error (const struct argp_state *__restrict __state, + const char *__restrict __fmt, ...) + __attribute__ ((__format__ (__printf__, 2, 3))); + +/* Similar to the standard gnu error-reporting function error(), but will + respect the ARGP_NO_EXIT and ARGP_NO_ERRS flags in STATE, and will print + to STATE->err_stream. This is useful for argument parsing code that is + shared between program startup (when exiting is desired) and runtime + option parsing (when typically an error code is returned instead). The + difference between this function and argp_error is that the latter is for + *parsing errors*, and the former is for other problems that occur during + parsing but don't reflect a (syntactic) problem with the input. */ +extern void argp_failure (const struct argp_state *__restrict __state, + int __status, int __errnum, + const char *__restrict __fmt, ...) + __attribute__ ((__format__ (__printf__, 4, 5))); +extern void __argp_failure (const struct argp_state *__restrict __state, + int __status, int __errnum, + const char *__restrict __fmt, ...) + __attribute__ ((__format__ (__printf__, 4, 5))); + +/* Returns true if the option OPT is a valid short option. */ +extern int _option_is_short (const struct argp_option *__opt) __THROW; +extern int __option_is_short (const struct argp_option *__opt) __THROW; + +/* Returns true if the option OPT is in fact the last (unused) entry in an + options array. */ +extern int _option_is_end (const struct argp_option *__opt) __THROW; +extern int __option_is_end (const struct argp_option *__opt) __THROW; + +/* Return the input field for ARGP in the parser corresponding to STATE; used + by the help routines. */ +extern void *_argp_input (const struct argp *__restrict __argp, + const struct argp_state *__restrict __state) + __THROW; +extern void *__argp_input (const struct argp *__restrict __argp, + const struct argp_state *__restrict __state) + __THROW; + +#ifdef __USE_EXTERN_INLINES + +# if !_LIBC +# define __argp_usage argp_usage +# define __argp_state_help argp_state_help +# define __option_is_short _option_is_short +# define __option_is_end _option_is_end +# endif + +# ifndef ARGP_EI +# define ARGP_EI extern __inline__ +# endif + +ARGP_EI void +__argp_usage (const struct argp_state *__state) +{ + __argp_state_help (__state, stderr, ARGP_HELP_STD_USAGE); +} + +ARGP_EI int +__NTH (__option_is_short (const struct argp_option *__opt)) +{ + if (__opt->flags & OPTION_DOC) + return 0; + else + { + int __key = __opt->key; + return __key > 0 && __key <= UCHAR_MAX && isprint (__key); + } +} + +ARGP_EI int +__NTH (__option_is_end (const struct argp_option *__opt)) +{ + return !__opt->key && !__opt->name && !__opt->doc && !__opt->group; +} + +# if !_LIBC +# undef __argp_usage +# undef __argp_state_help +# undef __option_is_short +# undef __option_is_end +# endif +#endif /* Use extern inlines. */ + +#ifdef __cplusplus +} +#endif + +#endif /* argp.h */ diff --git a/lib/asnprintf.c b/lib/asnprintf.c new file mode 100644 index 0000000..26c3988 --- /dev/null +++ b/lib/asnprintf.c @@ -0,0 +1,35 @@ +/* Formatted output to strings. + Copyright (C) 1999, 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "vasnprintf.h" + +#include <stdarg.h> + +char * +asnprintf (char *resultbuf, size_t *lengthp, const char *format, ...) +{ + va_list args; + char *result; + + va_start (args, format); + result = vasnprintf (resultbuf, lengthp, format, args); + va_end (args); + return result; +} diff --git a/lib/at-func.c b/lib/at-func.c new file mode 100644 index 0000000..f98c207 --- /dev/null +++ b/lib/at-func.c @@ -0,0 +1,86 @@ +/* Define an at-style functions like fstatat, unlinkat, fchownat, etc. + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#define CALL_FUNC(F) \ + (AT_FUNC_USE_F1_COND \ + ? AT_FUNC_F1 (F AT_FUNC_POST_FILE_ARGS) \ + : AT_FUNC_F2 (F AT_FUNC_POST_FILE_ARGS)) + +/* Call AT_FUNC_F1 or AT_FUNC_F2 (testing AT_FUNC_USE_F1_COND to + determine which) to operate on FILE, which is in the directory + open on descriptor FD. If possible, do it without changing the + working directory. Otherwise, resort to using save_cwd/fchdir, + then AT_FUNC_F?/restore_cwd. If either the save_cwd or the restore_cwd + fails, then give a diagnostic and exit nonzero. */ +int +AT_FUNC_NAME (int fd, char const *file AT_FUNC_POST_FILE_PARAM_DECLS) +{ + struct saved_cwd saved_cwd; + int saved_errno; + int err; + + if (fd == AT_FDCWD || IS_ABSOLUTE_FILE_NAME (file)) + return CALL_FUNC (file); + + { + char buf[OPENAT_BUFFER_SIZE]; + char *proc_file = openat_proc_name (buf, fd, file); + if (proc_file) + { + int proc_result = CALL_FUNC (proc_file); + int proc_errno = errno; + if (proc_file != buf) + free (proc_file); + /* If the syscall succeeds, or if it fails with an unexpected + errno value, then return right away. Otherwise, fall through + and resort to using save_cwd/restore_cwd. */ + if (0 <= proc_result) + return proc_result; + if (! EXPECTED_ERRNO (proc_errno)) + { + errno = proc_errno; + return proc_result; + } + } + } + + if (save_cwd (&saved_cwd) != 0) + openat_save_fail (errno); + + if (fchdir (fd) != 0) + { + saved_errno = errno; + free_cwd (&saved_cwd); + errno = saved_errno; + return -1; + } + + err = CALL_FUNC (file); + saved_errno = (err < 0 ? errno : 0); + + if (restore_cwd (&saved_cwd) != 0) + openat_restore_fail (errno); + + free_cwd (&saved_cwd); + + if (saved_errno) + errno = saved_errno; + return err; +} +#undef CALL_FUNC diff --git a/lib/backupfile.c b/lib/backupfile.c new file mode 100644 index 0000000..adfc0e5 --- /dev/null +++ b/lib/backupfile.c @@ -0,0 +1,364 @@ +/* backupfile.c -- make Emacs style backup file names + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and David MacKenzie. + Some algorithms adapted from GNU Emacs. */ + +#include <config.h> + +#include "backupfile.h" + +#include "argmatch.h" +#include "dirname.h" +#include "xalloc.h" + +#include <errno.h> +#include <stdbool.h> +#include <stdlib.h> +#include <string.h> + +#include <limits.h> + +#include <unistd.h> + +#include <dirent.h> +#ifndef _D_EXACT_NAMLEN +# define _D_EXACT_NAMLEN(dp) strlen ((dp)->d_name) +#endif +#if D_INO_IN_DIRENT +# define REAL_DIR_ENTRY(dp) ((dp)->d_ino != 0) +#else +# define REAL_DIR_ENTRY(dp) 1 +#endif + +#if ! (HAVE_PATHCONF && defined _PC_NAME_MAX) +# define pathconf(file, option) (errno = -1) +#endif + +#ifndef _POSIX_NAME_MAX +# define _POSIX_NAME_MAX 14 +#endif +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +#if defined _XOPEN_NAME_MAX +# define NAME_MAX_MINIMUM _XOPEN_NAME_MAX +#else +# define NAME_MAX_MINIMUM _POSIX_NAME_MAX +#endif + +#ifndef HAVE_DOS_FILE_NAMES +# define HAVE_DOS_FILE_NAMES 0 +#endif +#ifndef HAVE_LONG_FILE_NAMES +# define HAVE_LONG_FILE_NAMES 0 +#endif + +/* ISDIGIT differs from isdigit, as follows: + - Its arg may be any int or unsigned int; it need not be an unsigned char + or EOF. + - It's typically faster. + POSIX says that only '0' through '9' are digits. Prefer ISDIGIT to + ISDIGIT unless it's important to use the locale's definition + of `digit' even when the host does not conform to POSIX. */ +#define ISDIGIT(c) ((unsigned int) (c) - '0' <= 9) + +/* The results of opendir() in this file are not used with dirfd and fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef opendir +#undef closedir + +/* The extension added to file names to produce a simple (as opposed + to numbered) backup file name. */ +char const *simple_backup_suffix = "~"; + + +/* If FILE (which was of length FILELEN before an extension was + appended to it) is too long, replace the extension with the single + char E. If the result is still too long, remove the char just + before E. */ + +static void +check_extension (char *file, size_t filelen, char e) +{ + char *base = last_component (file); + size_t baselen = base_len (base); + size_t baselen_max = HAVE_LONG_FILE_NAMES ? 255 : NAME_MAX_MINIMUM; + + if (HAVE_DOS_FILE_NAMES || NAME_MAX_MINIMUM < baselen) + { + /* The new base name is long enough to require a pathconf check. */ + long name_max; + + /* Temporarily modify the buffer into its parent directory name, + invoke pathconf on the directory, and then restore the buffer. */ + char tmp[sizeof "."]; + memcpy (tmp, base, sizeof "."); + strcpy (base, "."); + errno = 0; + name_max = pathconf (file, _PC_NAME_MAX); + if (0 <= name_max || errno == 0) + { + long size = baselen_max = name_max; + if (name_max != size) + baselen_max = SIZE_MAX; + } + memcpy (base, tmp, sizeof "."); + } + + if (HAVE_DOS_FILE_NAMES && baselen_max <= 12) + { + /* Live within DOS's 8.3 limit. */ + char *dot = strchr (base, '.'); + if (!dot) + baselen_max = 8; + else + { + char const *second_dot = strchr (dot + 1, '.'); + baselen_max = (second_dot + ? second_dot - base + : dot + 1 - base + 3); + } + } + + if (baselen_max < baselen) + { + baselen = file + filelen - base; + if (baselen_max <= baselen) + baselen = baselen_max - 1; + base[baselen] = e; + base[baselen + 1] = '\0'; + } +} + +/* Returned values for NUMBERED_BACKUP. */ + +enum numbered_backup_result + { + /* The new backup name is the same length as an existing backup + name, so it's valid for that directory. */ + BACKUP_IS_SAME_LENGTH, + + /* Some backup names already exist, but the returned name is longer + than any of them, and its length should be checked. */ + BACKUP_IS_LONGER, + + /* There are no existing backup names. The new name's length + should be checked. */ + BACKUP_IS_NEW + }; + +/* *BUFFER contains a file name. Store into *BUFFER the next backup + name for the named file, with a version number greater than all the + existing numbered backups. Reallocate *BUFFER as necessary; its + initial allocated size is BUFFER_SIZE, which must be at least 4 + bytes longer than the file name to make room for the initially + appended ".~1". FILELEN is the length of the original file name. + The returned value indicates what kind of backup was found. If an + I/O or other read error occurs, use the highest backup number that + was found. */ + +static enum numbered_backup_result +numbered_backup (char **buffer, size_t buffer_size, size_t filelen) +{ + enum numbered_backup_result result = BACKUP_IS_NEW; + DIR *dirp; + struct dirent *dp; + char *buf = *buffer; + size_t versionlenmax = 1; + char *base = last_component (buf); + size_t base_offset = base - buf; + size_t baselen = base_len (base); + + /* Temporarily modify the buffer into its parent directory name, + open the directory, and then restore the buffer. */ + char tmp[sizeof "."]; + memcpy (tmp, base, sizeof "."); + strcpy (base, "."); + dirp = opendir (buf); + memcpy (base, tmp, sizeof "."); + strcpy (base + baselen, ".~1~"); + + if (!dirp) + return result; + + while ((dp = readdir (dirp)) != NULL) + { + char const *p; + char *q; + bool all_9s; + size_t versionlen; + size_t new_buflen; + + if (! REAL_DIR_ENTRY (dp) || _D_EXACT_NAMLEN (dp) < baselen + 4) + continue; + + if (memcmp (buf + base_offset, dp->d_name, baselen + 2) != 0) + continue; + + p = dp->d_name + baselen + 2; + + /* Check whether this file has a version number and if so, + whether it is larger. Use string operations rather than + integer arithmetic, to avoid problems with integer overflow. */ + + if (! ('1' <= *p && *p <= '9')) + continue; + all_9s = (*p == '9'); + for (versionlen = 1; ISDIGIT (p[versionlen]); versionlen++) + all_9s &= (p[versionlen] == '9'); + + if (! (p[versionlen] == '~' && !p[versionlen + 1] + && (versionlenmax < versionlen + || (versionlenmax == versionlen + && memcmp (buf + filelen + 2, p, versionlen) <= 0)))) + continue; + + /* This directory has the largest version number seen so far. + Append this highest numbered extension to the file name, + prepending '0' to the number if it is all 9s. */ + + versionlenmax = all_9s + versionlen; + result = (all_9s ? BACKUP_IS_LONGER : BACKUP_IS_SAME_LENGTH); + new_buflen = filelen + 2 + versionlenmax + 1; + if (buffer_size <= new_buflen) + { + buf = xnrealloc (buf, 2, new_buflen); + buffer_size = new_buflen * 2; + } + q = buf + filelen; + *q++ = '.'; + *q++ = '~'; + *q = '0'; + q += all_9s; + memcpy (q, p, versionlen + 2); + + /* Add 1 to the version number. */ + + q += versionlen; + while (*--q == '9') + *q = '0'; + ++*q; + } + + closedir (dirp); + *buffer = buf; + return result; +} + +/* Return the name of the new backup file for the existing file FILE, + allocated with malloc. Report an error and fail if out of memory. + Do not call this function if backup_type == no_backups. */ + +char * +find_backup_file_name (char const *file, enum backup_type backup_type) +{ + size_t filelen = strlen (file); + char *s; + size_t ssize; + bool simple = true; + + /* Allow room for simple or ".~N~" backups. The guess must be at + least sizeof ".~1~", but otherwise will be adjusted as needed. */ + size_t simple_backup_suffix_size = strlen (simple_backup_suffix) + 1; + size_t backup_suffix_size_guess = simple_backup_suffix_size; + enum { GUESS = sizeof ".~12345~" }; + if (backup_suffix_size_guess < GUESS) + backup_suffix_size_guess = GUESS; + + ssize = filelen + backup_suffix_size_guess + 1; + s = xmalloc (ssize); + memcpy (s, file, filelen + 1); + + if (backup_type != simple_backups) + switch (numbered_backup (&s, ssize, filelen)) + { + case BACKUP_IS_SAME_LENGTH: + return s; + + case BACKUP_IS_LONGER: + simple = false; + break; + + case BACKUP_IS_NEW: + simple = (backup_type == numbered_existing_backups); + break; + } + + if (simple) + memcpy (s + filelen, simple_backup_suffix, simple_backup_suffix_size); + check_extension (s, filelen, '~'); + return s; +} + +static char const * const backup_args[] = +{ + /* In a series of synonyms, present the most meaningful first, so + that argmatch_valid be more readable. */ + "none", "off", + "simple", "never", + "existing", "nil", + "numbered", "t", + NULL +}; + +static const enum backup_type backup_types[] = +{ + no_backups, no_backups, + simple_backups, simple_backups, + numbered_existing_backups, numbered_existing_backups, + numbered_backups, numbered_backups +}; + +/* Ensure that these two vectors have the same number of elements, + not counting the final NULL in the first one. */ +ARGMATCH_VERIFY (backup_args, backup_types); + +/* Return the type of backup specified by VERSION. + If VERSION is NULL or the empty string, return numbered_existing_backups. + If VERSION is invalid or ambiguous, fail with a diagnostic appropriate + for the specified CONTEXT. Unambiguous abbreviations are accepted. */ + +enum backup_type +get_version (char const *context, char const *version) +{ + if (version == 0 || *version == 0) + return numbered_existing_backups; + else + return XARGMATCH (context, version, backup_args, backup_types); +} + + +/* Return the type of backup specified by VERSION. + If VERSION is NULL, use the value of the envvar VERSION_CONTROL. + If the specified string is invalid or ambiguous, fail with a diagnostic + appropriate for the specified CONTEXT. + Unambiguous abbreviations are accepted. */ + +enum backup_type +xget_version (char const *context, char const *version) +{ + if (version && *version) + return get_version (context, version); + else + return get_version ("$VERSION_CONTROL", getenv ("VERSION_CONTROL")); +} diff --git a/lib/backupfile.h b/lib/backupfile.h new file mode 100644 index 0000000..7b44e58 --- /dev/null +++ b/lib/backupfile.h @@ -0,0 +1,61 @@ +/* backupfile.h -- declarations for making Emacs style backup file names + + Copyright (C) 1990, 1991, 1992, 1997, 1998, 1999, 2003, 2004 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef BACKUPFILE_H_ +# define BACKUPFILE_H_ + +# ifdef __cplusplus +extern "C" { +# endif + + +/* When to make backup files. */ +enum backup_type +{ + /* Never make backups. */ + no_backups, + + /* Make simple backups of every file. */ + simple_backups, + + /* Make numbered backups of files that already have numbered backups, + and simple backups of the others. */ + numbered_existing_backups, + + /* Make numbered backups of every file. */ + numbered_backups +}; + +# define VALID_BACKUP_TYPE(Type) \ + ((unsigned int) (Type) <= numbered_backups) + +extern char const *simple_backup_suffix; + +char *find_backup_file_name (char const *, enum backup_type); +enum backup_type get_version (char const *context, char const *arg); +enum backup_type xget_version (char const *context, char const *arg); +void addext (char *, char const *, int); + + +# ifdef __cplusplus +} +# endif + +#endif /* ! BACKUPFILE_H_ */ diff --git a/lib/basename.c b/lib/basename.c new file mode 100644 index 0000000..fbe17ff --- /dev/null +++ b/lib/basename.c @@ -0,0 +1,129 @@ +/* basename.c -- return the last element in a file name + + Copyright (C) 1990, 1998, 1999, 2000, 2001, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "dirname.h" + +#include <string.h> +#include "xalloc.h" +#include "xstrndup.h" + +/* Return the address of the last file name component of NAME. If + NAME has no relative file name components because it is a file + system root, return the empty string. */ + +char * +last_component (char const *name) +{ + char const *base = name + FILE_SYSTEM_PREFIX_LEN (name); + char const *p; + bool saw_slash = false; + + while (ISSLASH (*base)) + base++; + + for (p = base; *p; p++) + { + if (ISSLASH (*p)) + saw_slash = true; + else if (saw_slash) + { + base = p; + saw_slash = false; + } + } + + return (char *) base; +} + + +/* In general, we can't use the builtin `basename' function if available, + since it has different meanings in different environments. + In some environments the builtin `basename' modifies its argument. + + Return the last file name component of NAME, allocated with + xmalloc. On systems with drive letters, a leading "./" + distinguishes relative names that would otherwise look like a drive + letter. Unlike POSIX basename(), NAME cannot be NULL, + base_name("") returns "", and the first trailing slash is not + stripped. + + If lstat (NAME) would succeed, then { chdir (dir_name (NAME)); + lstat (base_name (NAME)); } will access the same file. Likewise, + if the sequence { chdir (dir_name (NAME)); + rename (base_name (NAME), "foo"); } succeeds, you have renamed NAME + to "foo" in the same directory NAME was in. */ + +char * +base_name (char const *name) +{ + char const *base = last_component (name); + size_t length; + + /* If there is no last component, then name is a file system root or the + empty string. */ + if (! *base) + return xstrndup (name, base_len (name)); + + /* Collapse a sequence of trailing slashes into one. */ + length = base_len (base); + if (ISSLASH (base[length])) + length++; + + /* On systems with drive letters, `a/b:c' must return `./b:c' rather + than `b:c' to avoid confusion with a drive letter. On systems + with pure POSIX semantics, this is not an issue. */ + if (FILE_SYSTEM_PREFIX_LEN (base)) + { + char *p = xmalloc (length + 3); + p[0] = '.'; + p[1] = '/'; + memcpy (p + 2, base, length); + p[length + 2] = '\0'; + return p; + } + + /* Finally, copy the basename. */ + return xstrndup (base, length); +} + +/* Return the length of the basename NAME. Typically NAME is the + value returned by base_name or last_component. Act like strlen + (NAME), except omit all trailing slashes. */ + +size_t +base_len (char const *name) +{ + size_t len; + size_t prefix_len = FILE_SYSTEM_PREFIX_LEN (name); + + for (len = strlen (name); 1 < len && ISSLASH (name[len - 1]); len--) + continue; + + if (DOUBLE_SLASH_IS_DISTINCT_ROOT && len == 1 + && ISSLASH (name[0]) && ISSLASH (name[1]) && ! name[2]) + return 2; + + if (FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE && prefix_len + && len == prefix_len && ISSLASH (name[prefix_len])) + return prefix_len + 1; + + return len; +} diff --git a/lib/canonicalize-lgpl.c b/lib/canonicalize-lgpl.c new file mode 100644 index 0000000..05ced11 --- /dev/null +++ b/lib/canonicalize-lgpl.c @@ -0,0 +1,353 @@ +/* Return the canonical absolute name of a given file. + Copyright (C) 1996-2003, 2005-2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Avoid a clash of our rpl_realpath() function with the prototype in + <stdlib.h> on Solaris 2.5.1. */ +#undef realpath + +#if !HAVE_CANONICALIZE_FILE_NAME || defined _LIBC + +#include <alloca.h> + +/* Specification. */ +#include "canonicalize.h" + +#include <stddef.h> +#include <stdlib.h> +#include <string.h> + +#if HAVE_UNISTD_H || defined _LIBC +# include <unistd.h> +#endif + +#include <limits.h> + +#if HAVE_SYS_PARAM_H || defined _LIBC +# include <sys/param.h> +#endif +#ifndef MAXSYMLINKS +# define MAXSYMLINKS 20 +#endif + +#include <sys/stat.h> + +#include <errno.h> +#ifndef _LIBC +# define __set_errno(e) errno = (e) +# ifndef ENAMETOOLONG +# define ENAMETOOLONG EINVAL +# endif +#endif + +#ifdef _LIBC +# include <shlib-compat.h> +#else +# define SHLIB_COMPAT(lib, introduced, obsoleted) 0 +# define versioned_symbol(lib, local, symbol, version) +# define compat_symbol(lib, local, symbol, version) +# define weak_alias(local, symbol) +# define __canonicalize_file_name canonicalize_file_name +# define __realpath rpl_realpath +# include "pathmax.h" +# include "allocsa.h" +# if HAVE_GETCWD +# ifdef VMS + /* We want the directory in Unix syntax, not in VMS syntax. */ +# define __getcwd(buf, max) getcwd (buf, max, 0) +# else +# define __getcwd getcwd +# endif +# else +# define __getcwd(buf, max) getwd (buf) +# endif +# define __readlink readlink + /* On systems without symbolic links, call stat() instead of lstat(). */ +# if !defined S_ISNLK && !HAVE_READLINK +# define lstat stat +# endif +#endif + +/* Return the canonical absolute name of file NAME. A canonical name + does not contain any `.', `..' components nor any repeated path + separators ('/') or symlinks. All path components must exist. If + RESOLVED is null, the result is malloc'd; otherwise, if the + canonical name is PATH_MAX chars or more, returns null with `errno' + set to ENAMETOOLONG; if the name fits in fewer than PATH_MAX chars, + returns the name in RESOLVED. If the name cannot be resolved and + RESOLVED is non-NULL, it contains the path of the first component + that cannot be resolved. If the path can be resolved, RESOLVED + holds the same value as the value returned. */ + +char * +__realpath (const char *name, char *resolved) +{ + char *rpath, *dest, *extra_buf = NULL; + const char *start, *end, *rpath_limit; + long int path_max; +#if HAVE_READLINK + int num_links = 0; +#endif + + if (name == NULL) + { + /* As per Single Unix Specification V2 we must return an error if + either parameter is a null pointer. We extend this to allow + the RESOLVED parameter to be NULL in case the we are expected to + allocate the room for the return value. */ + __set_errno (EINVAL); + return NULL; + } + + if (name[0] == '\0') + { + /* As per Single Unix Specification V2 we must return an error if + the name argument points to an empty string. */ + __set_errno (ENOENT); + return NULL; + } + +#ifdef PATH_MAX + path_max = PATH_MAX; +#else + path_max = pathconf (name, _PC_PATH_MAX); + if (path_max <= 0) + path_max = 1024; +#endif + + if (resolved == NULL) + { + rpath = malloc (path_max); + if (rpath == NULL) + return NULL; + } + else + rpath = resolved; + rpath_limit = rpath + path_max; + + if (name[0] != '/') + { + if (!__getcwd (rpath, path_max)) + { + rpath[0] = '\0'; + goto error; + } + dest = strchr (rpath, '\0'); + } + else + { + rpath[0] = '/'; + dest = rpath + 1; + } + + for (start = end = name; *start; start = end) + { +#ifdef _LIBC + struct stat64 st; +#else + struct stat st; +#endif + + /* Skip sequence of multiple path-separators. */ + while (*start == '/') + ++start; + + /* Find end of path component. */ + for (end = start; *end && *end != '/'; ++end) + /* Nothing. */; + + if (end - start == 0) + break; + else if (end - start == 1 && start[0] == '.') + /* nothing */; + else if (end - start == 2 && start[0] == '.' && start[1] == '.') + { + /* Back up to previous component, ignore if at root already. */ + if (dest > rpath + 1) + while ((--dest)[-1] != '/'); + } + else + { + size_t new_size; + + if (dest[-1] != '/') + *dest++ = '/'; + + if (dest + (end - start) >= rpath_limit) + { + ptrdiff_t dest_offset = dest - rpath; + char *new_rpath; + + if (resolved) + { + __set_errno (ENAMETOOLONG); + if (dest > rpath + 1) + dest--; + *dest = '\0'; + goto error; + } + new_size = rpath_limit - rpath; + if (end - start + 1 > path_max) + new_size += end - start + 1; + else + new_size += path_max; + new_rpath = (char *) realloc (rpath, new_size); + if (new_rpath == NULL) + goto error; + rpath = new_rpath; + rpath_limit = rpath + new_size; + + dest = rpath + dest_offset; + } + +#ifdef _LIBC + dest = __mempcpy (dest, start, end - start); +#else + memcpy (dest, start, end - start); + dest += end - start; +#endif + *dest = '\0'; + +#ifdef _LIBC + if (__lxstat64 (_STAT_VER, rpath, &st) < 0) +#else + if (lstat (rpath, &st) < 0) +#endif + goto error; + +#if HAVE_READLINK + if (S_ISLNK (st.st_mode)) + { + char *buf; + size_t len; + int n; + + if (++num_links > MAXSYMLINKS) + { + __set_errno (ELOOP); + goto error; + } + + buf = allocsa (path_max); + if (!buf) + { + errno = ENOMEM; + goto error; + } + + n = __readlink (rpath, buf, path_max); + if (n < 0) + { + int saved_errno = errno; + freesa (buf); + errno = saved_errno; + goto error; + } + buf[n] = '\0'; + + if (!extra_buf) + { + extra_buf = allocsa (path_max); + if (!extra_buf) + { + freesa (buf); + errno = ENOMEM; + goto error; + } + } + + len = strlen (end); + if ((long int) (n + len) >= path_max) + { + freesa (buf); + __set_errno (ENAMETOOLONG); + goto error; + } + + /* Careful here, end may be a pointer into extra_buf... */ + memmove (&extra_buf[n], end, len + 1); + name = end = memcpy (extra_buf, buf, n); + + if (buf[0] == '/') + dest = rpath + 1; /* It's an absolute symlink */ + else + /* Back up to previous component, ignore if at root already: */ + if (dest > rpath + 1) + while ((--dest)[-1] != '/'); + } +#endif + } + } + if (dest > rpath + 1 && dest[-1] == '/') + --dest; + *dest = '\0'; + + if (extra_buf) + freesa (extra_buf); + + return resolved ? memcpy (resolved, rpath, dest - rpath + 1) : rpath; + +error: + { + int saved_errno = errno; + if (extra_buf) + freesa (extra_buf); + if (resolved) + strcpy (resolved, rpath); + else + free (rpath); + errno = saved_errno; + } + return NULL; +} +#ifdef _LIBC +versioned_symbol (libc, __realpath, realpath, GLIBC_2_3); +#endif + + +#if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_3) +char * +__old_realpath (const char *name, char *resolved) +{ + if (resolved == NULL) + { + __set_errno (EINVAL); + return NULL; + } + + return __realpath (name, resolved); +} +compat_symbol (libc, __old_realpath, realpath, GLIBC_2_0); +#endif + + +char * +__canonicalize_file_name (const char *name) +{ + return __realpath (name, NULL); +} +weak_alias (__canonicalize_file_name, canonicalize_file_name) + +#else + +/* This declaration is solely to ensure that after preprocessing + this file is never empty. */ +typedef int dummy; + +#endif diff --git a/lib/canonicalize.h b/lib/canonicalize.h new file mode 100644 index 0000000..5c4d3f1 --- /dev/null +++ b/lib/canonicalize.h @@ -0,0 +1,54 @@ +/* Return the canonical absolute name of a given file. + Copyright (C) 1996-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef CANONICALIZE_H_ +# define CANONICALIZE_H_ + +# if GNULIB_CANONICALIZE +enum canonicalize_mode_t + { + /* All components must exist. */ + CAN_EXISTING = 0, + + /* All components excluding last one must exist. */ + CAN_ALL_BUT_LAST = 1, + + /* No requirements on components existence. */ + CAN_MISSING = 2 + }; +typedef enum canonicalize_mode_t canonicalize_mode_t; + +/* Return a malloc'd string containing the canonical absolute name of + the named file. This acts like canonicalize_file_name, except that + whether components must exist depends on the canonicalize_mode_t + argument. */ +char *canonicalize_filename_mode (const char *, canonicalize_mode_t); +# endif + +# if HAVE_DECL_CANONICALIZE_FILE_NAME +# include <stdlib.h> +# else +/* Return a malloc'd string containing the canonical absolute name of + the named file. If any file name component does not exist or is a + symlink to a nonexistent file, return NULL. A canonical name does + not contain any `.', `..' components nor any repeated file name + separators ('/') or symlinks. */ +char *canonicalize_file_name (const char *); +# endif + +#endif /* !CANONICALIZE_H_ */ diff --git a/lib/chdir-long.c b/lib/chdir-long.c new file mode 100644 index 0000000..3fc7ef2 --- /dev/null +++ b/lib/chdir-long.c @@ -0,0 +1,265 @@ +/* provide a chdir function that tries not to fail due to ENAMETOOLONG + Copyright (C) 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +#include "chdir-long.h" + +#include <fcntl.h> +#include <stdlib.h> +#include <stdbool.h> +#include <string.h> +#include <errno.h> +#include <stdio.h> +#include <assert.h> + +#include "openat.h" + +#ifndef PATH_MAX +# error "compile this file only if your system defines PATH_MAX" +#endif + +struct cd_buf +{ + int fd; +}; + +static inline void +cdb_init (struct cd_buf *cdb) +{ + cdb->fd = AT_FDCWD; +} + +static inline int +cdb_fchdir (struct cd_buf const *cdb) +{ + return fchdir (cdb->fd); +} + +static inline void +cdb_free (struct cd_buf const *cdb) +{ + if (0 <= cdb->fd) + { + bool close_fail = close (cdb->fd); + assert (! close_fail); + } +} + +/* Given a file descriptor of an open directory (or AT_FDCWD), CDB->fd, + try to open the CDB->fd-relative directory, DIR. If the open succeeds, + update CDB->fd with the resulting descriptor, close the incoming file + descriptor, and return zero. Upon failure, return -1 and set errno. */ +static int +cdb_advance_fd (struct cd_buf *cdb, char const *dir) +{ + int new_fd = openat (cdb->fd, dir, + O_RDONLY | O_DIRECTORY | O_NOCTTY | O_NONBLOCK); + if (new_fd < 0) + return -1; + + cdb_free (cdb); + cdb->fd = new_fd; + + return 0; +} + +/* Return a pointer to the first non-slash in S. */ +static inline char * +find_non_slash (char const *s) +{ + size_t n_slash = strspn (s, "/"); + return (char *) s + n_slash; +} + +/* This is a function much like chdir, but without the PATH_MAX limitation + on the length of the directory name. A significant difference is that + it must be able to modify (albeit only temporarily) the directory + name. It handles an arbitrarily long directory name by operating + on manageable portions of the name. On systems without the openat + syscall, this means changing the working directory to more and more + `distant' points along the long directory name and then restoring + the working directory. If any of those attempts to save or restore + the working directory fails, this function exits nonzero. + + Note that this function may still fail with errno == ENAMETOOLONG, but + only if the specified directory name contains a component that is long + enough to provoke such a failure all by itself (e.g. if the component + has length PATH_MAX or greater on systems that define PATH_MAX). */ + +int +chdir_long (char *dir) +{ + int e = chdir (dir); + if (e == 0 || errno != ENAMETOOLONG) + return e; + + { + size_t len = strlen (dir); + char *dir_end = dir + len; + struct cd_buf cdb; + size_t n_leading_slash; + + cdb_init (&cdb); + + /* If DIR is the empty string, then the chdir above + must have failed and set errno to ENOENT. */ + assert (0 < len); + assert (PATH_MAX <= len); + + /* Count leading slashes. */ + n_leading_slash = strspn (dir, "/"); + + /* Handle any leading slashes as well as any name that matches + the regular expression, m!^//hostname[/]*! . Handling this + prefix separately usually results in a single additional + cdb_advance_fd call, but it's worthwhile, since it makes the + code in the following loop cleaner. */ + if (n_leading_slash == 2) + { + int err; + /* Find next slash. + We already know that dir[2] is neither a slash nor '\0'. */ + char *slash = memchr (dir + 3, '/', dir_end - (dir + 3)); + if (slash == NULL) + { + errno = ENAMETOOLONG; + return -1; + } + *slash = '\0'; + err = cdb_advance_fd (&cdb, dir); + *slash = '/'; + if (err != 0) + goto Fail; + dir = find_non_slash (slash + 1); + } + else if (n_leading_slash) + { + if (cdb_advance_fd (&cdb, "/") != 0) + goto Fail; + dir += n_leading_slash; + } + + assert (*dir != '/'); + assert (dir <= dir_end); + + while (PATH_MAX <= dir_end - dir) + { + int err; + /* Find a slash that is PATH_MAX or fewer bytes away from dir. + I.e. see if there is a slash that will give us a name of + length PATH_MAX-1 or less. */ + char *slash = memrchr (dir, '/', PATH_MAX); + if (slash == NULL) + { + errno = ENAMETOOLONG; + return -1; + } + + *slash = '\0'; + assert (slash - dir < PATH_MAX); + err = cdb_advance_fd (&cdb, dir); + *slash = '/'; + if (err != 0) + goto Fail; + + dir = find_non_slash (slash + 1); + } + + if (dir < dir_end) + { + if (cdb_advance_fd (&cdb, dir) != 0) + goto Fail; + } + + if (cdb_fchdir (&cdb) != 0) + goto Fail; + + cdb_free (&cdb); + return 0; + + Fail: + { + int saved_errno = errno; + cdb_free (&cdb); + errno = saved_errno; + return -1; + } + } +} + +#if TEST_CHDIR + +# include <stdio.h> +# include "closeout.h" +# include "error.h" + +char *program_name; + +int +main (int argc, char *argv[]) +{ + char *line = NULL; + size_t n = 0; + int len; + + program_name = argv[0]; + atexit (close_stdout); + + len = getline (&line, &n, stdin); + if (len < 0) + { + int saved_errno = errno; + if (feof (stdin)) + exit (0); + + error (EXIT_FAILURE, saved_errno, + "reading standard input"); + } + else if (len == 0) + exit (0); + + if (line[len-1] == '\n') + line[len-1] = '\0'; + + if (chdir_long (line) != 0) + error (EXIT_FAILURE, errno, + "chdir_long failed: %s", line); + + if (argc <= 1) + { + /* Using `pwd' here makes sense only if it is a robust implementation, + like the one in coreutils after the 2004-04-19 changes. */ + char const *cmd = "pwd"; + execlp (cmd, (char *) NULL); + error (EXIT_FAILURE, errno, "%s", cmd); + } + + fclose (stdin); + fclose (stderr); + + exit (EXIT_SUCCESS); +} +#endif + +/* +Local Variables: +compile-command: "gcc -DTEST_CHDIR=1 -g -O -W -Wall chdir-long.c libcoreutils.a" +End: +*/ diff --git a/lib/chdir-long.h b/lib/chdir-long.h new file mode 100644 index 0000000..4852b40 --- /dev/null +++ b/lib/chdir-long.h @@ -0,0 +1,35 @@ +/* provide a chdir function that tries not to fail due to ENAMETOOLONG + Copyright (C) 2004, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <unistd.h> +#include <limits.h> + +#ifndef PATH_MAX +# ifdef MAXPATHLEN +# define PATH_MAX MAXPATHLEN +# endif +#endif + +/* On systems without PATH_MAX, presume that chdir accepts + arbitrarily long directory names. */ +#ifndef PATH_MAX +# define chdir_long(Dir) chdir (Dir) +#else +int chdir_long (char *dir); +#endif diff --git a/lib/chown.c b/lib/chown.c new file mode 100644 index 0000000..b7786f6 --- /dev/null +++ b/lib/chown.c @@ -0,0 +1,104 @@ +/* provide consistent interface to chown for systems that don't interpret + an ID of -1 as meaning `don't change the corresponding ID'. + + Copyright (C) 1997, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +/* Specification. */ +#include <unistd.h> + +#include <stdbool.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <errno.h> + +/* Below we refer to the system's chown(). */ +#undef chown + +/* The results of open() in this file are not used with fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef open +#undef close + +/* Provide a more-closely POSIX-conforming version of chown on + systems with one or both of the following problems: + - chown doesn't treat an ID of -1 as meaning + `don't change the corresponding ID'. + - chown doesn't dereference symlinks. */ + +int +rpl_chown (const char *file, uid_t uid, gid_t gid) +{ +#if CHOWN_FAILS_TO_HONOR_ID_OF_NEGATIVE_ONE + if (gid == (gid_t) -1 || uid == (uid_t) -1) + { + struct stat file_stats; + + /* Stat file to get id(s) that should remain unchanged. */ + if (stat (file, &file_stats)) + return -1; + + if (gid == (gid_t) -1) + gid = file_stats.st_gid; + + if (uid == (uid_t) -1) + uid = file_stats.st_uid; + } +#endif + +#if CHOWN_MODIFIES_SYMLINK + { + /* Handle the case in which the system-supplied chown function + does *not* follow symlinks. Instead, it changes permissions + on the symlink itself. To work around that, we open the + file (but this can fail due to lack of read or write permission) and + use fchown on the resulting descriptor. */ + int open_flags = O_NONBLOCK | O_NOCTTY; + int fd = open (file, O_RDONLY | open_flags); + if (0 <= fd + || (errno == EACCES + && 0 <= (fd = open (file, O_WRONLY | open_flags)))) + { + int result = fchown (fd, uid, gid); + int saved_errno = errno; + + /* POSIX says fchown can fail with errno == EINVAL on sockets, + so fall back on chown in that case. */ + struct stat sb; + bool fchown_socket_failure = + (result != 0 && saved_errno == EINVAL + && fstat (fd, &sb) == 0 && S_ISFIFO (sb.st_mode)); + + close (fd); + + if (! fchown_socket_failure) + { + errno = saved_errno; + return result; + } + } + else if (errno != EACCES) + return -1; + } +#endif + + return chown (file, uid, gid); +} diff --git a/lib/close-stream.c b/lib/close-stream.c new file mode 100644 index 0000000..72d0d68 --- /dev/null +++ b/lib/close-stream.c @@ -0,0 +1,76 @@ +/* Close a stream, with nicer error checking than fclose's. + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 2004, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "close-stream.h" + +#include <errno.h> +#include <stdbool.h> + +#include "__fpending.h" + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +/* Close STREAM. Return 0 if successful, EOF (setting errno) + otherwise. A failure might set errno to 0 if the error number + cannot be determined. + + If a program writes *anything* to STREAM, that program should close + STREAM and make sure that it succeeds before exiting. Otherwise, + suppose that you go to the extreme of checking the return status + of every function that does an explicit write to STREAM. The last + printf can succeed in writing to the internal stream buffer, and yet + the fclose(STREAM) could still fail (due e.g., to a disk full error) + when it tries to write out that buffered data. Thus, you would be + left with an incomplete output file and the offending program would + exit successfully. Even calling fflush is not always sufficient, + since some file systems (NFS and CODA) buffer written/flushed data + until an actual close call. + + Besides, it's wasteful to check the return value from every call + that writes to STREAM -- just let the internal stream state record + the failure. That's what the ferror test is checking below. */ + +int +close_stream (FILE *stream) +{ + bool some_pending = (__fpending (stream) != 0); + bool prev_fail = (ferror (stream) != 0); + bool fclose_fail = (fclose (stream) != 0); + + /* Return an error indication if there was a previous failure or if + fclose failed, with one exception: ignore an fclose failure if + there was no previous error, no data remains to be flushed, and + fclose failed with EBADF. That can happen when a program like cp + is invoked like this `cp a b >&-' (i.e., with standard output + closed) and doesn't generate any output (hence no previous error + and nothing to be flushed). */ + + if (prev_fail || (fclose_fail && (some_pending || errno != EBADF))) + { + if (! fclose_fail) + errno = 0; + return EOF; + } + + return 0; +} diff --git a/lib/close-stream.h b/lib/close-stream.h new file mode 100644 index 0000000..be3d419 --- /dev/null +++ b/lib/close-stream.h @@ -0,0 +1,2 @@ +#include <stdio.h> +int close_stream (FILE *stream); diff --git a/lib/closeout.c b/lib/closeout.c new file mode 100644 index 0000000..830f16f --- /dev/null +++ b/lib/closeout.c @@ -0,0 +1,86 @@ +/* Close standard output and standard error, exiting with a diagnostic on error. + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 2004, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "closeout.h" + +#include <errno.h> +#include <stdio.h> +#include <unistd.h> + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +#include "close-stream.h" +#include "error.h" +#include "exitfail.h" +#include "quotearg.h" + +static const char *file_name; + +/* Set the file name to be reported in the event an error is detected + by close_stdout. */ +void +close_stdout_set_file_name (const char *file) +{ + file_name = file; +} + +/* Close standard output. On error, issue a diagnostic and _exit + with status 'exit_failure'. + + Also close standard error. On error, _exit with status 'exit_failure'. + + Since close_stdout is commonly registered via 'atexit', POSIX + and the C standard both say that it should not call 'exit', + because the behavior is undefined if 'exit' is called more than + once. So it calls '_exit' instead of 'exit'. If close_stdout + is registered via atexit before other functions are registered, + the other functions can act before this _exit is invoked. + + Applications that use close_stdout should flush any streams + other than stdout and stderr before exiting, since the call to + _exit will bypass other buffer flushing. Applications should + be flushing and closing other streams anyway, to check for I/O + errors. Also, applications should not use tmpfile, since _exit + can bypass the removal of these files. + + It's important to detect such failures and exit nonzero because many + tools (most notably `make' and other build-management systems) depend + on being able to detect failure in other tools via their exit status. */ + +void +close_stdout (void) +{ + if (close_stream (stdout) != 0) + { + char const *write_error = _("write error"); + if (file_name) + error (0, errno, "%s: %s", quotearg_colon (file_name), + write_error); + else + error (0, errno, "%s", write_error); + + _exit (exit_failure); + } + + if (close_stream (stderr) != 0) + _exit (exit_failure); +} diff --git a/lib/closeout.h b/lib/closeout.h new file mode 100644 index 0000000..8bed23b --- /dev/null +++ b/lib/closeout.h @@ -0,0 +1,33 @@ +/* Close standard output and standard error. + + Copyright (C) 1998, 2000, 2003, 2004, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef CLOSEOUT_H +# define CLOSEOUT_H 1 + +# ifdef __cplusplus +extern "C" { +# endif + +void close_stdout_set_file_name (const char *file); +void close_stdout (void); + +# ifdef __cplusplus +} +# endif + +#endif diff --git a/lib/config.charset b/lib/config.charset new file mode 100755 index 0000000..148ea44 --- /dev/null +++ b/lib/config.charset @@ -0,0 +1,639 @@ +#! /bin/sh +# Output a system dependent table of character encoding aliases. +# +# Copyright (C) 2000-2004, 2006 Free Software Foundation, Inc. +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License along +# with this program; if not, write to the Free Software Foundation, +# Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. +# +# The table consists of lines of the form +# ALIAS CANONICAL +# +# ALIAS is the (system dependent) result of "nl_langinfo (CODESET)". +# ALIAS is compared in a case sensitive way. +# +# CANONICAL is the GNU canonical name for this character encoding. +# It must be an encoding supported by libiconv. Support by GNU libc is +# also desirable. CANONICAL is case insensitive. Usually an upper case +# MIME charset name is preferred. +# The current list of GNU canonical charset names is as follows. +# +# name MIME? used by which systems +# ASCII, ANSI_X3.4-1968 glibc solaris freebsd netbsd darwin +# ISO-8859-1 Y glibc aix hpux irix osf solaris freebsd netbsd darwin +# ISO-8859-2 Y glibc aix hpux irix osf solaris freebsd netbsd darwin +# ISO-8859-3 Y glibc solaris +# ISO-8859-4 Y osf solaris freebsd netbsd darwin +# ISO-8859-5 Y glibc aix hpux irix osf solaris freebsd netbsd darwin +# ISO-8859-6 Y glibc aix hpux solaris +# ISO-8859-7 Y glibc aix hpux irix osf solaris netbsd darwin +# ISO-8859-8 Y glibc aix hpux osf solaris +# ISO-8859-9 Y glibc aix hpux irix osf solaris darwin +# ISO-8859-13 glibc netbsd darwin +# ISO-8859-14 glibc +# ISO-8859-15 glibc aix osf solaris freebsd darwin +# KOI8-R Y glibc solaris freebsd netbsd darwin +# KOI8-U Y glibc freebsd netbsd darwin +# KOI8-T glibc +# CP437 dos +# CP775 dos +# CP850 aix osf dos +# CP852 dos +# CP855 dos +# CP856 aix +# CP857 dos +# CP861 dos +# CP862 dos +# CP864 dos +# CP865 dos +# CP866 freebsd netbsd darwin dos +# CP869 dos +# CP874 woe32 dos +# CP922 aix +# CP932 aix woe32 dos +# CP943 aix +# CP949 osf woe32 dos +# CP950 woe32 dos +# CP1046 aix +# CP1124 aix +# CP1125 dos +# CP1129 aix +# CP1250 woe32 +# CP1251 glibc solaris netbsd darwin woe32 +# CP1252 aix woe32 +# CP1253 woe32 +# CP1254 woe32 +# CP1255 glibc woe32 +# CP1256 woe32 +# CP1257 woe32 +# GB2312 Y glibc aix hpux irix solaris freebsd netbsd darwin +# EUC-JP Y glibc aix hpux irix osf solaris freebsd netbsd darwin +# EUC-KR Y glibc aix hpux irix osf solaris freebsd netbsd darwin +# EUC-TW glibc aix hpux irix osf solaris netbsd +# BIG5 Y glibc aix hpux osf solaris freebsd netbsd darwin +# BIG5-HKSCS glibc solaris +# GBK glibc aix osf solaris woe32 dos +# GB18030 glibc solaris netbsd +# SHIFT_JIS Y hpux osf solaris freebsd netbsd darwin +# JOHAB glibc solaris woe32 +# TIS-620 glibc aix hpux osf solaris +# VISCII Y glibc +# TCVN5712-1 glibc +# GEORGIAN-PS glibc +# HP-ROMAN8 hpux +# HP-ARABIC8 hpux +# HP-GREEK8 hpux +# HP-HEBREW8 hpux +# HP-TURKISH8 hpux +# HP-KANA8 hpux +# DEC-KANJI osf +# DEC-HANYU osf +# UTF-8 Y glibc aix hpux osf solaris netbsd darwin +# +# Note: Names which are not marked as being a MIME name should not be used in +# Internet protocols for information interchange (mail, news, etc.). +# +# Note: ASCII and ANSI_X3.4-1968 are synonymous canonical names. Applications +# must understand both names and treat them as equivalent. +# +# The first argument passed to this file is the canonical host specification, +# CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM +# or +# CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM + +host="$1" +os=`echo "$host" | sed -e 's/^[^-]*-[^-]*-\(.*\)$/\1/'` +echo "# This file contains a table of character encoding aliases," +echo "# suitable for operating system '${os}'." +echo "# It was automatically generated from config.charset." +# List of references, updated during installation: +echo "# Packages using this file: " +case "$os" in + linux-gnulibc1*) + # Linux libc5 doesn't have nl_langinfo(CODESET); therefore + # localcharset.c falls back to using the full locale name + # from the environment variables. + echo "C ASCII" + echo "POSIX ASCII" + for l in af af_ZA ca ca_ES da da_DK de de_AT de_BE de_CH de_DE de_LU \ + en en_AU en_BW en_CA en_DK en_GB en_IE en_NZ en_US en_ZA \ + en_ZW es es_AR es_BO es_CL es_CO es_DO es_EC es_ES es_GT \ + es_HN es_MX es_PA es_PE es_PY es_SV es_US es_UY es_VE et \ + et_EE eu eu_ES fi fi_FI fo fo_FO fr fr_BE fr_CA fr_CH fr_FR \ + fr_LU ga ga_IE gl gl_ES id id_ID in in_ID is is_IS it it_CH \ + it_IT kl kl_GL nl nl_BE nl_NL no no_NO pt pt_BR pt_PT sv \ + sv_FI sv_SE; do + echo "$l ISO-8859-1" + echo "$l.iso-8859-1 ISO-8859-1" + echo "$l.iso-8859-15 ISO-8859-15" + echo "$l.iso-8859-15@euro ISO-8859-15" + echo "$l@euro ISO-8859-15" + echo "$l.cp-437 CP437" + echo "$l.cp-850 CP850" + echo "$l.cp-1252 CP1252" + echo "$l.cp-1252@euro CP1252" + #echo "$l.atari-st ATARI-ST" # not a commonly used encoding + echo "$l.utf-8 UTF-8" + echo "$l.utf-8@euro UTF-8" + done + for l in cs cs_CZ hr hr_HR hu hu_HU pl pl_PL ro ro_RO sk sk_SK sl \ + sl_SI sr sr_CS sr_YU; do + echo "$l ISO-8859-2" + echo "$l.iso-8859-2 ISO-8859-2" + echo "$l.cp-852 CP852" + echo "$l.cp-1250 CP1250" + echo "$l.utf-8 UTF-8" + done + for l in mk mk_MK ru ru_RU; do + echo "$l ISO-8859-5" + echo "$l.iso-8859-5 ISO-8859-5" + echo "$l.koi8-r KOI8-R" + echo "$l.cp-866 CP866" + echo "$l.cp-1251 CP1251" + echo "$l.utf-8 UTF-8" + done + for l in ar ar_SA; do + echo "$l ISO-8859-6" + echo "$l.iso-8859-6 ISO-8859-6" + echo "$l.cp-864 CP864" + #echo "$l.cp-868 CP868" # not a commonly used encoding + echo "$l.cp-1256 CP1256" + echo "$l.utf-8 UTF-8" + done + for l in el el_GR gr gr_GR; do + echo "$l ISO-8859-7" + echo "$l.iso-8859-7 ISO-8859-7" + echo "$l.cp-869 CP869" + echo "$l.cp-1253 CP1253" + echo "$l.cp-1253@euro CP1253" + echo "$l.utf-8 UTF-8" + echo "$l.utf-8@euro UTF-8" + done + for l in he he_IL iw iw_IL; do + echo "$l ISO-8859-8" + echo "$l.iso-8859-8 ISO-8859-8" + echo "$l.cp-862 CP862" + echo "$l.cp-1255 CP1255" + echo "$l.utf-8 UTF-8" + done + for l in tr tr_TR; do + echo "$l ISO-8859-9" + echo "$l.iso-8859-9 ISO-8859-9" + echo "$l.cp-857 CP857" + echo "$l.cp-1254 CP1254" + echo "$l.utf-8 UTF-8" + done + for l in lt lt_LT lv lv_LV; do + #echo "$l BALTIC" # not a commonly used encoding, wrong encoding name + echo "$l ISO-8859-13" + done + for l in ru_UA uk uk_UA; do + echo "$l KOI8-U" + done + for l in zh zh_CN; do + #echo "$l GB_2312-80" # not a commonly used encoding, wrong encoding name + echo "$l GB2312" + done + for l in ja ja_JP ja_JP.EUC; do + echo "$l EUC-JP" + done + for l in ko ko_KR; do + echo "$l EUC-KR" + done + for l in th th_TH; do + echo "$l TIS-620" + done + for l in fa fa_IR; do + #echo "$l ISIRI-3342" # a broken encoding + echo "$l.utf-8 UTF-8" + done + ;; + linux* | *-gnu*) + # With glibc-2.1 or newer, we don't need any canonicalization, + # because glibc has iconv and both glibc and libiconv support all + # GNU canonical names directly. Therefore, the Makefile does not + # need to install the alias file at all. + # The following applies only to glibc-2.0.x and older libcs. + echo "ISO_646.IRV:1983 ASCII" + ;; + aix*) + echo "ISO8859-1 ISO-8859-1" + echo "ISO8859-2 ISO-8859-2" + echo "ISO8859-5 ISO-8859-5" + echo "ISO8859-6 ISO-8859-6" + echo "ISO8859-7 ISO-8859-7" + echo "ISO8859-8 ISO-8859-8" + echo "ISO8859-9 ISO-8859-9" + echo "ISO8859-15 ISO-8859-15" + echo "IBM-850 CP850" + echo "IBM-856 CP856" + echo "IBM-921 ISO-8859-13" + echo "IBM-922 CP922" + echo "IBM-932 CP932" + echo "IBM-943 CP943" + echo "IBM-1046 CP1046" + echo "IBM-1124 CP1124" + echo "IBM-1129 CP1129" + echo "IBM-1252 CP1252" + echo "IBM-eucCN GB2312" + echo "IBM-eucJP EUC-JP" + echo "IBM-eucKR EUC-KR" + echo "IBM-eucTW EUC-TW" + echo "big5 BIG5" + echo "GBK GBK" + echo "TIS-620 TIS-620" + echo "UTF-8 UTF-8" + ;; + hpux*) + echo "iso88591 ISO-8859-1" + echo "iso88592 ISO-8859-2" + echo "iso88595 ISO-8859-5" + echo "iso88596 ISO-8859-6" + echo "iso88597 ISO-8859-7" + echo "iso88598 ISO-8859-8" + echo "iso88599 ISO-8859-9" + echo "iso885915 ISO-8859-15" + echo "roman8 HP-ROMAN8" + echo "arabic8 HP-ARABIC8" + echo "greek8 HP-GREEK8" + echo "hebrew8 HP-HEBREW8" + echo "turkish8 HP-TURKISH8" + echo "kana8 HP-KANA8" + echo "tis620 TIS-620" + echo "big5 BIG5" + echo "eucJP EUC-JP" + echo "eucKR EUC-KR" + echo "eucTW EUC-TW" + echo "hp15CN GB2312" + #echo "ccdc ?" # what is this? + echo "SJIS SHIFT_JIS" + echo "utf8 UTF-8" + ;; + irix*) + echo "ISO8859-1 ISO-8859-1" + echo "ISO8859-2 ISO-8859-2" + echo "ISO8859-5 ISO-8859-5" + echo "ISO8859-7 ISO-8859-7" + echo "ISO8859-9 ISO-8859-9" + echo "eucCN GB2312" + echo "eucJP EUC-JP" + echo "eucKR EUC-KR" + echo "eucTW EUC-TW" + ;; + osf*) + echo "ISO8859-1 ISO-8859-1" + echo "ISO8859-2 ISO-8859-2" + echo "ISO8859-4 ISO-8859-4" + echo "ISO8859-5 ISO-8859-5" + echo "ISO8859-7 ISO-8859-7" + echo "ISO8859-8 ISO-8859-8" + echo "ISO8859-9 ISO-8859-9" + echo "ISO8859-15 ISO-8859-15" + echo "cp850 CP850" + echo "big5 BIG5" + echo "dechanyu DEC-HANYU" + echo "dechanzi GB2312" + echo "deckanji DEC-KANJI" + echo "deckorean EUC-KR" + echo "eucJP EUC-JP" + echo "eucKR EUC-KR" + echo "eucTW EUC-TW" + echo "GBK GBK" + echo "KSC5601 CP949" + echo "sdeckanji EUC-JP" + echo "SJIS SHIFT_JIS" + echo "TACTIS TIS-620" + echo "UTF-8 UTF-8" + ;; + solaris*) + echo "646 ASCII" + echo "ISO8859-1 ISO-8859-1" + echo "ISO8859-2 ISO-8859-2" + echo "ISO8859-3 ISO-8859-3" + echo "ISO8859-4 ISO-8859-4" + echo "ISO8859-5 ISO-8859-5" + echo "ISO8859-6 ISO-8859-6" + echo "ISO8859-7 ISO-8859-7" + echo "ISO8859-8 ISO-8859-8" + echo "ISO8859-9 ISO-8859-9" + echo "ISO8859-15 ISO-8859-15" + echo "koi8-r KOI8-R" + echo "ansi-1251 CP1251" + echo "BIG5 BIG5" + echo "Big5-HKSCS BIG5-HKSCS" + echo "gb2312 GB2312" + echo "GBK GBK" + echo "GB18030 GB18030" + echo "cns11643 EUC-TW" + echo "5601 EUC-KR" + echo "ko_KR.johap92 JOHAB" + echo "eucJP EUC-JP" + echo "PCK SHIFT_JIS" + echo "TIS620.2533 TIS-620" + #echo "sun_eu_greek ?" # what is this? + echo "UTF-8 UTF-8" + ;; + freebsd* | os2*) + # FreeBSD 4.2 doesn't have nl_langinfo(CODESET); therefore + # localcharset.c falls back to using the full locale name + # from the environment variables. + # Likewise for OS/2. OS/2 has XFree86 just like FreeBSD. Just + # reuse FreeBSD's locale data for OS/2. + echo "C ASCII" + echo "US-ASCII ASCII" + for l in la_LN lt_LN; do + echo "$l.ASCII ASCII" + done + for l in da_DK de_AT de_CH de_DE en_AU en_CA en_GB en_US es_ES \ + fi_FI fr_BE fr_CA fr_CH fr_FR is_IS it_CH it_IT la_LN \ + lt_LN nl_BE nl_NL no_NO pt_PT sv_SE; do + echo "$l.ISO_8859-1 ISO-8859-1" + echo "$l.DIS_8859-15 ISO-8859-15" + done + for l in cs_CZ hr_HR hu_HU la_LN lt_LN pl_PL sl_SI; do + echo "$l.ISO_8859-2 ISO-8859-2" + done + for l in la_LN lt_LT; do + echo "$l.ISO_8859-4 ISO-8859-4" + done + for l in ru_RU ru_SU; do + echo "$l.KOI8-R KOI8-R" + echo "$l.ISO_8859-5 ISO-8859-5" + echo "$l.CP866 CP866" + done + echo "uk_UA.KOI8-U KOI8-U" + echo "zh_TW.BIG5 BIG5" + echo "zh_TW.Big5 BIG5" + echo "zh_CN.EUC GB2312" + echo "ja_JP.EUC EUC-JP" + echo "ja_JP.SJIS SHIFT_JIS" + echo "ja_JP.Shift_JIS SHIFT_JIS" + echo "ko_KR.EUC EUC-KR" + ;; + netbsd*) + echo "646 ASCII" + echo "ISO8859-1 ISO-8859-1" + echo "ISO8859-2 ISO-8859-2" + echo "ISO8859-4 ISO-8859-4" + echo "ISO8859-5 ISO-8859-5" + echo "ISO8859-7 ISO-8859-7" + echo "ISO8859-13 ISO-8859-13" + echo "ISO8859-15 ISO-8859-15" + echo "eucCN GB2312" + echo "eucJP EUC-JP" + echo "eucKR EUC-KR" + echo "eucTW EUC-TW" + echo "BIG5 BIG5" + echo "SJIS SHIFT_JIS" + ;; + darwin[56]*) + # Darwin 6.8 doesn't have nl_langinfo(CODESET); therefore + # localcharset.c falls back to using the full locale name + # from the environment variables. + echo "C ASCII" + for l in en_AU en_CA en_GB en_US la_LN; do + echo "$l.US-ASCII ASCII" + done + for l in da_DK de_AT de_CH de_DE en_AU en_CA en_GB en_US es_ES \ + fi_FI fr_BE fr_CA fr_CH fr_FR is_IS it_CH it_IT nl_BE \ + nl_NL no_NO pt_PT sv_SE; do + echo "$l ISO-8859-1" + echo "$l.ISO8859-1 ISO-8859-1" + echo "$l.ISO8859-15 ISO-8859-15" + done + for l in la_LN; do + echo "$l.ISO8859-1 ISO-8859-1" + echo "$l.ISO8859-15 ISO-8859-15" + done + for l in cs_CZ hr_HR hu_HU la_LN pl_PL sl_SI; do + echo "$l.ISO8859-2 ISO-8859-2" + done + for l in la_LN lt_LT; do + echo "$l.ISO8859-4 ISO-8859-4" + done + for l in ru_RU; do + echo "$l.KOI8-R KOI8-R" + echo "$l.ISO8859-5 ISO-8859-5" + echo "$l.CP866 CP866" + done + for l in bg_BG; do + echo "$l.CP1251 CP1251" + done + echo "uk_UA.KOI8-U KOI8-U" + echo "zh_TW.BIG5 BIG5" + echo "zh_TW.Big5 BIG5" + echo "zh_CN.EUC GB2312" + echo "ja_JP.EUC EUC-JP" + echo "ja_JP.SJIS SHIFT_JIS" + echo "ko_KR.EUC EUC-KR" + ;; + darwin*) + # Darwin 7.5 has nl_langinfo(CODESET), but it is useless: + # - It returns the empty string when LANG is set to a locale of the + # form ll_CC, although ll_CC/LC_CTYPE is a symlink to an UTF-8 + # LC_CTYPE file. + # - The environment variables LANG, LC_CTYPE, LC_ALL are not set by + # the system; nl_langinfo(CODESET) returns "US-ASCII" in this case. + # - The documentation says: + # "... all code that calls BSD system routines should ensure + # that the const *char parameters of these routines are in UTF-8 + # encoding. All BSD system functions expect their string + # parameters to be in UTF-8 encoding and nothing else." + # It also says + # "An additional caveat is that string parameters for files, + # paths, and other file-system entities must be in canonical + # UTF-8. In a canonical UTF-8 Unicode string, all decomposable + # characters are decomposed ..." + # but this is not true: You can pass non-decomposed UTF-8 strings + # to file system functions, and it is the OS which will convert + # them to decomposed UTF-8 before accessing the file system. + # - The Apple Terminal application displays UTF-8 by default. + # - However, other applications are free to use different encodings: + # - xterm uses ISO-8859-1 by default. + # - TextEdit uses MacRoman by default. + # We prefer UTF-8 over decomposed UTF-8-MAC because one should + # minimize the use of decomposed Unicode. Unfortunately, through the + # Darwin file system, decomposed UTF-8 strings are leaked into user + # space nevertheless. + echo "* UTF-8" + ;; + beos*) + # BeOS has a single locale, and it has UTF-8 encoding. + echo "* UTF-8" + ;; + msdosdjgpp*) + # DJGPP 2.03 doesn't have nl_langinfo(CODESET); therefore + # localcharset.c falls back to using the full locale name + # from the environment variables. + echo "#" + echo "# The encodings given here may not all be correct." + echo "# If you find that the encoding given for your language and" + echo "# country is not the one your DOS machine actually uses, just" + echo "# correct it in this file, and send a mail to" + echo "# Juan Manuel Guerrero <juan.guerrero@gmx.de>" + echo "# and Bruno Haible <bruno@clisp.org>." + echo "#" + echo "C ASCII" + # ISO-8859-1 languages + echo "ca CP850" + echo "ca_ES CP850" + echo "da CP865" # not CP850 ?? + echo "da_DK CP865" # not CP850 ?? + echo "de CP850" + echo "de_AT CP850" + echo "de_CH CP850" + echo "de_DE CP850" + echo "en CP850" + echo "en_AU CP850" # not CP437 ?? + echo "en_CA CP850" + echo "en_GB CP850" + echo "en_NZ CP437" + echo "en_US CP437" + echo "en_ZA CP850" # not CP437 ?? + echo "es CP850" + echo "es_AR CP850" + echo "es_BO CP850" + echo "es_CL CP850" + echo "es_CO CP850" + echo "es_CR CP850" + echo "es_CU CP850" + echo "es_DO CP850" + echo "es_EC CP850" + echo "es_ES CP850" + echo "es_GT CP850" + echo "es_HN CP850" + echo "es_MX CP850" + echo "es_NI CP850" + echo "es_PA CP850" + echo "es_PY CP850" + echo "es_PE CP850" + echo "es_SV CP850" + echo "es_UY CP850" + echo "es_VE CP850" + echo "et CP850" + echo "et_EE CP850" + echo "eu CP850" + echo "eu_ES CP850" + echo "fi CP850" + echo "fi_FI CP850" + echo "fr CP850" + echo "fr_BE CP850" + echo "fr_CA CP850" + echo "fr_CH CP850" + echo "fr_FR CP850" + echo "ga CP850" + echo "ga_IE CP850" + echo "gd CP850" + echo "gd_GB CP850" + echo "gl CP850" + echo "gl_ES CP850" + echo "id CP850" # not CP437 ?? + echo "id_ID CP850" # not CP437 ?? + echo "is CP861" # not CP850 ?? + echo "is_IS CP861" # not CP850 ?? + echo "it CP850" + echo "it_CH CP850" + echo "it_IT CP850" + echo "lt CP775" + echo "lt_LT CP775" + echo "lv CP775" + echo "lv_LV CP775" + echo "nb CP865" # not CP850 ?? + echo "nb_NO CP865" # not CP850 ?? + echo "nl CP850" + echo "nl_BE CP850" + echo "nl_NL CP850" + echo "nn CP865" # not CP850 ?? + echo "nn_NO CP865" # not CP850 ?? + echo "no CP865" # not CP850 ?? + echo "no_NO CP865" # not CP850 ?? + echo "pt CP850" + echo "pt_BR CP850" + echo "pt_PT CP850" + echo "sv CP850" + echo "sv_SE CP850" + # ISO-8859-2 languages + echo "cs CP852" + echo "cs_CZ CP852" + echo "hr CP852" + echo "hr_HR CP852" + echo "hu CP852" + echo "hu_HU CP852" + echo "pl CP852" + echo "pl_PL CP852" + echo "ro CP852" + echo "ro_RO CP852" + echo "sk CP852" + echo "sk_SK CP852" + echo "sl CP852" + echo "sl_SI CP852" + echo "sq CP852" + echo "sq_AL CP852" + echo "sr CP852" # CP852 or CP866 or CP855 ?? + echo "sr_CS CP852" # CP852 or CP866 or CP855 ?? + echo "sr_YU CP852" # CP852 or CP866 or CP855 ?? + # ISO-8859-3 languages + echo "mt CP850" + echo "mt_MT CP850" + # ISO-8859-5 languages + echo "be CP866" + echo "be_BE CP866" + echo "bg CP866" # not CP855 ?? + echo "bg_BG CP866" # not CP855 ?? + echo "mk CP866" # not CP855 ?? + echo "mk_MK CP866" # not CP855 ?? + echo "ru CP866" + echo "ru_RU CP866" + echo "uk CP1125" + echo "uk_UA CP1125" + # ISO-8859-6 languages + echo "ar CP864" + echo "ar_AE CP864" + echo "ar_DZ CP864" + echo "ar_EG CP864" + echo "ar_IQ CP864" + echo "ar_IR CP864" + echo "ar_JO CP864" + echo "ar_KW CP864" + echo "ar_MA CP864" + echo "ar_OM CP864" + echo "ar_QA CP864" + echo "ar_SA CP864" + echo "ar_SY CP864" + # ISO-8859-7 languages + echo "el CP869" + echo "el_GR CP869" + # ISO-8859-8 languages + echo "he CP862" + echo "he_IL CP862" + # ISO-8859-9 languages + echo "tr CP857" + echo "tr_TR CP857" + # Japanese + echo "ja CP932" + echo "ja_JP CP932" + # Chinese + echo "zh_CN GBK" + echo "zh_TW CP950" # not CP938 ?? + # Korean + echo "kr CP949" # not CP934 ?? + echo "kr_KR CP949" # not CP934 ?? + # Thai + echo "th CP874" + echo "th_TH CP874" + # Other + echo "eo CP850" + echo "eo_EO CP850" + ;; +esac diff --git a/lib/creat-safer.c b/lib/creat-safer.c new file mode 100644 index 0000000..f4a2e59 --- /dev/null +++ b/lib/creat-safer.c @@ -0,0 +1,32 @@ +/* Invoke creat, but avoid some glitches. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "fcntl-safer.h" + +#include <fcntl.h> +#include "unistd-safer.h" + +int +creat_safer (char const *file, mode_t mode) +{ + return fd_safer (creat (file, mode)); +} diff --git a/lib/dirent_.h b/lib/dirent_.h new file mode 100644 index 0000000..94e44fc --- /dev/null +++ b/lib/dirent_.h @@ -0,0 +1,50 @@ +/* Wrapper around <dirent.h>. + Copyright (C) 2006-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_DIRENT_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <dirent.h> +#else +# include @ABSOLUTE_DIRENT_H@ +#endif + +#ifndef _GL_DIRENT_H +#define _GL_DIRENT_H + + +/* Declare overridden functions. */ + +#ifdef __cplusplus +extern "C" { +#endif + +#if @REPLACE_FCHDIR@ +# define opendir rpl_opendir +extern DIR * opendir (const char *); +# define closedir rpl_closedir +extern int closedir (DIR *); +#endif + +#ifdef __cplusplus +} +#endif + + +#endif /* _GL_DIRENT_H */ +#endif /* _GL_DIRENT_H */ diff --git a/lib/dirfd.c b/lib/dirfd.c new file mode 100644 index 0000000..06cb3c4 --- /dev/null +++ b/lib/dirfd.c @@ -0,0 +1,29 @@ +/* dirfd.c -- return the file descriptor associated with an open DIR* + + Copyright (C) 2001, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "dirfd.h" + +int +dirfd (DIR const *dir_p) +{ + return DIR_TO_FD (dir_p); +} diff --git a/lib/dirfd.h b/lib/dirfd.h new file mode 100644 index 0000000..05b7777 --- /dev/null +++ b/lib/dirfd.h @@ -0,0 +1,29 @@ +/* Declare dirfd, if necessary. + Copyright (C) 2001, 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + + Written by Jim Meyering. */ + +#include <sys/types.h> + +#include <dirent.h> + +#ifndef HAVE_DECL_DIRFD +"this configure-time declaration test was not run" +#endif +#if !HAVE_DECL_DIRFD && !defined dirfd +int dirfd (DIR const *); +#endif diff --git a/lib/dirname.c b/lib/dirname.c new file mode 100644 index 0000000..16552c6 --- /dev/null +++ b/lib/dirname.c @@ -0,0 +1,85 @@ +/* dirname.c -- return all but the last element in a file name + + Copyright (C) 1990, 1998, 2000, 2001, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "dirname.h" + +#include <string.h> +#include "xalloc.h" + +/* Return the length of the prefix of FILE that will be used by + dir_name. If FILE is in the working directory, this returns zero + even though `dir_name (FILE)' will return ".". Works properly even + if there are trailing slashes (by effectively ignoring them). */ + +size_t +dir_len (char const *file) +{ + size_t prefix_length = FILE_SYSTEM_PREFIX_LEN (file); + size_t length; + + /* Advance prefix_length beyond important leading slashes. */ + prefix_length += (prefix_length != 0 + ? (FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE + && ISSLASH (file[prefix_length])) + : (ISSLASH (file[0]) + ? ((DOUBLE_SLASH_IS_DISTINCT_ROOT + && ISSLASH (file[1]) && ! ISSLASH (file[2]) + ? 2 : 1)) + : 0)); + + /* Strip the basename and any redundant slashes before it. */ + for (length = last_component (file) - file; + prefix_length < length; length--) + if (! ISSLASH (file[length - 1])) + break; + return length; +} + + +/* In general, we can't use the builtin `dirname' function if available, + since it has different meanings in different environments. + In some environments the builtin `dirname' modifies its argument. + + Return the leading directories part of FILE, allocated with xmalloc. + Works properly even if there are trailing slashes (by effectively + ignoring them). Unlike POSIX dirname(), FILE cannot be NULL. + + If lstat (FILE) would succeed, then { chdir (dir_name (FILE)); + lstat (base_name (FILE)); } will access the same file. Likewise, + if the sequence { chdir (dir_name (FILE)); + rename (base_name (FILE), "foo"); } succeeds, you have renamed FILE + to "foo" in the same directory FILE was in. */ + +char * +dir_name (char const *file) +{ + size_t length = dir_len (file); + bool append_dot = (length == 0 + || (FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE + && length == FILE_SYSTEM_PREFIX_LEN (file) + && file[2] != '\0' && ! ISSLASH (file[2]))); + char *dir = xmalloc (length + append_dot + 1); + memcpy (dir, file, length); + if (append_dot) + dir[length++] = '.'; + dir[length] = '\0'; + return dir; +} diff --git a/lib/dirname.h b/lib/dirname.h new file mode 100644 index 0000000..91e7ed3 --- /dev/null +++ b/lib/dirname.h @@ -0,0 +1,70 @@ +/* Take file names apart into directory and base names. + + Copyright (C) 1998, 2001, 2003-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef DIRNAME_H_ +# define DIRNAME_H_ 1 + +# include <stdbool.h> +# include <stddef.h> + +# ifndef DIRECTORY_SEPARATOR +# define DIRECTORY_SEPARATOR '/' +# endif + +# ifndef ISSLASH +# define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR) +# endif + +# ifndef FILE_SYSTEM_PREFIX_LEN +# if FILE_SYSTEM_ACCEPTS_DRIVE_LETTER_PREFIX + /* This internal macro assumes ASCII, but all hosts that support drive + letters use ASCII. */ +# define _IS_DRIVE_LETTER(c) (((unsigned int) (c) | ('a' - 'A')) - 'a' \ + <= 'z' - 'a') +# define FILE_SYSTEM_PREFIX_LEN(Filename) \ + (_IS_DRIVE_LETTER ((Filename)[0]) && (Filename)[1] == ':' ? 2 : 0) +# else +# define FILE_SYSTEM_PREFIX_LEN(Filename) 0 +# endif +# endif + +# ifndef FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE +# define FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE 0 +# endif + +# ifndef DOUBLE_SLASH_IS_DISTINCT_ROOT +# define DOUBLE_SLASH_IS_DISTINCT_ROOT 0 +# endif + +# if FILE_SYSTEM_DRIVE_PREFIX_CAN_BE_RELATIVE +# define IS_ABSOLUTE_FILE_NAME(F) ISSLASH ((F)[FILE_SYSTEM_PREFIX_LEN (F)]) +# else +# define IS_ABSOLUTE_FILE_NAME(F) \ + (ISSLASH ((F)[0]) || 0 < FILE_SYSTEM_PREFIX_LEN (F)) +# endif +# define IS_RELATIVE_FILE_NAME(F) (! IS_ABSOLUTE_FILE_NAME (F)) + +char *base_name (char const *file); +char *dir_name (char const *file); +size_t base_len (char const *file); +size_t dir_len (char const *file); +char *last_component (char const *file); + +bool strip_trailing_slashes (char *file); + +#endif /* not DIRNAME_H_ */ diff --git a/lib/dup-safer.c b/lib/dup-safer.c new file mode 100644 index 0000000..7b12b61 --- /dev/null +++ b/lib/dup-safer.c @@ -0,0 +1,45 @@ +/* Invoke dup, but avoid some glitches. + + Copyright (C) 2001, 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "unistd-safer.h" + +#include <fcntl.h> + +#include <unistd.h> +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +/* Like dup, but do not return STDIN_FILENO, STDOUT_FILENO, or + STDERR_FILENO. */ + +int +dup_safer (int fd) +{ +#if defined F_DUPFD && !defined FCHDIR_REPLACEMENT + return fcntl (fd, F_DUPFD, STDERR_FILENO + 1); +#else + /* fd_safer calls us back, but eventually the recursion unwinds and + does the right thing. */ + return fd_safer (dup (fd)); +#endif +} diff --git a/lib/dup2.c b/lib/dup2.c new file mode 100644 index 0000000..8894481 --- /dev/null +++ b/lib/dup2.c @@ -0,0 +1,58 @@ +/* Duplicate an open file descriptor to a specified file descriptor. + + Copyright (C) 1999, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Paul Eggert */ + +#include <config.h> + +/* Specification. */ +#include <unistd.h> + +#include <errno.h> +#include <fcntl.h> + +#ifndef F_DUPFD +static int +dupfd (int fd, int desired_fd) +{ + int duplicated_fd = dup (fd); + if (duplicated_fd < 0 || duplicated_fd == desired_fd) + return duplicated_fd; + else + { + int r = dupfd (fd, desired_fd); + int e = errno; + close (duplicated_fd); + errno = e; + return r; + } +} +#endif + +int +dup2 (int fd, int desired_fd) +{ + if (fd == desired_fd) + return fd; + close (desired_fd); +#ifdef F_DUPFD + return fcntl (fd, F_DUPFD, desired_fd); +#else + return dupfd (fd, desired_fd); +#endif +} diff --git a/lib/error.c b/lib/error.c new file mode 100644 index 0000000..cf86343 --- /dev/null +++ b/lib/error.c @@ -0,0 +1,338 @@ +/* Error handler for noninteractive utilities + Copyright (C) 1990-1998, 2000-2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#if !_LIBC +# include <config.h> +#endif + +#include "error.h" + +#include <stdarg.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#if !_LIBC && ENABLE_NLS +# include "gettext.h" +#endif + +#ifdef _LIBC +# include <libintl.h> +# include <stdbool.h> +# include <stdint.h> +# include <wchar.h> +# define mbsrtowcs __mbsrtowcs +#endif + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +#ifndef _ +# define _(String) String +#endif + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +void (*error_print_progname) (void); + +/* This variable is incremented each time `error' is called. */ +unsigned int error_message_count; + +#ifdef _LIBC +/* In the GNU C library, there is a predefined variable for this. */ + +# define program_name program_invocation_name +# include <errno.h> +# include <limits.h> +# include <libio/libioP.h> + +/* In GNU libc we want do not want to use the common name `error' directly. + Instead make it a weak alias. */ +extern void __error (int status, int errnum, const char *message, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); +extern void __error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, + ...) + __attribute__ ((__format__ (__printf__, 5, 6)));; +# define error __error +# define error_at_line __error_at_line + +# include <libio/iolibio.h> +# define fflush(s) INTUSE(_IO_fflush) (s) +# undef putc +# define putc(c, fp) INTUSE(_IO_putc) (c, fp) + +# include <bits/libc-lock.h> + +#else /* not _LIBC */ + +# if !HAVE_DECL_STRERROR_R && STRERROR_R_CHAR_P +# ifndef HAVE_DECL_STRERROR_R +"this configure-time declaration test was not run" +# endif +char *strerror_r (); +# endif + +/* The calling program should define program_name and set it to the + name of the executing program. */ +extern char *program_name; + +# if HAVE_STRERROR_R || defined strerror_r +# define __strerror_r strerror_r +# endif /* HAVE_STRERROR_R || defined strerror_r */ +#endif /* not _LIBC */ + +static void +print_errno_message (int errnum) +{ + char const *s; + +#if defined HAVE_STRERROR_R || _LIBC + char errbuf[1024]; +# if STRERROR_R_CHAR_P || _LIBC + s = __strerror_r (errnum, errbuf, sizeof errbuf); +# else + if (__strerror_r (errnum, errbuf, sizeof errbuf) == 0) + s = errbuf; + else + s = 0; +# endif +#else + s = strerror (errnum); +#endif + +#if !_LIBC + if (! s) + s = _("Unknown system error"); +#endif + +#if _LIBC + __fxprintf (NULL, ": %s", s); +#else + fprintf (stderr, ": %s", s); +#endif +} + +static void +error_tail (int status, int errnum, const char *message, va_list args) +{ +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + { +# define ALLOCA_LIMIT 2000 + size_t len = strlen (message) + 1; + wchar_t *wmessage = NULL; + mbstate_t st; + size_t res; + const char *tmp; + bool use_malloc = false; + + while (1) + { + if (__libc_use_alloca (len * sizeof (wchar_t))) + wmessage = (wchar_t *) alloca (len * sizeof (wchar_t)); + else + { + if (!use_malloc) + wmessage = NULL; + + wchar_t *p = (wchar_t *) realloc (wmessage, + len * sizeof (wchar_t)); + if (p == NULL) + { + free (wmessage); + fputws_unlocked (L"out of memory\n", stderr); + return; + } + wmessage = p; + use_malloc = true; + } + + memset (&st, '\0', sizeof (st)); + tmp = message; + + res = mbsrtowcs (wmessage, &tmp, len, &st); + if (res != len) + break; + + if (__builtin_expect (len >= SIZE_MAX / 2, 0)) + { + /* This really should not happen if everything is fine. */ + res = (size_t) -1; + break; + } + + len *= 2; + } + + if (res == (size_t) -1) + { + /* The string cannot be converted. */ + if (use_malloc) + { + free (wmessage); + use_malloc = false; + } + wmessage = (wchar_t *) L"???"; + } + + __vfwprintf (stderr, wmessage, args); + + if (use_malloc) + free (wmessage); + } + else +#endif + vfprintf (stderr, message, args); + va_end (args); + + ++error_message_count; + if (errnum) + print_errno_message (errnum); +#if _LIBC + __fxprintf (NULL, "\n"); +#else + putc ('\n', stderr); +#endif + fflush (stderr); + if (status) + exit (status); +} + + +/* Print the program name and error message MESSAGE, which is a printf-style + format string with optional args. + If ERRNUM is nonzero, print its corresponding system error message. + Exit with status STATUS if it is nonzero. */ +void +error (int status, int errnum, const char *message, ...) +{ + va_list args; + +#if defined _LIBC && defined __libc_ptf_call + /* We do not want this call to be cut short by a thread + cancellation. Therefore disable cancellation for now. */ + int state = PTHREAD_CANCEL_ENABLE; + __libc_ptf_call (pthread_setcancelstate, (PTHREAD_CANCEL_DISABLE, &state), + 0); +#endif + + fflush (stdout); +#ifdef _LIBC + _IO_flockfile (stderr); +#endif + if (error_print_progname) + (*error_print_progname) (); + else + { +#if _LIBC + __fxprintf (NULL, "%s: ", program_name); +#else + fprintf (stderr, "%s: ", program_name); +#endif + } + + va_start (args, message); + error_tail (status, errnum, message, args); + +#ifdef _LIBC + _IO_funlockfile (stderr); +# ifdef __libc_ptf_call + __libc_ptf_call (pthread_setcancelstate, (state, NULL), 0); +# endif +#endif +} + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +int error_one_per_line; + +void +error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, ...) +{ + va_list args; + + if (error_one_per_line) + { + static const char *old_file_name; + static unsigned int old_line_number; + + if (old_line_number == line_number + && (file_name == old_file_name + || strcmp (old_file_name, file_name) == 0)) + /* Simply return and print nothing. */ + return; + + old_file_name = file_name; + old_line_number = line_number; + } + +#if defined _LIBC && defined __libc_ptf_call + /* We do not want this call to be cut short by a thread + cancellation. Therefore disable cancellation for now. */ + int state = PTHREAD_CANCEL_ENABLE; + __libc_ptf_call (pthread_setcancelstate, (PTHREAD_CANCEL_DISABLE, &state), + 0); +#endif + + fflush (stdout); +#ifdef _LIBC + _IO_flockfile (stderr); +#endif + if (error_print_progname) + (*error_print_progname) (); + else + { +#if _LIBC + __fxprintf (NULL, "%s:", program_name); +#else + fprintf (stderr, "%s:", program_name); +#endif + } + +#if _LIBC + __fxprintf (NULL, file_name != NULL ? "%s:%d: " : " ", + file_name, line_number); +#else + fprintf (stderr, file_name != NULL ? "%s:%d: " : " ", + file_name, line_number); +#endif + + va_start (args, message); + error_tail (status, errnum, message, args); + +#ifdef _LIBC + _IO_funlockfile (stderr); +# ifdef __libc_ptf_call + __libc_ptf_call (pthread_setcancelstate, (state, NULL), 0); +# endif +#endif +} + +#ifdef _LIBC +/* Make the weak alias. */ +# undef error +# undef error_at_line +weak_alias (__error, error) +weak_alias (__error_at_line, error_at_line) +#endif diff --git a/lib/error.h b/lib/error.h new file mode 100644 index 0000000..5a5f247 --- /dev/null +++ b/lib/error.h @@ -0,0 +1,66 @@ +/* Declaration for error-reporting function + Copyright (C) 1995, 1996, 1997, 2003, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _ERROR_H +#define _ERROR_H 1 + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* Print a message with `fprintf (stderr, FORMAT, ...)'; + if ERRNUM is nonzero, follow it with ": " and strerror (ERRNUM). + If STATUS is nonzero, terminate the program with `exit (STATUS)'. */ + +extern void error (int __status, int __errnum, const char *__format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); + +extern void error_at_line (int __status, int __errnum, const char *__fname, + unsigned int __lineno, const char *__format, ...) + __attribute__ ((__format__ (__printf__, 5, 6))); + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +extern void (*error_print_progname) (void); + +/* This variable is incremented each time `error' is called. */ +extern unsigned int error_message_count; + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +extern int error_one_per_line; + +#ifdef __cplusplus +} +#endif + +#endif /* error.h */ diff --git a/lib/exclude.c b/lib/exclude.c new file mode 100644 index 0000000..ed34d0f --- /dev/null +++ b/lib/exclude.c @@ -0,0 +1,275 @@ +/* exclude.c -- exclude file names + + Copyright (C) 1992, 1993, 1994, 1997, 1999, 2000, 2001, 2002, 2003, + 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#include <config.h> + +#include <stdbool.h> + +#include <ctype.h> +#include <errno.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "exclude.h" +#include "fnmatch.h" +#include "xalloc.h" +#include "verify.h" + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +/* Non-GNU systems lack these options, so we don't need to check them. */ +#ifndef FNM_CASEFOLD +# define FNM_CASEFOLD 0 +#endif +#ifndef FNM_EXTMATCH +# define FNM_EXTMATCH 0 +#endif +#ifndef FNM_LEADING_DIR +# define FNM_LEADING_DIR 0 +#endif + +verify (((EXCLUDE_ANCHORED | EXCLUDE_INCLUDE | EXCLUDE_WILDCARDS) + & (FNM_PATHNAME | FNM_NOESCAPE | FNM_PERIOD | FNM_LEADING_DIR + | FNM_CASEFOLD | FNM_EXTMATCH)) + == 0); + +/* An exclude pattern-options pair. The options are fnmatch options + ORed with EXCLUDE_* options. */ + +struct patopts + { + char const *pattern; + int options; + }; + +/* An exclude list, of pattern-options pairs. */ + +struct exclude + { + struct patopts *exclude; + size_t exclude_alloc; + size_t exclude_count; + }; + +/* Return a newly allocated and empty exclude list. */ + +struct exclude * +new_exclude (void) +{ + return xzalloc (sizeof *new_exclude ()); +} + +/* Free the storage associated with an exclude list. */ + +void +free_exclude (struct exclude *ex) +{ + free (ex->exclude); + free (ex); +} + +/* Return zero if PATTERN matches F, obeying OPTIONS, except that + (unlike fnmatch) wildcards are disabled in PATTERN. */ + +static int +fnmatch_no_wildcards (char const *pattern, char const *f, int options) +{ + if (! (options & FNM_LEADING_DIR)) + return ((options & FNM_CASEFOLD) + ? mbscasecmp (pattern, f) + : strcmp (pattern, f)); + else if (! (options & FNM_CASEFOLD)) + { + size_t patlen = strlen (pattern); + int r = strncmp (pattern, f, patlen); + if (! r) + { + r = f[patlen]; + if (r == '/') + r = 0; + } + return r; + } + else + { + /* Walk through a copy of F, seeing whether P matches any prefix + of F. + + FIXME: This is an O(N**2) algorithm; it should be O(N). + Also, the copy should not be necessary. However, fixing this + will probably involve a change to the mbs* API. */ + + char *fcopy = xstrdup (f); + char *p; + int r; + for (p = fcopy; ; *p++ = '/') + { + p = strchr (p, '/'); + if (p) + *p = '\0'; + r = mbscasecmp (pattern, fcopy); + if (!p || r <= 0) + break; + } + free (fcopy); + return r; + } +} + +bool +exclude_fnmatch (char const *pattern, char const *f, int options) +{ + int (*matcher) (char const *, char const *, int) = + (options & EXCLUDE_WILDCARDS + ? fnmatch + : fnmatch_no_wildcards); + bool matched = ((*matcher) (pattern, f, options) == 0); + char const *p; + + if (! (options & EXCLUDE_ANCHORED)) + for (p = f; *p && ! matched; p++) + if (*p == '/' && p[1] != '/') + matched = ((*matcher) (pattern, p + 1, options) == 0); + + return matched; +} + +/* Return true if EX excludes F. */ + +bool +excluded_file_name (struct exclude const *ex, char const *f) +{ + size_t exclude_count = ex->exclude_count; + + /* If no options are given, the default is to include. */ + if (exclude_count == 0) + return false; + else + { + struct patopts const *exclude = ex->exclude; + size_t i; + + /* Otherwise, the default is the opposite of the first option. */ + bool excluded = !! (exclude[0].options & EXCLUDE_INCLUDE); + + /* Scan through the options, seeing whether they change F from + excluded to included or vice versa. */ + for (i = 0; i < exclude_count; i++) + { + char const *pattern = exclude[i].pattern; + int options = exclude[i].options; + if (excluded == !! (options & EXCLUDE_INCLUDE)) + excluded ^= exclude_fnmatch (pattern, f, options); + } + + return excluded; + } +} + +/* Append to EX the exclusion PATTERN with OPTIONS. */ + +void +add_exclude (struct exclude *ex, char const *pattern, int options) +{ + struct patopts *patopts; + + if (ex->exclude_count == ex->exclude_alloc) + ex->exclude = x2nrealloc (ex->exclude, &ex->exclude_alloc, + sizeof *ex->exclude); + + patopts = &ex->exclude[ex->exclude_count++]; + patopts->pattern = pattern; + patopts->options = options; +} + +/* Use ADD_FUNC to append to EX the patterns in FILE_NAME, each with + OPTIONS. LINE_END terminates each pattern in the file. If + LINE_END is a space character, ignore trailing spaces and empty + lines in FILE. Return -1 on failure, 0 on success. */ + +int +add_exclude_file (void (*add_func) (struct exclude *, char const *, int), + struct exclude *ex, char const *file_name, int options, + char line_end) +{ + bool use_stdin = file_name[0] == '-' && !file_name[1]; + FILE *in; + char *buf = NULL; + char *p; + char const *pattern; + char const *lim; + size_t buf_alloc = 0; + size_t buf_count = 0; + int c; + int e = 0; + + if (use_stdin) + in = stdin; + else if (! (in = fopen (file_name, "r"))) + return -1; + + while ((c = getc (in)) != EOF) + { + if (buf_count == buf_alloc) + buf = x2realloc (buf, &buf_alloc); + buf[buf_count++] = c; + } + + if (ferror (in)) + e = errno; + + if (!use_stdin && fclose (in) != 0) + e = errno; + + buf = xrealloc (buf, buf_count + 1); + buf[buf_count] = line_end; + lim = buf + buf_count + ! (buf_count == 0 || buf[buf_count - 1] == line_end); + pattern = buf; + + for (p = buf; p < lim; p++) + if (*p == line_end) + { + char *pattern_end = p; + + if (isspace ((unsigned char) line_end)) + { + for (; ; pattern_end--) + if (pattern_end == pattern) + goto next_pattern; + else if (! isspace ((unsigned char) pattern_end[-1])) + break; + } + + *pattern_end = '\0'; + (*add_func) (ex, pattern, options); + + next_pattern: + pattern = p + 1; + } + + errno = e; + return e ? -1 : 0; +} diff --git a/lib/exclude.h b/lib/exclude.h new file mode 100644 index 0000000..203e384 --- /dev/null +++ b/lib/exclude.h @@ -0,0 +1,44 @@ +/* exclude.h -- declarations for excluding file names + + Copyright (C) 1992, 1993, 1994, 1997, 1999, 2001, 2002, 2003, 2005, + 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +/* Exclude options, which can be ORed with fnmatch options. */ + +/* Patterns must match the start of file names, instead of matching + anywhere after a '/'. */ +#define EXCLUDE_ANCHORED (1 << 30) + +/* Include instead of exclude. */ +#define EXCLUDE_INCLUDE (1 << 29) + +/* '?', '*', '[', and '\\' are special in patterns. Without this + option, these characters are ordinary and fnmatch is not used. */ +#define EXCLUDE_WILDCARDS (1 << 28) + +struct exclude; + +struct exclude *new_exclude (void); +void free_exclude (struct exclude *); +void add_exclude (struct exclude *, char const *, int); +int add_exclude_file (void (*) (struct exclude *, char const *, int), + struct exclude *, char const *, int, char); +bool excluded_file_name (struct exclude const *, char const *); +bool exclude_fnmatch (char const *pattern, char const *f, int options); diff --git a/lib/exitfail.c b/lib/exitfail.c new file mode 100644 index 0000000..373d325 --- /dev/null +++ b/lib/exitfail.c @@ -0,0 +1,26 @@ +/* Failure exit status + + Copyright (C) 2002, 2003, 2005, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "exitfail.h" + +#include <stdlib.h> + +int volatile exit_failure = EXIT_FAILURE; diff --git a/lib/exitfail.h b/lib/exitfail.h new file mode 100644 index 0000000..e46cf9c --- /dev/null +++ b/lib/exitfail.h @@ -0,0 +1,20 @@ +/* Failure exit status + + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +extern int volatile exit_failure; diff --git a/lib/fchdir.c b/lib/fchdir.c new file mode 100644 index 0000000..e3ec2fc --- /dev/null +++ b/lib/fchdir.c @@ -0,0 +1,279 @@ +/* fchdir replacement. + Copyright (C) 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <unistd.h> + +#include <errno.h> +#include <fcntl.h> +#include <stdarg.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <dirent.h> + +#include "canonicalize.h" +#include "dirfd.h" + +/* This replacement assumes that a directory is not renamed while opened + through a file descriptor. */ + +/* Array of file descriptors opened. If it points to a directory, it stores + info about this directory; otherwise it stores an errno value of ENOTDIR. */ +typedef struct +{ + char *name; /* Absolute name of the directory, or NULL. */ + int saved_errno; /* If name == NULL: The error code describing the failure + reason. */ +} dir_info_t; +static dir_info_t *dirs; +static size_t dirs_allocated; + +/* Try to ensure dirs has enough room for a slot at index fd. */ +static void +ensure_dirs_slot (size_t fd) +{ + if (fd >= dirs_allocated) + { + size_t new_allocated; + dir_info_t *new_dirs; + size_t i; + + new_allocated = 2 * dirs_allocated + 1; + if (new_allocated <= fd) + new_allocated = fd + 1; + new_dirs = + (dirs != NULL + ? (dir_info_t *) realloc (dirs, new_allocated * sizeof (dir_info_t)) + : (dir_info_t *) malloc (new_allocated * sizeof (dir_info_t))); + if (new_dirs != NULL) + { + for (i = dirs_allocated; i < new_allocated; i++) + { + new_dirs[i].name = NULL; + new_dirs[i].saved_errno = ENOTDIR; + } + dirs = new_dirs; + dirs_allocated = new_allocated; + } + } +} + +/* Override open() and close(), to keep track of the open file descriptors. */ + +int +close (int fd) +#undef close +{ + int retval = close (fd); + + if (retval >= 0 && fd >= 0 && fd < dirs_allocated) + { + if (dirs[fd].name != NULL) + free (dirs[fd].name); + dirs[fd].name = NULL; + dirs[fd].saved_errno = ENOTDIR; + } + return retval; +} + +int +open (const char *filename, int flags, ...) +#undef open +{ + mode_t mode; + int fd; + struct stat statbuf; + + mode = 0; + if (flags & O_CREAT) + { + va_list arg; + va_start (arg, flags); + + /* If mode_t is narrower than int, use the promoted type (int), + not mode_t. Use sizeof to guess whether mode_t is narrower; + we don't know of any practical counterexamples. */ + mode = (sizeof (mode_t) < sizeof (int) + ? va_arg (arg, int) + : va_arg (arg, mode_t)); + + va_end (arg); + } + fd = open (filename, flags, mode); + if (fd >= 0) + { + ensure_dirs_slot (fd); + if (fd < dirs_allocated + && fstat (fd, &statbuf) >= 0 && S_ISDIR (statbuf.st_mode)) + { + dirs[fd].name = canonicalize_file_name (filename); + if (dirs[fd].name == NULL) + dirs[fd].saved_errno = errno; + } + } + return fd; +} + +/* Override opendir() and closedir(), to keep track of the open file + descriptors. Needed because there is a function dirfd(). */ + +int +closedir (DIR *dp) +#undef closedir +{ + int fd = dirfd (dp); + int retval = closedir (dp); + + if (retval >= 0 && fd >= 0 && fd < dirs_allocated) + { + if (dirs[fd].name != NULL) + free (dirs[fd].name); + dirs[fd].name = NULL; + dirs[fd].saved_errno = ENOTDIR; + } + return retval; +} + +DIR * +opendir (const char *filename) +#undef opendir +{ + DIR *dp; + + dp = opendir (filename); + if (dp != NULL) + { + int fd = dirfd (dp); + if (fd >= 0) + { + ensure_dirs_slot (fd); + if (fd < dirs_allocated) + { + dirs[fd].name = canonicalize_file_name (filename); + if (dirs[fd].name == NULL) + dirs[fd].saved_errno = errno; + } + } + } + return dp; +} + +/* Override dup() and dup2(), to keep track of open file descriptors. */ + +int +dup (int oldfd) +#undef dup +{ + int newfd = dup (oldfd); + + if (oldfd >= 0 && newfd >= 0) + { + ensure_dirs_slot (newfd); + if (newfd < dirs_allocated) + { + if (oldfd < dirs_allocated) + { + if (dirs[oldfd].name != NULL) + { + dirs[newfd].name = strdup (dirs[oldfd].name); + if (dirs[newfd].name == NULL) + dirs[newfd].saved_errno = ENOMEM; + } + else + { + dirs[newfd].name = NULL; + dirs[newfd].saved_errno = dirs[oldfd].saved_errno; + } + } + else + { + dirs[newfd].name = NULL; + dirs[newfd].saved_errno = ENOMEM; + } + } + } + return newfd; +} + +int +dup2 (int oldfd, int newfd) +#undef dup2 +{ + int retval = dup2 (oldfd, newfd); + + if (retval >= 0 && oldfd >= 0 && newfd >= 0 && newfd != oldfd) + { + ensure_dirs_slot (newfd); + if (newfd < dirs_allocated) + { + if (oldfd < dirs_allocated) + { + if (dirs[oldfd].name != NULL) + { + dirs[newfd].name = strdup (dirs[oldfd].name); + if (dirs[newfd].name == NULL) + dirs[newfd].saved_errno = ENOMEM; + } + else + { + dirs[newfd].name = NULL; + dirs[newfd].saved_errno = dirs[oldfd].saved_errno; + } + } + else + { + dirs[newfd].name = NULL; + dirs[newfd].saved_errno = ENOMEM; + } + } + } + return retval; +} + +/* Implement fchdir() in terms of chdir(). */ + +int +fchdir (int fd) +{ + if (fd >= 0) + { + if (fd < dirs_allocated) + { + if (dirs[fd].name != NULL) + return chdir (dirs[fd].name); + else + { + errno = dirs[fd].saved_errno; + return -1; + } + } + else + { + errno = ENOMEM; + return -1; + } + } + else + { + errno = EBADF; + return -1; + } +} diff --git a/lib/fchmodat.c b/lib/fchmodat.c new file mode 100644 index 0000000..2598b70 --- /dev/null +++ b/lib/fchmodat.c @@ -0,0 +1,50 @@ +/* Change the protections of file relative to an open directory. + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +#include "openat.h" +#include "dirname.h" /* solely for definition of IS_ABSOLUTE_FILE_NAME */ +#include "save-cwd.h" +#include "openat-priv.h" + +#ifndef HAVE_LCHMOD +/* Use a different name, to avoid conflicting with any + system-supplied declaration. */ +# undef lchmod +# define lchmod lchmod_rpl +static int lchmod (char const *f, mode_t m) { errno = ENOSYS; return -1; } +#endif + +/* Solaris 10 has no function like this. + Invoke chmod or lchmod on file, FILE, using mode MODE, in the directory + open on descriptor FD. If possible, do it without changing the + working directory. Otherwise, resort to using save_cwd/fchdir, + then mkdir/restore_cwd. If either the save_cwd or the restore_cwd + fails, then give a diagnostic and exit nonzero. + Note that an attempt to use a FLAG value of AT_SYMLINK_NOFOLLOW + on a system without lchmod support causes this function to fail. */ + +#define AT_FUNC_NAME fchmodat +#define AT_FUNC_F1 lchmod +#define AT_FUNC_F2 chmod +#define AT_FUNC_USE_F1_COND flag == AT_SYMLINK_NOFOLLOW +#define AT_FUNC_POST_FILE_PARAM_DECLS , mode_t mode, int flag +#define AT_FUNC_POST_FILE_ARGS , mode +#include "at-func.c" diff --git a/lib/fchown-stub.c b/lib/fchown-stub.c new file mode 100644 index 0000000..6be750b --- /dev/null +++ b/lib/fchown-stub.c @@ -0,0 +1,16 @@ +#include <config.h> + +#include <sys/types.h> +#include <errno.h> + +/* A trivial substitute for `fchown'. + + DJGPP 2.03 and earlier (and perhaps later) don't have `fchown', + so we pretend no-one has permission for this operation. */ + +int +fchown (int fd, uid_t uid, gid_t gid) +{ + errno = EPERM; + return -1; +} diff --git a/lib/fchownat.c b/lib/fchownat.c new file mode 100644 index 0000000..801c92a --- /dev/null +++ b/lib/fchownat.c @@ -0,0 +1,50 @@ +/* This function serves as replacement for a missing fchownat function, + as well as a work around for the fchownat bug in glibc-2.4: + <http://lists.ubuntu.com/archives/ubuntu-users/2006-September/093218.html> + when the buggy fchownat-with-AT_SYMLINK_NOFOLLOW operates on a symlink, it + mistakenly affects the symlink referent, rather than the symlink itself. + + Copyright (C) 2006-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +#include "openat.h" + +#include <unistd.h> + +#include "dirname.h" /* solely for definition of IS_ABSOLUTE_FILE_NAME */ +#include "lchown.h" +#include "save-cwd.h" +#include "openat-priv.h" + +/* Replacement for Solaris' function by the same name. + Invoke chown or lchown on file, FILE, using OWNER and GROUP, in the + directory open on descriptor FD. If FLAG is AT_SYMLINK_NOFOLLOW, then + use lchown, otherwise, use chown. If possible, do it without changing + the working directory. Otherwise, resort to using save_cwd/fchdir, + then mkdir/restore_cwd. If either the save_cwd or the restore_cwd + fails, then give a diagnostic and exit nonzero. */ + +#define AT_FUNC_NAME fchownat +#define AT_FUNC_F1 lchown +#define AT_FUNC_F2 chown +#define AT_FUNC_USE_F1_COND flag == AT_SYMLINK_NOFOLLOW +#define AT_FUNC_POST_FILE_PARAM_DECLS , uid_t owner, gid_t group, int flag +#define AT_FUNC_POST_FILE_ARGS , owner, group +#include "at-func.c" diff --git a/lib/fcntl--.h b/lib/fcntl--.h new file mode 100644 index 0000000..51b869e --- /dev/null +++ b/lib/fcntl--.h @@ -0,0 +1,28 @@ +/* Like fcntl.h, but redefine some names to avoid glitches. + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <fcntl.h> +#include "fcntl-safer.h" + +#undef open +#define open open_safer + +#undef creat +#define creat creat_safer diff --git a/lib/fcntl-safer.h b/lib/fcntl-safer.h new file mode 100644 index 0000000..cab6aab --- /dev/null +++ b/lib/fcntl-safer.h @@ -0,0 +1,24 @@ +/* Invoke fcntl-like functions, but avoid some glitches. + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <sys/types.h> + +int open_safer (char const *, int, ...); +int creat_safer (char const *, mode_t); diff --git a/lib/fcntl_.h b/lib/fcntl_.h new file mode 100644 index 0000000..1330cde --- /dev/null +++ b/lib/fcntl_.h @@ -0,0 +1,125 @@ +/* Like <fcntl.h>, but with non-working flags defined to 0. + + Copyright (C) 2006-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Paul Eggert */ + +#ifndef _GL_FCNTL_H + +#include <sys/types.h> +#include <sys/stat.h> +#include <unistd.h> +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <fcntl.h> +#else +# include @ABSOLUTE_FCNTL_H@ +#endif + +#ifndef _GL_FCNTL_H +#define _GL_FCNTL_H + + +/* Declare overridden functions. */ + +#ifdef __cplusplus +extern "C" { +#endif + +#ifdef FCHDIR_REPLACEMENT +# define open rpl_open +extern int open (const char *, int, ...); +#endif + +#ifdef __cplusplus +} +#endif + + +/* Fix up the O_* macros. */ + +#if !defined O_DIRECT && defined O_DIRECTIO +/* Tru64 spells it `O_DIRECTIO'. */ +# define O_DIRECT O_DIRECTIO +#endif + +#ifndef O_DIRECT +# define O_DIRECT 0 +#endif + +#ifndef O_DIRECTORY +# define O_DIRECTORY 0 +#endif + +#ifndef O_DSYNC +# define O_DSYNC 0 +#endif + +#ifndef O_NDELAY +# define O_NDELAY 0 +#endif + +#ifndef O_NOATIME +# define O_NOATIME 0 +#endif + +#ifndef O_NONBLOCK +# define O_NONBLOCK O_NDELAY +#endif + +#ifndef O_NOCTTY +# define O_NOCTTY 0 +#endif + +#ifndef O_NOFOLLOW +# define O_NOFOLLOW 0 +#endif + +#ifndef O_NOLINKS +# define O_NOLINKS 0 +#endif + +#ifndef O_RSYNC +# define O_RSYNC 0 +#endif + +#ifndef O_SYNC +# define O_SYNC 0 +#endif + +/* For systems that distinguish between text and binary I/O. + O_BINARY is usually declared in fcntl.h */ +#if !defined O_BINARY && defined _O_BINARY + /* For MSC-compatible compilers. */ +# define O_BINARY _O_BINARY +# define O_TEXT _O_TEXT +#endif + +#ifdef __BEOS__ + /* BeOS 5 has O_BINARY and O_TEXT, but they have no effect. */ +# undef O_BINARY +# undef O_TEXT +#endif + +#ifndef O_BINARY +# define O_BINARY 0 +# define O_TEXT 0 +#endif + + +#endif /* _GL_FCNTL_H */ +#endif /* _GL_FCNTL_H */ diff --git a/lib/fd-safer.c b/lib/fd-safer.c new file mode 100644 index 0000000..256bfa4 --- /dev/null +++ b/lib/fd-safer.c @@ -0,0 +1,57 @@ +/* Return a safer copy of a file descriptor. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "unistd-safer.h" + +#include <errno.h> + +#include <unistd.h> +#ifndef STDIN_FILENO +# define STDIN_FILENO 0 +#endif +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +/* Return FD, unless FD would be a copy of standard input, output, or + error; in that case, return a duplicate of FD, closing FD. On + failure to duplicate, close FD, set errno, and return -1. Preserve + errno if FD is negative, so that the caller can always inspect + errno when the returned value is negative. + + This function is usefully wrapped around functions that return file + descriptors, e.g., fd_safer (open ("file", O_RDONLY)). */ + +int +fd_safer (int fd) +{ + if (STDIN_FILENO <= fd && fd <= STDERR_FILENO) + { + int f = dup_safer (fd); + int e = errno; + close (fd); + errno = e; + fd = f; + } + + return fd; +} diff --git a/lib/fileblocks.c b/lib/fileblocks.c new file mode 100644 index 0000000..4024d1e --- /dev/null +++ b/lib/fileblocks.c @@ -0,0 +1,75 @@ +/* Convert file size to number of blocks on System V-like machines. + + Copyright (C) 1990, 1997, 1998, 1999, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Brian L. Matthews, blm@6sceng.UUCP. */ + +#include <config.h> + +#include <sys/types.h> + +#if HAVE_SYS_PARAM_H +# include <sys/param.h> +#endif + +#if !HAVE_STRUCT_STAT_ST_BLOCKS && !defined _POSIX_SOURCE && defined BSIZE + +# include <unistd.h> + +# ifndef NINDIR + +# if defined __DJGPP__ +typedef long daddr_t; /* for disk address */ +# endif + +/* Some SysV's, like Irix, seem to lack this. Hope it's correct. */ +/* Number of inode pointers per indirect block. */ +# define NINDIR (BSIZE / sizeof (daddr_t)) +# endif /* !NINDIR */ + +/* Number of direct block addresses in an inode. */ +# define NDIR 10 + +/* Return the number of 512-byte blocks in a file of SIZE bytes. */ + +off_t +st_blocks (off_t size) +{ + off_t datablks = size / 512 + (size % 512 != 0); + off_t indrblks = 0; + + if (datablks > NDIR) + { + indrblks = (datablks - NDIR - 1) / NINDIR + 1; + + if (datablks > NDIR + NINDIR) + { + indrblks += (datablks - NDIR - NINDIR - 1) / (NINDIR * NINDIR) + 1; + + if (datablks > NDIR + NINDIR + NINDIR * NINDIR) + indrblks++; + } + } + + return datablks + indrblks; +} +#else +/* This declaration is solely to ensure that after preprocessing + this file is never empty. */ +typedef int textutils_fileblocks_unused; +#endif diff --git a/lib/float+.h b/lib/float+.h new file mode 100644 index 0000000..4de25a9 --- /dev/null +++ b/lib/float+.h @@ -0,0 +1,148 @@ +/* Supplemental information about the floating-point formats. + Copyright (C) 2007 Free Software Foundation, Inc. + Written by Bruno Haible <bruno@clisp.org>, 2007. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _FLOATPLUS_H +#define _FLOATPLUS_H + +#include <float.h> +#include <limits.h> + +/* Number of bits in the mantissa of a floating-point number, including the + "hidden bit". */ +#if FLT_RADIX == 2 +# define FLT_MANT_BIT FLT_MANT_DIG +# define DBL_MANT_BIT DBL_MANT_DIG +# define LDBL_MANT_BIT LDBL_MANT_DIG +#elif FLT_RADIX == 4 +# define FLT_MANT_BIT (FLT_MANT_DIG * 2) +# define DBL_MANT_BIT (DBL_MANT_DIG * 2) +# define LDBL_MANT_BIT (LDBL_MANT_DIG * 2) +#elif FLT_RADIX == 16 +# define FLT_MANT_BIT (FLT_MANT_DIG * 4) +# define DBL_MANT_BIT (DBL_MANT_DIG * 4) +# define LDBL_MANT_BIT (LDBL_MANT_DIG * 4) +#endif + +/* Bit mask that can be used to mask the exponent, as an unsigned number. */ +#define FLT_EXP_MASK ((FLT_MAX_EXP - FLT_MIN_EXP) | 7) +#define DBL_EXP_MASK ((DBL_MAX_EXP - DBL_MIN_EXP) | 7) +#define LDBL_EXP_MASK ((LDBL_MAX_EXP - LDBL_MIN_EXP) | 7) + +/* Number of bits used for the exponent of a floating-point number, including + the exponent's sign. */ +#define FLT_EXP_BIT \ + (FLT_EXP_MASK < 0x100 ? 8 : \ + FLT_EXP_MASK < 0x200 ? 9 : \ + FLT_EXP_MASK < 0x400 ? 10 : \ + FLT_EXP_MASK < 0x800 ? 11 : \ + FLT_EXP_MASK < 0x1000 ? 12 : \ + FLT_EXP_MASK < 0x2000 ? 13 : \ + FLT_EXP_MASK < 0x4000 ? 14 : \ + FLT_EXP_MASK < 0x8000 ? 15 : \ + FLT_EXP_MASK < 0x10000 ? 16 : \ + FLT_EXP_MASK < 0x20000 ? 17 : \ + FLT_EXP_MASK < 0x40000 ? 18 : \ + FLT_EXP_MASK < 0x80000 ? 19 : \ + FLT_EXP_MASK < 0x100000 ? 20 : \ + FLT_EXP_MASK < 0x200000 ? 21 : \ + FLT_EXP_MASK < 0x400000 ? 22 : \ + FLT_EXP_MASK < 0x800000 ? 23 : \ + FLT_EXP_MASK < 0x1000000 ? 24 : \ + FLT_EXP_MASK < 0x2000000 ? 25 : \ + FLT_EXP_MASK < 0x4000000 ? 26 : \ + FLT_EXP_MASK < 0x8000000 ? 27 : \ + FLT_EXP_MASK < 0x10000000 ? 28 : \ + FLT_EXP_MASK < 0x20000000 ? 29 : \ + FLT_EXP_MASK < 0x40000000 ? 30 : \ + FLT_EXP_MASK <= 0x7fffffff ? 31 : \ + 32) +#define DBL_EXP_BIT \ + (DBL_EXP_MASK < 0x100 ? 8 : \ + DBL_EXP_MASK < 0x200 ? 9 : \ + DBL_EXP_MASK < 0x400 ? 10 : \ + DBL_EXP_MASK < 0x800 ? 11 : \ + DBL_EXP_MASK < 0x1000 ? 12 : \ + DBL_EXP_MASK < 0x2000 ? 13 : \ + DBL_EXP_MASK < 0x4000 ? 14 : \ + DBL_EXP_MASK < 0x8000 ? 15 : \ + DBL_EXP_MASK < 0x10000 ? 16 : \ + DBL_EXP_MASK < 0x20000 ? 17 : \ + DBL_EXP_MASK < 0x40000 ? 18 : \ + DBL_EXP_MASK < 0x80000 ? 19 : \ + DBL_EXP_MASK < 0x100000 ? 20 : \ + DBL_EXP_MASK < 0x200000 ? 21 : \ + DBL_EXP_MASK < 0x400000 ? 22 : \ + DBL_EXP_MASK < 0x800000 ? 23 : \ + DBL_EXP_MASK < 0x1000000 ? 24 : \ + DBL_EXP_MASK < 0x2000000 ? 25 : \ + DBL_EXP_MASK < 0x4000000 ? 26 : \ + DBL_EXP_MASK < 0x8000000 ? 27 : \ + DBL_EXP_MASK < 0x10000000 ? 28 : \ + DBL_EXP_MASK < 0x20000000 ? 29 : \ + DBL_EXP_MASK < 0x40000000 ? 30 : \ + DBL_EXP_MASK <= 0x7fffffff ? 31 : \ + 32) +#define LDBL_EXP_BIT \ + (LDBL_EXP_MASK < 0x100 ? 8 : \ + LDBL_EXP_MASK < 0x200 ? 9 : \ + LDBL_EXP_MASK < 0x400 ? 10 : \ + LDBL_EXP_MASK < 0x800 ? 11 : \ + LDBL_EXP_MASK < 0x1000 ? 12 : \ + LDBL_EXP_MASK < 0x2000 ? 13 : \ + LDBL_EXP_MASK < 0x4000 ? 14 : \ + LDBL_EXP_MASK < 0x8000 ? 15 : \ + LDBL_EXP_MASK < 0x10000 ? 16 : \ + LDBL_EXP_MASK < 0x20000 ? 17 : \ + LDBL_EXP_MASK < 0x40000 ? 18 : \ + LDBL_EXP_MASK < 0x80000 ? 19 : \ + LDBL_EXP_MASK < 0x100000 ? 20 : \ + LDBL_EXP_MASK < 0x200000 ? 21 : \ + LDBL_EXP_MASK < 0x400000 ? 22 : \ + LDBL_EXP_MASK < 0x800000 ? 23 : \ + LDBL_EXP_MASK < 0x1000000 ? 24 : \ + LDBL_EXP_MASK < 0x2000000 ? 25 : \ + LDBL_EXP_MASK < 0x4000000 ? 26 : \ + LDBL_EXP_MASK < 0x8000000 ? 27 : \ + LDBL_EXP_MASK < 0x10000000 ? 28 : \ + LDBL_EXP_MASK < 0x20000000 ? 29 : \ + LDBL_EXP_MASK < 0x40000000 ? 30 : \ + LDBL_EXP_MASK <= 0x7fffffff ? 31 : \ + 32) + +/* Number of bits used for a floating-point number: the mantissa (not + counting the "hidden bit", since it may or may not be explicit), the + exponent, and the sign. */ +#define FLT_TOTAL_BIT ((FLT_MANT_BIT - 1) + FLT_EXP_BIT + 1) +#define DBL_TOTAL_BIT ((DBL_MANT_BIT - 1) + DBL_EXP_BIT + 1) +#define LDBL_TOTAL_BIT ((LDBL_MANT_BIT - 1) + LDBL_EXP_BIT + 1) + +/* Number of bytes used for a floating-point number. + This can be smaller than the 'sizeof'. For example, on i386 systems, + 'long double' most often have LDBL_MANT_BIT = 64, LDBL_EXP_BIT = 16, hence + LDBL_TOTAL_BIT = 80 bits, i.e. 10 bytes of consecutive memory, but + sizeof (long double) = 12 or = 16. */ +#define SIZEOF_FLT ((FLT_TOTAL_BIT + CHAR_BIT - 1) / CHAR_BIT) +#define SIZEOF_DBL ((DBL_TOTAL_BIT + CHAR_BIT - 1) / CHAR_BIT) +#define SIZEOF_LDBL ((LDBL_TOTAL_BIT + CHAR_BIT - 1) / CHAR_BIT) + +/* Verify that SIZEOF_FLT <= sizeof (float) etc. */ +typedef int verify_sizeof_flt[2 * (SIZEOF_FLT <= sizeof (float)) - 1]; +typedef int verify_sizeof_dbl[2 * (SIZEOF_DBL <= sizeof (double)) - 1]; +typedef int verify_sizeof_ldbl[2 * (SIZEOF_LDBL <= sizeof (long double)) - 1]; + +#endif /* _FLOATPLUS_H */ diff --git a/lib/float_.h b/lib/float_.h new file mode 100644 index 0000000..d898d85 --- /dev/null +++ b/lib/float_.h @@ -0,0 +1,63 @@ +/* A correct <float.h>. + + Copyright (C) 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_FLOAT_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <float.h> +#else +# include @ABSOLUTE_FLOAT_H@ +#endif + +#ifndef _GL_FLOAT_H +#define _GL_FLOAT_H + +/* 'long double' properties. */ +#if defined __i386__ && defined __BEOS__ +/* Number of mantissa units, in base FLT_RADIX. */ +# undef LDBL_MANT_DIG +# define LDBL_MANT_DIG 64 +/* Number of decimal digits that is sufficient for representing a number. */ +# undef LDBL_DIG +# define LDBL_DIG 18 +/* x-1 where x is the smallest representable number > 1. */ +# undef LDBL_EPSILON +# define LDBL_EPSILON 1.0842021724855044340E-19L +/* Minimum e such that FLT_RADIX^(e-1) is a normalized number. */ +# undef LDBL_MIN_EXP +# define LDBL_MIN_EXP (-16381) +/* Maximum e such that FLT_RADIX^(e-1) is a representable finite number. */ +# undef LDBL_MAX_EXP +# define LDBL_MAX_EXP 16384 +/* Minimum positive normalized number. */ +# undef LDBL_MIN +# define LDBL_MIN 3.3621031431120935063E-4932L +/* Maximum representable finite number. */ +# undef LDBL_MAX +# define LDBL_MAX 1.1897314953572317650E+4932L +/* Minimum e such that 10^e is in the range of normalized numbers. */ +# undef LDBL_MIN_10_EXP +# define LDBL_MIN_10_EXP (-4931) +/* Maximum e such that 10^e is in the range of representable finite numbers. */ +# undef LDBL_MAX_10_EXP +# define LDBL_MAX_10_EXP 4932 +#endif + +#endif /* _GL_FLOAT_H */ +#endif /* _GL_FLOAT_H */ diff --git a/lib/fnmatch.c b/lib/fnmatch.c new file mode 100644 index 0000000..02dd365 --- /dev/null +++ b/lib/fnmatch.c @@ -0,0 +1,354 @@ +/* Copyright (C) 1991,1992,1993,1996,1997,1998,1999,2000,2001,2002,2003,2004,2005,2006,2007 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +#endif + +/* Enable GNU extensions in fnmatch.h. */ +#ifndef _GNU_SOURCE +# define _GNU_SOURCE 1 +#endif + +#if ! defined __builtin_expect && __GNUC__ < 3 +# define __builtin_expect(expr, expected) (expr) +#endif + +#include <fnmatch.h> + +#include <alloca.h> +#include <assert.h> +#include <ctype.h> +#include <errno.h> +#include <stddef.h> +#include <stdbool.h> +#include <stdlib.h> +#include <string.h> + +#define WIDE_CHAR_SUPPORT \ + (HAVE_WCTYPE_H && HAVE_BTOWC && HAVE_ISWCTYPE \ + && HAVE_WMEMCHR && (HAVE_WMEMCPY || HAVE_WMEMPCPY)) + +/* For platform which support the ISO C amendement 1 functionality we + support user defined character classes. */ +#if defined _LIBC || WIDE_CHAR_SUPPORT +# include <wctype.h> +# include <wchar.h> +#endif + +/* We need some of the locale data (the collation sequence information) + but there is no interface to get this information in general. Therefore + we support a correct implementation only in glibc. */ +#ifdef _LIBC +# include "../locale/localeinfo.h" +# include "../locale/elem-hash.h" +# include "../locale/coll-lookup.h" +# include <shlib-compat.h> + +# define CONCAT(a,b) __CONCAT(a,b) +# define mbsrtowcs __mbsrtowcs +# define fnmatch __fnmatch +extern int fnmatch (const char *pattern, const char *string, int flags); +#endif + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* We often have to test for FNM_FILE_NAME and FNM_PERIOD being both set. */ +#define NO_LEADING_PERIOD(flags) \ + ((flags & (FNM_FILE_NAME | FNM_PERIOD)) == (FNM_FILE_NAME | FNM_PERIOD)) + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself, and have not detected a bug + in the library. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#if defined _LIBC || !defined __GNU_LIBRARY__ || !HAVE_FNMATCH_GNU + + +# if ! (defined isblank || HAVE_DECL_ISBLANK) +# define isblank(c) ((c) == ' ' || (c) == '\t') +# endif + +# define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) + +# if defined _LIBC || WIDE_CHAR_SUPPORT +/* The GNU C library provides support for user-defined character classes + and the functions from ISO C amendement 1. */ +# ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +# else +/* This shouldn't happen but some implementation might still have this + problem. Use a reasonable default value. */ +# define CHAR_CLASS_MAX_LENGTH 256 +# endif + +# ifdef _LIBC +# define IS_CHAR_CLASS(string) __wctype (string) +# else +# define IS_CHAR_CLASS(string) wctype (string) +# endif + +# ifdef _LIBC +# define ISWCTYPE(WC, WT) __iswctype (WC, WT) +# else +# define ISWCTYPE(WC, WT) iswctype (WC, WT) +# endif + +# if (HAVE_MBSTATE_T && HAVE_MBSRTOWCS) || _LIBC +/* In this case we are implementing the multibyte character handling. */ +# define HANDLE_MULTIBYTE 1 +# endif + +# else +# define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ + +# define IS_CHAR_CLASS(string) \ + (STREQ (string, "alpha") || STREQ (string, "upper") \ + || STREQ (string, "lower") || STREQ (string, "digit") \ + || STREQ (string, "alnum") || STREQ (string, "xdigit") \ + || STREQ (string, "space") || STREQ (string, "print") \ + || STREQ (string, "punct") || STREQ (string, "graph") \ + || STREQ (string, "cntrl") || STREQ (string, "blank")) +# endif + +/* Avoid depending on library functions or files + whose names are inconsistent. */ + +/* Global variable. */ +static int posixly_correct; + +# ifndef internal_function +/* Inside GNU libc we mark some function in a special way. In other + environments simply ignore the marking. */ +# define internal_function +# endif + +/* Note that this evaluates C many times. */ +# define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c)) +# define CHAR char +# define UCHAR unsigned char +# define INT int +# define FCT internal_fnmatch +# define EXT ext_match +# define END end_pattern +# define L_(CS) CS +# ifdef _LIBC +# define BTOWC(C) __btowc (C) +# else +# define BTOWC(C) btowc (C) +# endif +# define STRLEN(S) strlen (S) +# define STRCAT(D, S) strcat (D, S) +# ifdef _LIBC +# define MEMPCPY(D, S, N) __mempcpy (D, S, N) +# else +# if HAVE_MEMPCPY +# define MEMPCPY(D, S, N) mempcpy (D, S, N) +# else +# define MEMPCPY(D, S, N) ((void *) ((char *) memcpy (D, S, N) + (N))) +# endif +# endif +# define MEMCHR(S, C, N) memchr (S, C, N) +# define STRCOLL(S1, S2) strcoll (S1, S2) +# include "fnmatch_loop.c" + + +# if HANDLE_MULTIBYTE +# define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c)) +# define CHAR wchar_t +# define UCHAR wint_t +# define INT wint_t +# define FCT internal_fnwmatch +# define EXT ext_wmatch +# define END end_wpattern +# define L_(CS) L##CS +# define BTOWC(C) (C) +# ifdef _LIBC +# define STRLEN(S) __wcslen (S) +# define STRCAT(D, S) __wcscat (D, S) +# define MEMPCPY(D, S, N) __wmempcpy (D, S, N) +# else +# define STRLEN(S) wcslen (S) +# define STRCAT(D, S) wcscat (D, S) +# if HAVE_WMEMPCPY +# define MEMPCPY(D, S, N) wmempcpy (D, S, N) +# else +# define MEMPCPY(D, S, N) (wmemcpy (D, S, N) + (N)) +# endif +# endif +# define MEMCHR(S, C, N) wmemchr (S, C, N) +# define STRCOLL(S1, S2) wcscoll (S1, S2) +# define WIDE_CHAR_VERSION 1 + +# undef IS_CHAR_CLASS +/* We have to convert the wide character string in a multibyte string. But + we know that the character class names consist of alphanumeric characters + from the portable character set, and since the wide character encoding + for a member of the portable character set is the same code point as + its single-byte encoding, we can use a simplified method to convert the + string to a multibyte character string. */ +static wctype_t +is_char_class (const wchar_t *wcs) +{ + char s[CHAR_CLASS_MAX_LENGTH + 1]; + char *cp = s; + + do + { + /* Test for a printable character from the portable character set. */ +# ifdef _LIBC + if (*wcs < 0x20 || *wcs > 0x7e + || *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60) + return (wctype_t) 0; +# else + switch (*wcs) + { + case L' ': case L'!': case L'"': case L'#': case L'%': + case L'&': case L'\'': case L'(': case L')': case L'*': + case L'+': case L',': case L'-': case L'.': case L'/': + case L'0': case L'1': case L'2': case L'3': case L'4': + case L'5': case L'6': case L'7': case L'8': case L'9': + case L':': case L';': case L'<': case L'=': case L'>': + case L'?': + case L'A': case L'B': case L'C': case L'D': case L'E': + case L'F': case L'G': case L'H': case L'I': case L'J': + case L'K': case L'L': case L'M': case L'N': case L'O': + case L'P': case L'Q': case L'R': case L'S': case L'T': + case L'U': case L'V': case L'W': case L'X': case L'Y': + case L'Z': + case L'[': case L'\\': case L']': case L'^': case L'_': + case L'a': case L'b': case L'c': case L'd': case L'e': + case L'f': case L'g': case L'h': case L'i': case L'j': + case L'k': case L'l': case L'm': case L'n': case L'o': + case L'p': case L'q': case L'r': case L's': case L't': + case L'u': case L'v': case L'w': case L'x': case L'y': + case L'z': case L'{': case L'|': case L'}': case L'~': + break; + default: + return (wctype_t) 0; + } +# endif + + /* Avoid overrunning the buffer. */ + if (cp == s + CHAR_CLASS_MAX_LENGTH) + return (wctype_t) 0; + + *cp++ = (char) *wcs++; + } + while (*wcs != L'\0'); + + *cp = '\0'; + +# ifdef _LIBC + return __wctype (s); +# else + return wctype (s); +# endif +} +# define IS_CHAR_CLASS(string) is_char_class (string) + +# include "fnmatch_loop.c" +# endif + + +int +fnmatch (const char *pattern, const char *string, int flags) +{ +# if HANDLE_MULTIBYTE +# define ALLOCA_LIMIT 2000 + if (__builtin_expect (MB_CUR_MAX, 1) != 1) + { + mbstate_t ps; + size_t patsize; + size_t strsize; + size_t totsize; + wchar_t *wpattern; + wchar_t *wstring; + int res; + + /* Calculate the size needed to convert the strings to + wide characters. */ + memset (&ps, '\0', sizeof (ps)); + patsize = mbsrtowcs (NULL, &pattern, 0, &ps) + 1; + if (__builtin_expect (patsize != 0, 1)) + { + assert (mbsinit (&ps)); + strsize = mbsrtowcs (NULL, &string, 0, &ps) + 1; + if (__builtin_expect (strsize != 0, 1)) + { + assert (mbsinit (&ps)); + totsize = patsize + strsize; + if (__builtin_expect (! (patsize <= totsize + && totsize <= SIZE_MAX / sizeof (wchar_t)), + 0)) + { + errno = ENOMEM; + return -1; + } + + /* Allocate room for the wide characters. */ + if (__builtin_expect (totsize < ALLOCA_LIMIT, 1)) + wpattern = (wchar_t *) alloca (totsize * sizeof (wchar_t)); + else + { + wpattern = malloc (totsize * sizeof (wchar_t)); + if (__builtin_expect (! wpattern, 0)) + { + errno = ENOMEM; + return -1; + } + } + wstring = wpattern + patsize; + + /* Convert the strings into wide characters. */ + mbsrtowcs (wpattern, &pattern, patsize, &ps); + assert (mbsinit (&ps)); + mbsrtowcs (wstring, &string, strsize, &ps); + + res = internal_fnwmatch (wpattern, wstring, wstring + strsize - 1, + flags & FNM_PERIOD, flags); + + if (__builtin_expect (! (totsize < ALLOCA_LIMIT), 0)) + free (wpattern); + return res; + } + } + } + +# endif /* HANDLE_MULTIBYTE */ + + return internal_fnmatch (pattern, string, string + strlen (string), + flags & FNM_PERIOD, flags); +} + +# ifdef _LIBC +# undef fnmatch +versioned_symbol (libc, __fnmatch, fnmatch, GLIBC_2_2_3); +# if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_2_3) +strong_alias (__fnmatch, __fnmatch_old) +compat_symbol (libc, __fnmatch_old, fnmatch, GLIBC_2_0); +# endif +libc_hidden_ver (__fnmatch, fnmatch) +# endif + +#endif /* _LIBC or not __GNU_LIBRARY__. */ diff --git a/lib/fnmatch_.h b/lib/fnmatch_.h new file mode 100644 index 0000000..b086b45 --- /dev/null +++ b/lib/fnmatch_.h @@ -0,0 +1,65 @@ +/* Copyright (C) 1991, 1992, 1993, 1996, 1997, 1998, 1999, 2001, 2002, 2003, + 2005, 2007 Free Software Foundation, Inc. + + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _FNMATCH_H +#define _FNMATCH_H 1 + +#ifdef __cplusplus +extern "C" { +#endif + +/* We #undef these before defining them because some losing systems + (HP-UX A.08.07 for example) define these in <unistd.h>. */ +#undef FNM_PATHNAME +#undef FNM_NOESCAPE +#undef FNM_PERIOD + +/* Bits set in the FLAGS argument to `fnmatch'. */ +#define FNM_PATHNAME (1 << 0) /* No wildcard can ever match `/'. */ +#define FNM_NOESCAPE (1 << 1) /* Backslashes don't quote special chars. */ +#define FNM_PERIOD (1 << 2) /* Leading `.' is matched only explicitly. */ + +#if !defined _POSIX_C_SOURCE || _POSIX_C_SOURCE < 2 || defined _GNU_SOURCE +# define FNM_FILE_NAME FNM_PATHNAME /* Preferred GNU name. */ +# define FNM_LEADING_DIR (1 << 3) /* Ignore `/...' after a match. */ +# define FNM_CASEFOLD (1 << 4) /* Compare without regard to case. */ +# define FNM_EXTMATCH (1 << 5) /* Use ksh-like extended matching. */ +#endif + +/* Value returned by `fnmatch' if STRING does not match PATTERN. */ +#define FNM_NOMATCH 1 + +/* This value is returned if the implementation does not support + `fnmatch'. Since this is not the case here it will never be + returned but the conformance test suites still require the symbol + to be defined. */ +#ifdef _XOPEN_SOURCE +# define FNM_NOSYS (-1) +#endif + +/* Match NAME against the file name pattern PATTERN, + returning zero if it matches, FNM_NOMATCH if not. */ +extern int fnmatch (const char *__pattern, const char *__name, + int __flags); + +#ifdef __cplusplus +} +#endif + +#endif /* fnmatch.h */ diff --git a/lib/fnmatch_loop.c b/lib/fnmatch_loop.c new file mode 100644 index 0000000..d1008c2 --- /dev/null +++ b/lib/fnmatch_loop.c @@ -0,0 +1,1210 @@ +/* Copyright (C) 1991,1992,1993,1996,1997,1998,1999,2000,2001,2002,2003,2004,2005,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Match STRING against the file name pattern PATTERN, returning zero if + it matches, nonzero if not. */ +static int EXT (INT opt, const CHAR *pattern, const CHAR *string, + const CHAR *string_end, bool no_leading_period, int flags) + internal_function; +static const CHAR *END (const CHAR *patternp) internal_function; + +static int +internal_function +FCT (const CHAR *pattern, const CHAR *string, const CHAR *string_end, + bool no_leading_period, int flags) +{ + register const CHAR *p = pattern, *n = string; + register UCHAR c; +#ifdef _LIBC +# if WIDE_CHAR_VERSION + const char *collseq = (const char *) + _NL_CURRENT(LC_COLLATE, _NL_COLLATE_COLLSEQWC); +# else + const UCHAR *collseq = (const UCHAR *) + _NL_CURRENT(LC_COLLATE, _NL_COLLATE_COLLSEQMB); +# endif +#endif + + while ((c = *p++) != L_('\0')) + { + bool new_no_leading_period = false; + c = FOLD (c); + + switch (c) + { + case L_('?'): + if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') + { + int res; + + res = EXT (c, p, n, string_end, no_leading_period, + flags); + if (res != -1) + return res; + } + + if (n == string_end) + return FNM_NOMATCH; + else if (*n == L_('/') && (flags & FNM_FILE_NAME)) + return FNM_NOMATCH; + else if (*n == L_('.') && no_leading_period) + return FNM_NOMATCH; + break; + + case L_('\\'): + if (!(flags & FNM_NOESCAPE)) + { + c = *p++; + if (c == L_('\0')) + /* Trailing \ loses. */ + return FNM_NOMATCH; + c = FOLD (c); + } + if (n == string_end || FOLD ((UCHAR) *n) != c) + return FNM_NOMATCH; + break; + + case L_('*'): + if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') + { + int res; + + res = EXT (c, p, n, string_end, no_leading_period, + flags); + if (res != -1) + return res; + } + + if (n != string_end && *n == L_('.') && no_leading_period) + return FNM_NOMATCH; + + for (c = *p++; c == L_('?') || c == L_('*'); c = *p++) + { + if (*p == L_('(') && (flags & FNM_EXTMATCH) != 0) + { + const CHAR *endp = END (p); + if (endp != p) + { + /* This is a pattern. Skip over it. */ + p = endp; + continue; + } + } + + if (c == L_('?')) + { + /* A ? needs to match one character. */ + if (n == string_end) + /* There isn't another character; no match. */ + return FNM_NOMATCH; + else if (*n == L_('/') + && __builtin_expect (flags & FNM_FILE_NAME, 0)) + /* A slash does not match a wildcard under + FNM_FILE_NAME. */ + return FNM_NOMATCH; + else + /* One character of the string is consumed in matching + this ? wildcard, so *??? won't match if there are + less than three characters. */ + ++n; + } + } + + if (c == L_('\0')) + /* The wildcard(s) is/are the last element of the pattern. + If the name is a file name and contains another slash + this means it cannot match, unless the FNM_LEADING_DIR + flag is set. */ + { + int result = (flags & FNM_FILE_NAME) == 0 ? 0 : FNM_NOMATCH; + + if (flags & FNM_FILE_NAME) + { + if (flags & FNM_LEADING_DIR) + result = 0; + else + { + if (MEMCHR (n, L_('/'), string_end - n) == NULL) + result = 0; + } + } + + return result; + } + else + { + const CHAR *endp; + + endp = MEMCHR (n, (flags & FNM_FILE_NAME) ? L_('/') : L_('\0'), + string_end - n); + if (endp == NULL) + endp = string_end; + + if (c == L_('[') + || (__builtin_expect (flags & FNM_EXTMATCH, 0) != 0 + && (c == L_('@') || c == L_('+') || c == L_('!')) + && *p == L_('('))) + { + int flags2 = ((flags & FNM_FILE_NAME) + ? flags : (flags & ~FNM_PERIOD)); + bool no_leading_period2 = no_leading_period; + + for (--p; n < endp; ++n, no_leading_period2 = false) + if (FCT (p, n, string_end, no_leading_period2, flags2) + == 0) + return 0; + } + else if (c == L_('/') && (flags & FNM_FILE_NAME)) + { + while (n < string_end && *n != L_('/')) + ++n; + if (n < string_end && *n == L_('/') + && (FCT (p, n + 1, string_end, flags & FNM_PERIOD, flags) + == 0)) + return 0; + } + else + { + int flags2 = ((flags & FNM_FILE_NAME) + ? flags : (flags & ~FNM_PERIOD)); + int no_leading_period2 = no_leading_period; + + if (c == L_('\\') && !(flags & FNM_NOESCAPE)) + c = *p; + c = FOLD (c); + for (--p; n < endp; ++n, no_leading_period2 = false) + if (FOLD ((UCHAR) *n) == c + && (FCT (p, n, string_end, no_leading_period2, flags2) + == 0)) + return 0; + } + } + + /* If we come here no match is possible with the wildcard. */ + return FNM_NOMATCH; + + case L_('['): + { + /* Nonzero if the sense of the character class is inverted. */ + register bool not; + CHAR cold; + UCHAR fn; + + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + if (n == string_end) + return FNM_NOMATCH; + + if (*n == L_('.') && no_leading_period) + return FNM_NOMATCH; + + if (*n == L_('/') && (flags & FNM_FILE_NAME)) + /* `/' cannot be matched. */ + return FNM_NOMATCH; + + not = (*p == L_('!') || (posixly_correct < 0 && *p == L_('^'))); + if (not) + ++p; + + fn = FOLD ((UCHAR) *n); + + c = *p++; + for (;;) + { + if (!(flags & FNM_NOESCAPE) && c == L_('\\')) + { + if (*p == L_('\0')) + return FNM_NOMATCH; + c = FOLD ((UCHAR) *p); + ++p; + + goto normal_bracket; + } + else if (c == L_('[') && *p == L_(':')) + { + /* Leave room for the null. */ + CHAR str[CHAR_CLASS_MAX_LENGTH + 1]; + size_t c1 = 0; +#if defined _LIBC || WIDE_CHAR_SUPPORT + wctype_t wt; +#endif + const CHAR *startp = p; + + for (;;) + { + if (c1 == CHAR_CLASS_MAX_LENGTH) + /* The name is too long and therefore the pattern + is ill-formed. */ + return FNM_NOMATCH; + + c = *++p; + if (c == L_(':') && p[1] == L_(']')) + { + p += 2; + break; + } + if (c < L_('a') || c >= L_('z')) + { + /* This cannot possibly be a character class name. + Match it as a normal range. */ + p = startp; + c = L_('['); + goto normal_bracket; + } + str[c1++] = c; + } + str[c1] = L_('\0'); + +#if defined _LIBC || WIDE_CHAR_SUPPORT + wt = IS_CHAR_CLASS (str); + if (wt == 0) + /* Invalid character class name. */ + return FNM_NOMATCH; + +# if defined _LIBC && ! WIDE_CHAR_VERSION + /* The following code is glibc specific but does + there a good job in speeding up the code since + we can avoid the btowc() call. */ + if (_ISCTYPE ((UCHAR) *n, wt)) + goto matched; +# else + if (ISWCTYPE (BTOWC ((UCHAR) *n), wt)) + goto matched; +# endif +#else + if ((STREQ (str, L_("alnum")) && isalnum ((UCHAR) *n)) + || (STREQ (str, L_("alpha")) && isalpha ((UCHAR) *n)) + || (STREQ (str, L_("blank")) && isblank ((UCHAR) *n)) + || (STREQ (str, L_("cntrl")) && iscntrl ((UCHAR) *n)) + || (STREQ (str, L_("digit")) && isdigit ((UCHAR) *n)) + || (STREQ (str, L_("graph")) && isgraph ((UCHAR) *n)) + || (STREQ (str, L_("lower")) && islower ((UCHAR) *n)) + || (STREQ (str, L_("print")) && isprint ((UCHAR) *n)) + || (STREQ (str, L_("punct")) && ispunct ((UCHAR) *n)) + || (STREQ (str, L_("space")) && isspace ((UCHAR) *n)) + || (STREQ (str, L_("upper")) && isupper ((UCHAR) *n)) + || (STREQ (str, L_("xdigit")) && isxdigit ((UCHAR) *n))) + goto matched; +#endif + c = *p++; + } +#ifdef _LIBC + else if (c == L_('[') && *p == L_('=')) + { + UCHAR str[1]; + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + const CHAR *startp = p; + + c = *++p; + if (c == L_('\0')) + { + p = startp; + c = L_('['); + goto normal_bracket; + } + str[0] = c; + + c = *++p; + if (c != L_('=') || p[1] != L_(']')) + { + p = startp; + c = L_('['); + goto normal_bracket; + } + p += 2; + + if (nrules == 0) + { + if ((UCHAR) *n == str[0]) + goto matched; + } + else + { + const int32_t *table; +# if WIDE_CHAR_VERSION + const int32_t *weights; + const int32_t *extra; +# else + const unsigned char *weights; + const unsigned char *extra; +# endif + const int32_t *indirect; + int32_t idx; + const UCHAR *cp = (const UCHAR *) str; + + /* This #include defines a local function! */ +# if WIDE_CHAR_VERSION +# include <locale/weightwc.h> +# else +# include <locale/weight.h> +# endif + +# if WIDE_CHAR_VERSION + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); + weights = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); + extra = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); +# else + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); +# endif + + idx = findidx (&cp); + if (idx != 0) + { + /* We found a table entry. Now see whether the + character we are currently at has the same + equivalance class value. */ + int len = weights[idx]; + int32_t idx2; + const UCHAR *np = (const UCHAR *) n; + + idx2 = findidx (&np); + if (idx2 != 0 && len == weights[idx2]) + { + int cnt = 0; + + while (cnt < len + && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + goto matched; + } + } + } + + c = *p++; + } +#endif + else if (c == L_('\0')) + /* [ (unterminated) loses. */ + return FNM_NOMATCH; + else + { + bool is_range = false; + +#ifdef _LIBC + bool is_seqval = false; + + if (c == L_('[') && *p == L_('.')) + { + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + const CHAR *startp = p; + size_t c1 = 0; + + while (1) + { + c = *++p; + if (c == L_('.') && p[1] == L_(']')) + { + p += 2; + break; + } + if (c == '\0') + return FNM_NOMATCH; + ++c1; + } + + /* We have to handling the symbols differently in + ranges since then the collation sequence is + important. */ + is_range = *p == L_('-') && p[1] != L_('\0'); + + if (nrules == 0) + { + /* There are no names defined in the collation + data. Therefore we only accept the trivial + names consisting of the character itself. */ + if (c1 != 1) + return FNM_NOMATCH; + + if (!is_range && *n == startp[1]) + goto matched; + + cold = startp[1]; + c = *p++; + } + else + { + int32_t table_size; + const int32_t *symb_table; +# ifdef WIDE_CHAR_VERSION + char str[c1]; + size_t strcnt; +# else +# define str (startp + 1) +# endif + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + +# ifdef WIDE_CHAR_VERSION + /* We have to convert the name to a single-byte + string. This is possible since the names + consist of ASCII characters and the internal + representation is UCS4. */ + for (strcnt = 0; strcnt < c1; ++strcnt) + str[strcnt] = startp[1 + strcnt]; +# endif + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing table. */ + hash = elem_hash (str, c1); + + idx = 0; + elem = hash % table_size; + if (symb_table[2 * elem] != 0) + { + second = hash % (table_size - 2) + 1; + + do + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && (c1 + == extra[symb_table[2 * elem + 1]]) + && memcmp (str, + &extra[symb_table[2 * elem + + 1] + + 1], c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + while (symb_table[2 * elem] != 0); + } + + if (symb_table[2 * elem] != 0) + { + /* Compare the byte sequence but only if + this is not part of a range. */ +# ifdef WIDE_CHAR_VERSION + int32_t *wextra; + + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + + wextra = (int32_t *) &extra[idx + 4]; +# endif + + if (! is_range) + { +# ifdef WIDE_CHAR_VERSION + for (c1 = 0; + (int32_t) c1 < wextra[idx]; + ++c1) + if (n[c1] != wextra[1 + c1]) + break; + + if ((int32_t) c1 == wextra[idx]) + goto matched; +# else + for (c1 = 0; c1 < extra[idx]; ++c1) + if (n[c1] != extra[1 + c1]) + break; + + if (c1 == extra[idx]) + goto matched; +# endif + } + + /* Get the collation sequence value. */ + is_seqval = true; +# ifdef WIDE_CHAR_VERSION + cold = wextra[1 + wextra[idx]]; +# else + /* Adjust for the alignment. */ + idx += 1 + extra[idx]; + idx = (idx + 3) & ~4; + cold = *((int32_t *) &extra[idx]); +# endif + + c = *p++; + } + else if (c1 == 1) + { + /* No valid character. Match it as a + single byte. */ + if (!is_range && *n == str[0]) + goto matched; + + cold = str[0]; + c = *p++; + } + else + return FNM_NOMATCH; + } + } + else +# undef str +#endif + { + c = FOLD (c); + normal_bracket: + + /* We have to handling the symbols differently in + ranges since then the collation sequence is + important. */ + is_range = (*p == L_('-') && p[1] != L_('\0') + && p[1] != L_(']')); + + if (!is_range && c == fn) + goto matched; + +#if _LIBC + /* This is needed if we goto normal_bracket; from + outside of is_seqval's scope. */ + is_seqval = false; +#endif + + cold = c; + c = *p++; + } + + if (c == L_('-') && *p != L_(']')) + { +#if _LIBC + /* We have to find the collation sequence + value for C. Collation sequence is nothing + we can regularly access. The sequence + value is defined by the order in which the + definitions of the collation values for the + various characters appear in the source + file. A strange concept, nowhere + documented. */ + uint32_t fcollseq; + uint32_t lcollseq; + UCHAR cend = *p++; + +# ifdef WIDE_CHAR_VERSION + /* Search in the `names' array for the characters. */ + fcollseq = __collseq_table_lookup (collseq, fn); + if (fcollseq == ~((uint32_t) 0)) + /* XXX We don't know anything about the character + we are supposed to match. This means we are + failing. */ + goto range_not_matched; + + if (is_seqval) + lcollseq = cold; + else + lcollseq = __collseq_table_lookup (collseq, cold); +# else + fcollseq = collseq[fn]; + lcollseq = is_seqval ? cold : collseq[(UCHAR) cold]; +# endif + + is_seqval = false; + if (cend == L_('[') && *p == L_('.')) + { + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_NRULES); + const CHAR *startp = p; + size_t c1 = 0; + + while (1) + { + c = *++p; + if (c == L_('.') && p[1] == L_(']')) + { + p += 2; + break; + } + if (c == '\0') + return FNM_NOMATCH; + ++c1; + } + + if (nrules == 0) + { + /* There are no names defined in the + collation data. Therefore we only + accept the trivial names consisting + of the character itself. */ + if (c1 != 1) + return FNM_NOMATCH; + + cend = startp[1]; + } + else + { + int32_t table_size; + const int32_t *symb_table; +# ifdef WIDE_CHAR_VERSION + char str[c1]; + size_t strcnt; +# else +# define str (startp + 1) +# endif + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + +# ifdef WIDE_CHAR_VERSION + /* We have to convert the name to a single-byte + string. This is possible since the names + consist of ASCII characters and the internal + representation is UCS4. */ + for (strcnt = 0; strcnt < c1; ++strcnt) + str[strcnt] = startp[1 + strcnt]; +# endif + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing + table. */ + hash = elem_hash (str, c1); + + idx = 0; + elem = hash % table_size; + if (symb_table[2 * elem] != 0) + { + second = hash % (table_size - 2) + 1; + + do + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && (c1 + == extra[symb_table[2 * elem + 1]]) + && memcmp (str, + &extra[symb_table[2 * elem + 1] + + 1], c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + while (symb_table[2 * elem] != 0); + } + + if (symb_table[2 * elem] != 0) + { + /* Compare the byte sequence but only if + this is not part of a range. */ +# ifdef WIDE_CHAR_VERSION + int32_t *wextra; + + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~4; + + wextra = (int32_t *) &extra[idx + 4]; +# endif + /* Get the collation sequence value. */ + is_seqval = true; +# ifdef WIDE_CHAR_VERSION + cend = wextra[1 + wextra[idx]]; +# else + /* Adjust for the alignment. */ + idx += 1 + extra[idx]; + idx = (idx + 3) & ~4; + cend = *((int32_t *) &extra[idx]); +# endif + } + else if (symb_table[2 * elem] != 0 && c1 == 1) + { + cend = str[0]; + c = *p++; + } + else + return FNM_NOMATCH; + } +# undef str + } + else + { + if (!(flags & FNM_NOESCAPE) && cend == L_('\\')) + cend = *p++; + if (cend == L_('\0')) + return FNM_NOMATCH; + cend = FOLD (cend); + } + + /* XXX It is not entirely clear to me how to handle + characters which are not mentioned in the + collation specification. */ + if ( +# ifdef WIDE_CHAR_VERSION + lcollseq == 0xffffffff || +# endif + lcollseq <= fcollseq) + { + /* We have to look at the upper bound. */ + uint32_t hcollseq; + + if (is_seqval) + hcollseq = cend; + else + { +# ifdef WIDE_CHAR_VERSION + hcollseq = + __collseq_table_lookup (collseq, cend); + if (hcollseq == ~((uint32_t) 0)) + { + /* Hum, no information about the upper + bound. The matching succeeds if the + lower bound is matched exactly. */ + if (lcollseq != fcollseq) + goto range_not_matched; + + goto matched; + } +# else + hcollseq = collseq[cend]; +# endif + } + + if (lcollseq <= hcollseq && fcollseq <= hcollseq) + goto matched; + } +# ifdef WIDE_CHAR_VERSION + range_not_matched: +# endif +#else + /* We use a boring value comparison of the character + values. This is better than comparing using + `strcoll' since the latter would have surprising + and sometimes fatal consequences. */ + UCHAR cend = *p++; + + if (!(flags & FNM_NOESCAPE) && cend == L_('\\')) + cend = *p++; + if (cend == L_('\0')) + return FNM_NOMATCH; + + /* It is a range. */ + if (cold <= fn && fn <= cend) + goto matched; +#endif + + c = *p++; + } + } + + if (c == L_(']')) + break; + } + + if (!not) + return FNM_NOMATCH; + break; + + matched: + /* Skip the rest of the [...] that already matched. */ + do + { + ignore_next: + c = *p++; + + if (c == L_('\0')) + /* [... (unterminated) loses. */ + return FNM_NOMATCH; + + if (!(flags & FNM_NOESCAPE) && c == L_('\\')) + { + if (*p == L_('\0')) + return FNM_NOMATCH; + /* XXX 1003.2d11 is unclear if this is right. */ + ++p; + } + else if (c == L_('[') && *p == L_(':')) + { + int c1 = 0; + const CHAR *startp = p; + + while (1) + { + c = *++p; + if (++c1 == CHAR_CLASS_MAX_LENGTH) + return FNM_NOMATCH; + + if (*p == L_(':') && p[1] == L_(']')) + break; + + if (c < L_('a') || c >= L_('z')) + { + p = startp; + goto ignore_next; + } + } + p += 2; + c = *p++; + } + else if (c == L_('[') && *p == L_('=')) + { + c = *++p; + if (c == L_('\0')) + return FNM_NOMATCH; + c = *++p; + if (c != L_('=') || p[1] != L_(']')) + return FNM_NOMATCH; + p += 2; + c = *p++; + } + else if (c == L_('[') && *p == L_('.')) + { + ++p; + while (1) + { + c = *++p; + if (c == '\0') + return FNM_NOMATCH; + + if (*p == L_('.') && p[1] == L_(']')) + break; + } + p += 2; + c = *p++; + } + } + while (c != L_(']')); + if (not) + return FNM_NOMATCH; + } + break; + + case L_('+'): + case L_('@'): + case L_('!'): + if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') + { + int res; + + res = EXT (c, p, n, string_end, no_leading_period, flags); + if (res != -1) + return res; + } + goto normal_match; + + case L_('/'): + if (NO_LEADING_PERIOD (flags)) + { + if (n == string_end || c != (UCHAR) *n) + return FNM_NOMATCH; + + new_no_leading_period = true; + break; + } + /* FALLTHROUGH */ + default: + normal_match: + if (n == string_end || c != FOLD ((UCHAR) *n)) + return FNM_NOMATCH; + } + + no_leading_period = new_no_leading_period; + ++n; + } + + if (n == string_end) + return 0; + + if ((flags & FNM_LEADING_DIR) && n != string_end && *n == L_('/')) + /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz". */ + return 0; + + return FNM_NOMATCH; +} + + +static const CHAR * +internal_function +END (const CHAR *pattern) +{ + const CHAR *p = pattern; + + while (1) + if (*++p == L_('\0')) + /* This is an invalid pattern. */ + return pattern; + else if (*p == L_('[')) + { + /* Handle brackets special. */ + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + /* Skip the not sign. We have to recognize it because of a possibly + following ']'. */ + if (*++p == L_('!') || (posixly_correct < 0 && *p == L_('^'))) + ++p; + /* A leading ']' is recognized as such. */ + if (*p == L_(']')) + ++p; + /* Skip over all characters of the list. */ + while (*p != L_(']')) + if (*p++ == L_('\0')) + /* This is no valid pattern. */ + return pattern; + } + else if ((*p == L_('?') || *p == L_('*') || *p == L_('+') || *p == L_('@') + || *p == L_('!')) && p[1] == L_('(')) + p = END (p + 1); + else if (*p == L_(')')) + break; + + return p + 1; +} + + +static int +internal_function +EXT (INT opt, const CHAR *pattern, const CHAR *string, const CHAR *string_end, + bool no_leading_period, int flags) +{ + const CHAR *startp; + size_t level; + struct patternlist + { + struct patternlist *next; + CHAR str[1]; + } *list = NULL; + struct patternlist **lastp = &list; + size_t pattern_len = STRLEN (pattern); + const CHAR *p; + const CHAR *rs; + enum { ALLOCA_LIMIT = 8000 }; + + /* Parse the pattern. Store the individual parts in the list. */ + level = 0; + for (startp = p = pattern + 1; ; ++p) + if (*p == L_('\0')) + /* This is an invalid pattern. */ + return -1; + else if (*p == L_('[')) + { + /* Handle brackets special. */ + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + /* Skip the not sign. We have to recognize it because of a possibly + following ']'. */ + if (*++p == L_('!') || (posixly_correct < 0 && *p == L_('^'))) + ++p; + /* A leading ']' is recognized as such. */ + if (*p == L_(']')) + ++p; + /* Skip over all characters of the list. */ + while (*p != L_(']')) + if (*p++ == L_('\0')) + /* This is no valid pattern. */ + return -1; + } + else if ((*p == L_('?') || *p == L_('*') || *p == L_('+') || *p == L_('@') + || *p == L_('!')) && p[1] == L_('(')) + /* Remember the nesting level. */ + ++level; + else if (*p == L_(')')) + { + if (level-- == 0) + { + /* This means we found the end of the pattern. */ +#define NEW_PATTERN \ + struct patternlist *newp; \ + size_t plen; \ + size_t plensize; \ + size_t newpsize; \ + \ + plen = (opt == L_('?') || opt == L_('@') \ + ? pattern_len \ + : p - startp + 1); \ + plensize = plen * sizeof (CHAR); \ + newpsize = offsetof (struct patternlist, str) + plensize; \ + if ((size_t) -1 / sizeof (CHAR) < plen \ + || newpsize < offsetof (struct patternlist, str) \ + || ALLOCA_LIMIT <= newpsize) \ + return -1; \ + newp = (struct patternlist *) alloca (newpsize); \ + *((CHAR *) MEMPCPY (newp->str, startp, p - startp)) = L_('\0'); \ + newp->next = NULL; \ + *lastp = newp; \ + lastp = &newp->next + NEW_PATTERN; + break; + } + } + else if (*p == L_('|')) + { + if (level == 0) + { + NEW_PATTERN; + startp = p + 1; + } + } + assert (list != NULL); + assert (p[-1] == L_(')')); +#undef NEW_PATTERN + + switch (opt) + { + case L_('*'): + if (FCT (p, string, string_end, no_leading_period, flags) == 0) + return 0; + /* FALLTHROUGH */ + + case L_('+'): + do + { + for (rs = string; rs <= string_end; ++rs) + /* First match the prefix with the current pattern with the + current pattern. */ + if (FCT (list->str, string, rs, no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) == 0 + /* This was successful. Now match the rest with the rest + of the pattern. */ + && (FCT (p, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags), + flags & FNM_FILE_NAME + ? flags : flags & ~FNM_PERIOD) == 0 + /* This didn't work. Try the whole pattern. */ + || (rs != string + && FCT (pattern - 1, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags), + flags & FNM_FILE_NAME + ? flags : flags & ~FNM_PERIOD) == 0))) + /* It worked. Signal success. */ + return 0; + } + while ((list = list->next) != NULL); + + /* None of the patterns lead to a match. */ + return FNM_NOMATCH; + + case L_('?'): + if (FCT (p, string, string_end, no_leading_period, flags) == 0) + return 0; + /* FALLTHROUGH */ + + case L_('@'): + do + /* I cannot believe it but `strcat' is actually acceptable + here. Match the entire string with the prefix from the + pattern list and the rest of the pattern following the + pattern list. */ + if (FCT (STRCAT (list->str, p), string, string_end, + no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) == 0) + /* It worked. Signal success. */ + return 0; + while ((list = list->next) != NULL); + + /* None of the patterns lead to a match. */ + return FNM_NOMATCH; + + case L_('!'): + for (rs = string; rs <= string_end; ++rs) + { + struct patternlist *runp; + + for (runp = list; runp != NULL; runp = runp->next) + if (FCT (runp->str, string, rs, no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) == 0) + break; + + /* If none of the patterns matched see whether the rest does. */ + if (runp == NULL + && (FCT (p, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags), + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) + == 0)) + /* This is successful. */ + return 0; + } + + /* None of the patterns together with the rest of the pattern + lead to a match. */ + return FNM_NOMATCH; + + default: + assert (! "Invalid extended matching operator"); + break; + } + + return -1; +} + + +#undef FOLD +#undef CHAR +#undef UCHAR +#undef INT +#undef FCT +#undef EXT +#undef END +#undef MEMPCPY +#undef MEMCHR +#undef STRCOLL +#undef STRLEN +#undef STRCAT +#undef L_ +#undef BTOWC diff --git a/lib/fstatat.c b/lib/fstatat.c new file mode 100644 index 0000000..92a5164 --- /dev/null +++ b/lib/fstatat.c @@ -0,0 +1,57 @@ +/* Work around an fstatat bug on Solaris 9. + + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and Jim Meyering. */ + +#include <config.h> + +#define COMPILING_FSTATAT 1 +#include "openat.h" + +#include <errno.h> +#include <string.h> + +/* fstatat should always follow symbolic links that end in /, but on + Solaris 9 it doesn't if AT_SYMLINK_NOFOLLOW is specified. This is + the same problem that lstat.c addresses, so solve it in a similar + way. */ + +int +rpl_fstatat (int fd, char const *file, struct stat *st, int flag) +{ + int result = fstatat (fd, file, st, flag); + + if (result == 0 && (flag & AT_SYMLINK_NOFOLLOW) && S_ISLNK (st->st_mode) + && file[strlen (file) - 1] == '/') + { + /* FILE refers to a symbolic link and the name ends with a slash. + Get info about the link's referent. */ + result = fstatat (fd, file, st, flag & ~AT_SYMLINK_NOFOLLOW); + if (result == 0 && ! S_ISDIR (st->st_mode)) + { + /* fstatat succeeded and FILE references a non-directory. + But it was specified via a name including a trailing + slash. Fail with errno set to ENOTDIR to indicate the + contradiction. */ + errno = ENOTDIR; + return -1; + } + } + + return result; +} diff --git a/lib/ftruncate.c b/lib/ftruncate.c new file mode 100644 index 0000000..ff7d11b --- /dev/null +++ b/lib/ftruncate.c @@ -0,0 +1,90 @@ +/* ftruncate emulations that work on some System V's. + This file is in the public domain. */ + +#include <config.h> + +/* Specification. */ +#include <unistd.h> + +#include <sys/types.h> +#include <fcntl.h> + +#ifdef F_CHSIZE + +int +ftruncate (int fd, off_t length) +{ + return fcntl (fd, F_CHSIZE, length); +} + +#else /* not F_CHSIZE */ +# ifdef F_FREESP + +/* By William Kucharski <kucharsk@netcom.com>. */ + +# include <sys/stat.h> +# include <errno.h> + +int +ftruncate (int fd, off_t length) +{ + struct flock fl; + struct stat filebuf; + + if (fstat (fd, &filebuf) < 0) + return -1; + + if (filebuf.st_size < length) + { + /* Extend file length. */ + if (lseek (fd, (length - 1), SEEK_SET) < 0) + return -1; + + /* Write a "0" byte. */ + if (write (fd, "", 1) != 1) + return -1; + } + else + { + + /* Truncate length. */ + + fl.l_whence = 0; + fl.l_len = 0; + fl.l_start = length; + fl.l_type = F_WRLCK; /* write lock on file space */ + + /* This relies on the *undocumented* F_FREESP argument to fcntl, + which truncates the file so that it ends at the position + indicated by fl.l_start. Will minor miracles never cease? */ + + if (fcntl (fd, F_FREESP, &fl) < 0) + return -1; + } + + return 0; +} + +# else /* not F_CHSIZE nor F_FREESP */ +# if HAVE_CHSIZE /* native Windows, e.g. mingw */ + +int +ftruncate (int fd, off_t length) +{ + return chsize (fd, length); +} + +# else /* not F_CHSIZE nor F_FREESP nor HAVE_CHSIZE */ + +# include <errno.h> + +int +ftruncate (int fd, off_t length) +{ + errno = EIO; + return -1; +} + +# endif /* not HAVE_CHSIZE */ +# endif /* not F_FREESP */ +#endif /* not F_CHSIZE */ diff --git a/lib/full-write.c b/lib/full-write.c new file mode 100644 index 0000000..cc16872 --- /dev/null +++ b/lib/full-write.c @@ -0,0 +1,81 @@ +/* An interface to read and write that retries (if necessary) until complete. + + Copyright (C) 1993, 1994, 1997, 1998, 1999, 2000, 2001, 2002, 2003, + 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#ifdef FULL_READ +# include "full-read.h" +#else +# include "full-write.h" +#endif + +#include <errno.h> + +#ifdef FULL_READ +# include "safe-read.h" +# define safe_rw safe_read +# define full_rw full_read +# undef const +# define const /* empty */ +#else +# include "safe-write.h" +# define safe_rw safe_write +# define full_rw full_write +#endif + +#ifdef FULL_READ +/* Set errno to zero upon EOF. */ +# define ZERO_BYTE_TRANSFER_ERRNO 0 +#else +/* Some buggy drivers return 0 when one tries to write beyond + a device's end. (Example: Linux 1.2.13 on /dev/fd0.) + Set errno to ENOSPC so they get a sensible diagnostic. */ +# define ZERO_BYTE_TRANSFER_ERRNO ENOSPC +#endif + +/* Write(read) COUNT bytes at BUF to(from) descriptor FD, retrying if + interrupted or if a partial write(read) occurs. Return the number + of bytes transferred. + When writing, set errno if fewer than COUNT bytes are written. + When reading, if fewer than COUNT bytes are read, you must examine + errno to distinguish failure from EOF (errno == 0). */ +size_t +full_rw (int fd, const void *buf, size_t count) +{ + size_t total = 0; + const char *ptr = (const char *) buf; + + while (count > 0) + { + size_t n_rw = safe_rw (fd, ptr, count); + if (n_rw == (size_t) -1) + break; + if (n_rw == 0) + { + errno = ZERO_BYTE_TRANSFER_ERRNO; + break; + } + total += n_rw; + ptr += n_rw; + count -= n_rw; + } + + return total; +} diff --git a/lib/full-write.h b/lib/full-write.h new file mode 100644 index 0000000..d20d2fe --- /dev/null +++ b/lib/full-write.h @@ -0,0 +1,35 @@ +/* An interface to write() that writes all it is asked to write. + + Copyright (C) 2002-2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + + +#ifdef __cplusplus +extern "C" { +#endif + + +/* Write COUNT bytes at BUF to descriptor FD, retrying if interrupted + or if partial writes occur. Return the number of bytes successfully + written, setting errno if that is less than COUNT. */ +extern size_t full_write (int fd, const void *buf, size_t count); + + +#ifdef __cplusplus +} +#endif diff --git a/lib/getcwd.c b/lib/getcwd.c new file mode 100644 index 0000000..23b35de --- /dev/null +++ b/lib/getcwd.c @@ -0,0 +1,428 @@ +/* Copyright (C) 1991,92,93,94,95,96,97,98,99,2004,2005,2006,2007 Free Software + Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if !_LIBC +# include <config.h> +# include <unistd.h> +# include "dirfd.h" +#endif + +#include <errno.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <stdbool.h> +#include <stddef.h> + +#include <fcntl.h> /* For AT_FDCWD on Solaris 9. */ + +#ifndef __set_errno +# define __set_errno(val) (errno = (val)) +#endif + +#include <dirent.h> +#ifndef _D_EXACT_NAMLEN +# define _D_EXACT_NAMLEN(d) strlen ((d)->d_name) +#endif +#ifndef _D_ALLOC_NAMLEN +# define _D_ALLOC_NAMLEN(d) (_D_EXACT_NAMLEN (d) + 1) +#endif + +#include <unistd.h> +#include <stdlib.h> +#include <string.h> + +#if _LIBC +# ifndef mempcpy +# define mempcpy __mempcpy +# endif +#endif + +#include <limits.h> + +/* Work around a bug in Solaris 9 and 10: AT_FDCWD is positive. Its + value exceeds INT_MAX, so its use as an int doesn't conform to the + C standard, and GCC and Sun C complain in some cases. */ +#if 0 < AT_FDCWD && AT_FDCWD == 0xffd19553 +# undef AT_FDCWD +# define AT_FDCWD (-3041965) +#endif + +#ifdef ENAMETOOLONG +# define is_ENAMETOOLONG(x) ((x) == ENAMETOOLONG) +#else +# define is_ENAMETOOLONG(x) 0 +#endif + +#ifndef MAX +# define MAX(a, b) ((a) < (b) ? (b) : (a)) +#endif +#ifndef MIN +# define MIN(a, b) ((a) < (b) ? (a) : (b)) +#endif + +#ifndef PATH_MAX +# ifdef MAXPATHLEN +# define PATH_MAX MAXPATHLEN +# else +# define PATH_MAX 1024 +# endif +#endif + +#if D_INO_IN_DIRENT +# define MATCHING_INO(dp, ino) ((dp)->d_ino == (ino)) +#else +# define MATCHING_INO(dp, ino) true +#endif + +#if !_LIBC +# define __getcwd getcwd +# define __lstat lstat +# define __closedir closedir +# define __opendir opendir +# define __readdir readdir +#endif + +/* The results of opendir() in this file are not used with dirfd and fchdir, + therefore save some unnecessary recursion in fchdir.c. */ +#undef opendir +#undef closedir + +/* Get the name of the current working directory, and put it in SIZE + bytes of BUF. Returns NULL if the directory couldn't be determined or + SIZE was too small. If successful, returns BUF. In GNU, if BUF is + NULL, an array is allocated with `malloc'; the array is SIZE bytes long, + unless SIZE == 0, in which case it is as big as necessary. */ + +char * +__getcwd (char *buf, size_t size) +{ + /* Lengths of big file name components and entire file names, and a + deep level of file name nesting. These numbers are not upper + bounds; they are merely large values suitable for initial + allocations, designed to be large enough for most real-world + uses. */ + enum + { + BIG_FILE_NAME_COMPONENT_LENGTH = 255, + BIG_FILE_NAME_LENGTH = MIN (4095, PATH_MAX - 1), + DEEP_NESTING = 100 + }; + +#ifdef AT_FDCWD + int fd = AT_FDCWD; + bool fd_needs_closing = false; +#else + char dots[DEEP_NESTING * sizeof ".." + BIG_FILE_NAME_COMPONENT_LENGTH + 1]; + char *dotlist = dots; + size_t dotsize = sizeof dots; + size_t dotlen = 0; +#endif + DIR *dirstream = NULL; + dev_t rootdev, thisdev; + ino_t rootino, thisino; + char *dir; + register char *dirp; + struct stat st; + size_t allocated = size; + size_t used; + +#if HAVE_PARTLY_WORKING_GETCWD + /* The system getcwd works, except it sometimes fails when it + shouldn't, setting errno to ERANGE, ENAMETOOLONG, or ENOENT. If + AT_FDCWD is not defined, the algorithm below is O(N**2) and this + is much slower than the system getcwd (at least on GNU/Linux). + So trust the system getcwd's results unless they look + suspicious. + + Use the system getcwd even if we have openat support, since the + system getcwd works even when a parent is unreadable, while the + openat-based approach does not. */ + +# undef getcwd + dir = getcwd (buf, size); + if (dir || (errno != ERANGE && !is_ENAMETOOLONG (errno) && errno != ENOENT)) + return dir; +#endif + + if (size == 0) + { + if (buf != NULL) + { + __set_errno (EINVAL); + return NULL; + } + + allocated = BIG_FILE_NAME_LENGTH + 1; + } + + if (buf == NULL) + { + dir = malloc (allocated); + if (dir == NULL) + return NULL; + } + else + dir = buf; + + dirp = dir + allocated; + *--dirp = '\0'; + + if (__lstat (".", &st) < 0) + goto lose; + thisdev = st.st_dev; + thisino = st.st_ino; + + if (__lstat ("/", &st) < 0) + goto lose; + rootdev = st.st_dev; + rootino = st.st_ino; + + while (!(thisdev == rootdev && thisino == rootino)) + { + struct dirent *d; + dev_t dotdev; + ino_t dotino; + bool mount_point; + int parent_status; + size_t dirroom; + size_t namlen; + bool use_d_ino = true; + + /* Look at the parent directory. */ +#ifdef AT_FDCWD + fd = openat (fd, "..", O_RDONLY); + if (fd < 0) + goto lose; + fd_needs_closing = true; + parent_status = fstat (fd, &st); +#else + dotlist[dotlen++] = '.'; + dotlist[dotlen++] = '.'; + dotlist[dotlen] = '\0'; + parent_status = __lstat (dotlist, &st); +#endif + if (parent_status != 0) + goto lose; + + if (dirstream && __closedir (dirstream) != 0) + { + dirstream = NULL; + goto lose; + } + + /* Figure out if this directory is a mount point. */ + dotdev = st.st_dev; + dotino = st.st_ino; + mount_point = dotdev != thisdev; + + /* Search for the last directory. */ +#ifdef AT_FDCWD + dirstream = fdopendir (fd); + if (dirstream == NULL) + goto lose; + /* Reset fd. It may have been closed by fdopendir. */ + fd = dirfd (dirstream); + fd_needs_closing = false; +#else + dirstream = __opendir (dotlist); + if (dirstream == NULL) + goto lose; + dotlist[dotlen++] = '/'; +#endif + for (;;) + { + /* Clear errno to distinguish EOF from error if readdir returns + NULL. */ + __set_errno (0); + d = __readdir (dirstream); + + /* When we've iterated through all directory entries without finding + one with a matching d_ino, rewind the stream and consider each + name again, but this time, using lstat. This is necessary in a + chroot on at least one system (glibc-2.3.6 + linux 2.6.12), where + .., ../.., ../../.., etc. all had the same device number, yet the + d_ino values for entries in / did not match those obtained + via lstat. */ + if (d == NULL && errno == 0 && use_d_ino) + { + use_d_ino = false; + rewinddir (dirstream); + d = __readdir (dirstream); + } + + if (d == NULL) + { + if (errno == 0) + /* EOF on dirstream, which can mean e.g., that the current + directory has been removed. */ + __set_errno (ENOENT); + goto lose; + } + if (d->d_name[0] == '.' && + (d->d_name[1] == '\0' || + (d->d_name[1] == '.' && d->d_name[2] == '\0'))) + continue; + + if (use_d_ino) + { + bool match = (MATCHING_INO (d, thisino) || mount_point); + if (! match) + continue; + } + + { + int entry_status; +#ifdef AT_FDCWD + entry_status = fstatat (fd, d->d_name, &st, AT_SYMLINK_NOFOLLOW); +#else + /* Compute size needed for this file name, or for the file + name ".." in the same directory, whichever is larger. + Room for ".." might be needed the next time through + the outer loop. */ + size_t name_alloc = _D_ALLOC_NAMLEN (d); + size_t filesize = dotlen + MAX (sizeof "..", name_alloc); + + if (filesize < dotlen) + goto memory_exhausted; + + if (dotsize < filesize) + { + /* My, what a deep directory tree you have, Grandma. */ + size_t newsize = MAX (filesize, dotsize * 2); + size_t i; + if (newsize < dotsize) + goto memory_exhausted; + if (dotlist != dots) + free (dotlist); + dotlist = malloc (newsize); + if (dotlist == NULL) + goto lose; + dotsize = newsize; + + i = 0; + do + { + dotlist[i++] = '.'; + dotlist[i++] = '.'; + dotlist[i++] = '/'; + } + while (i < dotlen); + } + + memcpy (dotlist + dotlen, d->d_name, _D_ALLOC_NAMLEN (d)); + entry_status = __lstat (dotlist, &st); +#endif + /* We don't fail here if we cannot stat() a directory entry. + This can happen when (network) file systems fail. If this + entry is in fact the one we are looking for we will find + out soon as we reach the end of the directory without + having found anything. */ + if (entry_status == 0 && S_ISDIR (st.st_mode) + && st.st_dev == thisdev && st.st_ino == thisino) + break; + } + } + + dirroom = dirp - dir; + namlen = _D_EXACT_NAMLEN (d); + + if (dirroom <= namlen) + { + if (size != 0) + { + __set_errno (ERANGE); + goto lose; + } + else + { + char *tmp; + size_t oldsize = allocated; + + allocated += MAX (allocated, namlen); + if (allocated < oldsize + || ! (tmp = realloc (dir, allocated))) + goto memory_exhausted; + + /* Move current contents up to the end of the buffer. + This is guaranteed to be non-overlapping. */ + dirp = memcpy (tmp + allocated - (oldsize - dirroom), + tmp + dirroom, + oldsize - dirroom); + dir = tmp; + } + } + dirp -= namlen; + memcpy (dirp, d->d_name, namlen); + *--dirp = '/'; + + thisdev = dotdev; + thisino = dotino; + } + + if (dirstream && __closedir (dirstream) != 0) + { + dirstream = NULL; + goto lose; + } + + if (dirp == &dir[allocated - 1]) + *--dirp = '/'; + +#ifndef AT_FDCWD + if (dotlist != dots) + free (dotlist); +#endif + + used = dir + allocated - dirp; + memmove (dir, dirp, used); + + if (size == 0) + /* Ensure that the buffer is only as large as necessary. */ + buf = realloc (dir, used); + + if (buf == NULL) + /* Either buf was NULL all along, or `realloc' failed but + we still have the original string. */ + buf = dir; + + return buf; + + memory_exhausted: + __set_errno (ENOMEM); + lose: + { + int save = errno; + if (dirstream) + __closedir (dirstream); +#ifdef AT_FDCWD + if (fd_needs_closing) + close (fd); +#else + if (dotlist != dots) + free (dotlist); +#endif + if (buf == NULL) + free (dir); + __set_errno (save); + } + return NULL; +} + +#ifdef weak_alias +weak_alias (__getcwd, getcwd) +#endif diff --git a/lib/getdate.c b/lib/getdate.c new file mode 100644 index 0000000..dca83d8 --- /dev/null +++ b/lib/getdate.c @@ -0,0 +1,3289 @@ +/* A Bison parser, made by GNU Bison 2.3. */ + +/* Skeleton implementation for Bison's Yacc-like parsers in C + + Copyright (C) 1984, 1989, 1990, 2000, 2001, 2002, 2003, 2004, 2005, 2006 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, + Boston, MA 02110-1301, USA. */ + +/* As a special exception, you may create a larger work that contains + part or all of the Bison parser skeleton and distribute that work + under terms of your choice, so long as that work isn't itself a + parser generator using the skeleton or a modified version thereof + as a parser skeleton. Alternatively, if you modify or redistribute + the parser skeleton itself, you may (at your option) remove this + special exception, which will cause the skeleton and the resulting + Bison output files to be licensed under the GNU General Public + License without this special exception. + + This special exception was added by the Free Software Foundation in + version 2.2 of Bison. */ + +/* C LALR(1) parser skeleton written by Richard Stallman, by + simplifying the original so-called "semantic" parser. */ + +/* All symbols defined below should begin with yy or YY, to avoid + infringing on user name space. This should be done even for local + variables, as they might otherwise be expanded by user macros. + There are some unavoidable exceptions within include files to + define necessary library symbols; they are noted "INFRINGES ON + USER NAME SPACE" below. */ + +/* Identify Bison output. */ +#define YYBISON 1 + +/* Bison version. */ +#define YYBISON_VERSION "2.3" + +/* Skeleton name. */ +#define YYSKELETON_NAME "yacc.c" + +/* Pure parsers. */ +#define YYPURE 1 + +/* Using locations. */ +#define YYLSP_NEEDED 0 + + + +/* Tokens. */ +#ifndef YYTOKENTYPE +# define YYTOKENTYPE + /* Put the tokens into the symbol table, so that GDB and other debuggers + know about them. */ + enum yytokentype { + tAGO = 258, + tDST = 259, + tYEAR_UNIT = 260, + tMONTH_UNIT = 261, + tHOUR_UNIT = 262, + tMINUTE_UNIT = 263, + tSEC_UNIT = 264, + tDAY_UNIT = 265, + tDAY = 266, + tDAYZONE = 267, + tLOCAL_ZONE = 268, + tMERIDIAN = 269, + tMONTH = 270, + tORDINAL = 271, + tZONE = 272, + tSNUMBER = 273, + tUNUMBER = 274, + tSDECIMAL_NUMBER = 275, + tUDECIMAL_NUMBER = 276 + }; +#endif +/* Tokens. */ +#define tAGO 258 +#define tDST 259 +#define tYEAR_UNIT 260 +#define tMONTH_UNIT 261 +#define tHOUR_UNIT 262 +#define tMINUTE_UNIT 263 +#define tSEC_UNIT 264 +#define tDAY_UNIT 265 +#define tDAY 266 +#define tDAYZONE 267 +#define tLOCAL_ZONE 268 +#define tMERIDIAN 269 +#define tMONTH 270 +#define tORDINAL 271 +#define tZONE 272 +#define tSNUMBER 273 +#define tUNUMBER 274 +#define tSDECIMAL_NUMBER 275 +#define tUDECIMAL_NUMBER 276 + + + + +/* Copy the first part of user declarations. */ +#line 1 "getdate.y" + +/* Parse a string into an internal time stamp. + + Copyright (C) 1999, 2000, 2002, 2003, 2004, 2005, 2006, 2007 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Originally written by Steven M. Bellovin <smb@research.att.com> while + at the University of North Carolina at Chapel Hill. Later tweaked by + a couple of people on Usenet. Completely overhauled by Rich $alz + <rsalz@bbn.com> and Jim Berets <jberets@bbn.com> in August, 1990. + + Modified by Paul Eggert <eggert@twinsun.com> in August 1999 to do + the right thing about local DST. Also modified by Paul Eggert + <eggert@cs.ucla.edu> in February 2004 to support + nanosecond-resolution time stamps, and in October 2004 to support + TZ strings in dates. */ + +/* FIXME: Check for arithmetic overflow in all cases, not just + some of them. */ + +#include <config.h> + +#include "getdate.h" +#include "timespec.h" + +/* There's no need to extend the stack, so there's no need to involve + alloca. */ +#define YYSTACK_USE_ALLOCA 0 + +/* Tell Bison how much stack space is needed. 20 should be plenty for + this grammar, which is not right recursive. Beware setting it too + high, since that might cause problems on machines whose + implementations have lame stack-overflow checking. */ +#define YYMAXDEPTH 20 +#define YYINITDEPTH YYMAXDEPTH + +/* Since the code of getdate.y is not included in the Emacs executable + itself, there is no need to #define static in this file. Even if + the code were included in the Emacs executable, it probably + wouldn't do any harm to #undef it here; this will only cause + problems if we try to write to a static variable, which I don't + think this code needs to do. */ +#ifdef emacs +# undef static +#endif + +#include <ctype.h> +#include <limits.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "setenv.h" +#include "xalloc.h" + + +/* ISDIGIT differs from isdigit, as follows: + - Its arg may be any int or unsigned int; it need not be an unsigned char + or EOF. + - It's typically faster. + POSIX says that only '0' through '9' are digits. Prefer ISDIGIT to + isdigit unless it's important to use the locale's definition + of `digit' even when the host does not conform to POSIX. */ +#define ISDIGIT(c) ((unsigned int) (c) - '0' <= 9) + +#ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +#endif + +#ifndef ATTRIBUTE_UNUSED +# define ATTRIBUTE_UNUSED __attribute__ ((__unused__)) +#endif + +/* Shift A right by B bits portably, by dividing A by 2**B and + truncating towards minus infinity. A and B should be free of side + effects, and B should be in the range 0 <= B <= INT_BITS - 2, where + INT_BITS is the number of useful bits in an int. GNU code can + assume that INT_BITS is at least 32. + + ISO C99 says that A >> B is implementation-defined if A < 0. Some + implementations (e.g., UNICOS 9.0 on a Cray Y-MP EL) don't shift + right in the usual way when A < 0, so SHR falls back on division if + ordinary A >> B doesn't seem to be the usual signed shift. */ +#define SHR(a, b) \ + (-1 >> 1 == -1 \ + ? (a) >> (b) \ + : (a) / (1 << (b)) - ((a) % (1 << (b)) < 0)) + +#define EPOCH_YEAR 1970 +#define TM_YEAR_BASE 1900 + +#define HOUR(x) ((x) * 60) + +/* An integer value, and the number of digits in its textual + representation. */ +typedef struct +{ + bool negative; + long int value; + size_t digits; +} textint; + +/* An entry in the lexical lookup table. */ +typedef struct +{ + char const *name; + int type; + int value; +} table; + +/* Meridian: am, pm, or 24-hour style. */ +enum { MERam, MERpm, MER24 }; + +enum { BILLION = 1000000000, LOG10_BILLION = 9 }; + +/* Relative times. */ +typedef struct +{ + /* Relative year, month, day, hour, minutes, seconds, and nanoseconds. */ + long int year; + long int month; + long int day; + long int hour; + long int minutes; + long int seconds; + long int ns; +} relative_time; + +#if HAVE_COMPOUND_LITERALS +# define RELATIVE_TIME_0 ((relative_time) { 0, 0, 0, 0, 0, 0, 0 }) +#else +static relative_time const RELATIVE_TIME_0; +#endif + +/* Information passed to and from the parser. */ +typedef struct +{ + /* The input string remaining to be parsed. */ + const char *input; + + /* N, if this is the Nth Tuesday. */ + long int day_ordinal; + + /* Day of week; Sunday is 0. */ + int day_number; + + /* tm_isdst flag for the local zone. */ + int local_isdst; + + /* Time zone, in minutes east of UTC. */ + long int time_zone; + + /* Style used for time. */ + int meridian; + + /* Gregorian year, month, day, hour, minutes, seconds, and nanoseconds. */ + textint year; + long int month; + long int day; + long int hour; + long int minutes; + struct timespec seconds; /* includes nanoseconds */ + + /* Relative year, month, day, hour, minutes, seconds, and nanoseconds. */ + relative_time rel; + + /* Presence or counts of nonterminals of various flavors parsed so far. */ + bool timespec_seen; + bool rels_seen; + size_t dates_seen; + size_t days_seen; + size_t local_zones_seen; + size_t dsts_seen; + size_t times_seen; + size_t zones_seen; + + /* Table of local time zone abbrevations, terminated by a null entry. */ + table local_time_zone_table[3]; +} parser_control; + +union YYSTYPE; +static int yylex (union YYSTYPE *, parser_control *); +static int yyerror (parser_control const *, char const *); +static long int time_zone_hhmm (textint, long int); + + + +/* Enabling traces. */ +#ifndef YYDEBUG +# define YYDEBUG 0 +#endif + +/* Enabling verbose error messages. */ +#ifdef YYERROR_VERBOSE +# undef YYERROR_VERBOSE +# define YYERROR_VERBOSE 1 +#else +# define YYERROR_VERBOSE 0 +#endif + +/* Enabling the token table. */ +#ifndef YYTOKEN_TABLE +# define YYTOKEN_TABLE 0 +#endif + +#if ! defined YYSTYPE && ! defined YYSTYPE_IS_DECLARED +typedef union YYSTYPE +#line 214 "getdate.y" +{ + long int intval; + textint textintval; + struct timespec timespec; + relative_time rel; +} +/* Line 187 of yacc.c. */ +#line 348 "getdate.c" + YYSTYPE; +# define yystype YYSTYPE /* obsolescent; will be withdrawn */ +# define YYSTYPE_IS_DECLARED 1 +# define YYSTYPE_IS_TRIVIAL 1 +#endif + + + +/* Copy the second part of user declarations. */ + + +/* Line 216 of yacc.c. */ +#line 361 "getdate.c" + +#ifdef short +# undef short +#endif + +#ifdef YYTYPE_UINT8 +typedef YYTYPE_UINT8 yytype_uint8; +#else +typedef unsigned char yytype_uint8; +#endif + +#ifdef YYTYPE_INT8 +typedef YYTYPE_INT8 yytype_int8; +#elif (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +typedef signed char yytype_int8; +#else +typedef short int yytype_int8; +#endif + +#ifdef YYTYPE_UINT16 +typedef YYTYPE_UINT16 yytype_uint16; +#else +typedef unsigned short int yytype_uint16; +#endif + +#ifdef YYTYPE_INT16 +typedef YYTYPE_INT16 yytype_int16; +#else +typedef short int yytype_int16; +#endif + +#ifndef YYSIZE_T +# ifdef __SIZE_TYPE__ +# define YYSIZE_T __SIZE_TYPE__ +# elif defined size_t +# define YYSIZE_T size_t +# elif ! defined YYSIZE_T && (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +# include <stddef.h> /* INFRINGES ON USER NAME SPACE */ +# define YYSIZE_T size_t +# else +# define YYSIZE_T unsigned int +# endif +#endif + +#define YYSIZE_MAXIMUM ((YYSIZE_T) -1) + +#ifndef YY_ +# if YYENABLE_NLS +# if ENABLE_NLS +# include <libintl.h> /* INFRINGES ON USER NAME SPACE */ +# define YY_(msgid) dgettext ("bison-runtime", msgid) +# endif +# endif +# ifndef YY_ +# define YY_(msgid) msgid +# endif +#endif + +/* Suppress unused-variable warnings by "using" E. */ +#if ! defined lint || defined __GNUC__ +# define YYUSE(e) ((void) (e)) +#else +# define YYUSE(e) /* empty */ +#endif + +/* Identity function, used to suppress warnings about constant conditions. */ +#ifndef lint +# define YYID(n) (n) +#else +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static int +YYID (int i) +#else +static int +YYID (i) + int i; +#endif +{ + return i; +} +#endif + +#if ! defined yyoverflow || YYERROR_VERBOSE + +/* The parser invokes alloca or malloc; define the necessary symbols. */ + +# ifdef YYSTACK_USE_ALLOCA +# if YYSTACK_USE_ALLOCA +# ifdef __GNUC__ +# define YYSTACK_ALLOC __builtin_alloca +# elif defined __BUILTIN_VA_ARG_INCR +# include <alloca.h> /* INFRINGES ON USER NAME SPACE */ +# elif defined _AIX +# define YYSTACK_ALLOC __alloca +# elif defined _MSC_VER +# include <malloc.h> /* INFRINGES ON USER NAME SPACE */ +# define alloca _alloca +# else +# define YYSTACK_ALLOC alloca +# if ! defined _ALLOCA_H && ! defined _STDLIB_H && (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +# include <stdlib.h> /* INFRINGES ON USER NAME SPACE */ +# ifndef _STDLIB_H +# define _STDLIB_H 1 +# endif +# endif +# endif +# endif +# endif + +# ifdef YYSTACK_ALLOC + /* Pacify GCC's `empty if-body' warning. */ +# define YYSTACK_FREE(Ptr) do { /* empty */; } while (YYID (0)) +# ifndef YYSTACK_ALLOC_MAXIMUM + /* The OS might guarantee only one guard page at the bottom of the stack, + and a page size can be as small as 4096 bytes. So we cannot safely + invoke alloca (N) if N exceeds 4096. Use a slightly smaller number + to allow for a few compiler-allocated temporary stack slots. */ +# define YYSTACK_ALLOC_MAXIMUM 4032 /* reasonable circa 2006 */ +# endif +# else +# define YYSTACK_ALLOC YYMALLOC +# define YYSTACK_FREE YYFREE +# ifndef YYSTACK_ALLOC_MAXIMUM +# define YYSTACK_ALLOC_MAXIMUM YYSIZE_MAXIMUM +# endif +# if (defined __cplusplus && ! defined _STDLIB_H \ + && ! ((defined YYMALLOC || defined malloc) \ + && (defined YYFREE || defined free))) +# include <stdlib.h> /* INFRINGES ON USER NAME SPACE */ +# ifndef _STDLIB_H +# define _STDLIB_H 1 +# endif +# endif +# ifndef YYMALLOC +# define YYMALLOC malloc +# if ! defined malloc && ! defined _STDLIB_H && (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +void *malloc (YYSIZE_T); /* INFRINGES ON USER NAME SPACE */ +# endif +# endif +# ifndef YYFREE +# define YYFREE free +# if ! defined free && ! defined _STDLIB_H && (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +void free (void *); /* INFRINGES ON USER NAME SPACE */ +# endif +# endif +# endif +#endif /* ! defined yyoverflow || YYERROR_VERBOSE */ + + +#if (! defined yyoverflow \ + && (! defined __cplusplus \ + || (defined YYSTYPE_IS_TRIVIAL && YYSTYPE_IS_TRIVIAL))) + +/* A type that is properly aligned for any stack member. */ +union yyalloc +{ + yytype_int16 yyss; + YYSTYPE yyvs; + }; + +/* The size of the maximum gap between one aligned stack and the next. */ +# define YYSTACK_GAP_MAXIMUM (sizeof (union yyalloc) - 1) + +/* The size of an array large to enough to hold all stacks, each with + N elements. */ +# define YYSTACK_BYTES(N) \ + ((N) * (sizeof (yytype_int16) + sizeof (YYSTYPE)) \ + + YYSTACK_GAP_MAXIMUM) + +/* Copy COUNT objects from FROM to TO. The source and destination do + not overlap. */ +# ifndef YYCOPY +# if defined __GNUC__ && 1 < __GNUC__ +# define YYCOPY(To, From, Count) \ + __builtin_memcpy (To, From, (Count) * sizeof (*(From))) +# else +# define YYCOPY(To, From, Count) \ + do \ + { \ + YYSIZE_T yyi; \ + for (yyi = 0; yyi < (Count); yyi++) \ + (To)[yyi] = (From)[yyi]; \ + } \ + while (YYID (0)) +# endif +# endif + +/* Relocate STACK from its old location to the new one. The + local variables YYSIZE and YYSTACKSIZE give the old and new number of + elements in the stack, and YYPTR gives the new location of the + stack. Advance YYPTR to a properly aligned location for the next + stack. */ +# define YYSTACK_RELOCATE(Stack) \ + do \ + { \ + YYSIZE_T yynewbytes; \ + YYCOPY (&yyptr->Stack, Stack, yysize); \ + Stack = &yyptr->Stack; \ + yynewbytes = yystacksize * sizeof (*Stack) + YYSTACK_GAP_MAXIMUM; \ + yyptr += yynewbytes / sizeof (*yyptr); \ + } \ + while (YYID (0)) + +#endif + +/* YYFINAL -- State number of the termination state. */ +#define YYFINAL 12 +/* YYLAST -- Last index in YYTABLE. */ +#define YYLAST 91 + +/* YYNTOKENS -- Number of terminals. */ +#define YYNTOKENS 26 +/* YYNNTS -- Number of nonterminals. */ +#define YYNNTS 19 +/* YYNRULES -- Number of rules. */ +#define YYNRULES 78 +/* YYNRULES -- Number of states. */ +#define YYNSTATES 96 + +/* YYTRANSLATE(YYLEX) -- Bison symbol number corresponding to YYLEX. */ +#define YYUNDEFTOK 2 +#define YYMAXUTOK 276 + +#define YYTRANSLATE(YYX) \ + ((unsigned int) (YYX) <= YYMAXUTOK ? yytranslate[YYX] : YYUNDEFTOK) + +/* YYTRANSLATE[YYLEX] -- Bison symbol number corresponding to YYLEX. */ +static const yytype_uint8 yytranslate[] = +{ + 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 24, 2, 2, 25, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 23, 2, + 2, 2, 2, 2, 22, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 1, 2, 3, 4, + 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, + 15, 16, 17, 18, 19, 20, 21 +}; + +#if YYDEBUG +/* YYPRHS[YYN] -- Index of the first RHS symbol of rule number YYN in + YYRHS. */ +static const yytype_uint8 yyprhs[] = +{ + 0, 0, 3, 5, 7, 10, 11, 14, 16, 18, + 20, 22, 24, 26, 28, 31, 36, 42, 49, 57, + 59, 62, 64, 67, 71, 73, 76, 78, 81, 84, + 87, 91, 97, 101, 105, 109, 112, 117, 120, 124, + 127, 129, 132, 135, 137, 140, 143, 145, 148, 151, + 153, 156, 159, 161, 164, 167, 169, 172, 175, 178, + 181, 183, 185, 188, 191, 194, 197, 200, 203, 205, + 207, 209, 211, 213, 215, 217, 218, 221, 222 +}; + +/* YYRHS -- A `-1'-separated list of the rules' RHS. */ +static const yytype_int8 yyrhs[] = +{ + 27, 0, -1, 28, -1, 29, -1, 22, 39, -1, + -1, 29, 30, -1, 31, -1, 32, -1, 33, -1, + 35, -1, 34, -1, 36, -1, 42, -1, 19, 14, + -1, 19, 23, 19, 44, -1, 19, 23, 19, 18, + 43, -1, 19, 23, 19, 23, 41, 44, -1, 19, + 23, 19, 23, 41, 18, 43, -1, 13, -1, 13, + 4, -1, 17, -1, 17, 38, -1, 17, 18, 43, + -1, 12, -1, 17, 4, -1, 11, -1, 11, 24, + -1, 16, 11, -1, 19, 11, -1, 19, 25, 19, + -1, 19, 25, 19, 25, 19, -1, 19, 18, 18, + -1, 19, 15, 18, -1, 15, 18, 18, -1, 15, + 19, -1, 15, 19, 24, 19, -1, 19, 15, -1, + 19, 15, 19, -1, 37, 3, -1, 37, -1, 16, + 5, -1, 19, 5, -1, 5, -1, 16, 6, -1, + 19, 6, -1, 6, -1, 16, 10, -1, 19, 10, + -1, 10, -1, 16, 7, -1, 19, 7, -1, 7, + -1, 16, 8, -1, 19, 8, -1, 8, -1, 16, + 9, -1, 19, 9, -1, 20, 9, -1, 21, 9, + -1, 9, -1, 38, -1, 18, 5, -1, 18, 6, + -1, 18, 10, -1, 18, 7, -1, 18, 8, -1, + 18, 9, -1, 40, -1, 41, -1, 20, -1, 18, + -1, 21, -1, 19, -1, 19, -1, -1, 23, 19, + -1, -1, 14, -1 +}; + +/* YYRLINE[YYN] -- source line where rule number YYN was defined. */ +static const yytype_uint16 yyrline[] = +{ + 0, 240, 240, 241, 245, 252, 254, 258, 260, 262, + 264, 266, 268, 270, 274, 282, 290, 300, 307, 319, + 324, 332, 334, 344, 346, 348, 353, 358, 363, 368, + 376, 381, 401, 408, 416, 424, 429, 435, 440, 449, + 459, 472, 474, 476, 478, 480, 482, 484, 486, 488, + 490, 492, 494, 496, 498, 500, 502, 504, 506, 508, + 510, 512, 516, 518, 520, 522, 524, 526, 530, 530, + 533, 534, 539, 540, 545, 583, 584, 590, 591 +}; +#endif + +#if YYDEBUG || YYERROR_VERBOSE || YYTOKEN_TABLE +/* YYTNAME[SYMBOL-NUM] -- String name of the symbol SYMBOL-NUM. + First, the terminals, then, starting at YYNTOKENS, nonterminals. */ +static const char *const yytname[] = +{ + "$end", "error", "$undefined", "tAGO", "tDST", "tYEAR_UNIT", + "tMONTH_UNIT", "tHOUR_UNIT", "tMINUTE_UNIT", "tSEC_UNIT", "tDAY_UNIT", + "tDAY", "tDAYZONE", "tLOCAL_ZONE", "tMERIDIAN", "tMONTH", "tORDINAL", + "tZONE", "tSNUMBER", "tUNUMBER", "tSDECIMAL_NUMBER", "tUDECIMAL_NUMBER", + "'@'", "':'", "','", "'/'", "$accept", "spec", "timespec", "items", + "item", "time", "local_zone", "zone", "day", "date", "rel", "relunit", + "relunit_snumber", "seconds", "signed_seconds", "unsigned_seconds", + "number", "o_colon_minutes", "o_merid", 0 +}; +#endif + +# ifdef YYPRINT +/* YYTOKNUM[YYLEX-NUM] -- Internal token number corresponding to + token YYLEX-NUM. */ +static const yytype_uint16 yytoknum[] = +{ + 0, 256, 257, 258, 259, 260, 261, 262, 263, 264, + 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, + 275, 276, 64, 58, 44, 47 +}; +# endif + +/* YYR1[YYN] -- Symbol number of symbol that rule YYN derives. */ +static const yytype_uint8 yyr1[] = +{ + 0, 26, 27, 27, 28, 29, 29, 30, 30, 30, + 30, 30, 30, 30, 31, 31, 31, 31, 31, 32, + 32, 33, 33, 33, 33, 33, 34, 34, 34, 34, + 35, 35, 35, 35, 35, 35, 35, 35, 35, 36, + 36, 37, 37, 37, 37, 37, 37, 37, 37, 37, + 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, + 37, 37, 38, 38, 38, 38, 38, 38, 39, 39, + 40, 40, 41, 41, 42, 43, 43, 44, 44 +}; + +/* YYR2[YYN] -- Number of symbols composing right hand side of rule YYN. */ +static const yytype_uint8 yyr2[] = +{ + 0, 2, 1, 1, 2, 0, 2, 1, 1, 1, + 1, 1, 1, 1, 2, 4, 5, 6, 7, 1, + 2, 1, 2, 3, 1, 2, 1, 2, 2, 2, + 3, 5, 3, 3, 3, 2, 4, 2, 3, 2, + 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, + 2, 2, 1, 2, 2, 1, 2, 2, 2, 2, + 1, 1, 2, 2, 2, 2, 2, 2, 1, 1, + 1, 1, 1, 1, 1, 0, 2, 0, 1 +}; + +/* YYDEFACT[STATE-NAME] -- Default rule to reduce with in state + STATE-NUM when YYTABLE doesn't specify something else to do. Zero + means the default is an error. */ +static const yytype_uint8 yydefact[] = +{ + 5, 0, 0, 2, 3, 71, 73, 70, 72, 4, + 68, 69, 1, 43, 46, 52, 55, 60, 49, 26, + 24, 19, 0, 0, 21, 0, 74, 0, 0, 6, + 7, 8, 9, 11, 10, 12, 40, 61, 13, 27, + 20, 0, 35, 41, 44, 50, 53, 56, 47, 28, + 25, 75, 22, 62, 63, 65, 66, 67, 64, 42, + 45, 51, 54, 57, 48, 29, 14, 37, 0, 0, + 0, 58, 59, 39, 34, 0, 0, 23, 33, 38, + 32, 77, 30, 36, 76, 78, 75, 0, 15, 0, + 16, 77, 31, 75, 17, 18 +}; + +/* YYDEFGOTO[NTERM-NUM]. */ +static const yytype_int8 yydefgoto[] = +{ + -1, 2, 3, 4, 29, 30, 31, 32, 33, 34, + 35, 36, 37, 9, 10, 11, 38, 77, 88 +}; + +/* YYPACT[STATE-NUM] -- Index in YYTABLE of the portion describing + STATE-NUM. */ +#define YYPACT_NINF -79 +static const yytype_int8 yypact[] = +{ + -10, 47, 27, -79, 25, -79, -79, -79, -79, -79, + -79, -79, -79, -79, -79, -79, -79, -79, -79, 5, + -79, 59, 43, 42, 10, 49, -5, 62, 63, -79, + -79, -79, -79, -79, -79, -79, 70, -79, -79, -79, + -79, 56, 52, -79, -79, -79, -79, -79, -79, -79, + -79, 16, -79, -79, -79, -79, -79, -79, -79, -79, + -79, -79, -79, -79, -79, -79, -79, 51, 57, 58, + 60, -79, -79, -79, -79, 61, 64, -79, -79, -79, + -79, -7, 53, -79, -79, -79, 65, -2, -79, 66, + -79, 46, -79, 65, -79, -79 +}; + +/* YYPGOTO[NTERM-NUM]. */ +static const yytype_int8 yypgoto[] = +{ + -79, -79, -79, -79, -79, -79, -79, -79, -79, -79, + -79, -79, 67, -79, -79, -6, -79, -78, -9 +}; + +/* YYTABLE[YYPACT[STATE-NUM]]. What to do in state STATE-NUM. If + positive, shift that token. If negative, reduce the rule which + number is the opposite. If zero, do what YYDEFACT says. + If YYTABLE_NINF, syntax error. */ +#define YYTABLE_NINF -1 +static const yytype_uint8 yytable[] = +{ + 59, 60, 61, 62, 63, 64, 65, 85, 90, 66, + 67, 86, 1, 68, 50, 95, 87, 6, 69, 8, + 70, 53, 54, 55, 56, 57, 58, 12, 51, 39, + 13, 14, 15, 16, 17, 18, 19, 20, 21, 76, + 22, 23, 24, 25, 26, 27, 28, 43, 44, 45, + 46, 47, 48, 49, 53, 54, 55, 56, 57, 58, + 85, 41, 42, 40, 93, 5, 6, 7, 8, 78, + 79, 71, 72, 73, 74, 80, 75, 81, 89, 82, + 83, 91, 94, 84, 0, 92, 0, 0, 76, 0, + 0, 52 +}; + +static const yytype_int8 yycheck[] = +{ + 5, 6, 7, 8, 9, 10, 11, 14, 86, 14, + 15, 18, 22, 18, 4, 93, 23, 19, 23, 21, + 25, 5, 6, 7, 8, 9, 10, 0, 18, 24, + 5, 6, 7, 8, 9, 10, 11, 12, 13, 23, + 15, 16, 17, 18, 19, 20, 21, 5, 6, 7, + 8, 9, 10, 11, 5, 6, 7, 8, 9, 10, + 14, 18, 19, 4, 18, 18, 19, 20, 21, 18, + 19, 9, 9, 3, 18, 18, 24, 19, 25, 19, + 19, 87, 91, 19, -1, 19, -1, -1, 23, -1, + -1, 24 +}; + +/* YYSTOS[STATE-NUM] -- The (internal number of the) accessing + symbol of state STATE-NUM. */ +static const yytype_uint8 yystos[] = +{ + 0, 22, 27, 28, 29, 18, 19, 20, 21, 39, + 40, 41, 0, 5, 6, 7, 8, 9, 10, 11, + 12, 13, 15, 16, 17, 18, 19, 20, 21, 30, + 31, 32, 33, 34, 35, 36, 37, 38, 42, 24, + 4, 18, 19, 5, 6, 7, 8, 9, 10, 11, + 4, 18, 38, 5, 6, 7, 8, 9, 10, 5, + 6, 7, 8, 9, 10, 11, 14, 15, 18, 23, + 25, 9, 9, 3, 18, 24, 23, 43, 18, 19, + 18, 19, 19, 19, 19, 14, 18, 23, 44, 25, + 43, 41, 19, 18, 44, 43 +}; + +#define yyerrok (yyerrstatus = 0) +#define yyclearin (yychar = YYEMPTY) +#define YYEMPTY (-2) +#define YYEOF 0 + +#define YYACCEPT goto yyacceptlab +#define YYABORT goto yyabortlab +#define YYERROR goto yyerrorlab + + +/* Like YYERROR except do call yyerror. This remains here temporarily + to ease the transition to the new meaning of YYERROR, for GCC. + Once GCC version 2 has supplanted version 1, this can go. */ + +#define YYFAIL goto yyerrlab + +#define YYRECOVERING() (!!yyerrstatus) + +#define YYBACKUP(Token, Value) \ +do \ + if (yychar == YYEMPTY && yylen == 1) \ + { \ + yychar = (Token); \ + yylval = (Value); \ + yytoken = YYTRANSLATE (yychar); \ + YYPOPSTACK (1); \ + goto yybackup; \ + } \ + else \ + { \ + yyerror (pc, YY_("syntax error: cannot back up")); \ + YYERROR; \ + } \ +while (YYID (0)) + + +#define YYTERROR 1 +#define YYERRCODE 256 + + +/* YYLLOC_DEFAULT -- Set CURRENT to span from RHS[1] to RHS[N]. + If N is 0, then set CURRENT to the empty location which ends + the previous symbol: RHS[0] (always defined). */ + +#define YYRHSLOC(Rhs, K) ((Rhs)[K]) +#ifndef YYLLOC_DEFAULT +# define YYLLOC_DEFAULT(Current, Rhs, N) \ + do \ + if (YYID (N)) \ + { \ + (Current).first_line = YYRHSLOC (Rhs, 1).first_line; \ + (Current).first_column = YYRHSLOC (Rhs, 1).first_column; \ + (Current).last_line = YYRHSLOC (Rhs, N).last_line; \ + (Current).last_column = YYRHSLOC (Rhs, N).last_column; \ + } \ + else \ + { \ + (Current).first_line = (Current).last_line = \ + YYRHSLOC (Rhs, 0).last_line; \ + (Current).first_column = (Current).last_column = \ + YYRHSLOC (Rhs, 0).last_column; \ + } \ + while (YYID (0)) +#endif + + +/* YY_LOCATION_PRINT -- Print the location on the stream. + This macro was not mandated originally: define only if we know + we won't break user code: when these are the locations we know. */ + +#ifndef YY_LOCATION_PRINT +# if YYLTYPE_IS_TRIVIAL +# define YY_LOCATION_PRINT(File, Loc) \ + fprintf (File, "%d.%d-%d.%d", \ + (Loc).first_line, (Loc).first_column, \ + (Loc).last_line, (Loc).last_column) +# else +# define YY_LOCATION_PRINT(File, Loc) ((void) 0) +# endif +#endif + + +/* YYLEX -- calling `yylex' with the right arguments. */ + +#ifdef YYLEX_PARAM +# define YYLEX yylex (&yylval, YYLEX_PARAM) +#else +# define YYLEX yylex (&yylval, pc) +#endif + +/* Enable debugging if requested. */ +#if YYDEBUG + +# ifndef YYFPRINTF +# include <stdio.h> /* INFRINGES ON USER NAME SPACE */ +# define YYFPRINTF fprintf +# endif + +# define YYDPRINTF(Args) \ +do { \ + if (yydebug) \ + YYFPRINTF Args; \ +} while (YYID (0)) + +# define YY_SYMBOL_PRINT(Title, Type, Value, Location) \ +do { \ + if (yydebug) \ + { \ + YYFPRINTF (stderr, "%s ", Title); \ + yy_symbol_print (stderr, \ + Type, Value, pc); \ + YYFPRINTF (stderr, "\n"); \ + } \ +} while (YYID (0)) + + +/*--------------------------------. +| Print this symbol on YYOUTPUT. | +`--------------------------------*/ + +/*ARGSUSED*/ +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static void +yy_symbol_value_print (FILE *yyoutput, int yytype, YYSTYPE const * const yyvaluep, parser_control *pc) +#else +static void +yy_symbol_value_print (yyoutput, yytype, yyvaluep, pc) + FILE *yyoutput; + int yytype; + YYSTYPE const * const yyvaluep; + parser_control *pc; +#endif +{ + if (!yyvaluep) + return; + YYUSE (pc); +# ifdef YYPRINT + if (yytype < YYNTOKENS) + YYPRINT (yyoutput, yytoknum[yytype], *yyvaluep); +# else + YYUSE (yyoutput); +# endif + switch (yytype) + { + default: + break; + } +} + + +/*--------------------------------. +| Print this symbol on YYOUTPUT. | +`--------------------------------*/ + +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static void +yy_symbol_print (FILE *yyoutput, int yytype, YYSTYPE const * const yyvaluep, parser_control *pc) +#else +static void +yy_symbol_print (yyoutput, yytype, yyvaluep, pc) + FILE *yyoutput; + int yytype; + YYSTYPE const * const yyvaluep; + parser_control *pc; +#endif +{ + if (yytype < YYNTOKENS) + YYFPRINTF (yyoutput, "token %s (", yytname[yytype]); + else + YYFPRINTF (yyoutput, "nterm %s (", yytname[yytype]); + + yy_symbol_value_print (yyoutput, yytype, yyvaluep, pc); + YYFPRINTF (yyoutput, ")"); +} + +/*------------------------------------------------------------------. +| yy_stack_print -- Print the state stack from its BOTTOM up to its | +| TOP (included). | +`------------------------------------------------------------------*/ + +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static void +yy_stack_print (yytype_int16 *bottom, yytype_int16 *top) +#else +static void +yy_stack_print (bottom, top) + yytype_int16 *bottom; + yytype_int16 *top; +#endif +{ + YYFPRINTF (stderr, "Stack now"); + for (; bottom <= top; ++bottom) + YYFPRINTF (stderr, " %d", *bottom); + YYFPRINTF (stderr, "\n"); +} + +# define YY_STACK_PRINT(Bottom, Top) \ +do { \ + if (yydebug) \ + yy_stack_print ((Bottom), (Top)); \ +} while (YYID (0)) + + +/*------------------------------------------------. +| Report that the YYRULE is going to be reduced. | +`------------------------------------------------*/ + +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static void +yy_reduce_print (YYSTYPE *yyvsp, int yyrule, parser_control *pc) +#else +static void +yy_reduce_print (yyvsp, yyrule, pc) + YYSTYPE *yyvsp; + int yyrule; + parser_control *pc; +#endif +{ + int yynrhs = yyr2[yyrule]; + int yyi; + unsigned long int yylno = yyrline[yyrule]; + YYFPRINTF (stderr, "Reducing stack by rule %d (line %lu):\n", + yyrule - 1, yylno); + /* The symbols being reduced. */ + for (yyi = 0; yyi < yynrhs; yyi++) + { + fprintf (stderr, " $%d = ", yyi + 1); + yy_symbol_print (stderr, yyrhs[yyprhs[yyrule] + yyi], + &(yyvsp[(yyi + 1) - (yynrhs)]) + , pc); + fprintf (stderr, "\n"); + } +} + +# define YY_REDUCE_PRINT(Rule) \ +do { \ + if (yydebug) \ + yy_reduce_print (yyvsp, Rule, pc); \ +} while (YYID (0)) + +/* Nonzero means print parse trace. It is left uninitialized so that + multiple parsers can coexist. */ +int yydebug; +#else /* !YYDEBUG */ +# define YYDPRINTF(Args) +# define YY_SYMBOL_PRINT(Title, Type, Value, Location) +# define YY_STACK_PRINT(Bottom, Top) +# define YY_REDUCE_PRINT(Rule) +#endif /* !YYDEBUG */ + + +/* YYINITDEPTH -- initial size of the parser's stacks. */ +#ifndef YYINITDEPTH +# define YYINITDEPTH 200 +#endif + +/* YYMAXDEPTH -- maximum size the stacks can grow to (effective only + if the built-in stack extension method is used). + + Do not make this value too large; the results are undefined if + YYSTACK_ALLOC_MAXIMUM < YYSTACK_BYTES (YYMAXDEPTH) + evaluated with infinite-precision integer arithmetic. */ + +#ifndef YYMAXDEPTH +# define YYMAXDEPTH 10000 +#endif + + + +#if YYERROR_VERBOSE + +# ifndef yystrlen +# if defined __GLIBC__ && defined _STRING_H +# define yystrlen strlen +# else +/* Return the length of YYSTR. */ +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static YYSIZE_T +yystrlen (const char *yystr) +#else +static YYSIZE_T +yystrlen (yystr) + const char *yystr; +#endif +{ + YYSIZE_T yylen; + for (yylen = 0; yystr[yylen]; yylen++) + continue; + return yylen; +} +# endif +# endif + +# ifndef yystpcpy +# if defined __GLIBC__ && defined _STRING_H && defined _GNU_SOURCE +# define yystpcpy stpcpy +# else +/* Copy YYSRC to YYDEST, returning the address of the terminating '\0' in + YYDEST. */ +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static char * +yystpcpy (char *yydest, const char *yysrc) +#else +static char * +yystpcpy (yydest, yysrc) + char *yydest; + const char *yysrc; +#endif +{ + char *yyd = yydest; + const char *yys = yysrc; + + while ((*yyd++ = *yys++) != '\0') + continue; + + return yyd - 1; +} +# endif +# endif + +# ifndef yytnamerr +/* Copy to YYRES the contents of YYSTR after stripping away unnecessary + quotes and backslashes, so that it's suitable for yyerror. The + heuristic is that double-quoting is unnecessary unless the string + contains an apostrophe, a comma, or backslash (other than + backslash-backslash). YYSTR is taken from yytname. If YYRES is + null, do not copy; instead, return the length of what the result + would have been. */ +static YYSIZE_T +yytnamerr (char *yyres, const char *yystr) +{ + if (*yystr == '"') + { + YYSIZE_T yyn = 0; + char const *yyp = yystr; + + for (;;) + switch (*++yyp) + { + case '\'': + case ',': + goto do_not_strip_quotes; + + case '\\': + if (*++yyp != '\\') + goto do_not_strip_quotes; + /* Fall through. */ + default: + if (yyres) + yyres[yyn] = *yyp; + yyn++; + break; + + case '"': + if (yyres) + yyres[yyn] = '\0'; + return yyn; + } + do_not_strip_quotes: ; + } + + if (! yyres) + return yystrlen (yystr); + + return yystpcpy (yyres, yystr) - yyres; +} +# endif + +/* Copy into YYRESULT an error message about the unexpected token + YYCHAR while in state YYSTATE. Return the number of bytes copied, + including the terminating null byte. If YYRESULT is null, do not + copy anything; just return the number of bytes that would be + copied. As a special case, return 0 if an ordinary "syntax error" + message will do. Return YYSIZE_MAXIMUM if overflow occurs during + size calculation. */ +static YYSIZE_T +yysyntax_error (char *yyresult, int yystate, int yychar) +{ + int yyn = yypact[yystate]; + + if (! (YYPACT_NINF < yyn && yyn <= YYLAST)) + return 0; + else + { + int yytype = YYTRANSLATE (yychar); + YYSIZE_T yysize0 = yytnamerr (0, yytname[yytype]); + YYSIZE_T yysize = yysize0; + YYSIZE_T yysize1; + int yysize_overflow = 0; + enum { YYERROR_VERBOSE_ARGS_MAXIMUM = 5 }; + char const *yyarg[YYERROR_VERBOSE_ARGS_MAXIMUM]; + int yyx; + +# if 0 + /* This is so xgettext sees the translatable formats that are + constructed on the fly. */ + YY_("syntax error, unexpected %s"); + YY_("syntax error, unexpected %s, expecting %s"); + YY_("syntax error, unexpected %s, expecting %s or %s"); + YY_("syntax error, unexpected %s, expecting %s or %s or %s"); + YY_("syntax error, unexpected %s, expecting %s or %s or %s or %s"); +# endif + char *yyfmt; + char const *yyf; + static char const yyunexpected[] = "syntax error, unexpected %s"; + static char const yyexpecting[] = ", expecting %s"; + static char const yyor[] = " or %s"; + char yyformat[sizeof yyunexpected + + sizeof yyexpecting - 1 + + ((YYERROR_VERBOSE_ARGS_MAXIMUM - 2) + * (sizeof yyor - 1))]; + char const *yyprefix = yyexpecting; + + /* Start YYX at -YYN if negative to avoid negative indexes in + YYCHECK. */ + int yyxbegin = yyn < 0 ? -yyn : 0; + + /* Stay within bounds of both yycheck and yytname. */ + int yychecklim = YYLAST - yyn + 1; + int yyxend = yychecklim < YYNTOKENS ? yychecklim : YYNTOKENS; + int yycount = 1; + + yyarg[0] = yytname[yytype]; + yyfmt = yystpcpy (yyformat, yyunexpected); + + for (yyx = yyxbegin; yyx < yyxend; ++yyx) + if (yycheck[yyx + yyn] == yyx && yyx != YYTERROR) + { + if (yycount == YYERROR_VERBOSE_ARGS_MAXIMUM) + { + yycount = 1; + yysize = yysize0; + yyformat[sizeof yyunexpected - 1] = '\0'; + break; + } + yyarg[yycount++] = yytname[yyx]; + yysize1 = yysize + yytnamerr (0, yytname[yyx]); + yysize_overflow |= (yysize1 < yysize); + yysize = yysize1; + yyfmt = yystpcpy (yyfmt, yyprefix); + yyprefix = yyor; + } + + yyf = YY_(yyformat); + yysize1 = yysize + yystrlen (yyf); + yysize_overflow |= (yysize1 < yysize); + yysize = yysize1; + + if (yysize_overflow) + return YYSIZE_MAXIMUM; + + if (yyresult) + { + /* Avoid sprintf, as that infringes on the user's name space. + Don't have undefined behavior even if the translation + produced a string with the wrong number of "%s"s. */ + char *yyp = yyresult; + int yyi = 0; + while ((*yyp = *yyf) != '\0') + { + if (*yyp == '%' && yyf[1] == 's' && yyi < yycount) + { + yyp += yytnamerr (yyp, yyarg[yyi++]); + yyf += 2; + } + else + { + yyp++; + yyf++; + } + } + } + return yysize; + } +} +#endif /* YYERROR_VERBOSE */ + + +/*-----------------------------------------------. +| Release the memory associated to this symbol. | +`-----------------------------------------------*/ + +/*ARGSUSED*/ +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +static void +yydestruct (const char *yymsg, int yytype, YYSTYPE *yyvaluep, parser_control *pc) +#else +static void +yydestruct (yymsg, yytype, yyvaluep, pc) + const char *yymsg; + int yytype; + YYSTYPE *yyvaluep; + parser_control *pc; +#endif +{ + YYUSE (yyvaluep); + YYUSE (pc); + + if (!yymsg) + yymsg = "Deleting"; + YY_SYMBOL_PRINT (yymsg, yytype, yyvaluep, yylocationp); + + switch (yytype) + { + + default: + break; + } +} + + +/* Prevent warnings from -Wmissing-prototypes. */ + +#ifdef YYPARSE_PARAM +#if defined __STDC__ || defined __cplusplus +int yyparse (void *YYPARSE_PARAM); +#else +int yyparse (); +#endif +#else /* ! YYPARSE_PARAM */ +#if defined __STDC__ || defined __cplusplus +int yyparse (parser_control *pc); +#else +int yyparse (); +#endif +#endif /* ! YYPARSE_PARAM */ + + + + + + +/*----------. +| yyparse. | +`----------*/ + +#ifdef YYPARSE_PARAM +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +int +yyparse (void *YYPARSE_PARAM) +#else +int +yyparse (YYPARSE_PARAM) + void *YYPARSE_PARAM; +#endif +#else /* ! YYPARSE_PARAM */ +#if (defined __STDC__ || defined __C99__FUNC__ \ + || defined __cplusplus || defined _MSC_VER) +int +yyparse (parser_control *pc) +#else +int +yyparse (pc) + parser_control *pc; +#endif +#endif +{ + /* The look-ahead symbol. */ +int yychar; + +/* The semantic value of the look-ahead symbol. */ +YYSTYPE yylval; + +/* Number of syntax errors so far. */ +int yynerrs; + + int yystate; + int yyn; + int yyresult; + /* Number of tokens to shift before error messages enabled. */ + int yyerrstatus; + /* Look-ahead token as an internal (translated) token number. */ + int yytoken = 0; +#if YYERROR_VERBOSE + /* Buffer for error messages, and its allocated size. */ + char yymsgbuf[128]; + char *yymsg = yymsgbuf; + YYSIZE_T yymsg_alloc = sizeof yymsgbuf; +#endif + + /* Three stacks and their tools: + `yyss': related to states, + `yyvs': related to semantic values, + `yyls': related to locations. + + Refer to the stacks thru separate pointers, to allow yyoverflow + to reallocate them elsewhere. */ + + /* The state stack. */ + yytype_int16 yyssa[YYINITDEPTH]; + yytype_int16 *yyss = yyssa; + yytype_int16 *yyssp; + + /* The semantic value stack. */ + YYSTYPE yyvsa[YYINITDEPTH]; + YYSTYPE *yyvs = yyvsa; + YYSTYPE *yyvsp; + + + +#define YYPOPSTACK(N) (yyvsp -= (N), yyssp -= (N)) + + YYSIZE_T yystacksize = YYINITDEPTH; + + /* The variables used to return semantic value and location from the + action routines. */ + YYSTYPE yyval; + + + /* The number of symbols on the RHS of the reduced rule. + Keep to zero when no symbol should be popped. */ + int yylen = 0; + + YYDPRINTF ((stderr, "Starting parse\n")); + + yystate = 0; + yyerrstatus = 0; + yynerrs = 0; + yychar = YYEMPTY; /* Cause a token to be read. */ + + /* Initialize stack pointers. + Waste one element of value and location stack + so that they stay on the same level as the state stack. + The wasted elements are never initialized. */ + + yyssp = yyss; + yyvsp = yyvs; + + goto yysetstate; + +/*------------------------------------------------------------. +| yynewstate -- Push a new state, which is found in yystate. | +`------------------------------------------------------------*/ + yynewstate: + /* In all cases, when you get here, the value and location stacks + have just been pushed. So pushing a state here evens the stacks. */ + yyssp++; + + yysetstate: + *yyssp = yystate; + + if (yyss + yystacksize - 1 <= yyssp) + { + /* Get the current used size of the three stacks, in elements. */ + YYSIZE_T yysize = yyssp - yyss + 1; + +#ifdef yyoverflow + { + /* Give user a chance to reallocate the stack. Use copies of + these so that the &'s don't force the real ones into + memory. */ + YYSTYPE *yyvs1 = yyvs; + yytype_int16 *yyss1 = yyss; + + + /* Each stack pointer address is followed by the size of the + data in use in that stack, in bytes. This used to be a + conditional around just the two extra args, but that might + be undefined if yyoverflow is a macro. */ + yyoverflow (YY_("memory exhausted"), + &yyss1, yysize * sizeof (*yyssp), + &yyvs1, yysize * sizeof (*yyvsp), + + &yystacksize); + + yyss = yyss1; + yyvs = yyvs1; + } +#else /* no yyoverflow */ +# ifndef YYSTACK_RELOCATE + goto yyexhaustedlab; +# else + /* Extend the stack our own way. */ + if (YYMAXDEPTH <= yystacksize) + goto yyexhaustedlab; + yystacksize *= 2; + if (YYMAXDEPTH < yystacksize) + yystacksize = YYMAXDEPTH; + + { + yytype_int16 *yyss1 = yyss; + union yyalloc *yyptr = + (union yyalloc *) YYSTACK_ALLOC (YYSTACK_BYTES (yystacksize)); + if (! yyptr) + goto yyexhaustedlab; + YYSTACK_RELOCATE (yyss); + YYSTACK_RELOCATE (yyvs); + +# undef YYSTACK_RELOCATE + if (yyss1 != yyssa) + YYSTACK_FREE (yyss1); + } +# endif +#endif /* no yyoverflow */ + + yyssp = yyss + yysize - 1; + yyvsp = yyvs + yysize - 1; + + + YYDPRINTF ((stderr, "Stack size increased to %lu\n", + (unsigned long int) yystacksize)); + + if (yyss + yystacksize - 1 <= yyssp) + YYABORT; + } + + YYDPRINTF ((stderr, "Entering state %d\n", yystate)); + + goto yybackup; + +/*-----------. +| yybackup. | +`-----------*/ +yybackup: + + /* Do appropriate processing given the current state. Read a + look-ahead token if we need one and don't already have one. */ + + /* First try to decide what to do without reference to look-ahead token. */ + yyn = yypact[yystate]; + if (yyn == YYPACT_NINF) + goto yydefault; + + /* Not known => get a look-ahead token if don't already have one. */ + + /* YYCHAR is either YYEMPTY or YYEOF or a valid look-ahead symbol. */ + if (yychar == YYEMPTY) + { + YYDPRINTF ((stderr, "Reading a token: ")); + yychar = YYLEX; + } + + if (yychar <= YYEOF) + { + yychar = yytoken = YYEOF; + YYDPRINTF ((stderr, "Now at end of input.\n")); + } + else + { + yytoken = YYTRANSLATE (yychar); + YY_SYMBOL_PRINT ("Next token is", yytoken, &yylval, &yylloc); + } + + /* If the proper action on seeing token YYTOKEN is to reduce or to + detect an error, take that action. */ + yyn += yytoken; + if (yyn < 0 || YYLAST < yyn || yycheck[yyn] != yytoken) + goto yydefault; + yyn = yytable[yyn]; + if (yyn <= 0) + { + if (yyn == 0 || yyn == YYTABLE_NINF) + goto yyerrlab; + yyn = -yyn; + goto yyreduce; + } + + if (yyn == YYFINAL) + YYACCEPT; + + /* Count tokens shifted since error; after three, turn off error + status. */ + if (yyerrstatus) + yyerrstatus--; + + /* Shift the look-ahead token. */ + YY_SYMBOL_PRINT ("Shifting", yytoken, &yylval, &yylloc); + + /* Discard the shifted token unless it is eof. */ + if (yychar != YYEOF) + yychar = YYEMPTY; + + yystate = yyn; + *++yyvsp = yylval; + + goto yynewstate; + + +/*-----------------------------------------------------------. +| yydefault -- do the default action for the current state. | +`-----------------------------------------------------------*/ +yydefault: + yyn = yydefact[yystate]; + if (yyn == 0) + goto yyerrlab; + goto yyreduce; + + +/*-----------------------------. +| yyreduce -- Do a reduction. | +`-----------------------------*/ +yyreduce: + /* yyn is the number of a rule to reduce with. */ + yylen = yyr2[yyn]; + + /* If YYLEN is nonzero, implement the default value of the action: + `$$ = $1'. + + Otherwise, the following line sets YYVAL to garbage. + This behavior is undocumented and Bison + users should not rely upon it. Assigning to YYVAL + unconditionally makes the parser a bit smaller, and it avoids a + GCC warning that YYVAL may be used uninitialized. */ + yyval = yyvsp[1-yylen]; + + + YY_REDUCE_PRINT (yyn); + switch (yyn) + { + case 4: +#line 246 "getdate.y" + { + pc->seconds = (yyvsp[(2) - (2)].timespec); + pc->timespec_seen = true; + } + break; + + case 7: +#line 259 "getdate.y" + { pc->times_seen++; } + break; + + case 8: +#line 261 "getdate.y" + { pc->local_zones_seen++; } + break; + + case 9: +#line 263 "getdate.y" + { pc->zones_seen++; } + break; + + case 10: +#line 265 "getdate.y" + { pc->dates_seen++; } + break; + + case 11: +#line 267 "getdate.y" + { pc->days_seen++; } + break; + + case 12: +#line 269 "getdate.y" + { pc->rels_seen = true; } + break; + + case 14: +#line 275 "getdate.y" + { + pc->hour = (yyvsp[(1) - (2)].textintval).value; + pc->minutes = 0; + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = (yyvsp[(2) - (2)].intval); + } + break; + + case 15: +#line 283 "getdate.y" + { + pc->hour = (yyvsp[(1) - (4)].textintval).value; + pc->minutes = (yyvsp[(3) - (4)].textintval).value; + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = (yyvsp[(4) - (4)].intval); + } + break; + + case 16: +#line 291 "getdate.y" + { + pc->hour = (yyvsp[(1) - (5)].textintval).value; + pc->minutes = (yyvsp[(3) - (5)].textintval).value; + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = MER24; + pc->zones_seen++; + pc->time_zone = time_zone_hhmm ((yyvsp[(4) - (5)].textintval), (yyvsp[(5) - (5)].intval)); + } + break; + + case 17: +#line 301 "getdate.y" + { + pc->hour = (yyvsp[(1) - (6)].textintval).value; + pc->minutes = (yyvsp[(3) - (6)].textintval).value; + pc->seconds = (yyvsp[(5) - (6)].timespec); + pc->meridian = (yyvsp[(6) - (6)].intval); + } + break; + + case 18: +#line 308 "getdate.y" + { + pc->hour = (yyvsp[(1) - (7)].textintval).value; + pc->minutes = (yyvsp[(3) - (7)].textintval).value; + pc->seconds = (yyvsp[(5) - (7)].timespec); + pc->meridian = MER24; + pc->zones_seen++; + pc->time_zone = time_zone_hhmm ((yyvsp[(6) - (7)].textintval), (yyvsp[(7) - (7)].intval)); + } + break; + + case 19: +#line 320 "getdate.y" + { + pc->local_isdst = (yyvsp[(1) - (1)].intval); + pc->dsts_seen += (0 < (yyvsp[(1) - (1)].intval)); + } + break; + + case 20: +#line 325 "getdate.y" + { + pc->local_isdst = 1; + pc->dsts_seen += (0 < (yyvsp[(1) - (2)].intval)) + 1; + } + break; + + case 21: +#line 333 "getdate.y" + { pc->time_zone = (yyvsp[(1) - (1)].intval); } + break; + + case 22: +#line 335 "getdate.y" + { pc->time_zone = (yyvsp[(1) - (2)].intval); + pc->rel.ns += (yyvsp[(2) - (2)].rel).ns; + pc->rel.seconds += (yyvsp[(2) - (2)].rel).seconds; + pc->rel.minutes += (yyvsp[(2) - (2)].rel).minutes; + pc->rel.hour += (yyvsp[(2) - (2)].rel).hour; + pc->rel.day += (yyvsp[(2) - (2)].rel).day; + pc->rel.month += (yyvsp[(2) - (2)].rel).month; + pc->rel.year += (yyvsp[(2) - (2)].rel).year; + pc->rels_seen = true; } + break; + + case 23: +#line 345 "getdate.y" + { pc->time_zone = (yyvsp[(1) - (3)].intval) + time_zone_hhmm ((yyvsp[(2) - (3)].textintval), (yyvsp[(3) - (3)].intval)); } + break; + + case 24: +#line 347 "getdate.y" + { pc->time_zone = (yyvsp[(1) - (1)].intval) + 60; } + break; + + case 25: +#line 349 "getdate.y" + { pc->time_zone = (yyvsp[(1) - (2)].intval) + 60; } + break; + + case 26: +#line 354 "getdate.y" + { + pc->day_ordinal = 1; + pc->day_number = (yyvsp[(1) - (1)].intval); + } + break; + + case 27: +#line 359 "getdate.y" + { + pc->day_ordinal = 1; + pc->day_number = (yyvsp[(1) - (2)].intval); + } + break; + + case 28: +#line 364 "getdate.y" + { + pc->day_ordinal = (yyvsp[(1) - (2)].intval); + pc->day_number = (yyvsp[(2) - (2)].intval); + } + break; + + case 29: +#line 369 "getdate.y" + { + pc->day_ordinal = (yyvsp[(1) - (2)].textintval).value; + pc->day_number = (yyvsp[(2) - (2)].intval); + } + break; + + case 30: +#line 377 "getdate.y" + { + pc->month = (yyvsp[(1) - (3)].textintval).value; + pc->day = (yyvsp[(3) - (3)].textintval).value; + } + break; + + case 31: +#line 382 "getdate.y" + { + /* Interpret as YYYY/MM/DD if the first value has 4 or more digits, + otherwise as MM/DD/YY. + The goal in recognizing YYYY/MM/DD is solely to support legacy + machine-generated dates like those in an RCS log listing. If + you want portability, use the ISO 8601 format. */ + if (4 <= (yyvsp[(1) - (5)].textintval).digits) + { + pc->year = (yyvsp[(1) - (5)].textintval); + pc->month = (yyvsp[(3) - (5)].textintval).value; + pc->day = (yyvsp[(5) - (5)].textintval).value; + } + else + { + pc->month = (yyvsp[(1) - (5)].textintval).value; + pc->day = (yyvsp[(3) - (5)].textintval).value; + pc->year = (yyvsp[(5) - (5)].textintval); + } + } + break; + + case 32: +#line 402 "getdate.y" + { + /* ISO 8601 format. YYYY-MM-DD. */ + pc->year = (yyvsp[(1) - (3)].textintval); + pc->month = -(yyvsp[(2) - (3)].textintval).value; + pc->day = -(yyvsp[(3) - (3)].textintval).value; + } + break; + + case 33: +#line 409 "getdate.y" + { + /* e.g. 17-JUN-1992. */ + pc->day = (yyvsp[(1) - (3)].textintval).value; + pc->month = (yyvsp[(2) - (3)].intval); + pc->year.value = -(yyvsp[(3) - (3)].textintval).value; + pc->year.digits = (yyvsp[(3) - (3)].textintval).digits; + } + break; + + case 34: +#line 417 "getdate.y" + { + /* e.g. JUN-17-1992. */ + pc->month = (yyvsp[(1) - (3)].intval); + pc->day = -(yyvsp[(2) - (3)].textintval).value; + pc->year.value = -(yyvsp[(3) - (3)].textintval).value; + pc->year.digits = (yyvsp[(3) - (3)].textintval).digits; + } + break; + + case 35: +#line 425 "getdate.y" + { + pc->month = (yyvsp[(1) - (2)].intval); + pc->day = (yyvsp[(2) - (2)].textintval).value; + } + break; + + case 36: +#line 430 "getdate.y" + { + pc->month = (yyvsp[(1) - (4)].intval); + pc->day = (yyvsp[(2) - (4)].textintval).value; + pc->year = (yyvsp[(4) - (4)].textintval); + } + break; + + case 37: +#line 436 "getdate.y" + { + pc->day = (yyvsp[(1) - (2)].textintval).value; + pc->month = (yyvsp[(2) - (2)].intval); + } + break; + + case 38: +#line 441 "getdate.y" + { + pc->day = (yyvsp[(1) - (3)].textintval).value; + pc->month = (yyvsp[(2) - (3)].intval); + pc->year = (yyvsp[(3) - (3)].textintval); + } + break; + + case 39: +#line 450 "getdate.y" + { + pc->rel.ns -= (yyvsp[(1) - (2)].rel).ns; + pc->rel.seconds -= (yyvsp[(1) - (2)].rel).seconds; + pc->rel.minutes -= (yyvsp[(1) - (2)].rel).minutes; + pc->rel.hour -= (yyvsp[(1) - (2)].rel).hour; + pc->rel.day -= (yyvsp[(1) - (2)].rel).day; + pc->rel.month -= (yyvsp[(1) - (2)].rel).month; + pc->rel.year -= (yyvsp[(1) - (2)].rel).year; + } + break; + + case 40: +#line 460 "getdate.y" + { + pc->rel.ns += (yyvsp[(1) - (1)].rel).ns; + pc->rel.seconds += (yyvsp[(1) - (1)].rel).seconds; + pc->rel.minutes += (yyvsp[(1) - (1)].rel).minutes; + pc->rel.hour += (yyvsp[(1) - (1)].rel).hour; + pc->rel.day += (yyvsp[(1) - (1)].rel).day; + pc->rel.month += (yyvsp[(1) - (1)].rel).month; + pc->rel.year += (yyvsp[(1) - (1)].rel).year; + } + break; + + case 41: +#line 473 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = (yyvsp[(1) - (2)].intval); } + break; + + case 42: +#line 475 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 43: +#line 477 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = 1; } + break; + + case 44: +#line 479 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = (yyvsp[(1) - (2)].intval); } + break; + + case 45: +#line 481 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 46: +#line 483 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = 1; } + break; + + case 47: +#line 485 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).day = (yyvsp[(1) - (2)].intval) * (yyvsp[(2) - (2)].intval); } + break; + + case 48: +#line 487 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).day = (yyvsp[(1) - (2)].textintval).value * (yyvsp[(2) - (2)].intval); } + break; + + case 49: +#line 489 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).day = (yyvsp[(1) - (1)].intval); } + break; + + case 50: +#line 491 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = (yyvsp[(1) - (2)].intval); } + break; + + case 51: +#line 493 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 52: +#line 495 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = 1; } + break; + + case 53: +#line 497 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = (yyvsp[(1) - (2)].intval); } + break; + + case 54: +#line 499 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 55: +#line 501 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = 1; } + break; + + case 56: +#line 503 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[(1) - (2)].intval); } + break; + + case 57: +#line 505 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 58: +#line 507 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[(1) - (2)].timespec).tv_sec; (yyval.rel).ns = (yyvsp[(1) - (2)].timespec).tv_nsec; } + break; + + case 59: +#line 509 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[(1) - (2)].timespec).tv_sec; (yyval.rel).ns = (yyvsp[(1) - (2)].timespec).tv_nsec; } + break; + + case 60: +#line 511 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = 1; } + break; + + case 62: +#line 517 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 63: +#line 519 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 64: +#line 521 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).day = (yyvsp[(1) - (2)].textintval).value * (yyvsp[(2) - (2)].intval); } + break; + + case 65: +#line 523 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 66: +#line 525 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 67: +#line 527 "getdate.y" + { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[(1) - (2)].textintval).value; } + break; + + case 71: +#line 535 "getdate.y" + { (yyval.timespec).tv_sec = (yyvsp[(1) - (1)].textintval).value; (yyval.timespec).tv_nsec = 0; } + break; + + case 73: +#line 541 "getdate.y" + { (yyval.timespec).tv_sec = (yyvsp[(1) - (1)].textintval).value; (yyval.timespec).tv_nsec = 0; } + break; + + case 74: +#line 546 "getdate.y" + { + if (pc->dates_seen && ! pc->year.digits + && ! pc->rels_seen && (pc->times_seen || 2 < (yyvsp[(1) - (1)].textintval).digits)) + pc->year = (yyvsp[(1) - (1)].textintval); + else + { + if (4 < (yyvsp[(1) - (1)].textintval).digits) + { + pc->dates_seen++; + pc->day = (yyvsp[(1) - (1)].textintval).value % 100; + pc->month = ((yyvsp[(1) - (1)].textintval).value / 100) % 100; + pc->year.value = (yyvsp[(1) - (1)].textintval).value / 10000; + pc->year.digits = (yyvsp[(1) - (1)].textintval).digits - 4; + } + else + { + pc->times_seen++; + if ((yyvsp[(1) - (1)].textintval).digits <= 2) + { + pc->hour = (yyvsp[(1) - (1)].textintval).value; + pc->minutes = 0; + } + else + { + pc->hour = (yyvsp[(1) - (1)].textintval).value / 100; + pc->minutes = (yyvsp[(1) - (1)].textintval).value % 100; + } + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = MER24; + } + } + } + break; + + case 75: +#line 583 "getdate.y" + { (yyval.intval) = -1; } + break; + + case 76: +#line 585 "getdate.y" + { (yyval.intval) = (yyvsp[(2) - (2)].textintval).value; } + break; + + case 77: +#line 590 "getdate.y" + { (yyval.intval) = MER24; } + break; + + case 78: +#line 592 "getdate.y" + { (yyval.intval) = (yyvsp[(1) - (1)].intval); } + break; + + +/* Line 1267 of yacc.c. */ +#line 2149 "getdate.c" + default: break; + } + YY_SYMBOL_PRINT ("-> $$ =", yyr1[yyn], &yyval, &yyloc); + + YYPOPSTACK (yylen); + yylen = 0; + YY_STACK_PRINT (yyss, yyssp); + + *++yyvsp = yyval; + + + /* Now `shift' the result of the reduction. Determine what state + that goes to, based on the state we popped back to and the rule + number reduced by. */ + + yyn = yyr1[yyn]; + + yystate = yypgoto[yyn - YYNTOKENS] + *yyssp; + if (0 <= yystate && yystate <= YYLAST && yycheck[yystate] == *yyssp) + yystate = yytable[yystate]; + else + yystate = yydefgoto[yyn - YYNTOKENS]; + + goto yynewstate; + + +/*------------------------------------. +| yyerrlab -- here on detecting error | +`------------------------------------*/ +yyerrlab: + /* If not already recovering from an error, report this error. */ + if (!yyerrstatus) + { + ++yynerrs; +#if ! YYERROR_VERBOSE + yyerror (pc, YY_("syntax error")); +#else + { + YYSIZE_T yysize = yysyntax_error (0, yystate, yychar); + if (yymsg_alloc < yysize && yymsg_alloc < YYSTACK_ALLOC_MAXIMUM) + { + YYSIZE_T yyalloc = 2 * yysize; + if (! (yysize <= yyalloc && yyalloc <= YYSTACK_ALLOC_MAXIMUM)) + yyalloc = YYSTACK_ALLOC_MAXIMUM; + if (yymsg != yymsgbuf) + YYSTACK_FREE (yymsg); + yymsg = (char *) YYSTACK_ALLOC (yyalloc); + if (yymsg) + yymsg_alloc = yyalloc; + else + { + yymsg = yymsgbuf; + yymsg_alloc = sizeof yymsgbuf; + } + } + + if (0 < yysize && yysize <= yymsg_alloc) + { + (void) yysyntax_error (yymsg, yystate, yychar); + yyerror (pc, yymsg); + } + else + { + yyerror (pc, YY_("syntax error")); + if (yysize != 0) + goto yyexhaustedlab; + } + } +#endif + } + + + + if (yyerrstatus == 3) + { + /* If just tried and failed to reuse look-ahead token after an + error, discard it. */ + + if (yychar <= YYEOF) + { + /* Return failure if at end of input. */ + if (yychar == YYEOF) + YYABORT; + } + else + { + yydestruct ("Error: discarding", + yytoken, &yylval, pc); + yychar = YYEMPTY; + } + } + + /* Else will try to reuse look-ahead token after shifting the error + token. */ + goto yyerrlab1; + + +/*---------------------------------------------------. +| yyerrorlab -- error raised explicitly by YYERROR. | +`---------------------------------------------------*/ +yyerrorlab: + + /* Pacify compilers like GCC when the user code never invokes + YYERROR and the label yyerrorlab therefore never appears in user + code. */ + if (/*CONSTCOND*/ 0) + goto yyerrorlab; + + /* Do not reclaim the symbols of the rule which action triggered + this YYERROR. */ + YYPOPSTACK (yylen); + yylen = 0; + YY_STACK_PRINT (yyss, yyssp); + yystate = *yyssp; + goto yyerrlab1; + + +/*-------------------------------------------------------------. +| yyerrlab1 -- common code for both syntax error and YYERROR. | +`-------------------------------------------------------------*/ +yyerrlab1: + yyerrstatus = 3; /* Each real token shifted decrements this. */ + + for (;;) + { + yyn = yypact[yystate]; + if (yyn != YYPACT_NINF) + { + yyn += YYTERROR; + if (0 <= yyn && yyn <= YYLAST && yycheck[yyn] == YYTERROR) + { + yyn = yytable[yyn]; + if (0 < yyn) + break; + } + } + + /* Pop the current state because it cannot handle the error token. */ + if (yyssp == yyss) + YYABORT; + + + yydestruct ("Error: popping", + yystos[yystate], yyvsp, pc); + YYPOPSTACK (1); + yystate = *yyssp; + YY_STACK_PRINT (yyss, yyssp); + } + + if (yyn == YYFINAL) + YYACCEPT; + + *++yyvsp = yylval; + + + /* Shift the error token. */ + YY_SYMBOL_PRINT ("Shifting", yystos[yyn], yyvsp, yylsp); + + yystate = yyn; + goto yynewstate; + + +/*-------------------------------------. +| yyacceptlab -- YYACCEPT comes here. | +`-------------------------------------*/ +yyacceptlab: + yyresult = 0; + goto yyreturn; + +/*-----------------------------------. +| yyabortlab -- YYABORT comes here. | +`-----------------------------------*/ +yyabortlab: + yyresult = 1; + goto yyreturn; + +#ifndef yyoverflow +/*-------------------------------------------------. +| yyexhaustedlab -- memory exhaustion comes here. | +`-------------------------------------------------*/ +yyexhaustedlab: + yyerror (pc, YY_("memory exhausted")); + yyresult = 2; + /* Fall through. */ +#endif + +yyreturn: + if (yychar != YYEOF && yychar != YYEMPTY) + yydestruct ("Cleanup: discarding lookahead", + yytoken, &yylval, pc); + /* Do not reclaim the symbols of the rule which action triggered + this YYABORT or YYACCEPT. */ + YYPOPSTACK (yylen); + YY_STACK_PRINT (yyss, yyssp); + while (yyssp != yyss) + { + yydestruct ("Cleanup: popping", + yystos[*yyssp], yyvsp, pc); + YYPOPSTACK (1); + } +#ifndef yyoverflow + if (yyss != yyssa) + YYSTACK_FREE (yyss); +#endif +#if YYERROR_VERBOSE + if (yymsg != yymsgbuf) + YYSTACK_FREE (yymsg); +#endif + /* Make sure YYID is used. */ + return YYID (yyresult); +} + + +#line 595 "getdate.y" + + +static table const meridian_table[] = +{ + { "AM", tMERIDIAN, MERam }, + { "A.M.", tMERIDIAN, MERam }, + { "PM", tMERIDIAN, MERpm }, + { "P.M.", tMERIDIAN, MERpm }, + { NULL, 0, 0 } +}; + +static table const dst_table[] = +{ + { "DST", tDST, 0 } +}; + +static table const month_and_day_table[] = +{ + { "JANUARY", tMONTH, 1 }, + { "FEBRUARY", tMONTH, 2 }, + { "MARCH", tMONTH, 3 }, + { "APRIL", tMONTH, 4 }, + { "MAY", tMONTH, 5 }, + { "JUNE", tMONTH, 6 }, + { "JULY", tMONTH, 7 }, + { "AUGUST", tMONTH, 8 }, + { "SEPTEMBER",tMONTH, 9 }, + { "SEPT", tMONTH, 9 }, + { "OCTOBER", tMONTH, 10 }, + { "NOVEMBER", tMONTH, 11 }, + { "DECEMBER", tMONTH, 12 }, + { "SUNDAY", tDAY, 0 }, + { "MONDAY", tDAY, 1 }, + { "TUESDAY", tDAY, 2 }, + { "TUES", tDAY, 2 }, + { "WEDNESDAY",tDAY, 3 }, + { "WEDNES", tDAY, 3 }, + { "THURSDAY", tDAY, 4 }, + { "THUR", tDAY, 4 }, + { "THURS", tDAY, 4 }, + { "FRIDAY", tDAY, 5 }, + { "SATURDAY", tDAY, 6 }, + { NULL, 0, 0 } +}; + +static table const time_units_table[] = +{ + { "YEAR", tYEAR_UNIT, 1 }, + { "MONTH", tMONTH_UNIT, 1 }, + { "FORTNIGHT",tDAY_UNIT, 14 }, + { "WEEK", tDAY_UNIT, 7 }, + { "DAY", tDAY_UNIT, 1 }, + { "HOUR", tHOUR_UNIT, 1 }, + { "MINUTE", tMINUTE_UNIT, 1 }, + { "MIN", tMINUTE_UNIT, 1 }, + { "SECOND", tSEC_UNIT, 1 }, + { "SEC", tSEC_UNIT, 1 }, + { NULL, 0, 0 } +}; + +/* Assorted relative-time words. */ +static table const relative_time_table[] = +{ + { "TOMORROW", tDAY_UNIT, 1 }, + { "YESTERDAY",tDAY_UNIT, -1 }, + { "TODAY", tDAY_UNIT, 0 }, + { "NOW", tDAY_UNIT, 0 }, + { "LAST", tORDINAL, -1 }, + { "THIS", tORDINAL, 0 }, + { "NEXT", tORDINAL, 1 }, + { "FIRST", tORDINAL, 1 }, +/*{ "SECOND", tORDINAL, 2 }, */ + { "THIRD", tORDINAL, 3 }, + { "FOURTH", tORDINAL, 4 }, + { "FIFTH", tORDINAL, 5 }, + { "SIXTH", tORDINAL, 6 }, + { "SEVENTH", tORDINAL, 7 }, + { "EIGHTH", tORDINAL, 8 }, + { "NINTH", tORDINAL, 9 }, + { "TENTH", tORDINAL, 10 }, + { "ELEVENTH", tORDINAL, 11 }, + { "TWELFTH", tORDINAL, 12 }, + { "AGO", tAGO, 1 }, + { NULL, 0, 0 } +}; + +/* The universal time zone table. These labels can be used even for + time stamps that would not otherwise be valid, e.g., GMT time + stamps in London during summer. */ +static table const universal_time_zone_table[] = +{ + { "GMT", tZONE, HOUR ( 0) }, /* Greenwich Mean */ + { "UT", tZONE, HOUR ( 0) }, /* Universal (Coordinated) */ + { "UTC", tZONE, HOUR ( 0) }, + { NULL, 0, 0 } +}; + +/* The time zone table. This table is necessarily incomplete, as time + zone abbreviations are ambiguous; e.g. Australians interpret "EST" + as Eastern time in Australia, not as US Eastern Standard Time. + You cannot rely on getdate to handle arbitrary time zone + abbreviations; use numeric abbreviations like `-0500' instead. */ +static table const time_zone_table[] = +{ + { "WET", tZONE, HOUR ( 0) }, /* Western European */ + { "WEST", tDAYZONE, HOUR ( 0) }, /* Western European Summer */ + { "BST", tDAYZONE, HOUR ( 0) }, /* British Summer */ + { "ART", tZONE, -HOUR ( 3) }, /* Argentina */ + { "BRT", tZONE, -HOUR ( 3) }, /* Brazil */ + { "BRST", tDAYZONE, -HOUR ( 3) }, /* Brazil Summer */ + { "NST", tZONE, -(HOUR ( 3) + 30) }, /* Newfoundland Standard */ + { "NDT", tDAYZONE,-(HOUR ( 3) + 30) }, /* Newfoundland Daylight */ + { "AST", tZONE, -HOUR ( 4) }, /* Atlantic Standard */ + { "ADT", tDAYZONE, -HOUR ( 4) }, /* Atlantic Daylight */ + { "CLT", tZONE, -HOUR ( 4) }, /* Chile */ + { "CLST", tDAYZONE, -HOUR ( 4) }, /* Chile Summer */ + { "EST", tZONE, -HOUR ( 5) }, /* Eastern Standard */ + { "EDT", tDAYZONE, -HOUR ( 5) }, /* Eastern Daylight */ + { "CST", tZONE, -HOUR ( 6) }, /* Central Standard */ + { "CDT", tDAYZONE, -HOUR ( 6) }, /* Central Daylight */ + { "MST", tZONE, -HOUR ( 7) }, /* Mountain Standard */ + { "MDT", tDAYZONE, -HOUR ( 7) }, /* Mountain Daylight */ + { "PST", tZONE, -HOUR ( 8) }, /* Pacific Standard */ + { "PDT", tDAYZONE, -HOUR ( 8) }, /* Pacific Daylight */ + { "AKST", tZONE, -HOUR ( 9) }, /* Alaska Standard */ + { "AKDT", tDAYZONE, -HOUR ( 9) }, /* Alaska Daylight */ + { "HST", tZONE, -HOUR (10) }, /* Hawaii Standard */ + { "HAST", tZONE, -HOUR (10) }, /* Hawaii-Aleutian Standard */ + { "HADT", tDAYZONE, -HOUR (10) }, /* Hawaii-Aleutian Daylight */ + { "SST", tZONE, -HOUR (12) }, /* Samoa Standard */ + { "WAT", tZONE, HOUR ( 1) }, /* West Africa */ + { "CET", tZONE, HOUR ( 1) }, /* Central European */ + { "CEST", tDAYZONE, HOUR ( 1) }, /* Central European Summer */ + { "MET", tZONE, HOUR ( 1) }, /* Middle European */ + { "MEZ", tZONE, HOUR ( 1) }, /* Middle European */ + { "MEST", tDAYZONE, HOUR ( 1) }, /* Middle European Summer */ + { "MESZ", tDAYZONE, HOUR ( 1) }, /* Middle European Summer */ + { "EET", tZONE, HOUR ( 2) }, /* Eastern European */ + { "EEST", tDAYZONE, HOUR ( 2) }, /* Eastern European Summer */ + { "CAT", tZONE, HOUR ( 2) }, /* Central Africa */ + { "SAST", tZONE, HOUR ( 2) }, /* South Africa Standard */ + { "EAT", tZONE, HOUR ( 3) }, /* East Africa */ + { "MSK", tZONE, HOUR ( 3) }, /* Moscow */ + { "MSD", tDAYZONE, HOUR ( 3) }, /* Moscow Daylight */ + { "IST", tZONE, (HOUR ( 5) + 30) }, /* India Standard */ + { "SGT", tZONE, HOUR ( 8) }, /* Singapore */ + { "KST", tZONE, HOUR ( 9) }, /* Korea Standard */ + { "JST", tZONE, HOUR ( 9) }, /* Japan Standard */ + { "GST", tZONE, HOUR (10) }, /* Guam Standard */ + { "NZST", tZONE, HOUR (12) }, /* New Zealand Standard */ + { "NZDT", tDAYZONE, HOUR (12) }, /* New Zealand Daylight */ + { NULL, 0, 0 } +}; + +/* Military time zone table. */ +static table const military_table[] = +{ + { "A", tZONE, -HOUR ( 1) }, + { "B", tZONE, -HOUR ( 2) }, + { "C", tZONE, -HOUR ( 3) }, + { "D", tZONE, -HOUR ( 4) }, + { "E", tZONE, -HOUR ( 5) }, + { "F", tZONE, -HOUR ( 6) }, + { "G", tZONE, -HOUR ( 7) }, + { "H", tZONE, -HOUR ( 8) }, + { "I", tZONE, -HOUR ( 9) }, + { "K", tZONE, -HOUR (10) }, + { "L", tZONE, -HOUR (11) }, + { "M", tZONE, -HOUR (12) }, + { "N", tZONE, HOUR ( 1) }, + { "O", tZONE, HOUR ( 2) }, + { "P", tZONE, HOUR ( 3) }, + { "Q", tZONE, HOUR ( 4) }, + { "R", tZONE, HOUR ( 5) }, + { "S", tZONE, HOUR ( 6) }, + { "T", tZONE, HOUR ( 7) }, + { "U", tZONE, HOUR ( 8) }, + { "V", tZONE, HOUR ( 9) }, + { "W", tZONE, HOUR (10) }, + { "X", tZONE, HOUR (11) }, + { "Y", tZONE, HOUR (12) }, + { "Z", tZONE, HOUR ( 0) }, + { NULL, 0, 0 } +}; + + + +/* Convert a time zone expressed as HH:MM into an integer count of + minutes. If MM is negative, then S is of the form HHMM and needs + to be picked apart; otherwise, S is of the form HH. */ + +static long int +time_zone_hhmm (textint s, long int mm) +{ + if (mm < 0) + return (s.value / 100) * 60 + s.value % 100; + else + return s.value * 60 + (s.negative ? -mm : mm); +} + +static int +to_hour (long int hours, int meridian) +{ + switch (meridian) + { + default: /* Pacify GCC. */ + case MER24: + return 0 <= hours && hours < 24 ? hours : -1; + case MERam: + return 0 < hours && hours < 12 ? hours : hours == 12 ? 0 : -1; + case MERpm: + return 0 < hours && hours < 12 ? hours + 12 : hours == 12 ? 12 : -1; + } +} + +static long int +to_year (textint textyear) +{ + long int year = textyear.value; + + if (year < 0) + year = -year; + + /* XPG4 suggests that years 00-68 map to 2000-2068, and + years 69-99 map to 1969-1999. */ + else if (textyear.digits == 2) + year += year < 69 ? 2000 : 1900; + + return year; +} + +static table const * +lookup_zone (parser_control const *pc, char const *name) +{ + table const *tp; + + for (tp = universal_time_zone_table; tp->name; tp++) + if (strcmp (name, tp->name) == 0) + return tp; + + /* Try local zone abbreviations before those in time_zone_table, as + the local ones are more likely to be right. */ + for (tp = pc->local_time_zone_table; tp->name; tp++) + if (strcmp (name, tp->name) == 0) + return tp; + + for (tp = time_zone_table; tp->name; tp++) + if (strcmp (name, tp->name) == 0) + return tp; + + return NULL; +} + +#if ! HAVE_TM_GMTOFF +/* Yield the difference between *A and *B, + measured in seconds, ignoring leap seconds. + The body of this function is taken directly from the GNU C Library; + see src/strftime.c. */ +static long int +tm_diff (struct tm const *a, struct tm const *b) +{ + /* Compute intervening leap days correctly even if year is negative. + Take care to avoid int overflow in leap day calculations. */ + int a4 = SHR (a->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (a->tm_year & 3); + int b4 = SHR (b->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (b->tm_year & 3); + int a100 = a4 / 25 - (a4 % 25 < 0); + int b100 = b4 / 25 - (b4 % 25 < 0); + int a400 = SHR (a100, 2); + int b400 = SHR (b100, 2); + int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400); + long int ayear = a->tm_year; + long int years = ayear - b->tm_year; + long int days = (365 * years + intervening_leap_days + + (a->tm_yday - b->tm_yday)); + return (60 * (60 * (24 * days + (a->tm_hour - b->tm_hour)) + + (a->tm_min - b->tm_min)) + + (a->tm_sec - b->tm_sec)); +} +#endif /* ! HAVE_TM_GMTOFF */ + +static table const * +lookup_word (parser_control const *pc, char *word) +{ + char *p; + char *q; + size_t wordlen; + table const *tp; + bool period_found; + bool abbrev; + + /* Make it uppercase. */ + for (p = word; *p; p++) + { + unsigned char ch = *p; + *p = toupper (ch); + } + + for (tp = meridian_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + + /* See if we have an abbreviation for a month. */ + wordlen = strlen (word); + abbrev = wordlen == 3 || (wordlen == 4 && word[3] == '.'); + + for (tp = month_and_day_table; tp->name; tp++) + if ((abbrev ? strncmp (word, tp->name, 3) : strcmp (word, tp->name)) == 0) + return tp; + + if ((tp = lookup_zone (pc, word))) + return tp; + + if (strcmp (word, dst_table[0].name) == 0) + return dst_table; + + for (tp = time_units_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + + /* Strip off any plural and try the units table again. */ + if (word[wordlen - 1] == 'S') + { + word[wordlen - 1] = '\0'; + for (tp = time_units_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + word[wordlen - 1] = 'S'; /* For "this" in relative_time_table. */ + } + + for (tp = relative_time_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + + /* Military time zones. */ + if (wordlen == 1) + for (tp = military_table; tp->name; tp++) + if (word[0] == tp->name[0]) + return tp; + + /* Drop out any periods and try the time zone table again. */ + for (period_found = false, p = q = word; (*p = *q); q++) + if (*q == '.') + period_found = true; + else + p++; + if (period_found && (tp = lookup_zone (pc, word))) + return tp; + + return NULL; +} + +static int +yylex (YYSTYPE *lvalp, parser_control *pc) +{ + unsigned char c; + size_t count; + + for (;;) + { + while (c = *pc->input, isspace (c)) + pc->input++; + + if (ISDIGIT (c) || c == '-' || c == '+') + { + char const *p; + int sign; + unsigned long int value; + if (c == '-' || c == '+') + { + sign = c == '-' ? -1 : 1; + while (c = *++pc->input, isspace (c)) + continue; + if (! ISDIGIT (c)) + /* skip the '-' sign */ + continue; + } + else + sign = 0; + p = pc->input; + for (value = 0; ; value *= 10) + { + unsigned long int value1 = value + (c - '0'); + if (value1 < value) + return '?'; + value = value1; + c = *++p; + if (! ISDIGIT (c)) + break; + if (ULONG_MAX / 10 < value) + return '?'; + } + if ((c == '.' || c == ',') && ISDIGIT (p[1])) + { + time_t s; + int ns; + int digits; + unsigned long int value1; + + /* Check for overflow when converting value to time_t. */ + if (sign < 0) + { + s = - value; + if (0 < s) + return '?'; + value1 = -s; + } + else + { + s = value; + if (s < 0) + return '?'; + value1 = s; + } + if (value != value1) + return '?'; + + /* Accumulate fraction, to ns precision. */ + p++; + ns = *p++ - '0'; + for (digits = 2; digits <= LOG10_BILLION; digits++) + { + ns *= 10; + if (ISDIGIT (*p)) + ns += *p++ - '0'; + } + + /* Skip excess digits, truncating toward -Infinity. */ + if (sign < 0) + for (; ISDIGIT (*p); p++) + if (*p != '0') + { + ns++; + break; + } + while (ISDIGIT (*p)) + p++; + + /* Adjust to the timespec convention, which is that + tv_nsec is always a positive offset even if tv_sec is + negative. */ + if (sign < 0 && ns) + { + s--; + if (! (s < 0)) + return '?'; + ns = BILLION - ns; + } + + lvalp->timespec.tv_sec = s; + lvalp->timespec.tv_nsec = ns; + pc->input = p; + return sign ? tSDECIMAL_NUMBER : tUDECIMAL_NUMBER; + } + else + { + lvalp->textintval.negative = sign < 0; + if (sign < 0) + { + lvalp->textintval.value = - value; + if (0 < lvalp->textintval.value) + return '?'; + } + else + { + lvalp->textintval.value = value; + if (lvalp->textintval.value < 0) + return '?'; + } + lvalp->textintval.digits = p - pc->input; + pc->input = p; + return sign ? tSNUMBER : tUNUMBER; + } + } + + if (isalpha (c)) + { + char buff[20]; + char *p = buff; + table const *tp; + + do + { + if (p < buff + sizeof buff - 1) + *p++ = c; + c = *++pc->input; + } + while (isalpha (c) || c == '.'); + + *p = '\0'; + tp = lookup_word (pc, buff); + if (! tp) + return '?'; + lvalp->intval = tp->value; + return tp->type; + } + + if (c != '(') + return *pc->input++; + count = 0; + do + { + c = *pc->input++; + if (c == '\0') + return c; + if (c == '(') + count++; + else if (c == ')') + count--; + } + while (count != 0); + } +} + +/* Do nothing if the parser reports an error. */ +static int +yyerror (parser_control const *pc ATTRIBUTE_UNUSED, + char const *s ATTRIBUTE_UNUSED) +{ + return 0; +} + +/* If *TM0 is the old and *TM1 is the new value of a struct tm after + passing it to mktime, return true if it's OK that mktime returned T. + It's not OK if *TM0 has out-of-range members. */ + +static bool +mktime_ok (struct tm const *tm0, struct tm const *tm1, time_t t) +{ + if (t == (time_t) -1) + { + /* Guard against falsely reporting an error when parsing a time + stamp that happens to equal (time_t) -1, on a host that + supports such a time stamp. */ + tm1 = localtime (&t); + if (!tm1) + return false; + } + + return ! ((tm0->tm_sec ^ tm1->tm_sec) + | (tm0->tm_min ^ tm1->tm_min) + | (tm0->tm_hour ^ tm1->tm_hour) + | (tm0->tm_mday ^ tm1->tm_mday) + | (tm0->tm_mon ^ tm1->tm_mon) + | (tm0->tm_year ^ tm1->tm_year)); +} + +/* A reasonable upper bound for the size of ordinary TZ strings. + Use heap allocation if TZ's length exceeds this. */ +enum { TZBUFSIZE = 100 }; + +/* Return a copy of TZ, stored in TZBUF if it fits, and heap-allocated + otherwise. */ +static char * +get_tz (char tzbuf[TZBUFSIZE]) +{ + char *tz = getenv ("TZ"); + if (tz) + { + size_t tzsize = strlen (tz) + 1; + tz = (tzsize <= TZBUFSIZE + ? memcpy (tzbuf, tz, tzsize) + : xmemdup (tz, tzsize)); + } + return tz; +} + +/* Parse a date/time string, storing the resulting time value into *RESULT. + The string itself is pointed to by P. Return true if successful. + P can be an incomplete or relative time specification; if so, use + *NOW as the basis for the returned time. */ +bool +get_date (struct timespec *result, char const *p, struct timespec const *now) +{ + time_t Start; + long int Start_ns; + struct tm const *tmp; + struct tm tm; + struct tm tm0; + parser_control pc; + struct timespec gettime_buffer; + unsigned char c; + bool tz_was_altered = false; + char *tz0 = NULL; + char tz0buf[TZBUFSIZE]; + bool ok = true; + + if (! now) + { + gettime (&gettime_buffer); + now = &gettime_buffer; + } + + Start = now->tv_sec; + Start_ns = now->tv_nsec; + + tmp = localtime (&now->tv_sec); + if (! tmp) + return false; + + while (c = *p, isspace (c)) + p++; + + if (strncmp (p, "TZ=\"", 4) == 0) + { + char const *tzbase = p + 4; + size_t tzsize = 1; + char const *s; + + for (s = tzbase; *s; s++, tzsize++) + if (*s == '\\') + { + s++; + if (! (*s == '\\' || *s == '"')) + break; + } + else if (*s == '"') + { + char *z; + char *tz1; + char tz1buf[TZBUFSIZE]; + bool large_tz = TZBUFSIZE < tzsize; + bool setenv_ok; + tz0 = get_tz (tz0buf); + z = tz1 = large_tz ? xmalloc (tzsize) : tz1buf; + for (s = tzbase; *s != '"'; s++) + *z++ = *(s += *s == '\\'); + *z = '\0'; + setenv_ok = setenv ("TZ", tz1, 1) == 0; + if (large_tz) + free (tz1); + if (!setenv_ok) + goto fail; + tz_was_altered = true; + p = s + 1; + } + } + + pc.input = p; + pc.year.value = tmp->tm_year; + pc.year.value += TM_YEAR_BASE; + pc.year.digits = 0; + pc.month = tmp->tm_mon + 1; + pc.day = tmp->tm_mday; + pc.hour = tmp->tm_hour; + pc.minutes = tmp->tm_min; + pc.seconds.tv_sec = tmp->tm_sec; + pc.seconds.tv_nsec = Start_ns; + tm.tm_isdst = tmp->tm_isdst; + + pc.meridian = MER24; + pc.rel = RELATIVE_TIME_0; + pc.timespec_seen = false; + pc.rels_seen = false; + pc.dates_seen = 0; + pc.days_seen = 0; + pc.times_seen = 0; + pc.local_zones_seen = 0; + pc.dsts_seen = 0; + pc.zones_seen = 0; + +#if HAVE_STRUCT_TM_TM_ZONE + pc.local_time_zone_table[0].name = tmp->tm_zone; + pc.local_time_zone_table[0].type = tLOCAL_ZONE; + pc.local_time_zone_table[0].value = tmp->tm_isdst; + pc.local_time_zone_table[1].name = NULL; + + /* Probe the names used in the next three calendar quarters, looking + for a tm_isdst different from the one we already have. */ + { + int quarter; + for (quarter = 1; quarter <= 3; quarter++) + { + time_t probe = Start + quarter * (90 * 24 * 60 * 60); + struct tm const *probe_tm = localtime (&probe); + if (probe_tm && probe_tm->tm_zone + && probe_tm->tm_isdst != pc.local_time_zone_table[0].value) + { + { + pc.local_time_zone_table[1].name = probe_tm->tm_zone; + pc.local_time_zone_table[1].type = tLOCAL_ZONE; + pc.local_time_zone_table[1].value = probe_tm->tm_isdst; + pc.local_time_zone_table[2].name = NULL; + } + break; + } + } + } +#else +#if HAVE_TZNAME + { +# ifndef tzname + extern char *tzname[]; +# endif + int i; + for (i = 0; i < 2; i++) + { + pc.local_time_zone_table[i].name = tzname[i]; + pc.local_time_zone_table[i].type = tLOCAL_ZONE; + pc.local_time_zone_table[i].value = i; + } + pc.local_time_zone_table[i].name = NULL; + } +#else + pc.local_time_zone_table[0].name = NULL; +#endif +#endif + + if (pc.local_time_zone_table[0].name && pc.local_time_zone_table[1].name + && ! strcmp (pc.local_time_zone_table[0].name, + pc.local_time_zone_table[1].name)) + { + /* This locale uses the same abbrevation for standard and + daylight times. So if we see that abbreviation, we don't + know whether it's daylight time. */ + pc.local_time_zone_table[0].value = -1; + pc.local_time_zone_table[1].name = NULL; + } + + if (yyparse (&pc) != 0) + goto fail; + + if (pc.timespec_seen) + *result = pc.seconds; + else + { + if (1 < (pc.times_seen | pc.dates_seen | pc.days_seen | pc.dsts_seen + | (pc.local_zones_seen + pc.zones_seen))) + goto fail; + + tm.tm_year = to_year (pc.year) - TM_YEAR_BASE; + tm.tm_mon = pc.month - 1; + tm.tm_mday = pc.day; + if (pc.times_seen || (pc.rels_seen && ! pc.dates_seen && ! pc.days_seen)) + { + tm.tm_hour = to_hour (pc.hour, pc.meridian); + if (tm.tm_hour < 0) + goto fail; + tm.tm_min = pc.minutes; + tm.tm_sec = pc.seconds.tv_sec; + } + else + { + tm.tm_hour = tm.tm_min = tm.tm_sec = 0; + pc.seconds.tv_nsec = 0; + } + + /* Let mktime deduce tm_isdst if we have an absolute time stamp. */ + if (pc.dates_seen | pc.days_seen | pc.times_seen) + tm.tm_isdst = -1; + + /* But if the input explicitly specifies local time with or without + DST, give mktime that information. */ + if (pc.local_zones_seen) + tm.tm_isdst = pc.local_isdst; + + tm0 = tm; + + Start = mktime (&tm); + + if (! mktime_ok (&tm0, &tm, Start)) + { + if (! pc.zones_seen) + goto fail; + else + { + /* Guard against falsely reporting errors near the time_t + boundaries when parsing times in other time zones. For + example, suppose the input string "1969-12-31 23:00:00 -0100", + the current time zone is 8 hours ahead of UTC, and the min + time_t value is 1970-01-01 00:00:00 UTC. Then the min + localtime value is 1970-01-01 08:00:00, and mktime will + therefore fail on 1969-12-31 23:00:00. To work around the + problem, set the time zone to 1 hour behind UTC temporarily + by setting TZ="XXX1:00" and try mktime again. */ + + long int time_zone = pc.time_zone; + long int abs_time_zone = time_zone < 0 ? - time_zone : time_zone; + long int abs_time_zone_hour = abs_time_zone / 60; + int abs_time_zone_min = abs_time_zone % 60; + char tz1buf[sizeof "XXX+0:00" + + sizeof pc.time_zone * CHAR_BIT / 3]; + if (!tz_was_altered) + tz0 = get_tz (tz0buf); + sprintf (tz1buf, "XXX%s%ld:%02d", "-" + (time_zone < 0), + abs_time_zone_hour, abs_time_zone_min); + if (setenv ("TZ", tz1buf, 1) != 0) + goto fail; + tz_was_altered = true; + tm = tm0; + Start = mktime (&tm); + if (! mktime_ok (&tm0, &tm, Start)) + goto fail; + } + } + + if (pc.days_seen && ! pc.dates_seen) + { + tm.tm_mday += ((pc.day_number - tm.tm_wday + 7) % 7 + + 7 * (pc.day_ordinal - (0 < pc.day_ordinal))); + tm.tm_isdst = -1; + Start = mktime (&tm); + if (Start == (time_t) -1) + goto fail; + } + + if (pc.zones_seen) + { + long int delta = pc.time_zone * 60; + time_t t1; +#ifdef HAVE_TM_GMTOFF + delta -= tm.tm_gmtoff; +#else + time_t t = Start; + struct tm const *gmt = gmtime (&t); + if (! gmt) + goto fail; + delta -= tm_diff (&tm, gmt); +#endif + t1 = Start - delta; + if ((Start < t1) != (delta < 0)) + goto fail; /* time_t overflow */ + Start = t1; + } + + /* Add relative date. */ + if (pc.rel.year | pc.rel.month | pc.rel.day) + { + int year = tm.tm_year + pc.rel.year; + int month = tm.tm_mon + pc.rel.month; + int day = tm.tm_mday + pc.rel.day; + if (((year < tm.tm_year) ^ (pc.rel.year < 0)) + | ((month < tm.tm_mon) ^ (pc.rel.month < 0)) + | ((day < tm.tm_mday) ^ (pc.rel.day < 0))) + goto fail; + tm.tm_year = year; + tm.tm_mon = month; + tm.tm_mday = day; + tm.tm_hour = tm0.tm_hour; + tm.tm_min = tm0.tm_min; + tm.tm_sec = tm0.tm_sec; + tm.tm_isdst = tm0.tm_isdst; + Start = mktime (&tm); + if (Start == (time_t) -1) + goto fail; + } + + /* Add relative hours, minutes, and seconds. On hosts that support + leap seconds, ignore the possibility of leap seconds; e.g., + "+ 10 minutes" adds 600 seconds, even if one of them is a + leap second. Typically this is not what the user wants, but it's + too hard to do it the other way, because the time zone indicator + must be applied before relative times, and if mktime is applied + again the time zone will be lost. */ + { + long int sum_ns = pc.seconds.tv_nsec + pc.rel.ns; + long int normalized_ns = (sum_ns % BILLION + BILLION) % BILLION; + time_t t0 = Start; + long int d1 = 60 * 60 * pc.rel.hour; + time_t t1 = t0 + d1; + long int d2 = 60 * pc.rel.minutes; + time_t t2 = t1 + d2; + long int d3 = pc.rel.seconds; + time_t t3 = t2 + d3; + long int d4 = (sum_ns - normalized_ns) / BILLION; + time_t t4 = t3 + d4; + + if ((d1 / (60 * 60) ^ pc.rel.hour) + | (d2 / 60 ^ pc.rel.minutes) + | ((t1 < t0) ^ (d1 < 0)) + | ((t2 < t1) ^ (d2 < 0)) + | ((t3 < t2) ^ (d3 < 0)) + | ((t4 < t3) ^ (d4 < 0))) + goto fail; + + result->tv_sec = t4; + result->tv_nsec = normalized_ns; + } + } + + goto done; + + fail: + ok = false; + done: + if (tz_was_altered) + ok &= (tz0 ? setenv ("TZ", tz0, 1) : unsetenv ("TZ")) == 0; + if (tz0 != tz0buf) + free (tz0); + return ok; +} + +#if TEST + +int +main (int ac, char **av) +{ + char buff[BUFSIZ]; + + printf ("Enter date, or blank line to exit.\n\t> "); + fflush (stdout); + + buff[BUFSIZ - 1] = '\0'; + while (fgets (buff, BUFSIZ - 1, stdin) && buff[0]) + { + struct timespec d; + struct tm const *tm; + if (! get_date (&d, buff, NULL)) + printf ("Bad format - couldn't convert.\n"); + else if (! (tm = localtime (&d.tv_sec))) + { + long int sec = d.tv_sec; + printf ("localtime (%ld) failed\n", sec); + } + else + { + int ns = d.tv_nsec; + printf ("%04ld-%02d-%02d %02d:%02d:%02d.%09d\n", + tm->tm_year + 1900L, tm->tm_mon + 1, tm->tm_mday, + tm->tm_hour, tm->tm_min, tm->tm_sec, ns); + } + printf ("\t> "); + fflush (stdout); + } + return 0; +} +#endif /* TEST */ + diff --git a/lib/getdate.h b/lib/getdate.h new file mode 100644 index 0000000..142231e --- /dev/null +++ b/lib/getdate.h @@ -0,0 +1,23 @@ +/* Parse a string into an internal time stamp. + + Copyright (C) 1995, 1997, 1998, 2003, 2004, 2007 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stdbool.h> +#include <time.h> + +bool get_date (struct timespec *, char const *, struct timespec const *); diff --git a/lib/getdate.y b/lib/getdate.y new file mode 100644 index 0000000..cbf3ca1 --- /dev/null +++ b/lib/getdate.y @@ -0,0 +1,1520 @@ +%{ +/* Parse a string into an internal time stamp. + + Copyright (C) 1999, 2000, 2002, 2003, 2004, 2005, 2006, 2007 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Originally written by Steven M. Bellovin <smb@research.att.com> while + at the University of North Carolina at Chapel Hill. Later tweaked by + a couple of people on Usenet. Completely overhauled by Rich $alz + <rsalz@bbn.com> and Jim Berets <jberets@bbn.com> in August, 1990. + + Modified by Paul Eggert <eggert@twinsun.com> in August 1999 to do + the right thing about local DST. Also modified by Paul Eggert + <eggert@cs.ucla.edu> in February 2004 to support + nanosecond-resolution time stamps, and in October 2004 to support + TZ strings in dates. */ + +/* FIXME: Check for arithmetic overflow in all cases, not just + some of them. */ + +#include <config.h> + +#include "getdate.h" +#include "timespec.h" + +/* There's no need to extend the stack, so there's no need to involve + alloca. */ +#define YYSTACK_USE_ALLOCA 0 + +/* Tell Bison how much stack space is needed. 20 should be plenty for + this grammar, which is not right recursive. Beware setting it too + high, since that might cause problems on machines whose + implementations have lame stack-overflow checking. */ +#define YYMAXDEPTH 20 +#define YYINITDEPTH YYMAXDEPTH + +/* Since the code of getdate.y is not included in the Emacs executable + itself, there is no need to #define static in this file. Even if + the code were included in the Emacs executable, it probably + wouldn't do any harm to #undef it here; this will only cause + problems if we try to write to a static variable, which I don't + think this code needs to do. */ +#ifdef emacs +# undef static +#endif + +#include <ctype.h> +#include <limits.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "setenv.h" +#include "xalloc.h" + + +/* ISDIGIT differs from isdigit, as follows: + - Its arg may be any int or unsigned int; it need not be an unsigned char + or EOF. + - It's typically faster. + POSIX says that only '0' through '9' are digits. Prefer ISDIGIT to + isdigit unless it's important to use the locale's definition + of `digit' even when the host does not conform to POSIX. */ +#define ISDIGIT(c) ((unsigned int) (c) - '0' <= 9) + +#ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +#endif + +#ifndef ATTRIBUTE_UNUSED +# define ATTRIBUTE_UNUSED __attribute__ ((__unused__)) +#endif + +/* Shift A right by B bits portably, by dividing A by 2**B and + truncating towards minus infinity. A and B should be free of side + effects, and B should be in the range 0 <= B <= INT_BITS - 2, where + INT_BITS is the number of useful bits in an int. GNU code can + assume that INT_BITS is at least 32. + + ISO C99 says that A >> B is implementation-defined if A < 0. Some + implementations (e.g., UNICOS 9.0 on a Cray Y-MP EL) don't shift + right in the usual way when A < 0, so SHR falls back on division if + ordinary A >> B doesn't seem to be the usual signed shift. */ +#define SHR(a, b) \ + (-1 >> 1 == -1 \ + ? (a) >> (b) \ + : (a) / (1 << (b)) - ((a) % (1 << (b)) < 0)) + +#define EPOCH_YEAR 1970 +#define TM_YEAR_BASE 1900 + +#define HOUR(x) ((x) * 60) + +/* An integer value, and the number of digits in its textual + representation. */ +typedef struct +{ + bool negative; + long int value; + size_t digits; +} textint; + +/* An entry in the lexical lookup table. */ +typedef struct +{ + char const *name; + int type; + int value; +} table; + +/* Meridian: am, pm, or 24-hour style. */ +enum { MERam, MERpm, MER24 }; + +enum { BILLION = 1000000000, LOG10_BILLION = 9 }; + +/* Relative times. */ +typedef struct +{ + /* Relative year, month, day, hour, minutes, seconds, and nanoseconds. */ + long int year; + long int month; + long int day; + long int hour; + long int minutes; + long int seconds; + long int ns; +} relative_time; + +#if HAVE_COMPOUND_LITERALS +# define RELATIVE_TIME_0 ((relative_time) { 0, 0, 0, 0, 0, 0, 0 }) +#else +static relative_time const RELATIVE_TIME_0; +#endif + +/* Information passed to and from the parser. */ +typedef struct +{ + /* The input string remaining to be parsed. */ + const char *input; + + /* N, if this is the Nth Tuesday. */ + long int day_ordinal; + + /* Day of week; Sunday is 0. */ + int day_number; + + /* tm_isdst flag for the local zone. */ + int local_isdst; + + /* Time zone, in minutes east of UTC. */ + long int time_zone; + + /* Style used for time. */ + int meridian; + + /* Gregorian year, month, day, hour, minutes, seconds, and nanoseconds. */ + textint year; + long int month; + long int day; + long int hour; + long int minutes; + struct timespec seconds; /* includes nanoseconds */ + + /* Relative year, month, day, hour, minutes, seconds, and nanoseconds. */ + relative_time rel; + + /* Presence or counts of nonterminals of various flavors parsed so far. */ + bool timespec_seen; + bool rels_seen; + size_t dates_seen; + size_t days_seen; + size_t local_zones_seen; + size_t dsts_seen; + size_t times_seen; + size_t zones_seen; + + /* Table of local time zone abbrevations, terminated by a null entry. */ + table local_time_zone_table[3]; +} parser_control; + +union YYSTYPE; +static int yylex (union YYSTYPE *, parser_control *); +static int yyerror (parser_control const *, char const *); +static long int time_zone_hhmm (textint, long int); + +%} + +/* We want a reentrant parser, even if the TZ manipulation and the calls to + localtime and gmtime are not reentrant. */ +%pure-parser +%parse-param { parser_control *pc } +%lex-param { parser_control *pc } + +/* This grammar has 20 shift/reduce conflicts. */ +%expect 20 + +%union +{ + long int intval; + textint textintval; + struct timespec timespec; + relative_time rel; +} + +%token tAGO tDST + +%token tYEAR_UNIT tMONTH_UNIT tHOUR_UNIT tMINUTE_UNIT tSEC_UNIT +%token <intval> tDAY_UNIT + +%token <intval> tDAY tDAYZONE tLOCAL_ZONE tMERIDIAN +%token <intval> tMONTH tORDINAL tZONE + +%token <textintval> tSNUMBER tUNUMBER +%token <timespec> tSDECIMAL_NUMBER tUDECIMAL_NUMBER + +%type <intval> o_colon_minutes o_merid +%type <timespec> seconds signed_seconds unsigned_seconds + +%type <rel> relunit relunit_snumber + +%% + +spec: + timespec + | items + ; + +timespec: + '@' seconds + { + pc->seconds = $2; + pc->timespec_seen = true; + } + ; + +items: + /* empty */ + | items item + ; + +item: + time + { pc->times_seen++; } + | local_zone + { pc->local_zones_seen++; } + | zone + { pc->zones_seen++; } + | date + { pc->dates_seen++; } + | day + { pc->days_seen++; } + | rel + { pc->rels_seen = true; } + | number + ; + +time: + tUNUMBER tMERIDIAN + { + pc->hour = $1.value; + pc->minutes = 0; + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = $2; + } + | tUNUMBER ':' tUNUMBER o_merid + { + pc->hour = $1.value; + pc->minutes = $3.value; + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = $4; + } + | tUNUMBER ':' tUNUMBER tSNUMBER o_colon_minutes + { + pc->hour = $1.value; + pc->minutes = $3.value; + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = MER24; + pc->zones_seen++; + pc->time_zone = time_zone_hhmm ($4, $5); + } + | tUNUMBER ':' tUNUMBER ':' unsigned_seconds o_merid + { + pc->hour = $1.value; + pc->minutes = $3.value; + pc->seconds = $5; + pc->meridian = $6; + } + | tUNUMBER ':' tUNUMBER ':' unsigned_seconds tSNUMBER o_colon_minutes + { + pc->hour = $1.value; + pc->minutes = $3.value; + pc->seconds = $5; + pc->meridian = MER24; + pc->zones_seen++; + pc->time_zone = time_zone_hhmm ($6, $7); + } + ; + +local_zone: + tLOCAL_ZONE + { + pc->local_isdst = $1; + pc->dsts_seen += (0 < $1); + } + | tLOCAL_ZONE tDST + { + pc->local_isdst = 1; + pc->dsts_seen += (0 < $1) + 1; + } + ; + +zone: + tZONE + { pc->time_zone = $1; } + | tZONE relunit_snumber + { pc->time_zone = $1; + pc->rel.ns += $2.ns; + pc->rel.seconds += $2.seconds; + pc->rel.minutes += $2.minutes; + pc->rel.hour += $2.hour; + pc->rel.day += $2.day; + pc->rel.month += $2.month; + pc->rel.year += $2.year; + pc->rels_seen = true; } + | tZONE tSNUMBER o_colon_minutes + { pc->time_zone = $1 + time_zone_hhmm ($2, $3); } + | tDAYZONE + { pc->time_zone = $1 + 60; } + | tZONE tDST + { pc->time_zone = $1 + 60; } + ; + +day: + tDAY + { + pc->day_ordinal = 1; + pc->day_number = $1; + } + | tDAY ',' + { + pc->day_ordinal = 1; + pc->day_number = $1; + } + | tORDINAL tDAY + { + pc->day_ordinal = $1; + pc->day_number = $2; + } + | tUNUMBER tDAY + { + pc->day_ordinal = $1.value; + pc->day_number = $2; + } + ; + +date: + tUNUMBER '/' tUNUMBER + { + pc->month = $1.value; + pc->day = $3.value; + } + | tUNUMBER '/' tUNUMBER '/' tUNUMBER + { + /* Interpret as YYYY/MM/DD if the first value has 4 or more digits, + otherwise as MM/DD/YY. + The goal in recognizing YYYY/MM/DD is solely to support legacy + machine-generated dates like those in an RCS log listing. If + you want portability, use the ISO 8601 format. */ + if (4 <= $1.digits) + { + pc->year = $1; + pc->month = $3.value; + pc->day = $5.value; + } + else + { + pc->month = $1.value; + pc->day = $3.value; + pc->year = $5; + } + } + | tUNUMBER tSNUMBER tSNUMBER + { + /* ISO 8601 format. YYYY-MM-DD. */ + pc->year = $1; + pc->month = -$2.value; + pc->day = -$3.value; + } + | tUNUMBER tMONTH tSNUMBER + { + /* e.g. 17-JUN-1992. */ + pc->day = $1.value; + pc->month = $2; + pc->year.value = -$3.value; + pc->year.digits = $3.digits; + } + | tMONTH tSNUMBER tSNUMBER + { + /* e.g. JUN-17-1992. */ + pc->month = $1; + pc->day = -$2.value; + pc->year.value = -$3.value; + pc->year.digits = $3.digits; + } + | tMONTH tUNUMBER + { + pc->month = $1; + pc->day = $2.value; + } + | tMONTH tUNUMBER ',' tUNUMBER + { + pc->month = $1; + pc->day = $2.value; + pc->year = $4; + } + | tUNUMBER tMONTH + { + pc->day = $1.value; + pc->month = $2; + } + | tUNUMBER tMONTH tUNUMBER + { + pc->day = $1.value; + pc->month = $2; + pc->year = $3; + } + ; + +rel: + relunit tAGO + { + pc->rel.ns -= $1.ns; + pc->rel.seconds -= $1.seconds; + pc->rel.minutes -= $1.minutes; + pc->rel.hour -= $1.hour; + pc->rel.day -= $1.day; + pc->rel.month -= $1.month; + pc->rel.year -= $1.year; + } + | relunit + { + pc->rel.ns += $1.ns; + pc->rel.seconds += $1.seconds; + pc->rel.minutes += $1.minutes; + pc->rel.hour += $1.hour; + pc->rel.day += $1.day; + pc->rel.month += $1.month; + pc->rel.year += $1.year; + } + ; + +relunit: + tORDINAL tYEAR_UNIT + { $$ = RELATIVE_TIME_0; $$.year = $1; } + | tUNUMBER tYEAR_UNIT + { $$ = RELATIVE_TIME_0; $$.year = $1.value; } + | tYEAR_UNIT + { $$ = RELATIVE_TIME_0; $$.year = 1; } + | tORDINAL tMONTH_UNIT + { $$ = RELATIVE_TIME_0; $$.month = $1; } + | tUNUMBER tMONTH_UNIT + { $$ = RELATIVE_TIME_0; $$.month = $1.value; } + | tMONTH_UNIT + { $$ = RELATIVE_TIME_0; $$.month = 1; } + | tORDINAL tDAY_UNIT + { $$ = RELATIVE_TIME_0; $$.day = $1 * $2; } + | tUNUMBER tDAY_UNIT + { $$ = RELATIVE_TIME_0; $$.day = $1.value * $2; } + | tDAY_UNIT + { $$ = RELATIVE_TIME_0; $$.day = $1; } + | tORDINAL tHOUR_UNIT + { $$ = RELATIVE_TIME_0; $$.hour = $1; } + | tUNUMBER tHOUR_UNIT + { $$ = RELATIVE_TIME_0; $$.hour = $1.value; } + | tHOUR_UNIT + { $$ = RELATIVE_TIME_0; $$.hour = 1; } + | tORDINAL tMINUTE_UNIT + { $$ = RELATIVE_TIME_0; $$.minutes = $1; } + | tUNUMBER tMINUTE_UNIT + { $$ = RELATIVE_TIME_0; $$.minutes = $1.value; } + | tMINUTE_UNIT + { $$ = RELATIVE_TIME_0; $$.minutes = 1; } + | tORDINAL tSEC_UNIT + { $$ = RELATIVE_TIME_0; $$.seconds = $1; } + | tUNUMBER tSEC_UNIT + { $$ = RELATIVE_TIME_0; $$.seconds = $1.value; } + | tSDECIMAL_NUMBER tSEC_UNIT + { $$ = RELATIVE_TIME_0; $$.seconds = $1.tv_sec; $$.ns = $1.tv_nsec; } + | tUDECIMAL_NUMBER tSEC_UNIT + { $$ = RELATIVE_TIME_0; $$.seconds = $1.tv_sec; $$.ns = $1.tv_nsec; } + | tSEC_UNIT + { $$ = RELATIVE_TIME_0; $$.seconds = 1; } + | relunit_snumber + ; + +relunit_snumber: + tSNUMBER tYEAR_UNIT + { $$ = RELATIVE_TIME_0; $$.year = $1.value; } + | tSNUMBER tMONTH_UNIT + { $$ = RELATIVE_TIME_0; $$.month = $1.value; } + | tSNUMBER tDAY_UNIT + { $$ = RELATIVE_TIME_0; $$.day = $1.value * $2; } + | tSNUMBER tHOUR_UNIT + { $$ = RELATIVE_TIME_0; $$.hour = $1.value; } + | tSNUMBER tMINUTE_UNIT + { $$ = RELATIVE_TIME_0; $$.minutes = $1.value; } + | tSNUMBER tSEC_UNIT + { $$ = RELATIVE_TIME_0; $$.seconds = $1.value; } + ; + +seconds: signed_seconds | unsigned_seconds; + +signed_seconds: + tSDECIMAL_NUMBER + | tSNUMBER + { $$.tv_sec = $1.value; $$.tv_nsec = 0; } + ; + +unsigned_seconds: + tUDECIMAL_NUMBER + | tUNUMBER + { $$.tv_sec = $1.value; $$.tv_nsec = 0; } + ; + +number: + tUNUMBER + { + if (pc->dates_seen && ! pc->year.digits + && ! pc->rels_seen && (pc->times_seen || 2 < $1.digits)) + pc->year = $1; + else + { + if (4 < $1.digits) + { + pc->dates_seen++; + pc->day = $1.value % 100; + pc->month = ($1.value / 100) % 100; + pc->year.value = $1.value / 10000; + pc->year.digits = $1.digits - 4; + } + else + { + pc->times_seen++; + if ($1.digits <= 2) + { + pc->hour = $1.value; + pc->minutes = 0; + } + else + { + pc->hour = $1.value / 100; + pc->minutes = $1.value % 100; + } + pc->seconds.tv_sec = 0; + pc->seconds.tv_nsec = 0; + pc->meridian = MER24; + } + } + } + ; + +o_colon_minutes: + /* empty */ + { $$ = -1; } + | ':' tUNUMBER + { $$ = $2.value; } + ; + +o_merid: + /* empty */ + { $$ = MER24; } + | tMERIDIAN + { $$ = $1; } + ; + +%% + +static table const meridian_table[] = +{ + { "AM", tMERIDIAN, MERam }, + { "A.M.", tMERIDIAN, MERam }, + { "PM", tMERIDIAN, MERpm }, + { "P.M.", tMERIDIAN, MERpm }, + { NULL, 0, 0 } +}; + +static table const dst_table[] = +{ + { "DST", tDST, 0 } +}; + +static table const month_and_day_table[] = +{ + { "JANUARY", tMONTH, 1 }, + { "FEBRUARY", tMONTH, 2 }, + { "MARCH", tMONTH, 3 }, + { "APRIL", tMONTH, 4 }, + { "MAY", tMONTH, 5 }, + { "JUNE", tMONTH, 6 }, + { "JULY", tMONTH, 7 }, + { "AUGUST", tMONTH, 8 }, + { "SEPTEMBER",tMONTH, 9 }, + { "SEPT", tMONTH, 9 }, + { "OCTOBER", tMONTH, 10 }, + { "NOVEMBER", tMONTH, 11 }, + { "DECEMBER", tMONTH, 12 }, + { "SUNDAY", tDAY, 0 }, + { "MONDAY", tDAY, 1 }, + { "TUESDAY", tDAY, 2 }, + { "TUES", tDAY, 2 }, + { "WEDNESDAY",tDAY, 3 }, + { "WEDNES", tDAY, 3 }, + { "THURSDAY", tDAY, 4 }, + { "THUR", tDAY, 4 }, + { "THURS", tDAY, 4 }, + { "FRIDAY", tDAY, 5 }, + { "SATURDAY", tDAY, 6 }, + { NULL, 0, 0 } +}; + +static table const time_units_table[] = +{ + { "YEAR", tYEAR_UNIT, 1 }, + { "MONTH", tMONTH_UNIT, 1 }, + { "FORTNIGHT",tDAY_UNIT, 14 }, + { "WEEK", tDAY_UNIT, 7 }, + { "DAY", tDAY_UNIT, 1 }, + { "HOUR", tHOUR_UNIT, 1 }, + { "MINUTE", tMINUTE_UNIT, 1 }, + { "MIN", tMINUTE_UNIT, 1 }, + { "SECOND", tSEC_UNIT, 1 }, + { "SEC", tSEC_UNIT, 1 }, + { NULL, 0, 0 } +}; + +/* Assorted relative-time words. */ +static table const relative_time_table[] = +{ + { "TOMORROW", tDAY_UNIT, 1 }, + { "YESTERDAY",tDAY_UNIT, -1 }, + { "TODAY", tDAY_UNIT, 0 }, + { "NOW", tDAY_UNIT, 0 }, + { "LAST", tORDINAL, -1 }, + { "THIS", tORDINAL, 0 }, + { "NEXT", tORDINAL, 1 }, + { "FIRST", tORDINAL, 1 }, +/*{ "SECOND", tORDINAL, 2 }, */ + { "THIRD", tORDINAL, 3 }, + { "FOURTH", tORDINAL, 4 }, + { "FIFTH", tORDINAL, 5 }, + { "SIXTH", tORDINAL, 6 }, + { "SEVENTH", tORDINAL, 7 }, + { "EIGHTH", tORDINAL, 8 }, + { "NINTH", tORDINAL, 9 }, + { "TENTH", tORDINAL, 10 }, + { "ELEVENTH", tORDINAL, 11 }, + { "TWELFTH", tORDINAL, 12 }, + { "AGO", tAGO, 1 }, + { NULL, 0, 0 } +}; + +/* The universal time zone table. These labels can be used even for + time stamps that would not otherwise be valid, e.g., GMT time + stamps in London during summer. */ +static table const universal_time_zone_table[] = +{ + { "GMT", tZONE, HOUR ( 0) }, /* Greenwich Mean */ + { "UT", tZONE, HOUR ( 0) }, /* Universal (Coordinated) */ + { "UTC", tZONE, HOUR ( 0) }, + { NULL, 0, 0 } +}; + +/* The time zone table. This table is necessarily incomplete, as time + zone abbreviations are ambiguous; e.g. Australians interpret "EST" + as Eastern time in Australia, not as US Eastern Standard Time. + You cannot rely on getdate to handle arbitrary time zone + abbreviations; use numeric abbreviations like `-0500' instead. */ +static table const time_zone_table[] = +{ + { "WET", tZONE, HOUR ( 0) }, /* Western European */ + { "WEST", tDAYZONE, HOUR ( 0) }, /* Western European Summer */ + { "BST", tDAYZONE, HOUR ( 0) }, /* British Summer */ + { "ART", tZONE, -HOUR ( 3) }, /* Argentina */ + { "BRT", tZONE, -HOUR ( 3) }, /* Brazil */ + { "BRST", tDAYZONE, -HOUR ( 3) }, /* Brazil Summer */ + { "NST", tZONE, -(HOUR ( 3) + 30) }, /* Newfoundland Standard */ + { "NDT", tDAYZONE,-(HOUR ( 3) + 30) }, /* Newfoundland Daylight */ + { "AST", tZONE, -HOUR ( 4) }, /* Atlantic Standard */ + { "ADT", tDAYZONE, -HOUR ( 4) }, /* Atlantic Daylight */ + { "CLT", tZONE, -HOUR ( 4) }, /* Chile */ + { "CLST", tDAYZONE, -HOUR ( 4) }, /* Chile Summer */ + { "EST", tZONE, -HOUR ( 5) }, /* Eastern Standard */ + { "EDT", tDAYZONE, -HOUR ( 5) }, /* Eastern Daylight */ + { "CST", tZONE, -HOUR ( 6) }, /* Central Standard */ + { "CDT", tDAYZONE, -HOUR ( 6) }, /* Central Daylight */ + { "MST", tZONE, -HOUR ( 7) }, /* Mountain Standard */ + { "MDT", tDAYZONE, -HOUR ( 7) }, /* Mountain Daylight */ + { "PST", tZONE, -HOUR ( 8) }, /* Pacific Standard */ + { "PDT", tDAYZONE, -HOUR ( 8) }, /* Pacific Daylight */ + { "AKST", tZONE, -HOUR ( 9) }, /* Alaska Standard */ + { "AKDT", tDAYZONE, -HOUR ( 9) }, /* Alaska Daylight */ + { "HST", tZONE, -HOUR (10) }, /* Hawaii Standard */ + { "HAST", tZONE, -HOUR (10) }, /* Hawaii-Aleutian Standard */ + { "HADT", tDAYZONE, -HOUR (10) }, /* Hawaii-Aleutian Daylight */ + { "SST", tZONE, -HOUR (12) }, /* Samoa Standard */ + { "WAT", tZONE, HOUR ( 1) }, /* West Africa */ + { "CET", tZONE, HOUR ( 1) }, /* Central European */ + { "CEST", tDAYZONE, HOUR ( 1) }, /* Central European Summer */ + { "MET", tZONE, HOUR ( 1) }, /* Middle European */ + { "MEZ", tZONE, HOUR ( 1) }, /* Middle European */ + { "MEST", tDAYZONE, HOUR ( 1) }, /* Middle European Summer */ + { "MESZ", tDAYZONE, HOUR ( 1) }, /* Middle European Summer */ + { "EET", tZONE, HOUR ( 2) }, /* Eastern European */ + { "EEST", tDAYZONE, HOUR ( 2) }, /* Eastern European Summer */ + { "CAT", tZONE, HOUR ( 2) }, /* Central Africa */ + { "SAST", tZONE, HOUR ( 2) }, /* South Africa Standard */ + { "EAT", tZONE, HOUR ( 3) }, /* East Africa */ + { "MSK", tZONE, HOUR ( 3) }, /* Moscow */ + { "MSD", tDAYZONE, HOUR ( 3) }, /* Moscow Daylight */ + { "IST", tZONE, (HOUR ( 5) + 30) }, /* India Standard */ + { "SGT", tZONE, HOUR ( 8) }, /* Singapore */ + { "KST", tZONE, HOUR ( 9) }, /* Korea Standard */ + { "JST", tZONE, HOUR ( 9) }, /* Japan Standard */ + { "GST", tZONE, HOUR (10) }, /* Guam Standard */ + { "NZST", tZONE, HOUR (12) }, /* New Zealand Standard */ + { "NZDT", tDAYZONE, HOUR (12) }, /* New Zealand Daylight */ + { NULL, 0, 0 } +}; + +/* Military time zone table. */ +static table const military_table[] = +{ + { "A", tZONE, -HOUR ( 1) }, + { "B", tZONE, -HOUR ( 2) }, + { "C", tZONE, -HOUR ( 3) }, + { "D", tZONE, -HOUR ( 4) }, + { "E", tZONE, -HOUR ( 5) }, + { "F", tZONE, -HOUR ( 6) }, + { "G", tZONE, -HOUR ( 7) }, + { "H", tZONE, -HOUR ( 8) }, + { "I", tZONE, -HOUR ( 9) }, + { "K", tZONE, -HOUR (10) }, + { "L", tZONE, -HOUR (11) }, + { "M", tZONE, -HOUR (12) }, + { "N", tZONE, HOUR ( 1) }, + { "O", tZONE, HOUR ( 2) }, + { "P", tZONE, HOUR ( 3) }, + { "Q", tZONE, HOUR ( 4) }, + { "R", tZONE, HOUR ( 5) }, + { "S", tZONE, HOUR ( 6) }, + { "T", tZONE, HOUR ( 7) }, + { "U", tZONE, HOUR ( 8) }, + { "V", tZONE, HOUR ( 9) }, + { "W", tZONE, HOUR (10) }, + { "X", tZONE, HOUR (11) }, + { "Y", tZONE, HOUR (12) }, + { "Z", tZONE, HOUR ( 0) }, + { NULL, 0, 0 } +}; + + + +/* Convert a time zone expressed as HH:MM into an integer count of + minutes. If MM is negative, then S is of the form HHMM and needs + to be picked apart; otherwise, S is of the form HH. */ + +static long int +time_zone_hhmm (textint s, long int mm) +{ + if (mm < 0) + return (s.value / 100) * 60 + s.value % 100; + else + return s.value * 60 + (s.negative ? -mm : mm); +} + +static int +to_hour (long int hours, int meridian) +{ + switch (meridian) + { + default: /* Pacify GCC. */ + case MER24: + return 0 <= hours && hours < 24 ? hours : -1; + case MERam: + return 0 < hours && hours < 12 ? hours : hours == 12 ? 0 : -1; + case MERpm: + return 0 < hours && hours < 12 ? hours + 12 : hours == 12 ? 12 : -1; + } +} + +static long int +to_year (textint textyear) +{ + long int year = textyear.value; + + if (year < 0) + year = -year; + + /* XPG4 suggests that years 00-68 map to 2000-2068, and + years 69-99 map to 1969-1999. */ + else if (textyear.digits == 2) + year += year < 69 ? 2000 : 1900; + + return year; +} + +static table const * +lookup_zone (parser_control const *pc, char const *name) +{ + table const *tp; + + for (tp = universal_time_zone_table; tp->name; tp++) + if (strcmp (name, tp->name) == 0) + return tp; + + /* Try local zone abbreviations before those in time_zone_table, as + the local ones are more likely to be right. */ + for (tp = pc->local_time_zone_table; tp->name; tp++) + if (strcmp (name, tp->name) == 0) + return tp; + + for (tp = time_zone_table; tp->name; tp++) + if (strcmp (name, tp->name) == 0) + return tp; + + return NULL; +} + +#if ! HAVE_TM_GMTOFF +/* Yield the difference between *A and *B, + measured in seconds, ignoring leap seconds. + The body of this function is taken directly from the GNU C Library; + see src/strftime.c. */ +static long int +tm_diff (struct tm const *a, struct tm const *b) +{ + /* Compute intervening leap days correctly even if year is negative. + Take care to avoid int overflow in leap day calculations. */ + int a4 = SHR (a->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (a->tm_year & 3); + int b4 = SHR (b->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (b->tm_year & 3); + int a100 = a4 / 25 - (a4 % 25 < 0); + int b100 = b4 / 25 - (b4 % 25 < 0); + int a400 = SHR (a100, 2); + int b400 = SHR (b100, 2); + int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400); + long int ayear = a->tm_year; + long int years = ayear - b->tm_year; + long int days = (365 * years + intervening_leap_days + + (a->tm_yday - b->tm_yday)); + return (60 * (60 * (24 * days + (a->tm_hour - b->tm_hour)) + + (a->tm_min - b->tm_min)) + + (a->tm_sec - b->tm_sec)); +} +#endif /* ! HAVE_TM_GMTOFF */ + +static table const * +lookup_word (parser_control const *pc, char *word) +{ + char *p; + char *q; + size_t wordlen; + table const *tp; + bool period_found; + bool abbrev; + + /* Make it uppercase. */ + for (p = word; *p; p++) + { + unsigned char ch = *p; + *p = toupper (ch); + } + + for (tp = meridian_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + + /* See if we have an abbreviation for a month. */ + wordlen = strlen (word); + abbrev = wordlen == 3 || (wordlen == 4 && word[3] == '.'); + + for (tp = month_and_day_table; tp->name; tp++) + if ((abbrev ? strncmp (word, tp->name, 3) : strcmp (word, tp->name)) == 0) + return tp; + + if ((tp = lookup_zone (pc, word))) + return tp; + + if (strcmp (word, dst_table[0].name) == 0) + return dst_table; + + for (tp = time_units_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + + /* Strip off any plural and try the units table again. */ + if (word[wordlen - 1] == 'S') + { + word[wordlen - 1] = '\0'; + for (tp = time_units_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + word[wordlen - 1] = 'S'; /* For "this" in relative_time_table. */ + } + + for (tp = relative_time_table; tp->name; tp++) + if (strcmp (word, tp->name) == 0) + return tp; + + /* Military time zones. */ + if (wordlen == 1) + for (tp = military_table; tp->name; tp++) + if (word[0] == tp->name[0]) + return tp; + + /* Drop out any periods and try the time zone table again. */ + for (period_found = false, p = q = word; (*p = *q); q++) + if (*q == '.') + period_found = true; + else + p++; + if (period_found && (tp = lookup_zone (pc, word))) + return tp; + + return NULL; +} + +static int +yylex (YYSTYPE *lvalp, parser_control *pc) +{ + unsigned char c; + size_t count; + + for (;;) + { + while (c = *pc->input, isspace (c)) + pc->input++; + + if (ISDIGIT (c) || c == '-' || c == '+') + { + char const *p; + int sign; + unsigned long int value; + if (c == '-' || c == '+') + { + sign = c == '-' ? -1 : 1; + while (c = *++pc->input, isspace (c)) + continue; + if (! ISDIGIT (c)) + /* skip the '-' sign */ + continue; + } + else + sign = 0; + p = pc->input; + for (value = 0; ; value *= 10) + { + unsigned long int value1 = value + (c - '0'); + if (value1 < value) + return '?'; + value = value1; + c = *++p; + if (! ISDIGIT (c)) + break; + if (ULONG_MAX / 10 < value) + return '?'; + } + if ((c == '.' || c == ',') && ISDIGIT (p[1])) + { + time_t s; + int ns; + int digits; + unsigned long int value1; + + /* Check for overflow when converting value to time_t. */ + if (sign < 0) + { + s = - value; + if (0 < s) + return '?'; + value1 = -s; + } + else + { + s = value; + if (s < 0) + return '?'; + value1 = s; + } + if (value != value1) + return '?'; + + /* Accumulate fraction, to ns precision. */ + p++; + ns = *p++ - '0'; + for (digits = 2; digits <= LOG10_BILLION; digits++) + { + ns *= 10; + if (ISDIGIT (*p)) + ns += *p++ - '0'; + } + + /* Skip excess digits, truncating toward -Infinity. */ + if (sign < 0) + for (; ISDIGIT (*p); p++) + if (*p != '0') + { + ns++; + break; + } + while (ISDIGIT (*p)) + p++; + + /* Adjust to the timespec convention, which is that + tv_nsec is always a positive offset even if tv_sec is + negative. */ + if (sign < 0 && ns) + { + s--; + if (! (s < 0)) + return '?'; + ns = BILLION - ns; + } + + lvalp->timespec.tv_sec = s; + lvalp->timespec.tv_nsec = ns; + pc->input = p; + return sign ? tSDECIMAL_NUMBER : tUDECIMAL_NUMBER; + } + else + { + lvalp->textintval.negative = sign < 0; + if (sign < 0) + { + lvalp->textintval.value = - value; + if (0 < lvalp->textintval.value) + return '?'; + } + else + { + lvalp->textintval.value = value; + if (lvalp->textintval.value < 0) + return '?'; + } + lvalp->textintval.digits = p - pc->input; + pc->input = p; + return sign ? tSNUMBER : tUNUMBER; + } + } + + if (isalpha (c)) + { + char buff[20]; + char *p = buff; + table const *tp; + + do + { + if (p < buff + sizeof buff - 1) + *p++ = c; + c = *++pc->input; + } + while (isalpha (c) || c == '.'); + + *p = '\0'; + tp = lookup_word (pc, buff); + if (! tp) + return '?'; + lvalp->intval = tp->value; + return tp->type; + } + + if (c != '(') + return *pc->input++; + count = 0; + do + { + c = *pc->input++; + if (c == '\0') + return c; + if (c == '(') + count++; + else if (c == ')') + count--; + } + while (count != 0); + } +} + +/* Do nothing if the parser reports an error. */ +static int +yyerror (parser_control const *pc ATTRIBUTE_UNUSED, + char const *s ATTRIBUTE_UNUSED) +{ + return 0; +} + +/* If *TM0 is the old and *TM1 is the new value of a struct tm after + passing it to mktime, return true if it's OK that mktime returned T. + It's not OK if *TM0 has out-of-range members. */ + +static bool +mktime_ok (struct tm const *tm0, struct tm const *tm1, time_t t) +{ + if (t == (time_t) -1) + { + /* Guard against falsely reporting an error when parsing a time + stamp that happens to equal (time_t) -1, on a host that + supports such a time stamp. */ + tm1 = localtime (&t); + if (!tm1) + return false; + } + + return ! ((tm0->tm_sec ^ tm1->tm_sec) + | (tm0->tm_min ^ tm1->tm_min) + | (tm0->tm_hour ^ tm1->tm_hour) + | (tm0->tm_mday ^ tm1->tm_mday) + | (tm0->tm_mon ^ tm1->tm_mon) + | (tm0->tm_year ^ tm1->tm_year)); +} + +/* A reasonable upper bound for the size of ordinary TZ strings. + Use heap allocation if TZ's length exceeds this. */ +enum { TZBUFSIZE = 100 }; + +/* Return a copy of TZ, stored in TZBUF if it fits, and heap-allocated + otherwise. */ +static char * +get_tz (char tzbuf[TZBUFSIZE]) +{ + char *tz = getenv ("TZ"); + if (tz) + { + size_t tzsize = strlen (tz) + 1; + tz = (tzsize <= TZBUFSIZE + ? memcpy (tzbuf, tz, tzsize) + : xmemdup (tz, tzsize)); + } + return tz; +} + +/* Parse a date/time string, storing the resulting time value into *RESULT. + The string itself is pointed to by P. Return true if successful. + P can be an incomplete or relative time specification; if so, use + *NOW as the basis for the returned time. */ +bool +get_date (struct timespec *result, char const *p, struct timespec const *now) +{ + time_t Start; + long int Start_ns; + struct tm const *tmp; + struct tm tm; + struct tm tm0; + parser_control pc; + struct timespec gettime_buffer; + unsigned char c; + bool tz_was_altered = false; + char *tz0 = NULL; + char tz0buf[TZBUFSIZE]; + bool ok = true; + + if (! now) + { + gettime (&gettime_buffer); + now = &gettime_buffer; + } + + Start = now->tv_sec; + Start_ns = now->tv_nsec; + + tmp = localtime (&now->tv_sec); + if (! tmp) + return false; + + while (c = *p, isspace (c)) + p++; + + if (strncmp (p, "TZ=\"", 4) == 0) + { + char const *tzbase = p + 4; + size_t tzsize = 1; + char const *s; + + for (s = tzbase; *s; s++, tzsize++) + if (*s == '\\') + { + s++; + if (! (*s == '\\' || *s == '"')) + break; + } + else if (*s == '"') + { + char *z; + char *tz1; + char tz1buf[TZBUFSIZE]; + bool large_tz = TZBUFSIZE < tzsize; + bool setenv_ok; + tz0 = get_tz (tz0buf); + z = tz1 = large_tz ? xmalloc (tzsize) : tz1buf; + for (s = tzbase; *s != '"'; s++) + *z++ = *(s += *s == '\\'); + *z = '\0'; + setenv_ok = setenv ("TZ", tz1, 1) == 0; + if (large_tz) + free (tz1); + if (!setenv_ok) + goto fail; + tz_was_altered = true; + p = s + 1; + } + } + + pc.input = p; + pc.year.value = tmp->tm_year; + pc.year.value += TM_YEAR_BASE; + pc.year.digits = 0; + pc.month = tmp->tm_mon + 1; + pc.day = tmp->tm_mday; + pc.hour = tmp->tm_hour; + pc.minutes = tmp->tm_min; + pc.seconds.tv_sec = tmp->tm_sec; + pc.seconds.tv_nsec = Start_ns; + tm.tm_isdst = tmp->tm_isdst; + + pc.meridian = MER24; + pc.rel = RELATIVE_TIME_0; + pc.timespec_seen = false; + pc.rels_seen = false; + pc.dates_seen = 0; + pc.days_seen = 0; + pc.times_seen = 0; + pc.local_zones_seen = 0; + pc.dsts_seen = 0; + pc.zones_seen = 0; + +#if HAVE_STRUCT_TM_TM_ZONE + pc.local_time_zone_table[0].name = tmp->tm_zone; + pc.local_time_zone_table[0].type = tLOCAL_ZONE; + pc.local_time_zone_table[0].value = tmp->tm_isdst; + pc.local_time_zone_table[1].name = NULL; + + /* Probe the names used in the next three calendar quarters, looking + for a tm_isdst different from the one we already have. */ + { + int quarter; + for (quarter = 1; quarter <= 3; quarter++) + { + time_t probe = Start + quarter * (90 * 24 * 60 * 60); + struct tm const *probe_tm = localtime (&probe); + if (probe_tm && probe_tm->tm_zone + && probe_tm->tm_isdst != pc.local_time_zone_table[0].value) + { + { + pc.local_time_zone_table[1].name = probe_tm->tm_zone; + pc.local_time_zone_table[1].type = tLOCAL_ZONE; + pc.local_time_zone_table[1].value = probe_tm->tm_isdst; + pc.local_time_zone_table[2].name = NULL; + } + break; + } + } + } +#else +#if HAVE_TZNAME + { +# ifndef tzname + extern char *tzname[]; +# endif + int i; + for (i = 0; i < 2; i++) + { + pc.local_time_zone_table[i].name = tzname[i]; + pc.local_time_zone_table[i].type = tLOCAL_ZONE; + pc.local_time_zone_table[i].value = i; + } + pc.local_time_zone_table[i].name = NULL; + } +#else + pc.local_time_zone_table[0].name = NULL; +#endif +#endif + + if (pc.local_time_zone_table[0].name && pc.local_time_zone_table[1].name + && ! strcmp (pc.local_time_zone_table[0].name, + pc.local_time_zone_table[1].name)) + { + /* This locale uses the same abbrevation for standard and + daylight times. So if we see that abbreviation, we don't + know whether it's daylight time. */ + pc.local_time_zone_table[0].value = -1; + pc.local_time_zone_table[1].name = NULL; + } + + if (yyparse (&pc) != 0) + goto fail; + + if (pc.timespec_seen) + *result = pc.seconds; + else + { + if (1 < (pc.times_seen | pc.dates_seen | pc.days_seen | pc.dsts_seen + | (pc.local_zones_seen + pc.zones_seen))) + goto fail; + + tm.tm_year = to_year (pc.year) - TM_YEAR_BASE; + tm.tm_mon = pc.month - 1; + tm.tm_mday = pc.day; + if (pc.times_seen || (pc.rels_seen && ! pc.dates_seen && ! pc.days_seen)) + { + tm.tm_hour = to_hour (pc.hour, pc.meridian); + if (tm.tm_hour < 0) + goto fail; + tm.tm_min = pc.minutes; + tm.tm_sec = pc.seconds.tv_sec; + } + else + { + tm.tm_hour = tm.tm_min = tm.tm_sec = 0; + pc.seconds.tv_nsec = 0; + } + + /* Let mktime deduce tm_isdst if we have an absolute time stamp. */ + if (pc.dates_seen | pc.days_seen | pc.times_seen) + tm.tm_isdst = -1; + + /* But if the input explicitly specifies local time with or without + DST, give mktime that information. */ + if (pc.local_zones_seen) + tm.tm_isdst = pc.local_isdst; + + tm0 = tm; + + Start = mktime (&tm); + + if (! mktime_ok (&tm0, &tm, Start)) + { + if (! pc.zones_seen) + goto fail; + else + { + /* Guard against falsely reporting errors near the time_t + boundaries when parsing times in other time zones. For + example, suppose the input string "1969-12-31 23:00:00 -0100", + the current time zone is 8 hours ahead of UTC, and the min + time_t value is 1970-01-01 00:00:00 UTC. Then the min + localtime value is 1970-01-01 08:00:00, and mktime will + therefore fail on 1969-12-31 23:00:00. To work around the + problem, set the time zone to 1 hour behind UTC temporarily + by setting TZ="XXX1:00" and try mktime again. */ + + long int time_zone = pc.time_zone; + long int abs_time_zone = time_zone < 0 ? - time_zone : time_zone; + long int abs_time_zone_hour = abs_time_zone / 60; + int abs_time_zone_min = abs_time_zone % 60; + char tz1buf[sizeof "XXX+0:00" + + sizeof pc.time_zone * CHAR_BIT / 3]; + if (!tz_was_altered) + tz0 = get_tz (tz0buf); + sprintf (tz1buf, "XXX%s%ld:%02d", "-" + (time_zone < 0), + abs_time_zone_hour, abs_time_zone_min); + if (setenv ("TZ", tz1buf, 1) != 0) + goto fail; + tz_was_altered = true; + tm = tm0; + Start = mktime (&tm); + if (! mktime_ok (&tm0, &tm, Start)) + goto fail; + } + } + + if (pc.days_seen && ! pc.dates_seen) + { + tm.tm_mday += ((pc.day_number - tm.tm_wday + 7) % 7 + + 7 * (pc.day_ordinal - (0 < pc.day_ordinal))); + tm.tm_isdst = -1; + Start = mktime (&tm); + if (Start == (time_t) -1) + goto fail; + } + + if (pc.zones_seen) + { + long int delta = pc.time_zone * 60; + time_t t1; +#ifdef HAVE_TM_GMTOFF + delta -= tm.tm_gmtoff; +#else + time_t t = Start; + struct tm const *gmt = gmtime (&t); + if (! gmt) + goto fail; + delta -= tm_diff (&tm, gmt); +#endif + t1 = Start - delta; + if ((Start < t1) != (delta < 0)) + goto fail; /* time_t overflow */ + Start = t1; + } + + /* Add relative date. */ + if (pc.rel.year | pc.rel.month | pc.rel.day) + { + int year = tm.tm_year + pc.rel.year; + int month = tm.tm_mon + pc.rel.month; + int day = tm.tm_mday + pc.rel.day; + if (((year < tm.tm_year) ^ (pc.rel.year < 0)) + | ((month < tm.tm_mon) ^ (pc.rel.month < 0)) + | ((day < tm.tm_mday) ^ (pc.rel.day < 0))) + goto fail; + tm.tm_year = year; + tm.tm_mon = month; + tm.tm_mday = day; + tm.tm_hour = tm0.tm_hour; + tm.tm_min = tm0.tm_min; + tm.tm_sec = tm0.tm_sec; + tm.tm_isdst = tm0.tm_isdst; + Start = mktime (&tm); + if (Start == (time_t) -1) + goto fail; + } + + /* Add relative hours, minutes, and seconds. On hosts that support + leap seconds, ignore the possibility of leap seconds; e.g., + "+ 10 minutes" adds 600 seconds, even if one of them is a + leap second. Typically this is not what the user wants, but it's + too hard to do it the other way, because the time zone indicator + must be applied before relative times, and if mktime is applied + again the time zone will be lost. */ + { + long int sum_ns = pc.seconds.tv_nsec + pc.rel.ns; + long int normalized_ns = (sum_ns % BILLION + BILLION) % BILLION; + time_t t0 = Start; + long int d1 = 60 * 60 * pc.rel.hour; + time_t t1 = t0 + d1; + long int d2 = 60 * pc.rel.minutes; + time_t t2 = t1 + d2; + long int d3 = pc.rel.seconds; + time_t t3 = t2 + d3; + long int d4 = (sum_ns - normalized_ns) / BILLION; + time_t t4 = t3 + d4; + + if ((d1 / (60 * 60) ^ pc.rel.hour) + | (d2 / 60 ^ pc.rel.minutes) + | ((t1 < t0) ^ (d1 < 0)) + | ((t2 < t1) ^ (d2 < 0)) + | ((t3 < t2) ^ (d3 < 0)) + | ((t4 < t3) ^ (d4 < 0))) + goto fail; + + result->tv_sec = t4; + result->tv_nsec = normalized_ns; + } + } + + goto done; + + fail: + ok = false; + done: + if (tz_was_altered) + ok &= (tz0 ? setenv ("TZ", tz0, 1) : unsetenv ("TZ")) == 0; + if (tz0 != tz0buf) + free (tz0); + return ok; +} + +#if TEST + +int +main (int ac, char **av) +{ + char buff[BUFSIZ]; + + printf ("Enter date, or blank line to exit.\n\t> "); + fflush (stdout); + + buff[BUFSIZ - 1] = '\0'; + while (fgets (buff, BUFSIZ - 1, stdin) && buff[0]) + { + struct timespec d; + struct tm const *tm; + if (! get_date (&d, buff, NULL)) + printf ("Bad format - couldn't convert.\n"); + else if (! (tm = localtime (&d.tv_sec))) + { + long int sec = d.tv_sec; + printf ("localtime (%ld) failed\n", sec); + } + else + { + int ns = d.tv_nsec; + printf ("%04ld-%02d-%02d %02d:%02d:%02d.%09d\n", + tm->tm_year + 1900L, tm->tm_mon + 1, tm->tm_mday, + tm->tm_hour, tm->tm_min, tm->tm_sec, ns); + } + printf ("\t> "); + fflush (stdout); + } + return 0; +} +#endif /* TEST */ diff --git a/lib/getdelim.c b/lib/getdelim.c new file mode 100644 index 0000000..e0b8a70 --- /dev/null +++ b/lib/getdelim.c @@ -0,0 +1,126 @@ +/* getdelim.c --- Implementation of replacement getdelim function. + Copyright (C) 1994, 1996, 1997, 1998, 2001, 2003, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or + modify it under the terms of the GNU General Public License as + published by the Free Software Foundation; either version 2, or (at + your option) any later version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA + 02110-1301, USA. */ + +/* Ported from glibc by Simon Josefsson. */ + +#include <config.h> + +#include "getdelim.h" + +#include <limits.h> +#include <stdlib.h> +#include <errno.h> + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif +#ifndef SSIZE_MAX +# define SSIZE_MAX ((ssize_t) (SIZE_MAX / 2)) +#endif +#if !HAVE_FLOCKFILE +# undef flockfile +# define flockfile(x) ((void) 0) +#endif +#if !HAVE_FUNLOCKFILE +# undef funlockfile +# define funlockfile(x) ((void) 0) +#endif + +/* Read up to (and including) a DELIMITER from FP into *LINEPTR (and + NUL-terminate it). *LINEPTR is a pointer returned from malloc (or + NULL), pointing to *N characters of space. It is realloc'ed as + necessary. Returns the number of characters read (not including + the null terminator), or -1 on error or EOF. */ + +ssize_t +getdelim (char **lineptr, size_t *n, int delimiter, FILE *fp) +{ + ssize_t result; + size_t cur_len = 0; + + if (lineptr == NULL || n == NULL || fp == NULL) + { + errno = EINVAL; + return -1; + } + + flockfile (fp); + + if (*lineptr == NULL || *n == 0) + { + *n = 120; + *lineptr = (char *) malloc (*n); + if (*lineptr == NULL) + { + result = -1; + goto unlock_return; + } + } + + for (;;) + { + int i; + + i = getc (fp); + if (i == EOF) + { + result = -1; + break; + } + + /* Make enough space for len+1 (for final NUL) bytes. */ + if (cur_len + 1 >= *n) + { + size_t needed_max = + SSIZE_MAX < SIZE_MAX ? (size_t) SSIZE_MAX + 1 : SIZE_MAX; + size_t needed = 2 * *n + 1; /* Be generous. */ + char *new_lineptr; + + if (needed_max < needed) + needed = needed_max; + if (cur_len + 1 >= needed) + { + result = -1; + goto unlock_return; + } + + new_lineptr = (char *) realloc (*lineptr, needed); + if (new_lineptr == NULL) + { + result = -1; + goto unlock_return; + } + + *lineptr = new_lineptr; + *n = needed; + } + + (*lineptr)[cur_len] = i; + cur_len++; + + if (i == delimiter) + break; + } + (*lineptr)[cur_len] = '\0'; + result = cur_len ? cur_len : result; + + unlock_return: + funlockfile (fp); + return result; +} diff --git a/lib/getdelim.h b/lib/getdelim.h new file mode 100644 index 0000000..8bc6130 --- /dev/null +++ b/lib/getdelim.h @@ -0,0 +1,28 @@ +/* getdelim.h --- Prototype for replacement getdelim function. + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or + modify it under the terms of the GNU General Public License as + published by the Free Software Foundation; either version 2, or (at + your option) any later version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA + 02110-1301, USA. */ + +/* Written by Simon Josefsson. */ + +/* Get size_t, FILE, ssize_t. And getdelim, if available. */ +# include <stddef.h> +# include <stdio.h> +# include <sys/types.h> + +#if !HAVE_DECL_GETDELIM +ssize_t getdelim (char **lineptr, size_t *n, int delimiter, FILE *stream); +#endif /* !HAVE_GETDELIM */ diff --git a/lib/getline.c b/lib/getline.c new file mode 100644 index 0000000..c8c9244 --- /dev/null +++ b/lib/getline.c @@ -0,0 +1,30 @@ +/* getline.c --- Implementation of replacement getline function. + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or + modify it under the terms of the GNU General Public License as + published by the Free Software Foundation; either version 2, or (at + your option) any later version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA + 02110-1301, USA. */ + +/* Written by Simon Josefsson. */ + +#include <config.h> + +#include "getdelim.h" +#include "getline.h" + +ssize_t +getline (char **lineptr, size_t *n, FILE *stream) +{ + return getdelim (lineptr, n, '\n', stream); +} diff --git a/lib/getline.h b/lib/getline.h new file mode 100644 index 0000000..f0f61c4 --- /dev/null +++ b/lib/getline.h @@ -0,0 +1,28 @@ +/* getline.h --- Prototype for replacement getline function. + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or + modify it under the terms of the GNU General Public License as + published by the Free Software Foundation; either version 2, or (at + your option) any later version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA + 02110-1301, USA. */ + +/* Written by Simon Josefsson. */ + +/* Get size_t, FILE, ssize_t. And getline, if available. */ +# include <stddef.h> +# include <stdio.h> +# include <sys/types.h> + +#if !HAVE_DECL_GETLINE +ssize_t getline (char **lineptr, size_t *n, FILE *stream); +#endif /* !HAVE_GETLINE */ diff --git a/lib/getopt.c b/lib/getopt.c new file mode 100644 index 0000000..3580ad8 --- /dev/null +++ b/lib/getopt.c @@ -0,0 +1,1191 @@ +/* Getopt for GNU. + NOTE: getopt is now part of the C library, so if you don't know what + "Keep this file name-space clean" means, talk to drepper@gnu.org + before changing it! + Copyright (C) 1987,88,89,90,91,92,93,94,95,96,98,99,2000,2001,2002,2003,2004,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +#endif + +#include "getopt.h" + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> + +#ifdef __VMS +# include <unixlib.h> +#endif + +#ifdef _LIBC +# include <libintl.h> +#else +# include "gettext.h" +# define _(msgid) gettext (msgid) +#endif + +#if defined _LIBC && defined USE_IN_LIBIO +# include <wchar.h> +#endif + +#ifndef attribute_hidden +# define attribute_hidden +#endif + +/* Unlike standard Unix `getopt', functions like `getopt_long' + let the user intersperse the options with the other arguments. + + As `getopt_long' works, it permutes the elements of ARGV so that, + when it is done, all the options precede everything else. Thus + all application programs are extended to handle flexible argument order. + + Using `getopt' or setting the environment variable POSIXLY_CORRECT + disables permutation. + Then the application's behavior is completely standard. + + GNU application programs can use a third alternative mode in which + they can distinguish the relative order of options and other arguments. */ + +#include "getopt_int.h" + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +/* 1003.2 says this must be 1 before any call. */ +int optind = 1; + +/* Callers store zero here to inhibit the error message + for unrecognized options. */ + +int opterr = 1; + +/* Set to an option character which was unrecognized. + This must be initialized on some systems to avoid linking in the + system's own getopt implementation. */ + +int optopt = '?'; + +/* Keep a global copy of all internal members of getopt_data. */ + +static struct _getopt_data getopt_data; + + +#if defined HAVE_DECL_GETENV && !HAVE_DECL_GETENV +extern char *getenv (); +#endif + +#ifdef _LIBC +/* Stored original parameters. + XXX This is no good solution. We should rather copy the args so + that we can compare them later. But we must not use malloc(3). */ +extern int __libc_argc; +extern char **__libc_argv; + +/* Bash 2.0 gives us an environment variable containing flags + indicating ARGV elements that should not be considered arguments. */ + +# ifdef USE_NONOPTION_FLAGS +/* Defined in getopt_init.c */ +extern char *__getopt_nonoption_flags; +# endif + +# ifdef USE_NONOPTION_FLAGS +# define SWAP_FLAGS(ch1, ch2) \ + if (d->__nonoption_flags_len > 0) \ + { \ + char __tmp = __getopt_nonoption_flags[ch1]; \ + __getopt_nonoption_flags[ch1] = __getopt_nonoption_flags[ch2]; \ + __getopt_nonoption_flags[ch2] = __tmp; \ + } +# else +# define SWAP_FLAGS(ch1, ch2) +# endif +#else /* !_LIBC */ +# define SWAP_FLAGS(ch1, ch2) +#endif /* _LIBC */ + +/* Exchange two adjacent subsequences of ARGV. + One subsequence is elements [first_nonopt,last_nonopt) + which contains all the non-options that have been skipped so far. + The other is elements [last_nonopt,optind), which contains all + the options processed since those non-options were skipped. + + `first_nonopt' and `last_nonopt' are relocated so that they describe + the new indices of the non-options in ARGV after they are moved. */ + +static void +exchange (char **argv, struct _getopt_data *d) +{ + int bottom = d->__first_nonopt; + int middle = d->__last_nonopt; + int top = d->optind; + char *tem; + + /* Exchange the shorter segment with the far end of the longer segment. + That puts the shorter segment into the right place. + It leaves the longer segment in the right place overall, + but it consists of two parts that need to be swapped next. */ + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + /* First make sure the handling of the `__getopt_nonoption_flags' + string can work normally. Our top argument must be in the range + of the string. */ + if (d->__nonoption_flags_len > 0 && top >= d->__nonoption_flags_max_len) + { + /* We must extend the array. The user plays games with us and + presents new arguments. */ + char *new_str = malloc (top + 1); + if (new_str == NULL) + d->__nonoption_flags_len = d->__nonoption_flags_max_len = 0; + else + { + memset (__mempcpy (new_str, __getopt_nonoption_flags, + d->__nonoption_flags_max_len), + '\0', top + 1 - d->__nonoption_flags_max_len); + d->__nonoption_flags_max_len = top + 1; + __getopt_nonoption_flags = new_str; + } + } +#endif + + while (top > middle && middle > bottom) + { + if (top - middle > middle - bottom) + { + /* Bottom segment is the short one. */ + int len = middle - bottom; + register int i; + + /* Swap it with the top part of the top segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[top - (middle - bottom) + i]; + argv[top - (middle - bottom) + i] = tem; + SWAP_FLAGS (bottom + i, top - (middle - bottom) + i); + } + /* Exclude the moved bottom segment from further swapping. */ + top -= len; + } + else + { + /* Top segment is the short one. */ + int len = top - middle; + register int i; + + /* Swap it with the bottom part of the bottom segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[middle + i]; + argv[middle + i] = tem; + SWAP_FLAGS (bottom + i, middle + i); + } + /* Exclude the moved top segment from further swapping. */ + bottom += len; + } + } + + /* Update records for the slots the non-options now occupy. */ + + d->__first_nonopt += (d->optind - d->__last_nonopt); + d->__last_nonopt = d->optind; +} + +/* Initialize the internal data when the first call is made. */ + +static const char * +_getopt_initialize (int argc, char **argv, const char *optstring, + int posixly_correct, struct _getopt_data *d) +{ + /* Start processing options with ARGV-element 1 (since ARGV-element 0 + is the program name); the sequence of previously skipped + non-option ARGV-elements is empty. */ + + d->__first_nonopt = d->__last_nonopt = d->optind; + + d->__nextchar = NULL; + + d->__posixly_correct = posixly_correct || !!getenv ("POSIXLY_CORRECT"); + + /* Determine how to handle the ordering of options and nonoptions. */ + + if (optstring[0] == '-') + { + d->__ordering = RETURN_IN_ORDER; + ++optstring; + } + else if (optstring[0] == '+') + { + d->__ordering = REQUIRE_ORDER; + ++optstring; + } + else if (d->__posixly_correct) + d->__ordering = REQUIRE_ORDER; + else + d->__ordering = PERMUTE; + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + if (!d->__posixly_correct + && argc == __libc_argc && argv == __libc_argv) + { + if (d->__nonoption_flags_max_len == 0) + { + if (__getopt_nonoption_flags == NULL + || __getopt_nonoption_flags[0] == '\0') + d->__nonoption_flags_max_len = -1; + else + { + const char *orig_str = __getopt_nonoption_flags; + int len = d->__nonoption_flags_max_len = strlen (orig_str); + if (d->__nonoption_flags_max_len < argc) + d->__nonoption_flags_max_len = argc; + __getopt_nonoption_flags = + (char *) malloc (d->__nonoption_flags_max_len); + if (__getopt_nonoption_flags == NULL) + d->__nonoption_flags_max_len = -1; + else + memset (__mempcpy (__getopt_nonoption_flags, orig_str, len), + '\0', d->__nonoption_flags_max_len - len); + } + } + d->__nonoption_flags_len = d->__nonoption_flags_max_len; + } + else + d->__nonoption_flags_len = 0; +#endif + + return optstring; +} + +/* Scan elements of ARGV (whose length is ARGC) for option characters + given in OPTSTRING. + + If an element of ARGV starts with '-', and is not exactly "-" or "--", + then it is an option element. The characters of this element + (aside from the initial '-') are option characters. If `getopt' + is called repeatedly, it returns successively each of the option characters + from each of the option elements. + + If `getopt' finds another option character, it returns that character, + updating `optind' and `nextchar' so that the next call to `getopt' can + resume the scan with the following option character or ARGV-element. + + If there are no more option characters, `getopt' returns -1. + Then `optind' is the index in ARGV of the first ARGV-element + that is not an option. (The ARGV-elements have been permuted + so that those that are not options now come last.) + + OPTSTRING is a string containing the legitimate option characters. + If an option character is seen that is not listed in OPTSTRING, + return '?' after printing an error message. If you set `opterr' to + zero, the error message is suppressed but we still return '?'. + + If a char in OPTSTRING is followed by a colon, that means it wants an arg, + so the following text in the same ARGV-element, or the text of the following + ARGV-element, is returned in `optarg'. Two colons mean an option that + wants an optional arg; if there is text in the current ARGV-element, + it is returned in `optarg', otherwise `optarg' is set to zero. + + If OPTSTRING starts with `-' or `+', it requests different methods of + handling the non-option ARGV-elements. + See the comments about RETURN_IN_ORDER and REQUIRE_ORDER, above. + + Long-named options begin with `--' instead of `-'. + Their names may be abbreviated as long as the abbreviation is unique + or is an exact match for some defined option. If they have an + argument, it follows the option name in the same ARGV-element, separated + from the option name by a `=', or else the in next ARGV-element. + When `getopt' finds a long-named option, it returns 0 if that option's + `flag' field is nonzero, the value of the option's `val' field + if the `flag' field is zero. + + LONGOPTS is a vector of `struct option' terminated by an + element containing a name which is zero. + + LONGIND returns the index in LONGOPT of the long-named option found. + It is only valid when a long-named option has been found by the most + recent call. + + If LONG_ONLY is nonzero, '-' as well as '--' can introduce + long-named options. + + If POSIXLY_CORRECT is nonzero, behave as if the POSIXLY_CORRECT + environment variable were set. */ + +int +_getopt_internal_r (int argc, char **argv, const char *optstring, + const struct option *longopts, int *longind, + int long_only, int posixly_correct, struct _getopt_data *d) +{ + int print_errors = d->opterr; + if (optstring[0] == ':') + print_errors = 0; + + if (argc < 1) + return -1; + + d->optarg = NULL; + + if (d->optind == 0 || !d->__initialized) + { + if (d->optind == 0) + d->optind = 1; /* Don't scan ARGV[0], the program name. */ + optstring = _getopt_initialize (argc, argv, optstring, + posixly_correct, d); + d->__initialized = 1; + } + + /* Test whether ARGV[optind] points to a non-option argument. + Either it does not have option syntax, or there is an environment flag + from the shell indicating it is not an option. The later information + is only used when the used in the GNU libc. */ +#if defined _LIBC && defined USE_NONOPTION_FLAGS +# define NONOPTION_P (argv[d->optind][0] != '-' || argv[d->optind][1] == '\0' \ + || (d->optind < d->__nonoption_flags_len \ + && __getopt_nonoption_flags[d->optind] == '1')) +#else +# define NONOPTION_P (argv[d->optind][0] != '-' || argv[d->optind][1] == '\0') +#endif + + if (d->__nextchar == NULL || *d->__nextchar == '\0') + { + /* Advance to the next ARGV-element. */ + + /* Give FIRST_NONOPT & LAST_NONOPT rational values if OPTIND has been + moved back by the user (who may also have changed the arguments). */ + if (d->__last_nonopt > d->optind) + d->__last_nonopt = d->optind; + if (d->__first_nonopt > d->optind) + d->__first_nonopt = d->optind; + + if (d->__ordering == PERMUTE) + { + /* If we have just processed some options following some non-options, + exchange them so that the options come first. */ + + if (d->__first_nonopt != d->__last_nonopt + && d->__last_nonopt != d->optind) + exchange ((char **) argv, d); + else if (d->__last_nonopt != d->optind) + d->__first_nonopt = d->optind; + + /* Skip any additional non-options + and extend the range of non-options previously skipped. */ + + while (d->optind < argc && NONOPTION_P) + d->optind++; + d->__last_nonopt = d->optind; + } + + /* The special ARGV-element `--' means premature end of options. + Skip it like a null option, + then exchange with previous non-options as if it were an option, + then skip everything else like a non-option. */ + + if (d->optind != argc && !strcmp (argv[d->optind], "--")) + { + d->optind++; + + if (d->__first_nonopt != d->__last_nonopt + && d->__last_nonopt != d->optind) + exchange ((char **) argv, d); + else if (d->__first_nonopt == d->__last_nonopt) + d->__first_nonopt = d->optind; + d->__last_nonopt = argc; + + d->optind = argc; + } + + /* If we have done all the ARGV-elements, stop the scan + and back over any non-options that we skipped and permuted. */ + + if (d->optind == argc) + { + /* Set the next-arg-index to point at the non-options + that we previously skipped, so the caller will digest them. */ + if (d->__first_nonopt != d->__last_nonopt) + d->optind = d->__first_nonopt; + return -1; + } + + /* If we have come to a non-option and did not permute it, + either stop the scan or describe it to the caller and pass it by. */ + + if (NONOPTION_P) + { + if (d->__ordering == REQUIRE_ORDER) + return -1; + d->optarg = argv[d->optind++]; + return 1; + } + + /* We have found another option-ARGV-element. + Skip the initial punctuation. */ + + d->__nextchar = (argv[d->optind] + 1 + + (longopts != NULL && argv[d->optind][1] == '-')); + } + + /* Decode the current option-ARGV-element. */ + + /* Check whether the ARGV-element is a long option. + + If long_only and the ARGV-element has the form "-f", where f is + a valid short option, don't consider it an abbreviated form of + a long option that starts with f. Otherwise there would be no + way to give the -f short option. + + On the other hand, if there's a long option "fubar" and + the ARGV-element is "-fu", do consider that an abbreviation of + the long option, just like "--fu", and not "-f" with arg "u". + + This distinction seems to be the most useful approach. */ + + if (longopts != NULL + && (argv[d->optind][1] == '-' + || (long_only && (argv[d->optind][2] + || !strchr (optstring, argv[d->optind][1]))))) + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = -1; + int option_index; + + for (nameend = d->__nextchar; *nameend && *nameend != '='; nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, d->__nextchar, nameend - d->__nextchar)) + { + if ((unsigned int) (nameend - d->__nextchar) + == (unsigned int) strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else if (long_only + || pfound->has_arg != p->has_arg + || pfound->flag != p->flag + || pfound->val != p->val) + /* Second or later nonexact match found. */ + ambig = 1; + } + + if (ambig && !exact) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("%s: option `%s' is ambiguous\n"), + argv[0], argv[d->optind]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option `%s' is ambiguous\n"), + argv[0], argv[d->optind]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optind++; + d->optopt = 0; + return '?'; + } + + if (pfound != NULL) + { + option_index = indfound; + d->optind++; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + d->optarg = nameend + 1; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (argv[d->optind - 1][1] == '-') + { + /* --option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("\ +%s: option `--%s' doesn't allow an argument\n"), + argv[0], pfound->name); +#else + fprintf (stderr, _("\ +%s: option `--%s' doesn't allow an argument\n"), + argv[0], pfound->name); +#endif + } + else + { + /* +option or -option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("\ +%s: option `%c%s' doesn't allow an argument\n"), + argv[0], argv[d->optind - 1][0], + pfound->name); +#else + fprintf (stderr, _("\ +%s: option `%c%s' doesn't allow an argument\n"), + argv[0], argv[d->optind - 1][0], + pfound->name); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + + d->__nextchar += strlen (d->__nextchar); + + d->optopt = pfound->val; + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (d->optind < argc) + d->optarg = argv[d->optind++]; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optopt = pfound->val; + return optstring[0] == ':' ? ':' : '?'; + } + } + d->__nextchar += strlen (d->__nextchar); + if (longind != NULL) + *longind = option_index; + if (pfound->flag) + { + *(pfound->flag) = pfound->val; + return 0; + } + return pfound->val; + } + + /* Can't find it as a long option. If this is not getopt_long_only, + or the option starts with '--' or is not a valid short + option, then it's an error. + Otherwise interpret it as a short option. */ + if (!long_only || argv[d->optind][1] == '-' + || strchr (optstring, *d->__nextchar) == NULL) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (argv[d->optind][1] == '-') + { + /* --option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: unrecognized option `--%s'\n"), + argv[0], d->__nextchar); +#else + fprintf (stderr, _("%s: unrecognized option `--%s'\n"), + argv[0], d->__nextchar); +#endif + } + else + { + /* +option or -option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: unrecognized option `%c%s'\n"), + argv[0], argv[d->optind][0], d->__nextchar); +#else + fprintf (stderr, _("%s: unrecognized option `%c%s'\n"), + argv[0], argv[d->optind][0], d->__nextchar); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + d->__nextchar = (char *) ""; + d->optind++; + d->optopt = 0; + return '?'; + } + } + + /* Look at and handle the next short option-character. */ + + { + char c = *d->__nextchar++; + char *temp = strchr (optstring, c); + + /* Increment `optind' when we start to process its last character. */ + if (*d->__nextchar == '\0') + ++d->optind; + + if (temp == NULL || c == ':') + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (d->__posixly_correct) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: illegal option -- %c\n"), + argv[0], c); +#else + fprintf (stderr, _("%s: illegal option -- %c\n"), argv[0], c); +#endif + } + else + { +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: invalid option -- %c\n"), + argv[0], c); +#else + fprintf (stderr, _("%s: invalid option -- %c\n"), argv[0], c); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + d->optopt = c; + return '?'; + } + /* Convenience. Treat POSIX -W foo same as long option --foo */ + if (temp[0] == 'W' && temp[1] == ';') + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = 0; + int option_index; + + /* This is an option that requires an argument. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + d->optind++; + } + else if (d->optind == argc) + { + if (print_errors) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, + _("%s: option requires an argument -- %c\n"), + argv[0], c) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option requires an argument -- %c\n"), + argv[0], c); +#endif + } + d->optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + return c; + } + else + /* We already incremented `d->optind' once; + increment it again when taking next ARGV-elt as argument. */ + d->optarg = argv[d->optind++]; + + /* optarg is now the argument, see if it's in the + table of longopts. */ + + for (d->__nextchar = nameend = d->optarg; *nameend && *nameend != '='; + nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, d->__nextchar, nameend - d->__nextchar)) + { + if ((unsigned int) (nameend - d->__nextchar) == strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else + /* Second or later nonexact match found. */ + ambig = 1; + } + if (ambig && !exact) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("%s: option `-W %s' is ambiguous\n"), + argv[0], argv[d->optind]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option `-W %s' is ambiguous\n"), + argv[0], argv[d->optind]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optind++; + return '?'; + } + if (pfound != NULL) + { + option_index = indfound; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + d->optarg = nameend + 1; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[0], pfound->name) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[0], pfound->name); +#endif + } + + d->__nextchar += strlen (d->__nextchar); + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (d->optind < argc) + d->optarg = argv[d->optind++]; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + return optstring[0] == ':' ? ':' : '?'; + } + } + d->__nextchar += strlen (d->__nextchar); + if (longind != NULL) + *longind = option_index; + if (pfound->flag) + { + *(pfound->flag) = pfound->val; + return 0; + } + return pfound->val; + } + d->__nextchar = NULL; + return 'W'; /* Let the application handle it. */ + } + if (temp[1] == ':') + { + if (temp[2] == ':') + { + /* This is an option that accepts an argument optionally. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + d->optind++; + } + else + d->optarg = NULL; + d->__nextchar = NULL; + } + else + { + /* This is an option that requires an argument. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + d->optind++; + } + else if (d->optind == argc) + { + if (print_errors) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option requires an argument -- %c\n"), + argv[0], c) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + __fxprintf (NULL, "%s", buf); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option requires an argument -- %c\n"), + argv[0], c); +#endif + } + d->optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + } + else + /* We already incremented `optind' once; + increment it again when taking next ARGV-elt as argument. */ + d->optarg = argv[d->optind++]; + d->__nextchar = NULL; + } + } + return c; + } +} + +int +_getopt_internal (int argc, char **argv, const char *optstring, + const struct option *longopts, int *longind, + int long_only, int posixly_correct) +{ + int result; + + getopt_data.optind = optind; + getopt_data.opterr = opterr; + + result = _getopt_internal_r (argc, argv, optstring, longopts, longind, + long_only, posixly_correct, &getopt_data); + + optind = getopt_data.optind; + optarg = getopt_data.optarg; + optopt = getopt_data.optopt; + + return result; +} + +/* glibc gets a LSB-compliant getopt. + Standalone applications get a POSIX-compliant getopt. */ +#if _LIBC +enum { POSIXLY_CORRECT = 0 }; +#else +enum { POSIXLY_CORRECT = 1 }; +#endif + +int +getopt (int argc, char *const *argv, const char *optstring) +{ + return _getopt_internal (argc, (char **) argv, optstring, NULL, NULL, 0, + POSIXLY_CORRECT); +} + + +#ifdef TEST + +/* Compile with -DTEST to make an executable for use in testing + the above definition of `getopt'. */ + +int +main (int argc, char **argv) +{ + int c; + int digit_optind = 0; + + while (1) + { + int this_option_optind = optind ? optind : 1; + + c = getopt (argc, argv, "abc:d:0123456789"); + if (c == -1) + break; + + switch (c) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/lib/getopt1.c b/lib/getopt1.c new file mode 100644 index 0000000..cc0746e --- /dev/null +++ b/lib/getopt1.c @@ -0,0 +1,171 @@ +/* getopt_long and getopt_long_only entry points for GNU getopt. + Copyright (C) 1987,88,89,90,91,92,93,94,96,97,98,2004,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef _LIBC +# include <getopt.h> +#else +# include <config.h> +# include "getopt.h" +#endif +#include "getopt_int.h" + +#include <stdio.h> + +/* This needs to come after some library #include + to get __GNU_LIBRARY__ defined. */ +#ifdef __GNU_LIBRARY__ +#include <stdlib.h> +#endif + +#ifndef NULL +#define NULL 0 +#endif + +int +getopt_long (int argc, char *__getopt_argv_const *argv, const char *options, + const struct option *long_options, int *opt_index) +{ + return _getopt_internal (argc, (char **) argv, options, long_options, + opt_index, 0, 0); +} + +int +_getopt_long_r (int argc, char **argv, const char *options, + const struct option *long_options, int *opt_index, + struct _getopt_data *d) +{ + return _getopt_internal_r (argc, argv, options, long_options, opt_index, + 0, 0, d); +} + +/* Like getopt_long, but '-' as well as '--' can indicate a long option. + If an option that starts with '-' (not '--') doesn't match a long option, + but does match a short option, it is parsed as a short option + instead. */ + +int +getopt_long_only (int argc, char *__getopt_argv_const *argv, + const char *options, + const struct option *long_options, int *opt_index) +{ + return _getopt_internal (argc, (char **) argv, options, long_options, + opt_index, 1, 0); +} + +int +_getopt_long_only_r (int argc, char **argv, const char *options, + const struct option *long_options, int *opt_index, + struct _getopt_data *d) +{ + return _getopt_internal_r (argc, argv, options, long_options, opt_index, + 1, 0, d); +} + + +#ifdef TEST + +#include <stdio.h> + +int +main (int argc, char **argv) +{ + int c; + int digit_optind = 0; + + while (1) + { + int this_option_optind = optind ? optind : 1; + int option_index = 0; + static struct option long_options[] = + { + {"add", 1, 0, 0}, + {"append", 0, 0, 0}, + {"delete", 1, 0, 0}, + {"verbose", 0, 0, 0}, + {"create", 0, 0, 0}, + {"file", 1, 0, 0}, + {0, 0, 0, 0} + }; + + c = getopt_long (argc, argv, "abc:d:0123456789", + long_options, &option_index); + if (c == -1) + break; + + switch (c) + { + case 0: + printf ("option %s", long_options[option_index].name); + if (optarg) + printf (" with arg %s", optarg); + printf ("\n"); + break; + + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case 'd': + printf ("option d with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/lib/getopt_.h b/lib/getopt_.h new file mode 100644 index 0000000..615ef9a --- /dev/null +++ b/lib/getopt_.h @@ -0,0 +1,226 @@ +/* Declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004,2005,2006,2007 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GETOPT_H + +#ifndef __need_getopt +# define _GETOPT_H 1 +#endif + +/* Standalone applications should #define __GETOPT_PREFIX to an + identifier that prefixes the external functions and variables + defined in this header. When this happens, include the + headers that might declare getopt so that they will not cause + confusion if included after this file. Then systematically rename + identifiers so that they do not collide with the system functions + and variables. Renaming avoids problems with some compilers and + linkers. */ +#if defined __GETOPT_PREFIX && !defined __need_getopt +# include <stdlib.h> +# include <stdio.h> +# include <unistd.h> +# undef __need_getopt +# undef getopt +# undef getopt_long +# undef getopt_long_only +# undef optarg +# undef opterr +# undef optind +# undef optopt +# define __GETOPT_CONCAT(x, y) x ## y +# define __GETOPT_XCONCAT(x, y) __GETOPT_CONCAT (x, y) +# define __GETOPT_ID(y) __GETOPT_XCONCAT (__GETOPT_PREFIX, y) +# define getopt __GETOPT_ID (getopt) +# define getopt_long __GETOPT_ID (getopt_long) +# define getopt_long_only __GETOPT_ID (getopt_long_only) +# define optarg __GETOPT_ID (optarg) +# define opterr __GETOPT_ID (opterr) +# define optind __GETOPT_ID (optind) +# define optopt __GETOPT_ID (optopt) +#endif + +/* Standalone applications get correct prototypes for getopt_long and + getopt_long_only; they declare "char **argv". libc uses prototypes + with "char *const *argv" that are incorrect because getopt_long and + getopt_long_only can permute argv; this is required for backward + compatibility (e.g., for LSB 2.0.1). + + This used to be `#if defined __GETOPT_PREFIX && !defined __need_getopt', + but it caused redefinition warnings if both unistd.h and getopt.h were + included, since unistd.h includes getopt.h having previously defined + __need_getopt. + + The only place where __getopt_argv_const is used is in definitions + of getopt_long and getopt_long_only below, but these are visible + only if __need_getopt is not defined, so it is quite safe to rewrite + the conditional as follows: +*/ +#if !defined __need_getopt +# if defined __GETOPT_PREFIX +# define __getopt_argv_const /* empty */ +# else +# define __getopt_argv_const const +# endif +#endif + +/* If __GNU_LIBRARY__ is not already defined, either we are being used + standalone, or this is the first header included in the source file. + If we are being used with glibc, we need to include <features.h>, but + that does not exist if we are standalone. So: if __GNU_LIBRARY__ is + not defined, include <ctype.h>, which will pull in <features.h> for us + if it's from glibc. (Why ctype.h? It's guaranteed to exist and it + doesn't flood the namespace with stuff the way some other headers do.) */ +#if !defined __GNU_LIBRARY__ +# include <ctype.h> +#endif + +#ifndef __THROW +# ifndef __GNUC_PREREQ +# define __GNUC_PREREQ(maj, min) (0) +# endif +# if defined __cplusplus && __GNUC_PREREQ (2,8) +# define __THROW throw () +# else +# define __THROW +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +extern char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +extern int optind; + +/* Callers store zero here to inhibit the error message `getopt' prints + for unrecognized options. */ + +extern int opterr; + +/* Set to an option character which was unrecognized. */ + +extern int optopt; + +#ifndef __need_getopt +/* Describe the long-named options requested by the application. + The LONG_OPTIONS argument to getopt_long or getopt_long_only is a vector + of `struct option' terminated by an element containing a name which is + zero. + + The field `has_arg' is: + no_argument (or 0) if the option does not take an argument, + required_argument (or 1) if the option requires an argument, + optional_argument (or 2) if the option takes an optional argument. + + If the field `flag' is not NULL, it points to a variable that is set + to the value given in the field `val' when the option is found, but + left unchanged if the option is not found. + + To have a long-named option do something other than set an `int' to + a compiled-in constant, such as set a value from `optarg', set the + option's `flag' field to zero and its `val' field to a nonzero + value (the equivalent single-letter option character, if there is + one). For long options that have a zero `flag' field, `getopt' + returns the contents of the `val' field. */ + +struct option +{ + const char *name; + /* has_arg can't be an enum because some compilers complain about + type mismatches in all the code that assumes it is an int. */ + int has_arg; + int *flag; + int val; +}; + +/* Names for the values of the `has_arg' field of `struct option'. */ + +# define no_argument 0 +# define required_argument 1 +# define optional_argument 2 +#endif /* need getopt */ + + +/* Get definitions and prototypes for functions to process the + arguments in ARGV (ARGC of them, minus the program name) for + options given in OPTS. + + Return the option character from OPTS just read. Return -1 when + there are no more options. For unrecognized options, or options + missing arguments, `optopt' is set to the option letter, and '?' is + returned. + + The OPTS string is a list of characters which are recognized option + letters, optionally followed by colons, specifying that that letter + takes an argument, to be placed in `optarg'. + + If a letter in OPTS is followed by two colons, its argument is + optional. This behavior is specific to the GNU `getopt'. + + The argument `--' causes premature termination of argument + scanning, explicitly telling `getopt' that there are no more + options. + + If OPTS begins with `-', then non-option arguments are treated as + arguments to the option '\1'. This behavior is specific to the GNU + `getopt'. If OPTS begins with `+', or POSIXLY_CORRECT is set in + the environment, then do not permute arguments. */ + +extern int getopt (int ___argc, char *const *___argv, const char *__shortopts) + __THROW; + +#ifndef __need_getopt +extern int getopt_long (int ___argc, char *__getopt_argv_const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind) + __THROW; +extern int getopt_long_only (int ___argc, char *__getopt_argv_const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind) + __THROW; + +#endif + +#ifdef __cplusplus +} +#endif + +/* Make sure we later can get all the definitions and declarations. */ +#undef __need_getopt + +#endif /* getopt.h */ diff --git a/lib/getopt_int.h b/lib/getopt_int.h new file mode 100644 index 0000000..401579f --- /dev/null +++ b/lib/getopt_int.h @@ -0,0 +1,131 @@ +/* Internal declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GETOPT_INT_H +#define _GETOPT_INT_H 1 + +extern int _getopt_internal (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + int __long_only, int __posixly_correct); + + +/* Reentrant versions which can handle parsing multiple argument + vectors at the same time. */ + +/* Data type for reentrant functions. */ +struct _getopt_data +{ + /* These have exactly the same meaning as the corresponding global + variables, except that they are used for the reentrant + versions of getopt. */ + int optind; + int opterr; + int optopt; + char *optarg; + + /* Internal members. */ + + /* True if the internal members have been initialized. */ + int __initialized; + + /* The next char to be scanned in the option-element + in which the last option character we returned was found. + This allows us to pick up the scan where we left off. + + If this is zero, or a null string, it means resume the scan + by advancing to the next ARGV-element. */ + char *__nextchar; + + /* Describe how to deal with options that follow non-option ARGV-elements. + + If the caller did not specify anything, + the default is REQUIRE_ORDER if the environment variable + POSIXLY_CORRECT is defined, PERMUTE otherwise. + + REQUIRE_ORDER means don't recognize them as options; + stop option processing when the first non-option is seen. + This is what Unix does. + This mode of operation is selected by either setting the environment + variable POSIXLY_CORRECT, or using `+' as the first character + of the list of option characters, or by calling getopt. + + PERMUTE is the default. We permute the contents of ARGV as we + scan, so that eventually all the non-options are at the end. + This allows options to be given in any order, even with programs + that were not written to expect this. + + RETURN_IN_ORDER is an option available to programs that were + written to expect options and other ARGV-elements in any order + and that care about the ordering of the two. We describe each + non-option ARGV-element as if it were the argument of an option + with character code 1. Using `-' as the first character of the + list of option characters selects this mode of operation. + + The special argument `--' forces an end of option-scanning regardless + of the value of `ordering'. In the case of RETURN_IN_ORDER, only + `--' can cause `getopt' to return -1 with `optind' != ARGC. */ + + enum + { + REQUIRE_ORDER, PERMUTE, RETURN_IN_ORDER + } __ordering; + + /* If the POSIXLY_CORRECT environment variable is set + or getopt was called. */ + int __posixly_correct; + + + /* Handle permutation of arguments. */ + + /* Describe the part of ARGV that contains non-options that have + been skipped. `first_nonopt' is the index in ARGV of the first + of them; `last_nonopt' is the index after the last of them. */ + + int __first_nonopt; + int __last_nonopt; + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + int __nonoption_flags_max_len; + int __nonoption_flags_len; +# endif +}; + +/* The initializer is necessary to set OPTIND and OPTERR to their + default values and to clear the initialization flag. */ +#define _GETOPT_DATA_INITIALIZER { 1, 1 } + +extern int _getopt_internal_r (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + int __long_only, int __posixly_correct, + struct _getopt_data *__data); + +extern int _getopt_long_r (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + struct _getopt_data *__data); + +extern int _getopt_long_only_r (int ___argc, char **___argv, + const char *__shortopts, + const struct option *__longopts, + int *__longind, + struct _getopt_data *__data); + +#endif /* getopt_int.h */ diff --git a/lib/getpagesize.h b/lib/getpagesize.h new file mode 100644 index 0000000..8863336 --- /dev/null +++ b/lib/getpagesize.h @@ -0,0 +1,68 @@ +/* Emulate getpagesize on systems that lack it. + Copyright (C) 1999, 2000, 2004, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, + USA. */ + +#ifndef HAVE_GETPAGESIZE + +#include <unistd.h> + +#if !defined getpagesize && defined _SC_PAGESIZE +# if ! (defined __VMS && __VMS_VER < 70000000) +# define getpagesize() sysconf (_SC_PAGESIZE) +# endif +#endif + +#if !defined getpagesize && defined __VMS +# ifdef __ALPHA +# define getpagesize() 8192 +# else +# define getpagesize() 512 +# endif +#endif + +/* This is for BeOS. */ +#if !defined getpagesize && HAVE_OS_H +# include <OS.h> +# if defined B_PAGE_SIZE +# define getpagesize() B_PAGE_SIZE +# endif +#endif + +/* This is for AmigaOS4.0. */ +#if !defined getpagesize && defined __amigaos4__ +# define getpagesize() 2048 +#endif + +#if !defined getpagesize && HAVE_SYS_PARAM_H +# include <sys/param.h> +# ifdef EXEC_PAGESIZE +# define getpagesize() EXEC_PAGESIZE +# else +# ifdef NBPG +# ifndef CLSIZE +# define CLSIZE 1 +# endif +# define getpagesize() (NBPG * CLSIZE) +# else +# ifdef NBPC +# define getpagesize() NBPC +# endif +# endif +# endif +#endif + +#endif /* not HAVE_GETPAGESIZE */ diff --git a/lib/gettext.h b/lib/gettext.h new file mode 100644 index 0000000..9d76ec9 --- /dev/null +++ b/lib/gettext.h @@ -0,0 +1,270 @@ +/* Convenience header for conditional use of GNU <libintl.h>. + Copyright (C) 1995-1998, 2000-2002, 2004-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBGETTEXT_H +#define _LIBGETTEXT_H 1 + +/* NLS can be disabled through the configure --disable-nls option. */ +#if ENABLE_NLS + +/* Get declarations of GNU message catalog functions. */ +# include <libintl.h> + +/* You can set the DEFAULT_TEXT_DOMAIN macro to specify the domain used by + the gettext() and ngettext() macros. This is an alternative to calling + textdomain(), and is useful for libraries. */ +# ifdef DEFAULT_TEXT_DOMAIN +# undef gettext +# define gettext(Msgid) \ + dgettext (DEFAULT_TEXT_DOMAIN, Msgid) +# undef ngettext +# define ngettext(Msgid1, Msgid2, N) \ + dngettext (DEFAULT_TEXT_DOMAIN, Msgid1, Msgid2, N) +# endif + +#else + +/* Solaris /usr/include/locale.h includes /usr/include/libintl.h, which + chokes if dcgettext is defined as a macro. So include it now, to make + later inclusions of <locale.h> a NOP. We don't include <libintl.h> + as well because people using "gettext.h" will not include <libintl.h>, + and also including <libintl.h> would fail on SunOS 4, whereas <locale.h> + is OK. */ +#if defined(__sun) +# include <locale.h> +#endif + +/* Many header files from the libstdc++ coming with g++ 3.3 or newer include + <libintl.h>, which chokes if dcgettext is defined as a macro. So include + it now, to make later inclusions of <libintl.h> a NOP. */ +#if defined(__cplusplus) && defined(__GNUG__) && (__GNUC__ >= 3) +# include <cstdlib> +# if (__GLIBC__ >= 2) || _GLIBCXX_HAVE_LIBINTL_H +# include <libintl.h> +# endif +#endif + +/* Disabled NLS. + The casts to 'const char *' serve the purpose of producing warnings + for invalid uses of the value returned from these functions. + On pre-ANSI systems without 'const', the config.h file is supposed to + contain "#define const". */ +# define gettext(Msgid) ((const char *) (Msgid)) +# define dgettext(Domainname, Msgid) ((void) (Domainname), gettext (Msgid)) +# define dcgettext(Domainname, Msgid, Category) \ + ((void) (Category), dgettext (Domainname, Msgid)) +# define ngettext(Msgid1, Msgid2, N) \ + ((N) == 1 \ + ? ((void) (Msgid2), (const char *) (Msgid1)) \ + : ((void) (Msgid1), (const char *) (Msgid2))) +# define dngettext(Domainname, Msgid1, Msgid2, N) \ + ((void) (Domainname), ngettext (Msgid1, Msgid2, N)) +# define dcngettext(Domainname, Msgid1, Msgid2, N, Category) \ + ((void) (Category), dngettext(Domainname, Msgid1, Msgid2, N)) +# define textdomain(Domainname) ((const char *) (Domainname)) +# define bindtextdomain(Domainname, Dirname) \ + ((void) (Domainname), (const char *) (Dirname)) +# define bind_textdomain_codeset(Domainname, Codeset) \ + ((void) (Domainname), (const char *) (Codeset)) + +#endif + +/* A pseudo function call that serves as a marker for the automated + extraction of messages, but does not call gettext(). The run-time + translation is done at a different place in the code. + The argument, String, should be a literal string. Concatenated strings + and other string expressions won't work. + The macro's expansion is not parenthesized, so that it is suitable as + initializer for static 'char[]' or 'const char[]' variables. */ +#define gettext_noop(String) String + +/* The separator between msgctxt and msgid in a .mo file. */ +#define GETTEXT_CONTEXT_GLUE "\004" + +/* Pseudo function calls, taking a MSGCTXT and a MSGID instead of just a + MSGID. MSGCTXT and MSGID must be string literals. MSGCTXT should be + short and rarely need to change. + The letter 'p' stands for 'particular' or 'special'. */ +#ifdef DEFAULT_TEXT_DOMAIN +# define pgettext(Msgctxt, Msgid) \ + pgettext_aux (DEFAULT_TEXT_DOMAIN, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#else +# define pgettext(Msgctxt, Msgid) \ + pgettext_aux (NULL, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#endif +#define dpgettext(Domainname, Msgctxt, Msgid) \ + pgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#define dcpgettext(Domainname, Msgctxt, Msgid, Category) \ + pgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, Category) +#ifdef DEFAULT_TEXT_DOMAIN +# define npgettext(Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (DEFAULT_TEXT_DOMAIN, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#else +# define npgettext(Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (NULL, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#endif +#define dnpgettext(Domainname, Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#define dcnpgettext(Domainname, Msgctxt, Msgid, MsgidPlural, N, Category) \ + npgettext_aux (Domainname, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, Category) + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +pgettext_aux (const char *domain, + const char *msg_ctxt_id, const char *msgid, + int category) +{ + const char *translation = dcgettext (domain, msg_ctxt_id, category); + if (translation == msg_ctxt_id) + return msgid; + else + return translation; +} + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +npgettext_aux (const char *domain, + const char *msg_ctxt_id, const char *msgid, + const char *msgid_plural, unsigned long int n, + int category) +{ + const char *translation = + dcngettext (domain, msg_ctxt_id, msgid_plural, n, category); + if (translation == msg_ctxt_id || translation == msgid_plural) + return (n == 1 ? msgid : msgid_plural); + else + return translation; +} + +/* The same thing extended for non-constant arguments. Here MSGCTXT and MSGID + can be arbitrary expressions. But for string literals these macros are + less efficient than those above. */ + +#include <string.h> + +#define _LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS \ + (((__GNUC__ >= 3 || __GNUG__ >= 2) && !__STRICT_ANSI__) \ + /* || __STDC_VERSION__ >= 199901L */ ) + +#if !_LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS +#include <stdlib.h> +#endif + +#define pgettext_expr(Msgctxt, Msgid) \ + dcpgettext_expr (NULL, Msgctxt, Msgid, LC_MESSAGES) +#define dpgettext_expr(Domainname, Msgctxt, Msgid) \ + dcpgettext_expr (Domainname, Msgctxt, Msgid, LC_MESSAGES) + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +dcpgettext_expr (const char *domain, + const char *msgctxt, const char *msgid, + int category) +{ + size_t msgctxt_len = strlen (msgctxt) + 1; + size_t msgid_len = strlen (msgid) + 1; + const char *translation; +#if _LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + char msg_ctxt_id[msgctxt_len + msgid_len]; +#else + char buf[1024]; + char *msg_ctxt_id = + (msgctxt_len + msgid_len <= sizeof (buf) + ? buf + : (char *) malloc (msgctxt_len + msgid_len)); + if (msg_ctxt_id != NULL) +#endif + { + memcpy (msg_ctxt_id, msgctxt, msgctxt_len - 1); + msg_ctxt_id[msgctxt_len - 1] = '\004'; + memcpy (msg_ctxt_id + msgctxt_len, msgid, msgid_len); + translation = dcgettext (domain, msg_ctxt_id, category); +#if !_LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + if (msg_ctxt_id != buf) + free (msg_ctxt_id); +#endif + if (translation != msg_ctxt_id) + return translation; + } + return msgid; +} + +#define npgettext_expr(Msgctxt, Msgid, MsgidPlural, N) \ + dcnpgettext_expr (NULL, Msgctxt, Msgid, MsgidPlural, N, LC_MESSAGES) +#define dnpgettext_expr(Domainname, Msgctxt, Msgid, MsgidPlural, N) \ + dcnpgettext_expr (Domainname, Msgctxt, Msgid, MsgidPlural, N, LC_MESSAGES) + +#ifdef __GNUC__ +__inline +#else +#ifdef __cplusplus +inline +#endif +#endif +static const char * +dcnpgettext_expr (const char *domain, + const char *msgctxt, const char *msgid, + const char *msgid_plural, unsigned long int n, + int category) +{ + size_t msgctxt_len = strlen (msgctxt) + 1; + size_t msgid_len = strlen (msgid) + 1; + const char *translation; +#if _LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + char msg_ctxt_id[msgctxt_len + msgid_len]; +#else + char buf[1024]; + char *msg_ctxt_id = + (msgctxt_len + msgid_len <= sizeof (buf) + ? buf + : (char *) malloc (msgctxt_len + msgid_len)); + if (msg_ctxt_id != NULL) +#endif + { + memcpy (msg_ctxt_id, msgctxt, msgctxt_len - 1); + msg_ctxt_id[msgctxt_len - 1] = '\004'; + memcpy (msg_ctxt_id + msgctxt_len, msgid, msgid_len); + translation = dcngettext (domain, msg_ctxt_id, msgid_plural, n, category); +#if !_LIBGETTEXT_HAVE_VARIABLE_SIZE_ARRAYS + if (msg_ctxt_id != buf) + free (msg_ctxt_id); +#endif + if (!(translation == msg_ctxt_id || translation == msgid_plural)) + return translation; + } + return (n == 1 ? msgid : msgid_plural); +} + +#endif /* _LIBGETTEXT_H */ diff --git a/lib/gettime.c b/lib/gettime.c new file mode 100644 index 0000000..86bd325 --- /dev/null +++ b/lib/gettime.c @@ -0,0 +1,49 @@ +/* gettime -- get the system clock + + Copyright (C) 2002, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "timespec.h" + +#include <sys/time.h> + +/* Get the system time into *TS. */ + +void +gettime (struct timespec *ts) +{ +#if HAVE_NANOTIME + nanotime (ts); +#else + +# if defined CLOCK_REALTIME && HAVE_CLOCK_GETTIME + if (clock_gettime (CLOCK_REALTIME, ts) == 0) + return; +# endif + + { + struct timeval tv; + gettimeofday (&tv, NULL); + ts->tv_sec = tv.tv_sec; + ts->tv_nsec = tv.tv_usec * 1000; + } + +#endif +} diff --git a/lib/gettimeofday.c b/lib/gettimeofday.c new file mode 100644 index 0000000..bd5576c --- /dev/null +++ b/lib/gettimeofday.c @@ -0,0 +1,142 @@ +/* Provide gettimeofday for systems that don't have it or for which it's broken. + + Copyright (C) 2001, 2002, 2003, 2005, 2006, 2007 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +/* Specification. */ +#include <sys/time.h> + +#include <time.h> + +#if HAVE_SYS_TIMEB_H +# include <sys/timeb.h> +#endif + +#if GETTIMEOFDAY_CLOBBERS_LOCALTIME || TZSET_CLOBBERS_LOCALTIME + +/* Work around the bug in some systems whereby gettimeofday clobbers + the static buffer that localtime uses for its return value. The + gettimeofday function from Mac OS X 10.0.4 (i.e., Darwin 1.3.7) has + this problem. The tzset replacement is necessary for at least + Solaris 2.5, 2.5.1, and 2.6. */ + +static struct tm tm_zero_buffer; +static struct tm *localtime_buffer_addr = &tm_zero_buffer; + +/* This is a wrapper for localtime. It is used only on systems for which + gettimeofday clobbers the static buffer used for localtime's result. + + On the first call, record the address of the static buffer that + localtime uses for its result. */ + +struct tm * +localtime (time_t const *timep) +{ +#undef localtime + extern struct tm *localtime (time_t const *); + struct tm *tm = localtime (timep); + + if (localtime_buffer_addr == &tm_zero_buffer) + localtime_buffer_addr = tm; + + return tm; +} + +/* Same as above, since gmtime and localtime use the same buffer. */ +struct tm * +gmtime (time_t const *timep) +{ +#undef gmtime + extern struct tm *gmtime (time_t const *); + struct tm *tm = gmtime (timep); + + if (localtime_buffer_addr == &tm_zero_buffer) + localtime_buffer_addr = tm; + + return tm; +} + +#endif /* GETTIMEOFDAY_CLOBBERS_LOCALTIME || TZSET_CLOBBERS_LOCALTIME */ + +#if TZSET_CLOBBERS_LOCALTIME +/* This is a wrapper for tzset, for systems on which tzset may clobber + the static buffer used for localtime's result. */ +void +tzset (void) +{ +#undef tzset + extern void tzset (void); + + /* Save and restore the contents of the buffer used for localtime's + result around the call to tzset. */ + struct tm save = *localtime_buffer_addr; + tzset (); + *localtime_buffer_addr = save; +} +#endif + +/* This is a wrapper for gettimeofday. It is used only on systems + that lack this function, or whose implementation of this function + causes problems. */ + +int +rpl_gettimeofday (struct timeval *restrict tv, void *restrict tz) +{ +#undef gettimeofday +#if HAVE_GETTIMEOFDAY +# if GETTIMEOFDAY_CLOBBERS_LOCALTIME + /* Save and restore the contents of the buffer used for localtime's + result around the call to gettimeofday. */ + struct tm save = *localtime_buffer_addr; +# endif + + int result = gettimeofday (tv, tz); + +# if GETTIMEOFDAY_CLOBBERS_LOCALTIME + *localtime_buffer_addr = save; +# endif + + return result; + +#else + +# if HAVE__FTIME + + struct _timeb timebuf; + _ftime (&timebuf); + tv->tv_sec = timebuf.time; + tv->tv_usec = timebuf.millitm * 1000; + +# else + +# if !defined OK_TO_USE_1S_CLOCK +# error "Only 1-second nominal clock resolution found. Is that intended?" \ + "If so, compile with the -DOK_TO_USE_1S_CLOCK option." +# endif + tv->tv_sec = time (NULL); + tv->tv_usec = 0; + +# endif + + return 0; + +#endif +} diff --git a/lib/gnulib.mk b/lib/gnulib.mk new file mode 100644 index 0000000..77691d6 --- /dev/null +++ b/lib/gnulib.mk @@ -0,0 +1,1553 @@ +## DO NOT EDIT! GENERATED AUTOMATICALLY! +## Process this file with automake to produce Makefile.in. +# Copyright (C) 2004-2007 Free Software Foundation, Inc. +# +# This file is free software, distributed under the terms of the GNU +# General Public License. As a special exception to the GNU General +# Public License, this file may be distributed as part of a program +# that contains a configuration script generated by Autoconf, under +# the same distribution terms as the rest of that program. +# +# Generated by gnulib-tool. +# Reproduce by: gnulib-tool --import --dir=. --local-dir=gl --lib=libtar --source-base=.#bootmp/lib --m4-base=.#bootmp/m4 --doc-base=.#bootmp/doc --aux-dir=.#bootmp/build-aux --avoid=lock --avoid=size_max --avoid=xsize --no-libtool --macro-prefix=gl alloca argmatch argp backupfile closeout configmake dirname error exclude exitfail fileblocks fnmatch-gnu ftruncate full-write getdate getline getopt getpagesize gettext gettime hash human inttostr inttypes lchown localcharset memset mkdtemp modechange obstack quote quotearg rmdir rpmatch safe-read save-cwd savedir setenv stat-time stdbool stdint stpcpy strdup strerror strtol strtoul timespec unlinkdir unlocked-io utime utimens version-etc-fsf xalloc xalloc-die xgetcwd xstrtoumax + +AUTOMAKE_OPTIONS = 1.5 gnits + +noinst_HEADERS = +noinst_LIBRARIES = +noinst_LTLIBRARIES = +EXTRA_DIST = +BUILT_SOURCES = +SUFFIXES = +MOSTLYCLEANFILES = core *.stackdump +MOSTLYCLEANDIRS = +CLEANFILES = +DISTCLEANFILES = +MAINTAINERCLEANFILES = + +AM_CPPFLAGS = + +noinst_LIBRARIES += libtar.a + +libtar_a_SOURCES = +libtar_a_LIBADD = $(gl_LIBOBJS) +libtar_a_DEPENDENCIES = $(gl_LIBOBJS) +EXTRA_libtar_a_SOURCES = + +## begin gnulib module absolute-header + +# Use this preprocessor expression to decide whether #include_next works. +# Do not rely on a 'configure'-time test for this, since the expression +# might appear in an installed header, which is used by some other compiler. +HAVE_INCLUDE_NEXT = (__GNUC__ || 60000000 <= __DECC_VER) + +## end gnulib module absolute-header + +## begin gnulib module alloca + + +EXTRA_DIST += alloca.c + +EXTRA_libtar_a_SOURCES += alloca.c + +libtar_a_LIBADD += @ALLOCA@ +libtar_a_DEPENDENCIES += @ALLOCA@ +## end gnulib module alloca + +## begin gnulib module alloca-opt + +BUILT_SOURCES += $(ALLOCA_H) + +# We need the following in order to create <alloca.h> when the system +# doesn't have one that works with the given compiler. +alloca.h: alloca_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/alloca_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += alloca.h alloca.h-t + +EXTRA_DIST += alloca_.h + +## end gnulib module alloca-opt + +## begin gnulib module allocsa + +libtar_a_SOURCES += allocsa.h allocsa.c + +EXTRA_DIST += allocsa.valgrind + +## end gnulib module allocsa + +## begin gnulib module argmatch + + +EXTRA_DIST += argmatch.c argmatch.h + +EXTRA_libtar_a_SOURCES += argmatch.c + +## end gnulib module argmatch + +## begin gnulib module argp + +libtar_a_SOURCES += argp.h argp-ba.c argp-eexst.c \ + argp-fmtstream.c argp-fmtstream.h argp-fs-xinl.c argp-help.c \ + argp-namefrob.h argp-parse.c argp-pin.c argp-pv.c argp-pvh.c \ + argp-xinl.c + +## end gnulib module argp + +## begin gnulib module backupfile + + +EXTRA_DIST += backupfile.c backupfile.h + +EXTRA_libtar_a_SOURCES += backupfile.c + +## end gnulib module backupfile + +## begin gnulib module canonicalize-lgpl + + +EXTRA_DIST += canonicalize-lgpl.c canonicalize.h + +EXTRA_libtar_a_SOURCES += canonicalize-lgpl.c + +## end gnulib module canonicalize-lgpl + +## begin gnulib module chdir-long + + +EXTRA_DIST += chdir-long.c chdir-long.h + +EXTRA_libtar_a_SOURCES += chdir-long.c + +## end gnulib module chdir-long + +## begin gnulib module chown + + +EXTRA_DIST += chown.c fchown-stub.c + +EXTRA_libtar_a_SOURCES += chown.c fchown-stub.c + +## end gnulib module chown + +## begin gnulib module close-stream + + +EXTRA_DIST += close-stream.c close-stream.h + +EXTRA_libtar_a_SOURCES += close-stream.c + +## end gnulib module close-stream + +## begin gnulib module closeout + + +EXTRA_DIST += closeout.c closeout.h + +EXTRA_libtar_a_SOURCES += closeout.c + +## end gnulib module closeout + +## begin gnulib module configmake + +# Retrieve values of the variables through 'configure' followed by +# 'make', not directly through 'configure', so that a user who +# sets some of these variables consistently on the 'make' command +# line gets correct results. +# +# One advantage of this approach, compared to the classical +# approach of adding -DLIBDIR=\"$(libdir)\" etc. to AM_CPPFLAGS, +# is that it protects against the use of undefined variables. +# If, say, $(libdir) is not set in the Makefile, LIBDIR is not +# defined by this module, and code using LIBDIR gives a +# compilation error. +# +# Another advantage is that 'make' output is shorter. +# +# Listed in the same order as the GNU makefile conventions. +# The Automake-defined pkg* macros are appended, in the order +# listed in the Automake 1.10a+ documentation. +configmake.h: Makefile + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + echo '#define PREFIX "$(prefix)"'; \ + echo '#define EXEC_PREFIX "$(exec_prefix)"'; \ + echo '#define BINDIR "$(bindir)"'; \ + echo '#define SBINDIR "$(sbindir)"'; \ + echo '#define LIBEXECDIR "$(libexecdir)"'; \ + echo '#define DATAROOTDIR "$(datarootdir)"'; \ + echo '#define DATADIR "$(datadir)"'; \ + echo '#define SYSCONFDIR "$(sysconfdir)"'; \ + echo '#define SHAREDSTATEDIR "$(sharedstatedir)"'; \ + echo '#define LOCALSTATEDIR "$(localstatedir)"'; \ + echo '#define INCLUDEDIR "$(includedir)"'; \ + echo '#define OLDINCLUDEDIR "$(oldincludedir)"'; \ + echo '#define DOCDIR "$(docdir)"'; \ + echo '#define INFODIR "$(infodir)"'; \ + echo '#define HTMLDIR "$(htmldir)"'; \ + echo '#define DVIDIR "$(dvidir)"'; \ + echo '#define PDFDIR "$(pdfdir)"'; \ + echo '#define PSDIR "$(psdir)"'; \ + echo '#define LIBDIR "$(libdir)"'; \ + echo '#define LISPDIR "$(lispdir)"'; \ + echo '#define LOCALEDIR "$(localedir)"'; \ + echo '#define MANDIR "$(mandir)"'; \ + echo '#define MANEXT "$(manext)"'; \ + echo '#define PKGDATADIR "$(pkgdatadir)"'; \ + echo '#define PKGINCLUDEDIR "$(pkgincludedir)"'; \ + echo '#define PKGLIBDIR "$(pkglibdir)"'; \ + echo '#define PKGLIBEXECDIR "$(pkglibexecdir)"'; \ + } | sed '/""/d' > $@-t + mv $@-t $@ +BUILT_SOURCES += configmake.h +CLEANFILES += configmake.h configmake.h-t + +## end gnulib module configmake + +## begin gnulib module dirfd + + +EXTRA_DIST += dirfd.c dirfd.h + +EXTRA_libtar_a_SOURCES += dirfd.c + +## end gnulib module dirfd + +## begin gnulib module dirname + + +EXTRA_DIST += basename.c dirname.c dirname.h stripslash.c + +EXTRA_libtar_a_SOURCES += basename.c dirname.c stripslash.c + +## end gnulib module dirname + +## begin gnulib module dup2 + + +EXTRA_DIST += dup2.c + +EXTRA_libtar_a_SOURCES += dup2.c + +## end gnulib module dup2 + +## begin gnulib module error + + +EXTRA_DIST += error.c error.h + +EXTRA_libtar_a_SOURCES += error.c + +## end gnulib module error + +## begin gnulib module exclude + + +EXTRA_DIST += exclude.c exclude.h + +EXTRA_libtar_a_SOURCES += exclude.c + +## end gnulib module exclude + +## begin gnulib module exitfail + + +EXTRA_DIST += exitfail.c exitfail.h + +EXTRA_libtar_a_SOURCES += exitfail.c + +## end gnulib module exitfail + +## begin gnulib module fchdir + +BUILT_SOURCES += $(DIRENT_H) + +# We need the following in order to create <dirent.h> when the system +# doesn't have one that works with the given compiler. +dirent.h: dirent_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_DIRENT_H''@|$(ABSOLUTE_DIRENT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''REPLACE_FCHDIR''@|$(REPLACE_FCHDIR)|g' \ + < $(srcdir)/dirent_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += dirent.h dirent.h-t + +EXTRA_DIST += dirent_.h fchdir.c + +EXTRA_libtar_a_SOURCES += fchdir.c + +## end gnulib module fchdir + +## begin gnulib module fcntl + +BUILT_SOURCES += $(FCNTL_H) + +# We need the following in order to create <fcntl.h> when the system +# doesn't have one that works with the given compiler. +fcntl.h: fcntl_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_FCNTL_H''@|$(ABSOLUTE_FCNTL_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/fcntl_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += fcntl.h fcntl.h-t + +EXTRA_DIST += fcntl_.h + +## end gnulib module fcntl + +## begin gnulib module fcntl-safer + + +EXTRA_DIST += creat-safer.c fcntl--.h fcntl-safer.h open-safer.c + +EXTRA_libtar_a_SOURCES += creat-safer.c open-safer.c + +## end gnulib module fcntl-safer + +## begin gnulib module fileblocks + + +EXTRA_DIST += fileblocks.c + +EXTRA_libtar_a_SOURCES += fileblocks.c + +## end gnulib module fileblocks + +## begin gnulib module float + +BUILT_SOURCES += $(FLOAT_H) + +# We need the following in order to create <float.h> when the system +# doesn't have one that works with the given compiler. +float.h: float_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_FLOAT_H''@|$(ABSOLUTE_FLOAT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/float_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += float.h float.h-t + +EXTRA_DIST += float_.h + +## end gnulib module float + +## begin gnulib module fnmatch + +BUILT_SOURCES += $(FNMATCH_H) + +# We need the following in order to create <fnmatch.h> when the system +# doesn't have one that supports the required API. +fnmatch.h: fnmatch_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/fnmatch_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += fnmatch.h fnmatch.h-t + +EXTRA_DIST += fnmatch.c fnmatch_.h fnmatch_loop.c + +EXTRA_libtar_a_SOURCES += fnmatch.c fnmatch_loop.c + +## end gnulib module fnmatch + +## begin gnulib module fpending + + +EXTRA_DIST += __fpending.c __fpending.h + +EXTRA_libtar_a_SOURCES += __fpending.c + +## end gnulib module fpending + +## begin gnulib module ftruncate + + +EXTRA_DIST += ftruncate.c + +EXTRA_libtar_a_SOURCES += ftruncate.c + +## end gnulib module ftruncate + +## begin gnulib module full-write + +libtar_a_SOURCES += full-write.h full-write.c + +## end gnulib module full-write + +## begin gnulib module getcwd + + +EXTRA_DIST += getcwd.c + +EXTRA_libtar_a_SOURCES += getcwd.c + +## end gnulib module getcwd + +## begin gnulib module getdate + +libtar_a_SOURCES += getdate.y +BUILT_SOURCES += getdate.c +MAINTAINERCLEANFILES += getdate.c +EXTRA_DIST += getdate.c + +EXTRA_DIST += getdate.h + +## end gnulib module getdate + +## begin gnulib module getdelim + + +EXTRA_DIST += getdelim.c getdelim.h + +EXTRA_libtar_a_SOURCES += getdelim.c + +## end gnulib module getdelim + +## begin gnulib module getline + + +EXTRA_DIST += getline.c getline.h + +EXTRA_libtar_a_SOURCES += getline.c + +## end gnulib module getline + +## begin gnulib module getopt + +BUILT_SOURCES += $(GETOPT_H) + +# We need the following in order to create <getopt.h> when the system +# doesn't have one that works with the given compiler. +getopt.h: getopt_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + cat $(srcdir)/getopt_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += getopt.h getopt.h-t + +EXTRA_DIST += getopt.c getopt1.c getopt_.h getopt_int.h + +EXTRA_libtar_a_SOURCES += getopt.c getopt1.c + +## end gnulib module getopt + +## begin gnulib module getpagesize + + +EXTRA_DIST += getpagesize.h + +## end gnulib module getpagesize + +## begin gnulib module gettext + +# This is for those projects which use "gettextize --intl" to put a source-code +# copy of libintl into their package. In such projects, every Makefile.am needs +# -I$(top_builddir)/intl, so that <libintl.h> can be found in this directory. +# For the Makefile.ams in other directories it is the maintainer's +# responsibility; for the one from gnulib we do it here. +# This option has no effect when the user disables NLS (because then the intl +# directory contains no libintl.h file) or when the project does not use +# "gettextize --intl". +#AM_CPPFLAGS += -I$(top_builddir)/intl + +## end gnulib module gettext + +## begin gnulib module gettext-h + +libtar_a_SOURCES += gettext.h + +## end gnulib module gettext-h + +## begin gnulib module gettime + + +EXTRA_DIST += gettime.c + +EXTRA_libtar_a_SOURCES += gettime.c + +## end gnulib module gettime + +## begin gnulib module gettimeofday + + +EXTRA_DIST += gettimeofday.c + +EXTRA_libtar_a_SOURCES += gettimeofday.c + +## end gnulib module gettimeofday + +## begin gnulib module hash + + +EXTRA_DIST += hash.c hash.h + +EXTRA_libtar_a_SOURCES += hash.c + +## end gnulib module hash + +## begin gnulib module human + + +EXTRA_DIST += human.c human.h + +EXTRA_libtar_a_SOURCES += human.c + +## end gnulib module human + +## begin gnulib module intprops + + +EXTRA_DIST += intprops.h + +## end gnulib module intprops + +## begin gnulib module inttostr + + +EXTRA_DIST += imaxtostr.c inttostr.c inttostr.h offtostr.c uinttostr.c umaxtostr.c + +EXTRA_libtar_a_SOURCES += imaxtostr.c inttostr.c offtostr.c uinttostr.c umaxtostr.c + +## end gnulib module inttostr + +## begin gnulib module inttypes + +BUILT_SOURCES += $(INTTYPES_H) + +# We need the following in order to create <inttypes.h> when the system +# doesn't have one that works with the given compiler. +inttypes.h: inttypes_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_INTTYPES_H''@/$(HAVE_INTTYPES_H)/g' \ + -e 's|@''ABSOLUTE_INTTYPES_H''@|$(ABSOLUTE_INTTYPES_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''PRI_MACROS_BROKEN''@/$(PRI_MACROS_BROKEN)/g' \ + -e 's/@''HAVE_LONG_LONG_INT''@/$(HAVE_LONG_LONG_INT)/g' \ + -e 's/@''HAVE_UNSIGNED_LONG_LONG_INT''@/$(HAVE_UNSIGNED_LONG_LONG_INT)/g' \ + -e 's/@''PRIPTR_PREFIX''@/$(PRIPTR_PREFIX)/g' \ + -e 's/@''GNULIB_IMAXABS''@/$(GNULIB_IMAXABS)/g' \ + -e 's/@''GNULIB_IMAXDIV''@/$(GNULIB_IMAXDIV)/g' \ + -e 's/@''GNULIB_STRTOIMAX''@/$(GNULIB_STRTOIMAX)/g' \ + -e 's/@''GNULIB_STRTOUMAX''@/$(GNULIB_STRTOUMAX)/g' \ + -e 's/@''HAVE_DECL_IMAXABS''@/$(HAVE_DECL_IMAXABS)/g' \ + -e 's/@''HAVE_DECL_IMAXDIV''@/$(HAVE_DECL_IMAXDIV)/g' \ + -e 's/@''HAVE_DECL_STRTOIMAX''@/$(HAVE_DECL_STRTOIMAX)/g' \ + -e 's/@''HAVE_DECL_STRTOUMAX''@/$(HAVE_DECL_STRTOUMAX)/g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/inttypes_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += inttypes.h inttypes.h-t + +EXTRA_DIST += inttypes_.h + +## end gnulib module inttypes + +## begin gnulib module lchown + + +EXTRA_DIST += lchown.c lchown.h + +EXTRA_libtar_a_SOURCES += lchown.c + +## end gnulib module lchown + +## begin gnulib module link-warning + +LINK_WARNING_H=$(top_srcdir)/build-aux/link-warning.h + +## end gnulib module link-warning + +## begin gnulib module localcharset + +libtar_a_SOURCES += localcharset.h localcharset.c + +# We need the following in order to install a simple file in $(libdir) +# which is shared with other installed packages. We use a list of referencing +# packages so that "make uninstall" will remove the file if and only if it +# is not used by another installed package. +# On systems with glibc-2.1 or newer, the file is redundant, therefore we +# avoid installing it. + +all-local: charset.alias ref-add.sed ref-del.sed + +charset_alias = $(DESTDIR)$(libdir)/charset.alias +charset_tmp = $(DESTDIR)$(libdir)/charset.tmp +install-exec-local: all-local + test $(GLIBC21) != no || $(mkinstalldirs) $(DESTDIR)$(libdir) + if test -f $(charset_alias); then \ + sed -f ref-add.sed $(charset_alias) > $(charset_tmp) ; \ + $(INSTALL_DATA) $(charset_tmp) $(charset_alias) ; \ + rm -f $(charset_tmp) ; \ + else \ + if test $(GLIBC21) = no; then \ + sed -f ref-add.sed charset.alias > $(charset_tmp) ; \ + $(INSTALL_DATA) $(charset_tmp) $(charset_alias) ; \ + rm -f $(charset_tmp) ; \ + fi ; \ + fi + +uninstall-local: all-local + if test -f $(charset_alias); then \ + sed -f ref-del.sed $(charset_alias) > $(charset_tmp); \ + if grep '^# Packages using this file: $$' $(charset_tmp) \ + > /dev/null; then \ + rm -f $(charset_alias); \ + else \ + $(INSTALL_DATA) $(charset_tmp) $(charset_alias); \ + fi; \ + rm -f $(charset_tmp); \ + fi + +charset.alias: config.charset + rm -f t-$@ $@ + $(SHELL) $(srcdir)/config.charset '$(host)' > t-$@ + mv t-$@ $@ + +SUFFIXES += .sed .sin +.sin.sed: + rm -f t-$@ $@ + sed -e '/^#/d' -e 's/@''PACKAGE''@/$(PACKAGE)/g' $< > t-$@ + mv t-$@ $@ + +CLEANFILES += charset.alias ref-add.sed ref-del.sed + +EXTRA_DIST += config.charset ref-add.sin ref-del.sin + +## end gnulib module localcharset + +## begin gnulib module lstat + + +EXTRA_DIST += lstat.c lstat.h + +EXTRA_libtar_a_SOURCES += lstat.c + +## end gnulib module lstat + +## begin gnulib module malloc + + +EXTRA_DIST += malloc.c + +EXTRA_libtar_a_SOURCES += malloc.c + +## end gnulib module malloc + +## begin gnulib module mbchar + +libtar_a_SOURCES += mbchar.c + +EXTRA_DIST += mbchar.h + +## end gnulib module mbchar + +## begin gnulib module mbscasecmp + +libtar_a_SOURCES += mbscasecmp.c + +## end gnulib module mbscasecmp + +## begin gnulib module mbuiter + +libtar_a_SOURCES += mbuiter.h + +## end gnulib module mbuiter + +## begin gnulib module memchr + + +EXTRA_DIST += memchr.c + +EXTRA_libtar_a_SOURCES += memchr.c + +## end gnulib module memchr + +## begin gnulib module mempcpy + + +EXTRA_DIST += mempcpy.c + +EXTRA_libtar_a_SOURCES += mempcpy.c + +## end gnulib module mempcpy + +## begin gnulib module memrchr + + +EXTRA_DIST += memrchr.c + +EXTRA_libtar_a_SOURCES += memrchr.c + +## end gnulib module memrchr + +## begin gnulib module memset + + +EXTRA_DIST += memset.c + +EXTRA_libtar_a_SOURCES += memset.c + +## end gnulib module memset + +## begin gnulib module mkdtemp + + +EXTRA_DIST += mkdtemp.c + +EXTRA_libtar_a_SOURCES += mkdtemp.c + +## end gnulib module mkdtemp + +## begin gnulib module mktime + + +EXTRA_DIST += mktime.c + +EXTRA_libtar_a_SOURCES += mktime.c + +## end gnulib module mktime + +## begin gnulib module modechange + + +EXTRA_DIST += modechange.c modechange.h + +EXTRA_libtar_a_SOURCES += modechange.c + +## end gnulib module modechange + +## begin gnulib module obstack + + +EXTRA_DIST += obstack.c obstack.h + +EXTRA_libtar_a_SOURCES += obstack.c + +## end gnulib module obstack + +## begin gnulib module openat + + +EXTRA_DIST += at-func.c fchmodat.c fchownat.c fstatat.c mkdirat.c openat-priv.h openat-proc.c openat.c openat.h + +EXTRA_libtar_a_SOURCES += at-func.c fchmodat.c fchownat.c fstatat.c mkdirat.c openat-proc.c openat.c + +## end gnulib module openat + +## begin gnulib module openat-die + +libtar_a_SOURCES += openat-die.c + +## end gnulib module openat-die + +## begin gnulib module pathmax + + +EXTRA_DIST += pathmax.h + +## end gnulib module pathmax + +## begin gnulib module quote + + +EXTRA_DIST += quote.c quote.h + +EXTRA_libtar_a_SOURCES += quote.c + +## end gnulib module quote + +## begin gnulib module quotearg + + +EXTRA_DIST += quotearg.c quotearg.h + +EXTRA_libtar_a_SOURCES += quotearg.c + +## end gnulib module quotearg + +## begin gnulib module readlink + + +EXTRA_DIST += readlink.c + +EXTRA_libtar_a_SOURCES += readlink.c + +## end gnulib module readlink + +## begin gnulib module regex + + +EXTRA_DIST += regcomp.c regex.c regex.h regex_internal.c regex_internal.h regexec.c + +EXTRA_libtar_a_SOURCES += regcomp.c regex.c regex_internal.c regexec.c + +## end gnulib module regex + +## begin gnulib module rmdir + + +EXTRA_DIST += rmdir.c + +EXTRA_libtar_a_SOURCES += rmdir.c + +## end gnulib module rmdir + +## begin gnulib module rpmatch + + +EXTRA_DIST += rpmatch.c + +EXTRA_libtar_a_SOURCES += rpmatch.c + +## end gnulib module rpmatch + +## begin gnulib module safe-read + + +EXTRA_DIST += safe-read.c safe-read.h + +EXTRA_libtar_a_SOURCES += safe-read.c + +## end gnulib module safe-read + +## begin gnulib module safe-write + + +EXTRA_DIST += safe-write.c safe-write.h + +EXTRA_libtar_a_SOURCES += safe-write.c + +## end gnulib module safe-write + +## begin gnulib module same-inode + + +EXTRA_DIST += same-inode.h + +## end gnulib module same-inode + +## begin gnulib module save-cwd + + +EXTRA_DIST += save-cwd.c save-cwd.h + +EXTRA_libtar_a_SOURCES += save-cwd.c + +## end gnulib module save-cwd + +## begin gnulib module savedir + + +EXTRA_DIST += savedir.c savedir.h + +EXTRA_libtar_a_SOURCES += savedir.c + +## end gnulib module savedir + +## begin gnulib module setenv + + +EXTRA_DIST += setenv.c setenv.h unsetenv.c + +EXTRA_libtar_a_SOURCES += setenv.c unsetenv.c + +## end gnulib module setenv + +## begin gnulib module sleep + + +EXTRA_DIST += sleep.c + +EXTRA_libtar_a_SOURCES += sleep.c + +## end gnulib module sleep + +## begin gnulib module stat-macros + + +EXTRA_DIST += stat-macros.h + +## end gnulib module stat-macros + +## begin gnulib module stat-time + + +EXTRA_DIST += stat-time.h + +## end gnulib module stat-time + +## begin gnulib module stdbool + +BUILT_SOURCES += $(STDBOOL_H) + +# We need the following in order to create <stdbool.h> when the system +# doesn't have one that works. +stdbool.h: stdbool_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE__BOOL''@/$(HAVE__BOOL)/g' < $(srcdir)/stdbool_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdbool.h stdbool.h-t + +EXTRA_DIST += stdbool_.h + +## end gnulib module stdbool + +## begin gnulib module stdint + +BUILT_SOURCES += $(STDINT_H) + +# We need the following in order to create <stdint.h> when the system +# doesn't have one that works with the given compiler. +stdint.h: stdint_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_STDINT_H''@/$(HAVE_STDINT_H)/g' \ + -e 's|@''ABSOLUTE_STDINT_H''@|$(ABSOLUTE_STDINT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''HAVE_SYS_TYPES_H''@/$(HAVE_SYS_TYPES_H)/g' \ + -e 's/@''HAVE_INTTYPES_H''@/$(HAVE_INTTYPES_H)/g' \ + -e 's/@''HAVE_SYS_INTTYPES_H''@/$(HAVE_SYS_INTTYPES_H)/g' \ + -e 's/@''HAVE_SYS_BITYPES_H''@/$(HAVE_SYS_BITYPES_H)/g' \ + -e 's/@''HAVE_LONG_LONG_INT''@/$(HAVE_LONG_LONG_INT)/g' \ + -e 's/@''HAVE_UNSIGNED_LONG_LONG_INT''@/$(HAVE_UNSIGNED_LONG_LONG_INT)/g' \ + -e 's/@''BITSIZEOF_PTRDIFF_T''@/$(BITSIZEOF_PTRDIFF_T)/g' \ + -e 's/@''PTRDIFF_T_SUFFIX''@/$(PTRDIFF_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_SIG_ATOMIC_T''@/$(BITSIZEOF_SIG_ATOMIC_T)/g' \ + -e 's/@''HAVE_SIGNED_SIG_ATOMIC_T''@/$(HAVE_SIGNED_SIG_ATOMIC_T)/g' \ + -e 's/@''SIG_ATOMIC_T_SUFFIX''@/$(SIG_ATOMIC_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_SIZE_T''@/$(BITSIZEOF_SIZE_T)/g' \ + -e 's/@''SIZE_T_SUFFIX''@/$(SIZE_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_WCHAR_T''@/$(BITSIZEOF_WCHAR_T)/g' \ + -e 's/@''HAVE_SIGNED_WCHAR_T''@/$(HAVE_SIGNED_WCHAR_T)/g' \ + -e 's/@''WCHAR_T_SUFFIX''@/$(WCHAR_T_SUFFIX)/g' \ + -e 's/@''BITSIZEOF_WINT_T''@/$(BITSIZEOF_WINT_T)/g' \ + -e 's/@''HAVE_SIGNED_WINT_T''@/$(HAVE_SIGNED_WINT_T)/g' \ + -e 's/@''WINT_T_SUFFIX''@/$(WINT_T_SUFFIX)/g' \ + < $(srcdir)/stdint_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdint.h stdint.h-t + +EXTRA_DIST += stdint_.h + +## end gnulib module stdint + +## begin gnulib module stdio + +BUILT_SOURCES += stdio.h + +# We need the following in order to create <stdio.h> when the system +# doesn't have one that works with the given compiler. +stdio.h: stdio_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_STDIO_H''@|$(ABSOLUTE_STDIO_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_FPRINTF_POSIX''@|$(GNULIB_FPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_PRINTF_POSIX''@|$(GNULIB_PRINTF_POSIX)|g' \ + -e 's|@''GNULIB_SNPRINTF''@|$(GNULIB_SNPRINTF)|g' \ + -e 's|@''GNULIB_SPRINTF_POSIX''@|$(GNULIB_SPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VFPRINTF_POSIX''@|$(GNULIB_VFPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VPRINTF_POSIX''@|$(GNULIB_VPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VSNPRINTF''@|$(GNULIB_VSNPRINTF)|g' \ + -e 's|@''GNULIB_VSPRINTF_POSIX''@|$(GNULIB_VSPRINTF_POSIX)|g' \ + -e 's|@''GNULIB_VASPRINTF''@|$(GNULIB_VASPRINTF)|g' \ + -e 's|@''GNULIB_FSEEK''@|$(GNULIB_FSEEK)|g' \ + -e 's|@''GNULIB_FSEEKO''@|$(GNULIB_FSEEKO)|g' \ + -e 's|@''GNULIB_FTELL''@|$(GNULIB_FTELL)|g' \ + -e 's|@''GNULIB_FTELLO''@|$(GNULIB_FTELLO)|g' \ + -e 's|@''GNULIB_FFLUSH''@|$(GNULIB_FFLUSH)|g' \ + -e 's|@''REPLACE_FPRINTF''@|$(REPLACE_FPRINTF)|g' \ + -e 's|@''REPLACE_VFPRINTF''@|$(REPLACE_VFPRINTF)|g' \ + -e 's|@''REPLACE_PRINTF''@|$(REPLACE_PRINTF)|g' \ + -e 's|@''REPLACE_VPRINTF''@|$(REPLACE_VPRINTF)|g' \ + -e 's|@''REPLACE_SNPRINTF''@|$(REPLACE_SNPRINTF)|g' \ + -e 's|@''HAVE_DECL_SNPRINTF''@|$(HAVE_DECL_SNPRINTF)|g' \ + -e 's|@''REPLACE_VSNPRINTF''@|$(REPLACE_VSNPRINTF)|g' \ + -e 's|@''HAVE_DECL_VSNPRINTF''@|$(HAVE_DECL_VSNPRINTF)|g' \ + -e 's|@''REPLACE_SPRINTF''@|$(REPLACE_SPRINTF)|g' \ + -e 's|@''REPLACE_VSPRINTF''@|$(REPLACE_VSPRINTF)|g' \ + -e 's|@''HAVE_VASPRINTF''@|$(HAVE_VASPRINTF)|g' \ + -e 's|@''REPLACE_VASPRINTF''@|$(REPLACE_VASPRINTF)|g' \ + -e 's|@''REPLACE_FSEEKO''@|$(REPLACE_FSEEKO)|g' \ + -e 's|@''REPLACE_FSEEK''@|$(REPLACE_FSEEK)|g' \ + -e 's|@''REPLACE_FTELLO''@|$(REPLACE_FTELLO)|g' \ + -e 's|@''REPLACE_FTELL''@|$(REPLACE_FTELL)|g' \ + -e 's|@''REPLACE_FFLUSH''@|$(REPLACE_FFLUSH)|g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/stdio_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdio.h stdio.h-t + +EXTRA_DIST += stdio_.h + +## end gnulib module stdio + +## begin gnulib module stdlib + +BUILT_SOURCES += stdlib.h + +# We need the following in order to create <stdlib.h> when the system +# doesn't have one that works with the given compiler. +stdlib.h: stdlib_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_STDLIB_H''@|$(ABSOLUTE_STDLIB_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_GETSUBOPT''@|$(GNULIB_GETSUBOPT)|g' \ + -e 's|@''GNULIB_MKDTEMP''@|$(GNULIB_MKDTEMP)|g' \ + -e 's|@''GNULIB_MKSTEMP''@|$(GNULIB_MKSTEMP)|g' \ + -e 's|@''HAVE_GETSUBOPT''@|$(HAVE_GETSUBOPT)|g' \ + -e 's|@''HAVE_MKDTEMP''@|$(HAVE_MKDTEMP)|g' \ + -e 's|@''REPLACE_MKSTEMP''@|$(REPLACE_MKSTEMP)|g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/stdlib_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdlib.h stdlib.h-t + +EXTRA_DIST += stdlib_.h + +## end gnulib module stdlib + +## begin gnulib module stpcpy + + +EXTRA_DIST += stpcpy.c + +EXTRA_libtar_a_SOURCES += stpcpy.c + +## end gnulib module stpcpy + +## begin gnulib module strcase + + +EXTRA_DIST += strcasecmp.c strncasecmp.c + +EXTRA_libtar_a_SOURCES += strcasecmp.c strncasecmp.c + +## end gnulib module strcase + +## begin gnulib module strchrnul + + +EXTRA_DIST += strchrnul.c + +EXTRA_libtar_a_SOURCES += strchrnul.c + +## end gnulib module strchrnul + +## begin gnulib module strdup + + +EXTRA_DIST += strdup.c + +EXTRA_libtar_a_SOURCES += strdup.c + +## end gnulib module strdup + +## begin gnulib module strerror + + +EXTRA_DIST += strerror.c + +EXTRA_libtar_a_SOURCES += strerror.c + +## end gnulib module strerror + +## begin gnulib module string + +BUILT_SOURCES += string.h + +# We need the following in order to create <string.h> when the system +# doesn't have one that works with the given compiler. +string.h: string_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@''ABSOLUTE_STRING_H''@|$(ABSOLUTE_STRING_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_MBSLEN''@|$(GNULIB_MBSLEN)|g' \ + -e 's|@''GNULIB_MBSCHR''@|$(GNULIB_MBSCHR)|g' \ + -e 's|@''GNULIB_MBSRCHR''@|$(GNULIB_MBSRCHR)|g' \ + -e 's|@''GNULIB_MBSSTR''@|$(GNULIB_MBSSTR)|g' \ + -e 's|@''GNULIB_MBSCASECMP''@|$(GNULIB_MBSCASECMP)|g' \ + -e 's|@''GNULIB_MBSNCASECMP''@|$(GNULIB_MBSNCASECMP)|g' \ + -e 's|@''GNULIB_MBSPCASECMP''@|$(GNULIB_MBSPCASECMP)|g' \ + -e 's|@''GNULIB_MBSCASESTR''@|$(GNULIB_MBSCASESTR)|g' \ + -e 's|@''GNULIB_MBSCSPN''@|$(GNULIB_MBSCSPN)|g' \ + -e 's|@''GNULIB_MBSPBRK''@|$(GNULIB_MBSPBRK)|g' \ + -e 's|@''GNULIB_MBSSPN''@|$(GNULIB_MBSSPN)|g' \ + -e 's|@''GNULIB_MBSSEP''@|$(GNULIB_MBSSEP)|g' \ + -e 's|@''GNULIB_MBSTOK_R''@|$(GNULIB_MBSTOK_R)|g' \ + -e 's|@''GNULIB_MEMMEM''@|$(GNULIB_MEMMEM)|g' \ + -e 's|@''GNULIB_MEMPCPY''@|$(GNULIB_MEMPCPY)|g' \ + -e 's|@''GNULIB_MEMRCHR''@|$(GNULIB_MEMRCHR)|g' \ + -e 's|@''GNULIB_STPCPY''@|$(GNULIB_STPCPY)|g' \ + -e 's|@''GNULIB_STPNCPY''@|$(GNULIB_STPNCPY)|g' \ + -e 's|@''GNULIB_STRCHRNUL''@|$(GNULIB_STRCHRNUL)|g' \ + -e 's|@''GNULIB_STRDUP''@|$(GNULIB_STRDUP)|g' \ + -e 's|@''GNULIB_STRNDUP''@|$(GNULIB_STRNDUP)|g' \ + -e 's|@''GNULIB_STRNLEN''@|$(GNULIB_STRNLEN)|g' \ + -e 's|@''GNULIB_STRPBRK''@|$(GNULIB_STRPBRK)|g' \ + -e 's|@''GNULIB_STRSEP''@|$(GNULIB_STRSEP)|g' \ + -e 's|@''GNULIB_STRCASESTR''@|$(GNULIB_STRCASESTR)|g' \ + -e 's|@''GNULIB_STRTOK_R''@|$(GNULIB_STRTOK_R)|g' \ + -e 's|@''HAVE_DECL_MEMMEM''@|$(HAVE_DECL_MEMMEM)|g' \ + -e 's|@''HAVE_MEMPCPY''@|$(HAVE_MEMPCPY)|g' \ + -e 's|@''HAVE_DECL_MEMRCHR''@|$(HAVE_DECL_MEMRCHR)|g' \ + -e 's|@''HAVE_STPCPY''@|$(HAVE_STPCPY)|g' \ + -e 's|@''HAVE_STPNCPY''@|$(HAVE_STPNCPY)|g' \ + -e 's|@''HAVE_STRCASECMP''@|$(HAVE_STRCASECMP)|g' \ + -e 's|@''HAVE_DECL_STRNCASECMP''@|$(HAVE_DECL_STRNCASECMP)|g' \ + -e 's|@''HAVE_STRCHRNUL''@|$(HAVE_STRCHRNUL)|g' \ + -e 's|@''HAVE_DECL_STRDUP''@|$(HAVE_DECL_STRDUP)|g' \ + -e 's|@''HAVE_STRNDUP''@|$(HAVE_STRNDUP)|g' \ + -e 's|@''HAVE_DECL_STRNDUP''@|$(HAVE_DECL_STRNDUP)|g' \ + -e 's|@''HAVE_DECL_STRNLEN''@|$(HAVE_DECL_STRNLEN)|g' \ + -e 's|@''HAVE_STRPBRK''@|$(HAVE_STRPBRK)|g' \ + -e 's|@''HAVE_STRSEP''@|$(HAVE_STRSEP)|g' \ + -e 's|@''HAVE_STRCASESTR''@|$(HAVE_STRCASESTR)|g' \ + -e 's|@''HAVE_DECL_STRTOK_R''@|$(HAVE_DECL_STRTOK_R)|g' \ + -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \ + < $(srcdir)/string_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += string.h string.h-t + +EXTRA_DIST += string_.h + +## end gnulib module string + +## begin gnulib module strndup + + +EXTRA_DIST += strndup.c + +EXTRA_libtar_a_SOURCES += strndup.c + +## end gnulib module strndup + +## begin gnulib module strnlen + + +EXTRA_DIST += strnlen.c + +EXTRA_libtar_a_SOURCES += strnlen.c + +## end gnulib module strnlen + +## begin gnulib module strnlen1 + +libtar_a_SOURCES += strnlen1.h strnlen1.c + +## end gnulib module strnlen1 + +## begin gnulib module strtoimax + + +EXTRA_DIST += strtoimax.c + +EXTRA_libtar_a_SOURCES += strtoimax.c + +## end gnulib module strtoimax + +## begin gnulib module strtol + + +EXTRA_DIST += strtol.c + +EXTRA_libtar_a_SOURCES += strtol.c + +## end gnulib module strtol + +## begin gnulib module strtoll + + +EXTRA_DIST += strtoll.c + +EXTRA_libtar_a_SOURCES += strtoll.c + +## end gnulib module strtoll + +## begin gnulib module strtoul + + +EXTRA_DIST += strtoul.c + +EXTRA_libtar_a_SOURCES += strtoul.c + +## end gnulib module strtoul + +## begin gnulib module strtoull + + +EXTRA_DIST += strtoull.c + +EXTRA_libtar_a_SOURCES += strtoull.c + +## end gnulib module strtoull + +## begin gnulib module strtoumax + + +EXTRA_DIST += strtoumax.c + +EXTRA_libtar_a_SOURCES += strtoumax.c + +## end gnulib module strtoumax + +## begin gnulib module sys_stat + +BUILT_SOURCES += $(SYS_STAT_H) + +# We need the following in order to create <sys/stat.h> when the system +# has one that is incomplete. +sys/stat.h: sys_stat_.h + @MKDIR_P@ sys + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_SYS_STAT_H''@|$(ABSOLUTE_SYS_STAT_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''HAVE_IO_H''@|$(HAVE_IO_H)|g' \ + -e 's|@''HAVE_LSTAT''@|$(HAVE_LSTAT)|g' \ + -e 's|@''HAVE_DECL_MKDIR''@|$(HAVE_DECL_MKDIR)|g' \ + < $(srcdir)/sys_stat_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += sys/stat.h sys/stat.h-t +MOSTLYCLEANDIRS += sys + +EXTRA_DIST += sys_stat_.h + +## end gnulib module sys_stat + +## begin gnulib module sys_time + +BUILT_SOURCES += $(SYS_TIME_H) + +# We need the following in order to create <sys/time.h> when the system +# doesn't have one that works with the given compiler. +sys/time.h: sys_time_.h + @MKDIR_P@ sys + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_SYS_TIME_H''@/$(HAVE_SYS_TIME_H)/g' \ + -e 's|@''ABSOLUTE_SYS_TIME_H''@|$(ABSOLUTE_SYS_TIME_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''REPLACE_GETTIMEOFDAY''@/$(REPLACE_GETTIMEOFDAY)/g' \ + -e 's/@''HAVE_STRUCT_TIMEVAL''@/$(HAVE_STRUCT_TIMEVAL)/g' \ + < $(srcdir)/sys_time_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += sys/time.h sys/time.h-t + +EXTRA_DIST += sys_time_.h + +## end gnulib module sys_time + +## begin gnulib module sysexits + +BUILT_SOURCES += $(SYSEXITS_H) + +# We need the following in order to create <sysexits.h> when the system +# doesn't have one that works with the given compiler. +sysexits.h: sysexits_.h + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''HAVE_SYSEXITS_H''@|$(HAVE_SYSEXITS_H)|g' \ + -e 's|@''ABSOLUTE_SYSEXITS_H''@|$(ABSOLUTE_SYSEXITS_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/sysexits_.h; \ + } > $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += sysexits.h sysexits.h-t + +EXTRA_DIST += sysexits_.h + +## end gnulib module sysexits + +## begin gnulib module tempname + + +EXTRA_DIST += tempname.c tempname.h + +EXTRA_libtar_a_SOURCES += tempname.c + +## end gnulib module tempname + +## begin gnulib module time + +BUILT_SOURCES += time.h + +# We need the following in order to create <time.h> when the system +# doesn't have one that works with the given compiler. +time.h: time_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \ + sed -e 's|@ABSOLUTE_TIME_H''@|$(ABSOLUTE_TIME_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@REPLACE_LOCALTIME_R''@|$(REPLACE_LOCALTIME_R)|g' \ + -e 's|@REPLACE_NANOSLEEP''@|$(REPLACE_NANOSLEEP)|g' \ + -e 's|@REPLACE_STRPTIME''@|$(REPLACE_STRPTIME)|g' \ + -e 's|@REPLACE_TIMEGM''@|$(REPLACE_TIMEGM)|g' \ + -e 's|@SYS_TIME_H_DEFINES_STRUCT_TIMESPEC''@|$(SYS_TIME_H_DEFINES_STRUCT_TIMESPEC)|g' \ + -e 's|@TIME_H_DEFINES_STRUCT_TIMESPEC''@|$(TIME_H_DEFINES_STRUCT_TIMESPEC)|g' \ + < $(srcdir)/time_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += time.h time.h-t + +EXTRA_DIST += time_.h + +## end gnulib module time + +## begin gnulib module time_r + + +EXTRA_DIST += time_r.c + +EXTRA_libtar_a_SOURCES += time_r.c + +## end gnulib module time_r + +## begin gnulib module timespec + + +EXTRA_DIST += timespec.h + +## end gnulib module timespec + +## begin gnulib module unistd + +BUILT_SOURCES += unistd.h + +# We need the following in order to create an empty placeholder for +# <unistd.h> when the system doesn't have one. +unistd.h: unistd_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''HAVE_UNISTD_H''@|$(HAVE_UNISTD_H)|g' \ + -e 's|@''ABSOLUTE_UNISTD_H''@|$(ABSOLUTE_UNISTD_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's|@''GNULIB_CHOWN''@|$(GNULIB_CHOWN)|g' \ + -e 's|@''GNULIB_DUP2''@|$(GNULIB_DUP2)|g' \ + -e 's|@''GNULIB_FCHDIR''@|$(GNULIB_FCHDIR)|g' \ + -e 's|@''GNULIB_FTRUNCATE''@|$(GNULIB_FTRUNCATE)|g' \ + -e 's|@''GNULIB_GETCWD''@|$(GNULIB_GETCWD)|g' \ + -e 's|@''GNULIB_GETLOGIN_R''@|$(GNULIB_GETLOGIN_R)|g' \ + -e 's|@''GNULIB_LSEEK''@|$(GNULIB_LSEEK)|g' \ + -e 's|@''GNULIB_READLINK''@|$(GNULIB_READLINK)|g' \ + -e 's|@''GNULIB_SLEEP''@|$(GNULIB_SLEEP)|g' \ + -e 's|@''HAVE_DUP2''@|$(HAVE_DUP2)|g' \ + -e 's|@''HAVE_FTRUNCATE''@|$(HAVE_FTRUNCATE)|g' \ + -e 's|@''HAVE_READLINK''@|$(HAVE_READLINK)|g' \ + -e 's|@''HAVE_SLEEP''@|$(HAVE_SLEEP)|g' \ + -e 's|@''HAVE_DECL_GETLOGIN_R''@|$(HAVE_DECL_GETLOGIN_R)|g' \ + -e 's|@''REPLACE_CHOWN''@|$(REPLACE_CHOWN)|g' \ + -e 's|@''REPLACE_FCHDIR''@|$(REPLACE_FCHDIR)|g' \ + -e 's|@''REPLACE_GETCWD''@|$(REPLACE_GETCWD)|g' \ + -e 's|@''REPLACE_LSEEK''@|$(REPLACE_LSEEK)|g' \ + < $(srcdir)/unistd_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += unistd.h unistd.h-t + +EXTRA_DIST += unistd_.h + +## end gnulib module unistd + +## begin gnulib module unistd-safer + + +EXTRA_DIST += dup-safer.c fd-safer.c pipe-safer.c unistd--.h unistd-safer.h + +EXTRA_libtar_a_SOURCES += dup-safer.c fd-safer.c pipe-safer.c + +## end gnulib module unistd-safer + +## begin gnulib module unlinkdir + + +EXTRA_DIST += unlinkdir.c unlinkdir.h + +EXTRA_libtar_a_SOURCES += unlinkdir.c + +## end gnulib module unlinkdir + +## begin gnulib module unlocked-io + + +EXTRA_DIST += unlocked-io.h + +## end gnulib module unlocked-io + +## begin gnulib module utime + + +EXTRA_DIST += utime.c + +EXTRA_libtar_a_SOURCES += utime.c + +## end gnulib module utime + +## begin gnulib module utimens + + +EXTRA_DIST += utimens.c utimens.h + +EXTRA_libtar_a_SOURCES += utimens.c + +## end gnulib module utimens + +## begin gnulib module vasnprintf + + +EXTRA_DIST += asnprintf.c float+.h printf-args.c printf-args.h printf-parse.c printf-parse.h vasnprintf.c vasnprintf.h + +EXTRA_libtar_a_SOURCES += asnprintf.c printf-args.c printf-parse.c vasnprintf.c + +## end gnulib module vasnprintf + +## begin gnulib module verify + +libtar_a_SOURCES += verify.h + +## end gnulib module verify + +## begin gnulib module version-etc + +libtar_a_SOURCES += version-etc.h version-etc.c + +## end gnulib module version-etc + +## begin gnulib module version-etc-fsf + +libtar_a_SOURCES += version-etc-fsf.c + +## end gnulib module version-etc-fsf + +## begin gnulib module vsnprintf + + +EXTRA_DIST += vsnprintf.c + +EXTRA_libtar_a_SOURCES += vsnprintf.c + +## end gnulib module vsnprintf + +## begin gnulib module wchar + +BUILT_SOURCES += $(WCHAR_H) + +# We need the following in order to create <wchar.h> when the system +# version does not work standalone. +wchar.h: wchar_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's|@''ABSOLUTE_WCHAR_H''@|$(ABSOLUTE_WCHAR_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + < $(srcdir)/wchar_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += wchar.h wchar.h-t + +EXTRA_DIST += wchar_.h + +## end gnulib module wchar + +## begin gnulib module wctype + +BUILT_SOURCES += $(WCTYPE_H) + +# We need the following in order to create <wctype.h> when the system +# doesn't have one that works with the given compiler. +wctype.h: wctype_.h + rm -f $@-t $@ + { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \ + sed -e 's/@''HAVE_WCTYPE_H''@/$(HAVE_WCTYPE_H)/g' \ + -e 's|@''ABSOLUTE_WCTYPE_H''@|$(ABSOLUTE_WCTYPE_H)|g' \ + -e 's/@''HAVE_INCLUDE_NEXT''@/$(HAVE_INCLUDE_NEXT)/g' \ + -e 's/@''HAVE_ISWCNTRL''@/$(HAVE_ISWCNTRL)/g' \ + -e 's/@''HAVE_WINT_T''@/$(HAVE_WINT_T)/g' \ + < $(srcdir)/wctype_.h; \ + } > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += wctype.h wctype.h-t + +EXTRA_DIST += wctype_.h + +## end gnulib module wctype + +## begin gnulib module wcwidth + +libtar_a_SOURCES += wcwidth.h + +## end gnulib module wcwidth + +## begin gnulib module xalloc + + +EXTRA_DIST += xalloc.h xmalloc.c + +EXTRA_libtar_a_SOURCES += xmalloc.c + +## end gnulib module xalloc + +## begin gnulib module xalloc-die + +libtar_a_SOURCES += xalloc-die.c + +## end gnulib module xalloc-die + +## begin gnulib module xgetcwd + + +EXTRA_DIST += xgetcwd.c xgetcwd.h + +EXTRA_libtar_a_SOURCES += xgetcwd.c + +## end gnulib module xgetcwd + +## begin gnulib module xstrndup + +libtar_a_SOURCES += xstrndup.h xstrndup.c + +## end gnulib module xstrndup + +## begin gnulib module xstrtol + + +EXTRA_DIST += xstrtol.c xstrtol.h xstrtoul.c + +EXTRA_libtar_a_SOURCES += xstrtol.c xstrtoul.c + +## end gnulib module xstrtol + +## begin gnulib module xstrtoumax + +libtar_a_SOURCES += xstrtoumax.c + +## end gnulib module xstrtoumax + + +mostlyclean-local: mostlyclean-generic + @for dir in '' $(MOSTLYCLEANDIRS); do \ + if test -n "$$dir" && test -d $$dir; then \ + echo "rmdir $$dir"; rmdir $$dir; \ + fi; \ + done diff --git a/lib/hash.c b/lib/hash.c new file mode 100644 index 0000000..f4ab12f --- /dev/null +++ b/lib/hash.c @@ -0,0 +1,1048 @@ +/* hash - hashing table processing. + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2006 Free + Software Foundation, Inc. + + Written by Jim Meyering, 1992. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* A generic hash table package. */ + +/* Define USE_OBSTACK to 1 if you want the allocator to use obstacks instead + of malloc. If you change USE_OBSTACK, you have to recompile! */ + +#include <config.h> + +#include "hash.h" +#include "xalloc.h" + +#include <limits.h> +#include <stdio.h> +#include <stdlib.h> + +#if USE_OBSTACK +# include "obstack.h" +# ifndef obstack_chunk_alloc +# define obstack_chunk_alloc malloc +# endif +# ifndef obstack_chunk_free +# define obstack_chunk_free free +# endif +#endif + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +struct hash_table + { + /* The array of buckets starts at BUCKET and extends to BUCKET_LIMIT-1, + for a possibility of N_BUCKETS. Among those, N_BUCKETS_USED buckets + are not empty, there are N_ENTRIES active entries in the table. */ + struct hash_entry *bucket; + struct hash_entry const *bucket_limit; + size_t n_buckets; + size_t n_buckets_used; + size_t n_entries; + + /* Tuning arguments, kept in a physicaly separate structure. */ + const Hash_tuning *tuning; + + /* Three functions are given to `hash_initialize', see the documentation + block for this function. In a word, HASHER randomizes a user entry + into a number up from 0 up to some maximum minus 1; COMPARATOR returns + true if two user entries compare equally; and DATA_FREER is the cleanup + function for a user entry. */ + Hash_hasher hasher; + Hash_comparator comparator; + Hash_data_freer data_freer; + + /* A linked list of freed struct hash_entry structs. */ + struct hash_entry *free_entry_list; + +#if USE_OBSTACK + /* Whenever obstacks are used, it is possible to allocate all overflowed + entries into a single stack, so they all can be freed in a single + operation. It is not clear if the speedup is worth the trouble. */ + struct obstack entry_stack; +#endif + }; + +/* A hash table contains many internal entries, each holding a pointer to + some user provided data (also called a user entry). An entry indistinctly + refers to both the internal entry and its associated user entry. A user + entry contents may be hashed by a randomization function (the hashing + function, or just `hasher' for short) into a number (or `slot') between 0 + and the current table size. At each slot position in the hash table, + starts a linked chain of entries for which the user data all hash to this + slot. A bucket is the collection of all entries hashing to the same slot. + + A good `hasher' function will distribute entries rather evenly in buckets. + In the ideal case, the length of each bucket is roughly the number of + entries divided by the table size. Finding the slot for a data is usually + done in constant time by the `hasher', and the later finding of a precise + entry is linear in time with the size of the bucket. Consequently, a + larger hash table size (that is, a larger number of buckets) is prone to + yielding shorter chains, *given* the `hasher' function behaves properly. + + Long buckets slow down the lookup algorithm. One might use big hash table + sizes in hope to reduce the average length of buckets, but this might + become inordinate, as unused slots in the hash table take some space. The + best bet is to make sure you are using a good `hasher' function (beware + that those are not that easy to write! :-), and to use a table size + larger than the actual number of entries. */ + +/* If an insertion makes the ratio of nonempty buckets to table size larger + than the growth threshold (a number between 0.0 and 1.0), then increase + the table size by multiplying by the growth factor (a number greater than + 1.0). The growth threshold defaults to 0.8, and the growth factor + defaults to 1.414, meaning that the table will have doubled its size + every second time 80% of the buckets get used. */ +#define DEFAULT_GROWTH_THRESHOLD 0.8 +#define DEFAULT_GROWTH_FACTOR 1.414 + +/* If a deletion empties a bucket and causes the ratio of used buckets to + table size to become smaller than the shrink threshold (a number between + 0.0 and 1.0), then shrink the table by multiplying by the shrink factor (a + number greater than the shrink threshold but smaller than 1.0). The shrink + threshold and factor default to 0.0 and 1.0, meaning that the table never + shrinks. */ +#define DEFAULT_SHRINK_THRESHOLD 0.0 +#define DEFAULT_SHRINK_FACTOR 1.0 + +/* Use this to initialize or reset a TUNING structure to + some sensible values. */ +static const Hash_tuning default_tuning = + { + DEFAULT_SHRINK_THRESHOLD, + DEFAULT_SHRINK_FACTOR, + DEFAULT_GROWTH_THRESHOLD, + DEFAULT_GROWTH_FACTOR, + false + }; + +/* Information and lookup. */ + +/* The following few functions provide information about the overall hash + table organization: the number of entries, number of buckets and maximum + length of buckets. */ + +/* Return the number of buckets in the hash table. The table size, the total + number of buckets (used plus unused), or the maximum number of slots, are + the same quantity. */ + +size_t +hash_get_n_buckets (const Hash_table *table) +{ + return table->n_buckets; +} + +/* Return the number of slots in use (non-empty buckets). */ + +size_t +hash_get_n_buckets_used (const Hash_table *table) +{ + return table->n_buckets_used; +} + +/* Return the number of active entries. */ + +size_t +hash_get_n_entries (const Hash_table *table) +{ + return table->n_entries; +} + +/* Return the length of the longest chain (bucket). */ + +size_t +hash_get_max_bucket_length (const Hash_table *table) +{ + struct hash_entry const *bucket; + size_t max_bucket_length = 0; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + if (bucket->data) + { + struct hash_entry const *cursor = bucket; + size_t bucket_length = 1; + + while (cursor = cursor->next, cursor) + bucket_length++; + + if (bucket_length > max_bucket_length) + max_bucket_length = bucket_length; + } + } + + return max_bucket_length; +} + +/* Do a mild validation of a hash table, by traversing it and checking two + statistics. */ + +bool +hash_table_ok (const Hash_table *table) +{ + struct hash_entry const *bucket; + size_t n_buckets_used = 0; + size_t n_entries = 0; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + if (bucket->data) + { + struct hash_entry const *cursor = bucket; + + /* Count bucket head. */ + n_buckets_used++; + n_entries++; + + /* Count bucket overflow. */ + while (cursor = cursor->next, cursor) + n_entries++; + } + } + + if (n_buckets_used == table->n_buckets_used && n_entries == table->n_entries) + return true; + + return false; +} + +void +hash_print_statistics (const Hash_table *table, FILE *stream) +{ + size_t n_entries = hash_get_n_entries (table); + size_t n_buckets = hash_get_n_buckets (table); + size_t n_buckets_used = hash_get_n_buckets_used (table); + size_t max_bucket_length = hash_get_max_bucket_length (table); + + fprintf (stream, "# entries: %lu\n", (unsigned long int) n_entries); + fprintf (stream, "# buckets: %lu\n", (unsigned long int) n_buckets); + fprintf (stream, "# buckets used: %lu (%.2f%%)\n", + (unsigned long int) n_buckets_used, + (100.0 * n_buckets_used) / n_buckets); + fprintf (stream, "max bucket length: %lu\n", + (unsigned long int) max_bucket_length); +} + +/* If ENTRY matches an entry already in the hash table, return the + entry from the table. Otherwise, return NULL. */ + +void * +hash_lookup (const Hash_table *table, const void *entry) +{ + struct hash_entry const *bucket + = table->bucket + table->hasher (entry, table->n_buckets); + struct hash_entry const *cursor; + + if (! (bucket < table->bucket_limit)) + abort (); + + if (bucket->data == NULL) + return NULL; + + for (cursor = bucket; cursor; cursor = cursor->next) + if (table->comparator (entry, cursor->data)) + return cursor->data; + + return NULL; +} + +/* Walking. */ + +/* The functions in this page traverse the hash table and process the + contained entries. For the traversal to work properly, the hash table + should not be resized nor modified while any particular entry is being + processed. In particular, entries should not be added or removed. */ + +/* Return the first data in the table, or NULL if the table is empty. */ + +void * +hash_get_first (const Hash_table *table) +{ + struct hash_entry const *bucket; + + if (table->n_entries == 0) + return NULL; + + for (bucket = table->bucket; ; bucket++) + if (! (bucket < table->bucket_limit)) + abort (); + else if (bucket->data) + return bucket->data; +} + +/* Return the user data for the entry following ENTRY, where ENTRY has been + returned by a previous call to either `hash_get_first' or `hash_get_next'. + Return NULL if there are no more entries. */ + +void * +hash_get_next (const Hash_table *table, const void *entry) +{ + struct hash_entry const *bucket + = table->bucket + table->hasher (entry, table->n_buckets); + struct hash_entry const *cursor; + + if (! (bucket < table->bucket_limit)) + abort (); + + /* Find next entry in the same bucket. */ + for (cursor = bucket; cursor; cursor = cursor->next) + if (cursor->data == entry && cursor->next) + return cursor->next->data; + + /* Find first entry in any subsequent bucket. */ + while (++bucket < table->bucket_limit) + if (bucket->data) + return bucket->data; + + /* None found. */ + return NULL; +} + +/* Fill BUFFER with pointers to active user entries in the hash table, then + return the number of pointers copied. Do not copy more than BUFFER_SIZE + pointers. */ + +size_t +hash_get_entries (const Hash_table *table, void **buffer, + size_t buffer_size) +{ + size_t counter = 0; + struct hash_entry const *bucket; + struct hash_entry const *cursor; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + if (bucket->data) + { + for (cursor = bucket; cursor; cursor = cursor->next) + { + if (counter >= buffer_size) + return counter; + buffer[counter++] = cursor->data; + } + } + } + + return counter; +} + +/* Call a PROCESSOR function for each entry of a hash table, and return the + number of entries for which the processor function returned success. A + pointer to some PROCESSOR_DATA which will be made available to each call to + the processor function. The PROCESSOR accepts two arguments: the first is + the user entry being walked into, the second is the value of PROCESSOR_DATA + as received. The walking continue for as long as the PROCESSOR function + returns nonzero. When it returns zero, the walking is interrupted. */ + +size_t +hash_do_for_each (const Hash_table *table, Hash_processor processor, + void *processor_data) +{ + size_t counter = 0; + struct hash_entry const *bucket; + struct hash_entry const *cursor; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + if (bucket->data) + { + for (cursor = bucket; cursor; cursor = cursor->next) + { + if (!(*processor) (cursor->data, processor_data)) + return counter; + counter++; + } + } + } + + return counter; +} + +/* Allocation and clean-up. */ + +/* Return a hash index for a NUL-terminated STRING between 0 and N_BUCKETS-1. + This is a convenience routine for constructing other hashing functions. */ + +#if USE_DIFF_HASH + +/* About hashings, Paul Eggert writes to me (FP), on 1994-01-01: "Please see + B. J. McKenzie, R. Harries & T. Bell, Selecting a hashing algorithm, + Software--practice & experience 20, 2 (Feb 1990), 209-224. Good hash + algorithms tend to be domain-specific, so what's good for [diffutils'] io.c + may not be good for your application." */ + +size_t +hash_string (const char *string, size_t n_buckets) +{ +# define ROTATE_LEFT(Value, Shift) \ + ((Value) << (Shift) | (Value) >> ((sizeof (size_t) * CHAR_BIT) - (Shift))) +# define HASH_ONE_CHAR(Value, Byte) \ + ((Byte) + ROTATE_LEFT (Value, 7)) + + size_t value = 0; + unsigned char ch; + + for (; (ch = *string); string++) + value = HASH_ONE_CHAR (value, ch); + return value % n_buckets; + +# undef ROTATE_LEFT +# undef HASH_ONE_CHAR +} + +#else /* not USE_DIFF_HASH */ + +/* This one comes from `recode', and performs a bit better than the above as + per a few experiments. It is inspired from a hashing routine found in the + very old Cyber `snoop', itself written in typical Greg Mansfield style. + (By the way, what happened to this excellent man? Is he still alive?) */ + +size_t +hash_string (const char *string, size_t n_buckets) +{ + size_t value = 0; + unsigned char ch; + + for (; (ch = *string); string++) + value = (value * 31 + ch) % n_buckets; + return value; +} + +#endif /* not USE_DIFF_HASH */ + +/* Return true if CANDIDATE is a prime number. CANDIDATE should be an odd + number at least equal to 11. */ + +static bool +is_prime (size_t candidate) +{ + size_t divisor = 3; + size_t square = divisor * divisor; + + while (square < candidate && (candidate % divisor)) + { + divisor++; + square += 4 * divisor; + divisor++; + } + + return (candidate % divisor ? true : false); +} + +/* Round a given CANDIDATE number up to the nearest prime, and return that + prime. Primes lower than 10 are merely skipped. */ + +static size_t +next_prime (size_t candidate) +{ + /* Skip small primes. */ + if (candidate < 10) + candidate = 10; + + /* Make it definitely odd. */ + candidate |= 1; + + while (!is_prime (candidate)) + candidate += 2; + + return candidate; +} + +void +hash_reset_tuning (Hash_tuning *tuning) +{ + *tuning = default_tuning; +} + +/* For the given hash TABLE, check the user supplied tuning structure for + reasonable values, and return true if there is no gross error with it. + Otherwise, definitively reset the TUNING field to some acceptable default + in the hash table (that is, the user loses the right of further modifying + tuning arguments), and return false. */ + +static bool +check_tuning (Hash_table *table) +{ + const Hash_tuning *tuning = table->tuning; + + /* Be a bit stricter than mathematics would require, so that + rounding errors in size calculations do not cause allocations to + fail to grow or shrink as they should. The smallest allocation + is 11 (due to next_prime's algorithm), so an epsilon of 0.1 + should be good enough. */ + float epsilon = 0.1f; + + if (epsilon < tuning->growth_threshold + && tuning->growth_threshold < 1 - epsilon + && 1 + epsilon < tuning->growth_factor + && 0 <= tuning->shrink_threshold + && tuning->shrink_threshold + epsilon < tuning->shrink_factor + && tuning->shrink_factor <= 1 + && tuning->shrink_threshold + epsilon < tuning->growth_threshold) + return true; + + table->tuning = &default_tuning; + return false; +} + +/* Allocate and return a new hash table, or NULL upon failure. The initial + number of buckets is automatically selected so as to _guarantee_ that you + may insert at least CANDIDATE different user entries before any growth of + the hash table size occurs. So, if have a reasonably tight a-priori upper + bound on the number of entries you intend to insert in the hash table, you + may save some table memory and insertion time, by specifying it here. If + the IS_N_BUCKETS field of the TUNING structure is true, the CANDIDATE + argument has its meaning changed to the wanted number of buckets. + + TUNING points to a structure of user-supplied values, in case some fine + tuning is wanted over the default behavior of the hasher. If TUNING is + NULL, the default tuning parameters are used instead. + + The user-supplied HASHER function should be provided. It accepts two + arguments ENTRY and TABLE_SIZE. It computes, by hashing ENTRY contents, a + slot number for that entry which should be in the range 0..TABLE_SIZE-1. + This slot number is then returned. + + The user-supplied COMPARATOR function should be provided. It accepts two + arguments pointing to user data, it then returns true for a pair of entries + that compare equal, or false otherwise. This function is internally called + on entries which are already known to hash to the same bucket index. + + The user-supplied DATA_FREER function, when not NULL, may be later called + with the user data as an argument, just before the entry containing the + data gets freed. This happens from within `hash_free' or `hash_clear'. + You should specify this function only if you want these functions to free + all of your `data' data. This is typically the case when your data is + simply an auxiliary struct that you have malloc'd to aggregate several + values. */ + +Hash_table * +hash_initialize (size_t candidate, const Hash_tuning *tuning, + Hash_hasher hasher, Hash_comparator comparator, + Hash_data_freer data_freer) +{ + Hash_table *table; + + if (hasher == NULL || comparator == NULL) + return NULL; + + table = malloc (sizeof *table); + if (table == NULL) + return NULL; + + if (!tuning) + tuning = &default_tuning; + table->tuning = tuning; + if (!check_tuning (table)) + { + /* Fail if the tuning options are invalid. This is the only occasion + when the user gets some feedback about it. Once the table is created, + if the user provides invalid tuning options, we silently revert to + using the defaults, and ignore further request to change the tuning + options. */ + goto fail; + } + + if (!tuning->is_n_buckets) + { + float new_candidate = candidate / tuning->growth_threshold; + if (SIZE_MAX <= new_candidate) + goto fail; + candidate = new_candidate; + } + + if (xalloc_oversized (candidate, sizeof *table->bucket)) + goto fail; + table->n_buckets = next_prime (candidate); + if (xalloc_oversized (table->n_buckets, sizeof *table->bucket)) + goto fail; + + table->bucket = calloc (table->n_buckets, sizeof *table->bucket); + table->bucket_limit = table->bucket + table->n_buckets; + table->n_buckets_used = 0; + table->n_entries = 0; + + table->hasher = hasher; + table->comparator = comparator; + table->data_freer = data_freer; + + table->free_entry_list = NULL; +#if USE_OBSTACK + obstack_init (&table->entry_stack); +#endif + return table; + + fail: + free (table); + return NULL; +} + +/* Make all buckets empty, placing any chained entries on the free list. + Apply the user-specified function data_freer (if any) to the datas of any + affected entries. */ + +void +hash_clear (Hash_table *table) +{ + struct hash_entry *bucket; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + if (bucket->data) + { + struct hash_entry *cursor; + struct hash_entry *next; + + /* Free the bucket overflow. */ + for (cursor = bucket->next; cursor; cursor = next) + { + if (table->data_freer) + (*table->data_freer) (cursor->data); + cursor->data = NULL; + + next = cursor->next; + /* Relinking is done one entry at a time, as it is to be expected + that overflows are either rare or short. */ + cursor->next = table->free_entry_list; + table->free_entry_list = cursor; + } + + /* Free the bucket head. */ + if (table->data_freer) + (*table->data_freer) (bucket->data); + bucket->data = NULL; + bucket->next = NULL; + } + } + + table->n_buckets_used = 0; + table->n_entries = 0; +} + +/* Reclaim all storage associated with a hash table. If a data_freer + function has been supplied by the user when the hash table was created, + this function applies it to the data of each entry before freeing that + entry. */ + +void +hash_free (Hash_table *table) +{ + struct hash_entry *bucket; + struct hash_entry *cursor; + struct hash_entry *next; + + /* Call the user data_freer function. */ + if (table->data_freer && table->n_entries) + { + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + if (bucket->data) + { + for (cursor = bucket; cursor; cursor = cursor->next) + { + (*table->data_freer) (cursor->data); + } + } + } + } + +#if USE_OBSTACK + + obstack_free (&table->entry_stack, NULL); + +#else + + /* Free all bucket overflowed entries. */ + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + for (cursor = bucket->next; cursor; cursor = next) + { + next = cursor->next; + free (cursor); + } + } + + /* Also reclaim the internal list of previously freed entries. */ + for (cursor = table->free_entry_list; cursor; cursor = next) + { + next = cursor->next; + free (cursor); + } + +#endif + + /* Free the remainder of the hash table structure. */ + free (table->bucket); + free (table); +} + +/* Insertion and deletion. */ + +/* Get a new hash entry for a bucket overflow, possibly by reclying a + previously freed one. If this is not possible, allocate a new one. */ + +static struct hash_entry * +allocate_entry (Hash_table *table) +{ + struct hash_entry *new; + + if (table->free_entry_list) + { + new = table->free_entry_list; + table->free_entry_list = new->next; + } + else + { +#if USE_OBSTACK + new = obstack_alloc (&table->entry_stack, sizeof *new); +#else + new = malloc (sizeof *new); +#endif + } + + return new; +} + +/* Free a hash entry which was part of some bucket overflow, + saving it for later recycling. */ + +static void +free_entry (Hash_table *table, struct hash_entry *entry) +{ + entry->data = NULL; + entry->next = table->free_entry_list; + table->free_entry_list = entry; +} + +/* This private function is used to help with insertion and deletion. When + ENTRY matches an entry in the table, return a pointer to the corresponding + user data and set *BUCKET_HEAD to the head of the selected bucket. + Otherwise, return NULL. When DELETE is true and ENTRY matches an entry in + the table, unlink the matching entry. */ + +static void * +hash_find_entry (Hash_table *table, const void *entry, + struct hash_entry **bucket_head, bool delete) +{ + struct hash_entry *bucket + = table->bucket + table->hasher (entry, table->n_buckets); + struct hash_entry *cursor; + + if (! (bucket < table->bucket_limit)) + abort (); + + *bucket_head = bucket; + + /* Test for empty bucket. */ + if (bucket->data == NULL) + return NULL; + + /* See if the entry is the first in the bucket. */ + if ((*table->comparator) (entry, bucket->data)) + { + void *data = bucket->data; + + if (delete) + { + if (bucket->next) + { + struct hash_entry *next = bucket->next; + + /* Bump the first overflow entry into the bucket head, then save + the previous first overflow entry for later recycling. */ + *bucket = *next; + free_entry (table, next); + } + else + { + bucket->data = NULL; + } + } + + return data; + } + + /* Scan the bucket overflow. */ + for (cursor = bucket; cursor->next; cursor = cursor->next) + { + if ((*table->comparator) (entry, cursor->next->data)) + { + void *data = cursor->next->data; + + if (delete) + { + struct hash_entry *next = cursor->next; + + /* Unlink the entry to delete, then save the freed entry for later + recycling. */ + cursor->next = next->next; + free_entry (table, next); + } + + return data; + } + } + + /* No entry found. */ + return NULL; +} + +/* For an already existing hash table, change the number of buckets through + specifying CANDIDATE. The contents of the hash table are preserved. The + new number of buckets is automatically selected so as to _guarantee_ that + the table may receive at least CANDIDATE different user entries, including + those already in the table, before any other growth of the hash table size + occurs. If TUNING->IS_N_BUCKETS is true, then CANDIDATE specifies the + exact number of buckets desired. */ + +bool +hash_rehash (Hash_table *table, size_t candidate) +{ + Hash_table *new_table; + struct hash_entry *bucket; + struct hash_entry *cursor; + struct hash_entry *next; + + new_table = hash_initialize (candidate, table->tuning, table->hasher, + table->comparator, table->data_freer); + if (new_table == NULL) + return false; + + /* Merely reuse the extra old space into the new table. */ +#if USE_OBSTACK + obstack_free (&new_table->entry_stack, NULL); + new_table->entry_stack = table->entry_stack; +#endif + new_table->free_entry_list = table->free_entry_list; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + if (bucket->data) + for (cursor = bucket; cursor; cursor = next) + { + void *data = cursor->data; + struct hash_entry *new_bucket + = (new_table->bucket + + new_table->hasher (data, new_table->n_buckets)); + + if (! (new_bucket < new_table->bucket_limit)) + abort (); + + next = cursor->next; + + if (new_bucket->data) + { + if (cursor == bucket) + { + /* Allocate or recycle an entry, when moving from a bucket + header into a bucket overflow. */ + struct hash_entry *new_entry = allocate_entry (new_table); + + if (new_entry == NULL) + return false; + + new_entry->data = data; + new_entry->next = new_bucket->next; + new_bucket->next = new_entry; + } + else + { + /* Merely relink an existing entry, when moving from a + bucket overflow into a bucket overflow. */ + cursor->next = new_bucket->next; + new_bucket->next = cursor; + } + } + else + { + /* Free an existing entry, when moving from a bucket + overflow into a bucket header. Also take care of the + simple case of moving from a bucket header into a bucket + header. */ + new_bucket->data = data; + new_table->n_buckets_used++; + if (cursor != bucket) + free_entry (new_table, cursor); + } + } + + free (table->bucket); + table->bucket = new_table->bucket; + table->bucket_limit = new_table->bucket_limit; + table->n_buckets = new_table->n_buckets; + table->n_buckets_used = new_table->n_buckets_used; + table->free_entry_list = new_table->free_entry_list; + /* table->n_entries already holds its value. */ +#if USE_OBSTACK + table->entry_stack = new_table->entry_stack; +#endif + free (new_table); + + return true; +} + +/* If ENTRY matches an entry already in the hash table, return the pointer + to the entry from the table. Otherwise, insert ENTRY and return ENTRY. + Return NULL if the storage required for insertion cannot be allocated. */ + +void * +hash_insert (Hash_table *table, const void *entry) +{ + void *data; + struct hash_entry *bucket; + + /* The caller cannot insert a NULL entry. */ + if (! entry) + abort (); + + /* If there's a matching entry already in the table, return that. */ + if ((data = hash_find_entry (table, entry, &bucket, false)) != NULL) + return data; + + /* ENTRY is not matched, it should be inserted. */ + + if (bucket->data) + { + struct hash_entry *new_entry = allocate_entry (table); + + if (new_entry == NULL) + return NULL; + + /* Add ENTRY in the overflow of the bucket. */ + + new_entry->data = (void *) entry; + new_entry->next = bucket->next; + bucket->next = new_entry; + table->n_entries++; + return (void *) entry; + } + + /* Add ENTRY right in the bucket head. */ + + bucket->data = (void *) entry; + table->n_entries++; + table->n_buckets_used++; + + /* If the growth threshold of the buckets in use has been reached, increase + the table size and rehash. There's no point in checking the number of + entries: if the hashing function is ill-conditioned, rehashing is not + likely to improve it. */ + + if (table->n_buckets_used + > table->tuning->growth_threshold * table->n_buckets) + { + /* Check more fully, before starting real work. If tuning arguments + became invalid, the second check will rely on proper defaults. */ + check_tuning (table); + if (table->n_buckets_used + > table->tuning->growth_threshold * table->n_buckets) + { + const Hash_tuning *tuning = table->tuning; + float candidate = + (tuning->is_n_buckets + ? (table->n_buckets * tuning->growth_factor) + : (table->n_buckets * tuning->growth_factor + * tuning->growth_threshold)); + + if (SIZE_MAX <= candidate) + return NULL; + + /* If the rehash fails, arrange to return NULL. */ + if (!hash_rehash (table, candidate)) + entry = NULL; + } + } + + return (void *) entry; +} + +/* If ENTRY is already in the table, remove it and return the just-deleted + data (the user may want to deallocate its storage). If ENTRY is not in the + table, don't modify the table and return NULL. */ + +void * +hash_delete (Hash_table *table, const void *entry) +{ + void *data; + struct hash_entry *bucket; + + data = hash_find_entry (table, entry, &bucket, true); + if (!data) + return NULL; + + table->n_entries--; + if (!bucket->data) + { + table->n_buckets_used--; + + /* If the shrink threshold of the buckets in use has been reached, + rehash into a smaller table. */ + + if (table->n_buckets_used + < table->tuning->shrink_threshold * table->n_buckets) + { + /* Check more fully, before starting real work. If tuning arguments + became invalid, the second check will rely on proper defaults. */ + check_tuning (table); + if (table->n_buckets_used + < table->tuning->shrink_threshold * table->n_buckets) + { + const Hash_tuning *tuning = table->tuning; + size_t candidate = + (tuning->is_n_buckets + ? table->n_buckets * tuning->shrink_factor + : (table->n_buckets * tuning->shrink_factor + * tuning->growth_threshold)); + + hash_rehash (table, candidate); + } + } + } + + return data; +} + +/* Testing. */ + +#if TESTING + +void +hash_print (const Hash_table *table) +{ + struct hash_entry const *bucket; + + for (bucket = table->bucket; bucket < table->bucket_limit; bucket++) + { + struct hash_entry *cursor; + + if (bucket) + printf ("%lu:\n", (unsigned long int) (bucket - table->bucket)); + + for (cursor = bucket; cursor; cursor = cursor->next) + { + char const *s = cursor->data; + /* FIXME */ + if (s) + printf (" %s\n", s); + } + } +} + +#endif /* TESTING */ diff --git a/lib/hash.h b/lib/hash.h new file mode 100644 index 0000000..ab63a86 --- /dev/null +++ b/lib/hash.h @@ -0,0 +1,88 @@ +/* hash - hashing table processing. + Copyright (C) 1998, 1999, 2001, 2003 Free Software Foundation, Inc. + Written by Jim Meyering <meyering@ascend.com>, 1998. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* A generic hash table package. */ + +/* Make sure USE_OBSTACK is defined to 1 if you want the allocator to use + obstacks instead of malloc, and recompile `hash.c' with same setting. */ + +#ifndef HASH_H_ +# define HASH_H_ + +# include <stdio.h> +# include <stdbool.h> + +typedef size_t (*Hash_hasher) (const void *, size_t); +typedef bool (*Hash_comparator) (const void *, const void *); +typedef void (*Hash_data_freer) (void *); +typedef bool (*Hash_processor) (void *, void *); + +struct hash_entry + { + void *data; + struct hash_entry *next; + }; + +struct hash_tuning + { + /* This structure is mainly used for `hash_initialize', see the block + documentation of `hash_reset_tuning' for more complete comments. */ + + float shrink_threshold; /* ratio of used buckets to trigger a shrink */ + float shrink_factor; /* ratio of new smaller size to original size */ + float growth_threshold; /* ratio of used buckets to trigger a growth */ + float growth_factor; /* ratio of new bigger size to original size */ + bool is_n_buckets; /* if CANDIDATE really means table size */ + }; + +typedef struct hash_tuning Hash_tuning; + +struct hash_table; + +typedef struct hash_table Hash_table; + +/* Information and lookup. */ +size_t hash_get_n_buckets (const Hash_table *); +size_t hash_get_n_buckets_used (const Hash_table *); +size_t hash_get_n_entries (const Hash_table *); +size_t hash_get_max_bucket_length (const Hash_table *); +bool hash_table_ok (const Hash_table *); +void hash_print_statistics (const Hash_table *, FILE *); +void *hash_lookup (const Hash_table *, const void *); + +/* Walking. */ +void *hash_get_first (const Hash_table *); +void *hash_get_next (const Hash_table *, const void *); +size_t hash_get_entries (const Hash_table *, void **, size_t); +size_t hash_do_for_each (const Hash_table *, Hash_processor, void *); + +/* Allocation and clean-up. */ +size_t hash_string (const char *, size_t); +void hash_reset_tuning (Hash_tuning *); +Hash_table *hash_initialize (size_t, const Hash_tuning *, + Hash_hasher, Hash_comparator, + Hash_data_freer); +void hash_clear (Hash_table *); +void hash_free (Hash_table *); + +/* Insertion and deletion. */ +bool hash_rehash (Hash_table *, size_t); +void *hash_insert (Hash_table *, const void *); +void *hash_delete (Hash_table *, const void *); + +#endif diff --git a/lib/human.c b/lib/human.c new file mode 100644 index 0000000..ecf4c97 --- /dev/null +++ b/lib/human.c @@ -0,0 +1,479 @@ +/* human.c -- print human readable file size + + Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, + 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and Larry McVoy. */ + +#include <config.h> + +#include "human.h" + +#include <locale.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +#include <argmatch.h> +#include <error.h> +#include <intprops.h> +#include <xstrtol.h> + +/* The maximum length of a suffix like "KiB". */ +#define HUMAN_READABLE_SUFFIX_LENGTH_MAX 3 + +static const char power_letter[] = +{ + 0, /* not used */ + 'K', /* kibi ('k' for kilo is a special case) */ + 'M', /* mega or mebi */ + 'G', /* giga or gibi */ + 'T', /* tera or tebi */ + 'P', /* peta or pebi */ + 'E', /* exa or exbi */ + 'Z', /* zetta or 2**70 */ + 'Y' /* yotta or 2**80 */ +}; + + +/* If INEXACT_STYLE is not human_round_to_nearest, and if easily + possible, adjust VALUE according to the style. */ + +static long double +adjust_value (int inexact_style, long double value) +{ + /* Do not use the floorl or ceill functions, as that would mean + checking for their presence and possibly linking with the + standard math library, which is a porting pain. So leave the + value alone if it is too large to easily round. */ + if (inexact_style != human_round_to_nearest && value < UINTMAX_MAX) + { + uintmax_t u = value; + value = u + (inexact_style == human_ceiling && u != value); + } + + return value; +} + +/* Group the digits of NUMBER according to the grouping rules of the + current locale. NUMBER contains NUMBERLEN digits. Modify the + bytes pointed to by NUMBER in place, subtracting 1 from NUMBER for + each byte inserted. Return the starting address of the modified + number. + + To group the digits, use GROUPING and THOUSANDS_SEP as in `struct + lconv' from <locale.h>. */ + +static char * +group_number (char *number, size_t numberlen, + char const *grouping, char const *thousands_sep) +{ + register char *d; + size_t grouplen = SIZE_MAX; + size_t thousands_seplen = strlen (thousands_sep); + size_t i = numberlen; + + /* The maximum possible value for NUMBERLEN is the number of digits + in the square of the largest uintmax_t, so double the size needed. */ + char buf[2 * INT_STRLEN_BOUND (uintmax_t) + 1]; + + memcpy (buf, number, numberlen); + d = number + numberlen; + + for (;;) + { + unsigned char g = *grouping; + + if (g) + { + grouplen = g < CHAR_MAX ? g : i; + grouping++; + } + + if (i < grouplen) + grouplen = i; + + d -= grouplen; + i -= grouplen; + memcpy (d, buf + i, grouplen); + if (i == 0) + return d; + + d -= thousands_seplen; + memcpy (d, thousands_sep, thousands_seplen); + } +} + +/* Convert N to a human readable format in BUF, using the options OPTS. + + N is expressed in units of FROM_BLOCK_SIZE. FROM_BLOCK_SIZE must + be nonnegative. + + Use units of TO_BLOCK_SIZE in the output number. TO_BLOCK_SIZE + must be positive. + + Use (OPTS & (human_round_to_nearest | human_floor | human_ceiling)) + to determine whether to take the ceiling or floor of any result + that cannot be expressed exactly. + + If (OPTS & human_group_digits), group the thousands digits + according to the locale, e.g., `1,000,000' in an American English + locale. + + If (OPTS & human_autoscale), deduce the output block size + automatically; TO_BLOCK_SIZE must be 1 but it has no effect on the + output. Use powers of 1024 if (OPTS & human_base_1024), and powers + of 1000 otherwise. For example, assuming powers of 1024, 8500 + would be converted to 8.3, 133456345 to 127, 56990456345 to 53, and + so on. Numbers smaller than the power aren't modified. + human_autoscale is normally used together with human_SI. + + If (OPTS & human_space_before_unit), use a space to separate the + number from any suffix that is appended as described below. + + If (OPTS & human_SI), append an SI prefix indicating which power is + being used. If in addition (OPTS & human_B), append "B" (if base + 1000) or "iB" (if base 1024) to the SI prefix. When ((OPTS & + human_SI) && ! (OPTS & human_autoscale)), TO_BLOCK_SIZE must be a + power of 1024 or of 1000, depending on (OPTS & + human_base_1024). */ + +char * +human_readable (uintmax_t n, char *buf, int opts, + uintmax_t from_block_size, uintmax_t to_block_size) +{ + int inexact_style = + opts & (human_round_to_nearest | human_floor | human_ceiling); + unsigned int base = opts & human_base_1024 ? 1024 : 1000; + uintmax_t amt; + int tenths; + int exponent = -1; + int exponent_max = sizeof power_letter - 1; + char *p; + char *psuffix; + char const *integerlim; + + /* 0 means adjusted N == AMT.TENTHS; + 1 means AMT.TENTHS < adjusted N < AMT.TENTHS + 0.05; + 2 means adjusted N == AMT.TENTHS + 0.05; + 3 means AMT.TENTHS + 0.05 < adjusted N < AMT.TENTHS + 0.1. */ + int rounding; + + char const *decimal_point = "."; + size_t decimal_pointlen = 1; + char const *grouping = ""; + char const *thousands_sep = ""; + struct lconv const *l = localeconv (); + size_t pointlen = strlen (l->decimal_point); + if (0 < pointlen && pointlen <= MB_LEN_MAX) + { + decimal_point = l->decimal_point; + decimal_pointlen = pointlen; + } + grouping = l->grouping; + if (strlen (l->thousands_sep) <= MB_LEN_MAX) + thousands_sep = l->thousands_sep; + + psuffix = buf + LONGEST_HUMAN_READABLE - HUMAN_READABLE_SUFFIX_LENGTH_MAX; + p = psuffix; + + /* Adjust AMT out of FROM_BLOCK_SIZE units and into TO_BLOCK_SIZE + units. If this can be done exactly with integer arithmetic, do + not use floating point operations. */ + if (to_block_size <= from_block_size) + { + if (from_block_size % to_block_size == 0) + { + uintmax_t multiplier = from_block_size / to_block_size; + amt = n * multiplier; + if (amt / multiplier == n) + { + tenths = 0; + rounding = 0; + goto use_integer_arithmetic; + } + } + } + else if (from_block_size != 0 && to_block_size % from_block_size == 0) + { + uintmax_t divisor = to_block_size / from_block_size; + uintmax_t r10 = (n % divisor) * 10; + uintmax_t r2 = (r10 % divisor) * 2; + amt = n / divisor; + tenths = r10 / divisor; + rounding = r2 < divisor ? 0 < r2 : 2 + (divisor < r2); + goto use_integer_arithmetic; + } + + { + /* Either the result cannot be computed easily using uintmax_t, + or from_block_size is zero. Fall back on floating point. + FIXME: This can yield answers that are slightly off. */ + + long double dto_block_size = to_block_size; + long double damt = n * (from_block_size / dto_block_size); + size_t buflen; + size_t nonintegerlen; + + if (! (opts & human_autoscale)) + { + sprintf (buf, "%.0Lf", adjust_value (inexact_style, damt)); + buflen = strlen (buf); + nonintegerlen = 0; + } + else + { + long double e = 1; + exponent = 0; + + do + { + e *= base; + exponent++; + } + while (e * base <= damt && exponent < exponent_max); + + damt /= e; + + sprintf (buf, "%.1Lf", adjust_value (inexact_style, damt)); + buflen = strlen (buf); + nonintegerlen = decimal_pointlen + 1; + + if (1 + nonintegerlen + ! (opts & human_base_1024) < buflen + || ((opts & human_suppress_point_zero) + && buf[buflen - 1] == '0')) + { + sprintf (buf, "%.0Lf", + adjust_value (inexact_style, damt * 10) / 10); + buflen = strlen (buf); + nonintegerlen = 0; + } + } + + p = psuffix - buflen; + memmove (p, buf, buflen); + integerlim = p + buflen - nonintegerlen; + } + goto do_grouping; + + use_integer_arithmetic: + { + /* The computation can be done exactly, with integer arithmetic. + + Use power of BASE notation if requested and if adjusted AMT is + large enough. */ + + if (opts & human_autoscale) + { + exponent = 0; + + if (base <= amt) + { + do + { + unsigned int r10 = (amt % base) * 10 + tenths; + unsigned int r2 = (r10 % base) * 2 + (rounding >> 1); + amt /= base; + tenths = r10 / base; + rounding = (r2 < base + ? (r2 + rounding) != 0 + : 2 + (base < r2 + rounding)); + exponent++; + } + while (base <= amt && exponent < exponent_max); + + if (amt < 10) + { + if (inexact_style == human_round_to_nearest + ? 2 < rounding + (tenths & 1) + : inexact_style == human_ceiling && 0 < rounding) + { + tenths++; + rounding = 0; + + if (tenths == 10) + { + amt++; + tenths = 0; + } + } + + if (amt < 10 + && (tenths || ! (opts & human_suppress_point_zero))) + { + *--p = '0' + tenths; + p -= decimal_pointlen; + memcpy (p, decimal_point, decimal_pointlen); + tenths = rounding = 0; + } + } + } + } + + if (inexact_style == human_round_to_nearest + ? 5 < tenths + (0 < rounding + (amt & 1)) + : inexact_style == human_ceiling && 0 < tenths + rounding) + { + amt++; + + if ((opts & human_autoscale) + && amt == base && exponent < exponent_max) + { + exponent++; + if (! (opts & human_suppress_point_zero)) + { + *--p = '0'; + p -= decimal_pointlen; + memcpy (p, decimal_point, decimal_pointlen); + } + amt = 1; + } + } + + integerlim = p; + + do + { + int digit = amt % 10; + *--p = digit + '0'; + } + while ((amt /= 10) != 0); + } + + do_grouping: + if (opts & human_group_digits) + p = group_number (p, integerlim - p, grouping, thousands_sep); + + if (opts & human_SI) + { + if (exponent < 0) + { + uintmax_t power; + exponent = 0; + for (power = 1; power < to_block_size; power *= base) + if (++exponent == exponent_max) + break; + } + + if ((exponent | (opts & human_B)) && (opts & human_space_before_unit)) + *psuffix++ = ' '; + + if (exponent) + *psuffix++ = (! (opts & human_base_1024) && exponent == 1 + ? 'k' + : power_letter[exponent]); + + if (opts & human_B) + { + if ((opts & human_base_1024) && exponent) + *psuffix++ = 'i'; + *psuffix++ = 'B'; + } + } + + *psuffix = '\0'; + + return p; +} + + +/* The default block size used for output. This number may change in + the future as disks get larger. */ +#ifndef DEFAULT_BLOCK_SIZE +# define DEFAULT_BLOCK_SIZE 1024 +#endif + +static char const *const block_size_args[] = { "human-readable", "si", 0 }; +static int const block_size_opts[] = + { + human_autoscale + human_SI + human_base_1024, + human_autoscale + human_SI + }; + +static uintmax_t +default_block_size (void) +{ + return getenv ("POSIXLY_CORRECT") ? 512 : DEFAULT_BLOCK_SIZE; +} + +static strtol_error +humblock (char const *spec, uintmax_t *block_size, int *options) +{ + int i; + int opts = 0; + + if (! spec + && ! (spec = getenv ("BLOCK_SIZE")) + && ! (spec = getenv ("BLOCKSIZE"))) + *block_size = default_block_size (); + else + { + if (*spec == '\'') + { + opts |= human_group_digits; + spec++; + } + + if (0 <= (i = ARGMATCH (spec, block_size_args, block_size_opts))) + { + opts |= block_size_opts[i]; + *block_size = 1; + } + else + { + char *ptr; + strtol_error e = xstrtoumax (spec, &ptr, 0, block_size, + "eEgGkKmMpPtTyYzZ0"); + if (e != LONGINT_OK) + { + *options = 0; + return e; + } + for (; ! ('0' <= *spec && *spec <= '9'); spec++) + if (spec == ptr) + { + opts |= human_SI; + if (ptr[-1] == 'B') + opts |= human_B; + if (ptr[-1] != 'B' || ptr[-2] == 'i') + opts |= human_base_1024; + break; + } + } + } + + *options = opts; + return LONGINT_OK; +} + +int +human_options (char const *spec, bool report_errors, uintmax_t *block_size) +{ + int opts; + strtol_error e = humblock (spec, block_size, &opts); + if (*block_size == 0) + { + *block_size = default_block_size (); + e = LONGINT_INVALID; + } + if (e != LONGINT_OK && report_errors) + STRTOL_FATAL_ERROR (spec, _("block size"), e); + return opts; +} diff --git a/lib/human.h b/lib/human.h new file mode 100644 index 0000000..44b8b36 --- /dev/null +++ b/lib/human.h @@ -0,0 +1,83 @@ +/* human.h -- print human readable file size + + Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, + 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and Larry McVoy. */ + +#ifndef HUMAN_H_ +# define HUMAN_H_ 1 + +# include <limits.h> +# include <stdbool.h> +# include <stdint.h> +# include <unistd.h> + +/* A conservative bound on the maximum length of a human-readable string. + The output can be the square of the largest uintmax_t, so double + its size before converting to a bound. + log10 (2.0) < 146/485. Add 1 for integer division truncation. + Also, the output can have a thousands separator between every digit, + so multiply by MB_LEN_MAX + 1 and then subtract MB_LEN_MAX. + Append 1 for a space before the suffix. + Finally, append 3, the maximum length of a suffix. */ +# define LONGEST_HUMAN_READABLE \ + ((2 * sizeof (uintmax_t) * CHAR_BIT * 146 / 485 + 1) * (MB_LEN_MAX + 1) \ + - MB_LEN_MAX + 1 + 3) + +/* Options for human_readable. */ +enum +{ + /* Unless otherwise specified these options may be ORed together. */ + + /* The following three options are mutually exclusive. */ + /* Round to plus infinity (default). */ + human_ceiling = 0, + /* Round to nearest, ties to even. */ + human_round_to_nearest = 1, + /* Round to minus infinity. */ + human_floor = 2, + + /* Group digits together, e.g. `1,000,000'. This uses the + locale-defined grouping; the traditional C locale does not group, + so this has effect only if some other locale is in use. */ + human_group_digits = 4, + + /* When autoscaling, suppress ".0" at end. */ + human_suppress_point_zero = 8, + + /* Scale output and use SI-style units, ignoring the output block size. */ + human_autoscale = 16, + + /* Prefer base 1024 to base 1000. */ + human_base_1024 = 32, + + /* Prepend " " before unit symbol. */ + human_space_before_unit = 64, + + /* Append SI prefix, e.g. "k" or "M". */ + human_SI = 128, + + /* Append "B" (if base 1000) or "iB" (if base 1024) to SI prefix. */ + human_B = 256 +}; + +char *human_readable (uintmax_t, char *, int, uintmax_t, uintmax_t); + +int human_options (char const *, bool, uintmax_t *); + +#endif /* HUMAN_H_ */ diff --git a/lib/imaxtostr.c b/lib/imaxtostr.c new file mode 100644 index 0000000..5e87ad5 --- /dev/null +++ b/lib/imaxtostr.c @@ -0,0 +1,3 @@ +#define inttostr imaxtostr +#define inttype intmax_t +#include "inttostr.c" diff --git a/lib/intprops.h b/lib/intprops.h new file mode 100644 index 0000000..34f971c --- /dev/null +++ b/lib/intprops.h @@ -0,0 +1,78 @@ +/* intprops.h -- properties of integer types + + Copyright (C) 2001, 2002, 2003, 2004, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <limits.h> + +/* The extra casts in the following macros work around compiler bugs, + e.g., in Cray C 5.0.3.0. */ + +/* True if the arithmetic type T is an integer type. bool counts as + an integer. */ +#define TYPE_IS_INTEGER(t) ((t) 1.5 == 1) + +/* True if negative values of the signed integer type T use two's + complement, ones' complement, or signed magnitude representation, + respectively. Much GNU code assumes two's complement, but some + people like to be portable to all possible C hosts. */ +#define TYPE_TWOS_COMPLEMENT(t) ((t) ~ (t) 0 == (t) -1) +#define TYPE_ONES_COMPLEMENT(t) ((t) ~ (t) 0 == 0) +#define TYPE_SIGNED_MAGNITUDE(t) ((t) ~ (t) 0 < (t) -1) + +/* True if the arithmetic type T is signed. */ +#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) + +/* The maximum and minimum values for the integer type T. These + macros have undefined behavior if T is signed and has padding bits. + If this is a problem for you, please let us know how to fix it for + your host. */ +#define TYPE_MINIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) 0 \ + : TYPE_SIGNED_MAGNITUDE (t) \ + ? ~ (t) 0 \ + : ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1))) +#define TYPE_MAXIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) -1 \ + : ~ (~ (t) 0 << (sizeof (t) * CHAR_BIT - 1)))) + +/* Return zero if T can be determined to be an unsigned type. + Otherwise, return 1. + When compiling with GCC, INT_STRLEN_BOUND uses this macro to obtain a + tighter bound. Otherwise, it overestimates the true bound by one byte + when applied to unsigned types of size 2, 4, 16, ... bytes. + The symbol signed_type_or_expr__ is private to this header file. */ +#if __GNUC__ >= 2 +# define signed_type_or_expr__(t) TYPE_SIGNED (__typeof__ (t)) +#else +# define signed_type_or_expr__(t) 1 +#endif + +/* Bound on length of the string representing an integer type or expression T. + Subtract 1 for the sign bit if T is signed; log10 (2.0) < 146/485; + add 1 for integer division truncation; add 1 more for a minus sign + if needed. */ +#define INT_STRLEN_BOUND(t) \ + ((sizeof (t) * CHAR_BIT - signed_type_or_expr__ (t)) * 146 / 485 \ + + signed_type_or_expr__ (t) + 1) + +/* Bound on buffer size needed to represent an integer type or expression T, + including the terminating null. */ +#define INT_BUFSIZE_BOUND(t) (INT_STRLEN_BOUND (t) + 1) diff --git a/lib/inttostr.c b/lib/inttostr.c new file mode 100644 index 0000000..246d658 --- /dev/null +++ b/lib/inttostr.c @@ -0,0 +1,51 @@ +/* inttostr.c -- convert integers to printable strings + + Copyright (C) 2001, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert */ + +#include <config.h> + +#include "inttostr.h" + +/* Convert I to a printable string in BUF, which must be at least + INT_BUFSIZE_BOUND (INTTYPE) bytes long. Return the address of the + printable string, which need not start at BUF. */ + +char * +inttostr (inttype i, char *buf) +{ + char *p = buf + INT_STRLEN_BOUND (inttype); + *p = 0; + + if (i < 0) + { + do + *--p = '0' - i % 10; + while ((i /= 10) != 0); + + *--p = '-'; + } + else + { + do + *--p = '0' + i % 10; + while ((i /= 10) != 0); + } + + return p; +} diff --git a/lib/inttostr.h b/lib/inttostr.h new file mode 100644 index 0000000..31258ca --- /dev/null +++ b/lib/inttostr.h @@ -0,0 +1,30 @@ +/* inttostr.h -- convert integers to printable strings + + Copyright (C) 2001, 2002, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert */ + +#include <stdint.h> +#include <sys/types.h> + +#include "intprops.h" + +char *offtostr (off_t, char *); +char *imaxtostr (intmax_t, char *); +char *umaxtostr (uintmax_t, char *); +char *uinttostr (unsigned int, char *); diff --git a/lib/inttypes_.h b/lib/inttypes_.h new file mode 100644 index 0000000..6d78325 --- /dev/null +++ b/lib/inttypes_.h @@ -0,0 +1,1100 @@ +/* Copyright (C) 2006-2007 Free Software Foundation, Inc. + Written by Paul Eggert, Bruno Haible, Derek Price. + This file is part of gnulib. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2.1, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU Lesser General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* + * ISO C 99 <inttypes.h> for platforms that lack it. + * <http://www.opengroup.org/susv3xbd/inttypes.h.html> + */ + +/* Include the original <inttypes.h> if it exists, and if this file + has not been included yet or if this file includes gnulib stdint.h + which in turn includes this file. + The include_next requires a split double-inclusion guard. */ +#if ! defined INTTYPES_H || defined _GL_JUST_INCLUDE_ABSOLUTE_INTTYPES_H +# if @HAVE_INTTYPES_H@ +# if @HAVE_INCLUDE_NEXT@ +# include_next <inttypes.h> +# else +# include @ABSOLUTE_INTTYPES_H@ +# endif +# endif +#endif + +#if ! defined INTTYPES_H && ! defined _GL_JUST_INCLUDE_ABSOLUTE_INTTYPES_H +#define INTTYPES_H + +/* Include <stdint.h> or the gnulib replacement. */ +#include <stdint.h> +/* Get CHAR_BIT. */ +#include <limits.h> + +#if !(INT_MIN == INT32_MIN && INT_MAX == INT32_MAX) +# error "This file assumes that 'int' has exactly 32 bits. Please report your platform and compiler to <bug-gnulib@gnu.org>." +#endif + +/* The definition of GL_LINK_WARNING is copied here. */ + +/* 7.8.1 Macros for format specifiers */ + +#if ! defined __cplusplus || defined __STDC_FORMAT_MACROS + +# if defined _TNS_R_TARGET + /* Tandem NonStop R series and compatible platforms released before + July 2005 support %Ld but not %lld. */ +# define _LONG_LONG_FORMAT_PREFIX "L" +# else +# define _LONG_LONG_FORMAT_PREFIX "ll" +# endif + +# if !defined PRId8 || @PRI_MACROS_BROKEN@ +# undef PRId8 +# ifdef INT8_MAX +# define PRId8 "d" +# endif +# endif +# if !defined PRIi8 || @PRI_MACROS_BROKEN@ +# undef PRIi8 +# ifdef INT8_MAX +# define PRIi8 "i" +# endif +# endif +# if !defined PRIo8 || @PRI_MACROS_BROKEN@ +# undef PRIo8 +# ifdef UINT8_MAX +# define PRIo8 "o" +# endif +# endif +# if !defined PRIu8 || @PRI_MACROS_BROKEN@ +# undef PRIu8 +# ifdef UINT8_MAX +# define PRIu8 "u" +# endif +# endif +# if !defined PRIx8 || @PRI_MACROS_BROKEN@ +# undef PRIx8 +# ifdef UINT8_MAX +# define PRIx8 "x" +# endif +# endif +# if !defined PRIX8 || @PRI_MACROS_BROKEN@ +# undef PRIX8 +# ifdef UINT8_MAX +# define PRIX8 "X" +# endif +# endif +# if !defined PRId16 || @PRI_MACROS_BROKEN@ +# undef PRId16 +# ifdef INT16_MAX +# define PRId16 "d" +# endif +# endif +# if !defined PRIi16 || @PRI_MACROS_BROKEN@ +# undef PRIi16 +# ifdef INT16_MAX +# define PRIi16 "i" +# endif +# endif +# if !defined PRIo16 || @PRI_MACROS_BROKEN@ +# undef PRIo16 +# ifdef UINT16_MAX +# define PRIo16 "o" +# endif +# endif +# if !defined PRIu16 || @PRI_MACROS_BROKEN@ +# undef PRIu16 +# ifdef UINT16_MAX +# define PRIu16 "u" +# endif +# endif +# if !defined PRIx16 || @PRI_MACROS_BROKEN@ +# undef PRIx16 +# ifdef UINT16_MAX +# define PRIx16 "x" +# endif +# endif +# if !defined PRIX16 || @PRI_MACROS_BROKEN@ +# undef PRIX16 +# ifdef UINT16_MAX +# define PRIX16 "X" +# endif +# endif +# if !defined PRId32 || @PRI_MACROS_BROKEN@ +# undef PRId32 +# ifdef INT32_MAX +# define PRId32 "d" +# endif +# endif +# if !defined PRIi32 || @PRI_MACROS_BROKEN@ +# undef PRIi32 +# ifdef INT32_MAX +# define PRIi32 "i" +# endif +# endif +# if !defined PRIo32 || @PRI_MACROS_BROKEN@ +# undef PRIo32 +# ifdef UINT32_MAX +# define PRIo32 "o" +# endif +# endif +# if !defined PRIu32 || @PRI_MACROS_BROKEN@ +# undef PRIu32 +# ifdef UINT32_MAX +# define PRIu32 "u" +# endif +# endif +# if !defined PRIx32 || @PRI_MACROS_BROKEN@ +# undef PRIx32 +# ifdef UINT32_MAX +# define PRIx32 "x" +# endif +# endif +# if !defined PRIX32 || @PRI_MACROS_BROKEN@ +# undef PRIX32 +# ifdef UINT32_MAX +# define PRIX32 "X" +# endif +# endif +# ifdef INT64_MAX +# if INT64_MAX == LONG_MAX +# define _PRI64_PREFIX "l" +# elif defined _MSC_VER || defined __MINGW32__ +# define _PRI64_PREFIX "I64" +# elif @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define _PRI64_PREFIX _LONG_LONG_FORMAT_PREFIX +# endif +# if !defined PRId64 || @PRI_MACROS_BROKEN@ +# undef PRId64 +# define PRId64 _PRI64_PREFIX "d" +# endif +# if !defined PRIi64 || @PRI_MACROS_BROKEN@ +# undef PRIi64 +# define PRIi64 _PRI64_PREFIX "i" +# endif +# endif +# ifdef UINT64_MAX +# if UINT64_MAX == ULONG_MAX +# define _PRIu64_PREFIX "l" +# elif defined _MSC_VER || defined __MINGW32__ +# define _PRIu64_PREFIX "I64" +# elif @HAVE_UNSIGNED_LONG_LONG_INT@ && ULONG_MAX >> 31 == 1 +# define _PRIu64_PREFIX _LONG_LONG_FORMAT_PREFIX +# endif +# if !defined PRIo64 || @PRI_MACROS_BROKEN@ +# undef PRIo64 +# define PRIo64 _PRIu64_PREFIX "o" +# endif +# if !defined PRIu64 || @PRI_MACROS_BROKEN@ +# undef PRIu64 +# define PRIu64 _PRIu64_PREFIX "u" +# endif +# if !defined PRIx64 || @PRI_MACROS_BROKEN@ +# undef PRIx64 +# define PRIx64 _PRIu64_PREFIX "x" +# endif +# if !defined PRIX64 || @PRI_MACROS_BROKEN@ +# undef PRIX64 +# define PRIX64 _PRIu64_PREFIX "X" +# endif +# endif + +# if !defined PRIdLEAST8 || @PRI_MACROS_BROKEN@ +# undef PRIdLEAST8 +# define PRIdLEAST8 "d" +# endif +# if !defined PRIiLEAST8 || @PRI_MACROS_BROKEN@ +# undef PRIiLEAST8 +# define PRIiLEAST8 "i" +# endif +# if !defined PRIoLEAST8 || @PRI_MACROS_BROKEN@ +# undef PRIoLEAST8 +# define PRIoLEAST8 "o" +# endif +# if !defined PRIuLEAST8 || @PRI_MACROS_BROKEN@ +# undef PRIuLEAST8 +# define PRIuLEAST8 "u" +# endif +# if !defined PRIxLEAST8 || @PRI_MACROS_BROKEN@ +# undef PRIxLEAST8 +# define PRIxLEAST8 "x" +# endif +# if !defined PRIXLEAST8 || @PRI_MACROS_BROKEN@ +# undef PRIXLEAST8 +# define PRIXLEAST8 "X" +# endif +# if !defined PRIdLEAST16 || @PRI_MACROS_BROKEN@ +# undef PRIdLEAST16 +# define PRIdLEAST16 "d" +# endif +# if !defined PRIiLEAST16 || @PRI_MACROS_BROKEN@ +# undef PRIiLEAST16 +# define PRIiLEAST16 "i" +# endif +# if !defined PRIoLEAST16 || @PRI_MACROS_BROKEN@ +# undef PRIoLEAST16 +# define PRIoLEAST16 "o" +# endif +# if !defined PRIuLEAST16 || @PRI_MACROS_BROKEN@ +# undef PRIuLEAST16 +# define PRIuLEAST16 "u" +# endif +# if !defined PRIxLEAST16 || @PRI_MACROS_BROKEN@ +# undef PRIxLEAST16 +# define PRIxLEAST16 "x" +# endif +# if !defined PRIXLEAST16 || @PRI_MACROS_BROKEN@ +# undef PRIXLEAST16 +# define PRIXLEAST16 "X" +# endif +# if !defined PRIdLEAST32 || @PRI_MACROS_BROKEN@ +# undef PRIdLEAST32 +# define PRIdLEAST32 "d" +# endif +# if !defined PRIiLEAST32 || @PRI_MACROS_BROKEN@ +# undef PRIiLEAST32 +# define PRIiLEAST32 "i" +# endif +# if !defined PRIoLEAST32 || @PRI_MACROS_BROKEN@ +# undef PRIoLEAST32 +# define PRIoLEAST32 "o" +# endif +# if !defined PRIuLEAST32 || @PRI_MACROS_BROKEN@ +# undef PRIuLEAST32 +# define PRIuLEAST32 "u" +# endif +# if !defined PRIxLEAST32 || @PRI_MACROS_BROKEN@ +# undef PRIxLEAST32 +# define PRIxLEAST32 "x" +# endif +# if !defined PRIXLEAST32 || @PRI_MACROS_BROKEN@ +# undef PRIXLEAST32 +# define PRIXLEAST32 "X" +# endif +# ifdef INT64_MAX +# if !defined PRIdLEAST64 || @PRI_MACROS_BROKEN@ +# undef PRIdLEAST64 +# define PRIdLEAST64 PRId64 +# endif +# if !defined PRIiLEAST64 || @PRI_MACROS_BROKEN@ +# undef PRIiLEAST64 +# define PRIiLEAST64 PRIi64 +# endif +# endif +# ifdef UINT64_MAX +# if !defined PRIoLEAST64 || @PRI_MACROS_BROKEN@ +# undef PRIoLEAST64 +# define PRIoLEAST64 PRIo64 +# endif +# if !defined PRIuLEAST64 || @PRI_MACROS_BROKEN@ +# undef PRIuLEAST64 +# define PRIuLEAST64 PRIu64 +# endif +# if !defined PRIxLEAST64 || @PRI_MACROS_BROKEN@ +# undef PRIxLEAST64 +# define PRIxLEAST64 PRIx64 +# endif +# if !defined PRIXLEAST64 || @PRI_MACROS_BROKEN@ +# undef PRIXLEAST64 +# define PRIXLEAST64 PRIX64 +# endif +# endif + +# if !defined PRIdFAST8 || @PRI_MACROS_BROKEN@ +# undef PRIdFAST8 +# if INT_FAST8_MAX > INT32_MAX +# define PRIdFAST8 PRId64 +# else +# define PRIdFAST8 "d" +# endif +# endif +# if !defined PRIiFAST8 || @PRI_MACROS_BROKEN@ +# undef PRIiFAST8 +# if INT_FAST8_MAX > INT32_MAX +# define PRIiFAST8 PRIi64 +# else +# define PRIiFAST8 "i" +# endif +# endif +# if !defined PRIoFAST8 || @PRI_MACROS_BROKEN@ +# undef PRIoFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define PRIoFAST8 PRIo64 +# else +# define PRIoFAST8 "o" +# endif +# endif +# if !defined PRIuFAST8 || @PRI_MACROS_BROKEN@ +# undef PRIuFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define PRIuFAST8 PRIu64 +# else +# define PRIuFAST8 "u" +# endif +# endif +# if !defined PRIxFAST8 || @PRI_MACROS_BROKEN@ +# undef PRIxFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define PRIxFAST8 PRIx64 +# else +# define PRIxFAST8 "x" +# endif +# endif +# if !defined PRIXFAST8 || @PRI_MACROS_BROKEN@ +# undef PRIXFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define PRIXFAST8 PRIX64 +# else +# define PRIXFAST8 "X" +# endif +# endif +# if !defined PRIdFAST16 || @PRI_MACROS_BROKEN@ +# undef PRIdFAST16 +# if INT_FAST16_MAX > INT32_MAX +# define PRIdFAST16 PRId64 +# else +# define PRIdFAST16 "d" +# endif +# endif +# if !defined PRIiFAST16 || @PRI_MACROS_BROKEN@ +# undef PRIiFAST16 +# if INT_FAST16_MAX > INT32_MAX +# define PRIiFAST16 PRIi64 +# else +# define PRIiFAST16 "i" +# endif +# endif +# if !defined PRIoFAST16 || @PRI_MACROS_BROKEN@ +# undef PRIoFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define PRIoFAST16 PRIo64 +# else +# define PRIoFAST16 "o" +# endif +# endif +# if !defined PRIuFAST16 || @PRI_MACROS_BROKEN@ +# undef PRIuFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define PRIuFAST16 PRIu64 +# else +# define PRIuFAST16 "u" +# endif +# endif +# if !defined PRIxFAST16 || @PRI_MACROS_BROKEN@ +# undef PRIxFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define PRIxFAST16 PRIx64 +# else +# define PRIxFAST16 "x" +# endif +# endif +# if !defined PRIXFAST16 || @PRI_MACROS_BROKEN@ +# undef PRIXFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define PRIXFAST16 PRIX64 +# else +# define PRIXFAST16 "X" +# endif +# endif +# if !defined PRIdFAST32 || @PRI_MACROS_BROKEN@ +# undef PRIdFAST32 +# if INT_FAST32_MAX > INT32_MAX +# define PRIdFAST32 PRId64 +# else +# define PRIdFAST32 "d" +# endif +# endif +# if !defined PRIiFAST32 || @PRI_MACROS_BROKEN@ +# undef PRIiFAST32 +# if INT_FAST32_MAX > INT32_MAX +# define PRIiFAST32 PRIi64 +# else +# define PRIiFAST32 "i" +# endif +# endif +# if !defined PRIoFAST32 || @PRI_MACROS_BROKEN@ +# undef PRIoFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define PRIoFAST32 PRIo64 +# else +# define PRIoFAST32 "o" +# endif +# endif +# if !defined PRIuFAST32 || @PRI_MACROS_BROKEN@ +# undef PRIuFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define PRIuFAST32 PRIu64 +# else +# define PRIuFAST32 "u" +# endif +# endif +# if !defined PRIxFAST32 || @PRI_MACROS_BROKEN@ +# undef PRIxFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define PRIxFAST32 PRIx64 +# else +# define PRIxFAST32 "x" +# endif +# endif +# if !defined PRIXFAST32 || @PRI_MACROS_BROKEN@ +# undef PRIXFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define PRIXFAST32 PRIX64 +# else +# define PRIXFAST32 "X" +# endif +# endif +# ifdef INT64_MAX +# if !defined PRIdFAST64 || @PRI_MACROS_BROKEN@ +# undef PRIdFAST64 +# define PRIdFAST64 PRId64 +# endif +# if !defined PRIiFAST64 || @PRI_MACROS_BROKEN@ +# undef PRIiFAST64 +# define PRIiFAST64 PRIi64 +# endif +# endif +# ifdef UINT64_MAX +# if !defined PRIoFAST64 || @PRI_MACROS_BROKEN@ +# undef PRIoFAST64 +# define PRIoFAST64 PRIo64 +# endif +# if !defined PRIuFAST64 || @PRI_MACROS_BROKEN@ +# undef PRIuFAST64 +# define PRIuFAST64 PRIu64 +# endif +# if !defined PRIxFAST64 || @PRI_MACROS_BROKEN@ +# undef PRIxFAST64 +# define PRIxFAST64 PRIx64 +# endif +# if !defined PRIXFAST64 || @PRI_MACROS_BROKEN@ +# undef PRIXFAST64 +# define PRIXFAST64 PRIX64 +# endif +# endif + +# if !defined PRIdMAX || @PRI_MACROS_BROKEN@ +# undef PRIdMAX +# if INTMAX_MAX > INT32_MAX +# define PRIdMAX PRId64 +# else +# define PRIdMAX "ld" +# endif +# endif +# if !defined PRIiMAX || @PRI_MACROS_BROKEN@ +# undef PRIiMAX +# if INTMAX_MAX > INT32_MAX +# define PRIiMAX PRIi64 +# else +# define PRIiMAX "li" +# endif +# endif +# if !defined PRIoMAX || @PRI_MACROS_BROKEN@ +# undef PRIoMAX +# if UINTMAX_MAX > UINT32_MAX +# define PRIoMAX PRIo64 +# else +# define PRIoMAX "lo" +# endif +# endif +# if !defined PRIuMAX || @PRI_MACROS_BROKEN@ +# undef PRIuMAX +# if UINTMAX_MAX > UINT32_MAX +# define PRIuMAX PRIu64 +# else +# define PRIuMAX "lu" +# endif +# endif +# if !defined PRIxMAX || @PRI_MACROS_BROKEN@ +# undef PRIxMAX +# if UINTMAX_MAX > UINT32_MAX +# define PRIxMAX PRIx64 +# else +# define PRIxMAX "lx" +# endif +# endif +# if !defined PRIXMAX || @PRI_MACROS_BROKEN@ +# undef PRIXMAX +# if UINTMAX_MAX > UINT32_MAX +# define PRIXMAX PRIX64 +# else +# define PRIXMAX "lX" +# endif +# endif + +# if !defined PRIdPTR || @PRI_MACROS_BROKEN@ +# undef PRIdPTR +# ifdef INTPTR_MAX +# define PRIdPTR @PRIPTR_PREFIX@ "d" +# endif +# endif +# if !defined PRIiPTR || @PRI_MACROS_BROKEN@ +# undef PRIiPTR +# ifdef INTPTR_MAX +# define PRIiPTR @PRIPTR_PREFIX@ "i" +# endif +# endif +# if !defined PRIoPTR || @PRI_MACROS_BROKEN@ +# undef PRIoPTR +# ifdef UINTPTR_MAX +# define PRIoPTR @PRIPTR_PREFIX@ "o" +# endif +# endif +# if !defined PRIuPTR || @PRI_MACROS_BROKEN@ +# undef PRIuPTR +# ifdef UINTPTR_MAX +# define PRIuPTR @PRIPTR_PREFIX@ "u" +# endif +# endif +# if !defined PRIxPTR || @PRI_MACROS_BROKEN@ +# undef PRIxPTR +# ifdef UINTPTR_MAX +# define PRIxPTR @PRIPTR_PREFIX@ "x" +# endif +# endif +# if !defined PRIXPTR || @PRI_MACROS_BROKEN@ +# undef PRIXPTR +# ifdef UINTPTR_MAX +# define PRIXPTR @PRIPTR_PREFIX@ "X" +# endif +# endif + +# if !defined SCNd8 || @PRI_MACROS_BROKEN@ +# undef SCNd8 +# ifdef INT8_MAX +# define SCNd8 "hhd" +# endif +# endif +# if !defined SCNi8 || @PRI_MACROS_BROKEN@ +# undef SCNi8 +# ifdef INT8_MAX +# define SCNi8 "hhi" +# endif +# endif +# if !defined SCNo8 || @PRI_MACROS_BROKEN@ +# undef SCNo8 +# ifdef UINT8_MAX +# define SCNo8 "hho" +# endif +# endif +# if !defined SCNu8 || @PRI_MACROS_BROKEN@ +# undef SCNu8 +# ifdef UINT8_MAX +# define SCNu8 "hhu" +# endif +# endif +# if !defined SCNx8 || @PRI_MACROS_BROKEN@ +# undef SCNx8 +# ifdef UINT8_MAX +# define SCNx8 "hhx" +# endif +# endif +# if !defined SCNd16 || @PRI_MACROS_BROKEN@ +# undef SCNd16 +# ifdef INT16_MAX +# define SCNd16 "hd" +# endif +# endif +# if !defined SCNi16 || @PRI_MACROS_BROKEN@ +# undef SCNi16 +# ifdef INT16_MAX +# define SCNi16 "hi" +# endif +# endif +# if !defined SCNo16 || @PRI_MACROS_BROKEN@ +# undef SCNo16 +# ifdef UINT16_MAX +# define SCNo16 "ho" +# endif +# endif +# if !defined SCNu16 || @PRI_MACROS_BROKEN@ +# undef SCNu16 +# ifdef UINT16_MAX +# define SCNu16 "hu" +# endif +# endif +# if !defined SCNx16 || @PRI_MACROS_BROKEN@ +# undef SCNx16 +# ifdef UINT16_MAX +# define SCNx16 "hx" +# endif +# endif +# if !defined SCNd32 || @PRI_MACROS_BROKEN@ +# undef SCNd32 +# ifdef INT32_MAX +# define SCNd32 "d" +# endif +# endif +# if !defined SCNi32 || @PRI_MACROS_BROKEN@ +# undef SCNi32 +# ifdef INT32_MAX +# define SCNi32 "i" +# endif +# endif +# if !defined SCNo32 || @PRI_MACROS_BROKEN@ +# undef SCNo32 +# ifdef UINT32_MAX +# define SCNo32 "o" +# endif +# endif +# if !defined SCNu32 || @PRI_MACROS_BROKEN@ +# undef SCNu32 +# ifdef UINT32_MAX +# define SCNu32 "u" +# endif +# endif +# if !defined SCNx32 || @PRI_MACROS_BROKEN@ +# undef SCNx32 +# ifdef UINT32_MAX +# define SCNx32 "x" +# endif +# endif +# ifdef INT64_MAX +# if INT64_MAX == LONG_MAX +# define _SCN64_PREFIX "l" +# elif defined _MSC_VER || defined __MINGW32__ +# define _SCN64_PREFIX "I64" +# elif @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define _SCN64_PREFIX _LONG_LONG_FORMAT_PREFIX +# endif +# if !defined SCNd64 || @PRI_MACROS_BROKEN@ +# undef SCNd64 +# define SCNd64 _SCN64_PREFIX "d" +# endif +# if !defined SCNi64 || @PRI_MACROS_BROKEN@ +# undef SCNi64 +# define SCNi64 _SCN64_PREFIX "i" +# endif +# endif +# ifdef UINT64_MAX +# if UINT64_MAX == ULONG_MAX +# define _SCNu64_PREFIX "l" +# elif defined _MSC_VER || defined __MINGW32__ +# define _SCNu64_PREFIX "I64" +# elif @HAVE_UNSIGNED_LONG_LONG_INT@ && ULONG_MAX >> 31 == 1 +# define _SCNu64_PREFIX _LONG_LONG_FORMAT_PREFIX +# endif +# if !defined SCNo64 || @PRI_MACROS_BROKEN@ +# undef SCNo64 +# define SCNo64 _SCNu64_PREFIX "o" +# endif +# if !defined SCNu64 || @PRI_MACROS_BROKEN@ +# undef SCNu64 +# define SCNu64 _SCNu64_PREFIX "u" +# endif +# if !defined SCNx64 || @PRI_MACROS_BROKEN@ +# undef SCNx64 +# define SCNx64 _SCNu64_PREFIX "x" +# endif +# endif + +# if !defined SCNdLEAST8 || @PRI_MACROS_BROKEN@ +# undef SCNdLEAST8 +# define SCNdLEAST8 "hhd" +# endif +# if !defined SCNiLEAST8 || @PRI_MACROS_BROKEN@ +# undef SCNiLEAST8 +# define SCNiLEAST8 "hhi" +# endif +# if !defined SCNoLEAST8 || @PRI_MACROS_BROKEN@ +# undef SCNoLEAST8 +# define SCNoLEAST8 "hho" +# endif +# if !defined SCNuLEAST8 || @PRI_MACROS_BROKEN@ +# undef SCNuLEAST8 +# define SCNuLEAST8 "hhu" +# endif +# if !defined SCNxLEAST8 || @PRI_MACROS_BROKEN@ +# undef SCNxLEAST8 +# define SCNxLEAST8 "hhx" +# endif +# if !defined SCNdLEAST16 || @PRI_MACROS_BROKEN@ +# undef SCNdLEAST16 +# define SCNdLEAST16 "hd" +# endif +# if !defined SCNiLEAST16 || @PRI_MACROS_BROKEN@ +# undef SCNiLEAST16 +# define SCNiLEAST16 "hi" +# endif +# if !defined SCNoLEAST16 || @PRI_MACROS_BROKEN@ +# undef SCNoLEAST16 +# define SCNoLEAST16 "ho" +# endif +# if !defined SCNuLEAST16 || @PRI_MACROS_BROKEN@ +# undef SCNuLEAST16 +# define SCNuLEAST16 "hu" +# endif +# if !defined SCNxLEAST16 || @PRI_MACROS_BROKEN@ +# undef SCNxLEAST16 +# define SCNxLEAST16 "hx" +# endif +# if !defined SCNdLEAST32 || @PRI_MACROS_BROKEN@ +# undef SCNdLEAST32 +# define SCNdLEAST32 "d" +# endif +# if !defined SCNiLEAST32 || @PRI_MACROS_BROKEN@ +# undef SCNiLEAST32 +# define SCNiLEAST32 "i" +# endif +# if !defined SCNoLEAST32 || @PRI_MACROS_BROKEN@ +# undef SCNoLEAST32 +# define SCNoLEAST32 "o" +# endif +# if !defined SCNuLEAST32 || @PRI_MACROS_BROKEN@ +# undef SCNuLEAST32 +# define SCNuLEAST32 "u" +# endif +# if !defined SCNxLEAST32 || @PRI_MACROS_BROKEN@ +# undef SCNxLEAST32 +# define SCNxLEAST32 "x" +# endif +# ifdef INT64_MAX +# if !defined SCNdLEAST64 || @PRI_MACROS_BROKEN@ +# undef SCNdLEAST64 +# define SCNdLEAST64 SCNd64 +# endif +# if !defined SCNiLEAST64 || @PRI_MACROS_BROKEN@ +# undef SCNiLEAST64 +# define SCNiLEAST64 SCNi64 +# endif +# endif +# ifdef UINT64_MAX +# if !defined SCNoLEAST64 || @PRI_MACROS_BROKEN@ +# undef SCNoLEAST64 +# define SCNoLEAST64 SCNo64 +# endif +# if !defined SCNuLEAST64 || @PRI_MACROS_BROKEN@ +# undef SCNuLEAST64 +# define SCNuLEAST64 SCNu64 +# endif +# if !defined SCNxLEAST64 || @PRI_MACROS_BROKEN@ +# undef SCNxLEAST64 +# define SCNxLEAST64 SCNx64 +# endif +# endif + +# if !defined SCNdFAST8 || @PRI_MACROS_BROKEN@ +# undef SCNdFAST8 +# if INT_FAST8_MAX > INT32_MAX +# define SCNdFAST8 SCNd64 +# elif INT_FAST8_MAX == 0x7fff +# define SCNdFAST8 "hd" +# elif INT_FAST8_MAX == 0x7f +# define SCNdFAST8 "hhd" +# else +# define SCNdFAST8 "d" +# endif +# endif +# if !defined SCNiFAST8 || @PRI_MACROS_BROKEN@ +# undef SCNiFAST8 +# if INT_FAST8_MAX > INT32_MAX +# define SCNiFAST8 SCNi64 +# elif INT_FAST8_MAX == 0x7fff +# define SCNiFAST8 "hi" +# elif INT_FAST8_MAX == 0x7f +# define SCNiFAST8 "hhi" +# else +# define SCNiFAST8 "i" +# endif +# endif +# if !defined SCNoFAST8 || @PRI_MACROS_BROKEN@ +# undef SCNoFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define SCNoFAST8 SCNo64 +# elif UINT_FAST8_MAX == 0xffff +# define SCNoFAST8 "ho" +# elif UINT_FAST8_MAX == 0xff +# define SCNoFAST8 "hho" +# else +# define SCNoFAST8 "o" +# endif +# endif +# if !defined SCNuFAST8 || @PRI_MACROS_BROKEN@ +# undef SCNuFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define SCNuFAST8 SCNu64 +# elif UINT_FAST8_MAX == 0xffff +# define SCNuFAST8 "hu" +# elif UINT_FAST8_MAX == 0xff +# define SCNuFAST8 "hhu" +# else +# define SCNuFAST8 "u" +# endif +# endif +# if !defined SCNxFAST8 || @PRI_MACROS_BROKEN@ +# undef SCNxFAST8 +# if UINT_FAST8_MAX > UINT32_MAX +# define SCNxFAST8 SCNx64 +# elif UINT_FAST8_MAX == 0xffff +# define SCNxFAST8 "hx" +# elif UINT_FAST8_MAX == 0xff +# define SCNxFAST8 "hhx" +# else +# define SCNxFAST8 "x" +# endif +# endif +# if !defined SCNdFAST16 || @PRI_MACROS_BROKEN@ +# undef SCNdFAST16 +# if INT_FAST16_MAX > INT32_MAX +# define SCNdFAST16 SCNd64 +# elif INT_FAST16_MAX == 0x7fff +# define SCNdFAST16 "hd" +# else +# define SCNdFAST16 "d" +# endif +# endif +# if !defined SCNiFAST16 || @PRI_MACROS_BROKEN@ +# undef SCNiFAST16 +# if INT_FAST16_MAX > INT32_MAX +# define SCNiFAST16 SCNi64 +# elif INT_FAST16_MAX == 0x7fff +# define SCNiFAST16 "hi" +# else +# define SCNiFAST16 "i" +# endif +# endif +# if !defined SCNoFAST16 || @PRI_MACROS_BROKEN@ +# undef SCNoFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define SCNoFAST16 SCNo64 +# elif UINT_FAST16_MAX == 0xffff +# define SCNoFAST16 "ho" +# else +# define SCNoFAST16 "o" +# endif +# endif +# if !defined SCNuFAST16 || @PRI_MACROS_BROKEN@ +# undef SCNuFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define SCNuFAST16 SCNu64 +# elif UINT_FAST16_MAX == 0xffff +# define SCNuFAST16 "hu" +# else +# define SCNuFAST16 "u" +# endif +# endif +# if !defined SCNxFAST16 || @PRI_MACROS_BROKEN@ +# undef SCNxFAST16 +# if UINT_FAST16_MAX > UINT32_MAX +# define SCNxFAST16 SCNx64 +# elif UINT_FAST16_MAX == 0xffff +# define SCNxFAST16 "hx" +# else +# define SCNxFAST16 "x" +# endif +# endif +# if !defined SCNdFAST32 || @PRI_MACROS_BROKEN@ +# undef SCNdFAST32 +# if INT_FAST32_MAX > INT32_MAX +# define SCNdFAST32 SCNd64 +# else +# define SCNdFAST32 "d" +# endif +# endif +# if !defined SCNiFAST32 || @PRI_MACROS_BROKEN@ +# undef SCNiFAST32 +# if INT_FAST32_MAX > INT32_MAX +# define SCNiFAST32 SCNi64 +# else +# define SCNiFAST32 "i" +# endif +# endif +# if !defined SCNoFAST32 || @PRI_MACROS_BROKEN@ +# undef SCNoFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define SCNoFAST32 SCNo64 +# else +# define SCNoFAST32 "o" +# endif +# endif +# if !defined SCNuFAST32 || @PRI_MACROS_BROKEN@ +# undef SCNuFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define SCNuFAST32 SCNu64 +# else +# define SCNuFAST32 "u" +# endif +# endif +# if !defined SCNxFAST32 || @PRI_MACROS_BROKEN@ +# undef SCNxFAST32 +# if UINT_FAST32_MAX > UINT32_MAX +# define SCNxFAST32 SCNx64 +# else +# define SCNxFAST32 "x" +# endif +# endif +# ifdef INT64_MAX +# if !defined SCNdFAST64 || @PRI_MACROS_BROKEN@ +# undef SCNdFAST64 +# define SCNdFAST64 SCNd64 +# endif +# if !defined SCNiFAST64 || @PRI_MACROS_BROKEN@ +# undef SCNiFAST64 +# define SCNiFAST64 SCNi64 +# endif +# endif +# ifdef UINT64_MAX +# if !defined SCNoFAST64 || @PRI_MACROS_BROKEN@ +# undef SCNoFAST64 +# define SCNoFAST64 SCNo64 +# endif +# if !defined SCNuFAST64 || @PRI_MACROS_BROKEN@ +# undef SCNuFAST64 +# define SCNuFAST64 SCNu64 +# endif +# if !defined SCNxFAST64 || @PRI_MACROS_BROKEN@ +# undef SCNxFAST64 +# define SCNxFAST64 SCNx64 +# endif +# endif + +# if !defined SCNdMAX || @PRI_MACROS_BROKEN@ +# undef SCNdMAX +# if INTMAX_MAX > INT32_MAX +# define SCNdMAX SCNd64 +# else +# define SCNdMAX "ld" +# endif +# endif +# if !defined SCNiMAX || @PRI_MACROS_BROKEN@ +# undef SCNiMAX +# if INTMAX_MAX > INT32_MAX +# define SCNiMAX SCNi64 +# else +# define SCNiMAX "li" +# endif +# endif +# if !defined SCNoMAX || @PRI_MACROS_BROKEN@ +# undef SCNoMAX +# if UINTMAX_MAX > UINT32_MAX +# define SCNoMAX SCNo64 +# else +# define SCNoMAX "lo" +# endif +# endif +# if !defined SCNuMAX || @PRI_MACROS_BROKEN@ +# undef SCNuMAX +# if UINTMAX_MAX > UINT32_MAX +# define SCNuMAX SCNu64 +# else +# define SCNuMAX "lu" +# endif +# endif +# if !defined SCNxMAX || @PRI_MACROS_BROKEN@ +# undef SCNxMAX +# if UINTMAX_MAX > UINT32_MAX +# define SCNxMAX SCNx64 +# else +# define SCNxMAX "lx" +# endif +# endif + +# if !defined SCNdPTR || @PRI_MACROS_BROKEN@ +# undef SCNdPTR +# ifdef INTPTR_MAX +# define SCNdPTR @PRIPTR_PREFIX@ "d" +# endif +# endif +# if !defined SCNiPTR || @PRI_MACROS_BROKEN@ +# undef SCNiPTR +# ifdef INTPTR_MAX +# define SCNiPTR @PRIPTR_PREFIX@ "i" +# endif +# endif +# if !defined SCNoPTR || @PRI_MACROS_BROKEN@ +# undef SCNoPTR +# ifdef UINTPTR_MAX +# define SCNoPTR @PRIPTR_PREFIX@ "o" +# endif +# endif +# if !defined SCNuPTR || @PRI_MACROS_BROKEN@ +# undef SCNuPTR +# ifdef UINTPTR_MAX +# define SCNuPTR @PRIPTR_PREFIX@ "u" +# endif +# endif +# if !defined SCNxPTR || @PRI_MACROS_BROKEN@ +# undef SCNxPTR +# ifdef UINTPTR_MAX +# define SCNxPTR @PRIPTR_PREFIX@ "x" +# endif +# endif + +#endif + +/* 7.8.2 Functions for greatest-width integer types */ + +#ifdef __cplusplus +extern "C" { +#endif + +#if @GNULIB_IMAXABS@ +# if !@HAVE_DECL_IMAXABS@ +extern intmax_t imaxabs (intmax_t); +# endif +#elif defined GNULIB_POSIXCHECK +# undef imaxabs +# define imaxabs(a) \ + (GL_LINK_WARNING ("imaxabs is unportable - " \ + "use gnulib module imaxabs for portability"), \ + imaxabs (a)) +#endif + +#if @GNULIB_IMAXDIV@ +# if !@HAVE_DECL_IMAXDIV@ +typedef struct { intmax_t quot; intmax_t rem; } imaxdiv_t; +extern imaxdiv_t imaxdiv (intmax_t, intmax_t); +# endif +#elif defined GNULIB_POSIXCHECK +# undef imaxdiv +# define imaxdiv(a,b) \ + (GL_LINK_WARNING ("imaxdiv is unportable - " \ + "use gnulib module imaxdiv for portability"), \ + imaxdiv (a, b)) +#endif + +#if @GNULIB_STRTOIMAX@ +# if !@HAVE_DECL_STRTOIMAX@ +extern intmax_t strtoimax (const char *, char **, int); +# endif +#elif defined GNULIB_POSIXCHECK +# undef strtoimax +# define strtoimax(p,e,b) \ + (GL_LINK_WARNING ("strtoimax is unportable - " \ + "use gnulib module strtoimax for portability"), \ + strtoimax (p, e, b)) +#endif + +#if @GNULIB_STRTOUMAX@ +# if !@HAVE_DECL_STRTOUMAX@ +extern uintmax_t strtoumax (const char *, char **, int); +# endif +#elif defined GNULIB_POSIXCHECK +# undef strtoumax +# define strtoumax(p,e,b) \ + (GL_LINK_WARNING ("strtoumax is unportable - " \ + "use gnulib module strtoumax for portability"), \ + strtoumax (p, e, b)) +#endif + +/* Don't bother defining or declaring wcstoimax and wcstoumax, since + wide-character functions like this are hardly ever useful. */ + +#ifdef __cplusplus +} +#endif + +#endif /* INTTYPES_H */ diff --git a/lib/lchown.c b/lib/lchown.c new file mode 100644 index 0000000..fa0826e --- /dev/null +++ b/lib/lchown.c @@ -0,0 +1,47 @@ +/* Provide a stub lchown function for systems that lack it. + + Copyright (C) 1998, 1999, 2002, 2004, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +#include "lchown.h" + +#include <sys/stat.h> + +/* Work just like chown, except when FILE is a symbolic link. + In that case, set errno to EOPNOTSUPP and return -1. + But if autoconf tests determined that chown modifies + symlinks, then just call chown. */ + +int +lchown (const char *file, uid_t uid, gid_t gid) +{ +#if ! CHOWN_MODIFIES_SYMLINK + struct stat stats; + + if (lstat (file, &stats) == 0 && S_ISLNK (stats.st_mode)) + { + errno = EOPNOTSUPP; + return -1; + } +#endif + + return chown (file, uid, gid); +} diff --git a/lib/lchown.h b/lib/lchown.h new file mode 100644 index 0000000..6816d1f --- /dev/null +++ b/lib/lchown.h @@ -0,0 +1,42 @@ +/* Declare a replacement for lchown on hosts that lack it. + + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <errno.h> +#include <sys/types.h> +#include <unistd.h> + +#if HAVE_DECL_LCHOWN +# if ! HAVE_LCHOWN +# undef lchown +# define lchown rpl_chown +# endif +#else +int lchown (char const *, uid_t, gid_t); +#endif + +/* Some systems don't have EOPNOTSUPP. */ +#ifndef EOPNOTSUPP +# ifdef ENOTSUP +# define EOPNOTSUPP ENOTSUP +# else +/* Some systems don't have ENOTSUP either. */ +# define EOPNOTSUPP EINVAL +# endif +#endif diff --git a/lib/localcharset.c b/lib/localcharset.c new file mode 100644 index 0000000..a0f7cca --- /dev/null +++ b/lib/localcharset.c @@ -0,0 +1,460 @@ +/* Determine a canonical name for the current locale's character encoding. + + Copyright (C) 2000-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Bruno Haible <bruno@clisp.org>. */ + +#include <config.h> + +/* Specification. */ +#include "localcharset.h" + +#include <stddef.h> +#include <stdio.h> +#include <string.h> +#include <stdlib.h> + +#if defined _WIN32 || defined __WIN32__ +# define WIN32_NATIVE +#endif + +#if defined __EMX__ +/* Assume EMX program runs on OS/2, even if compiled under DOS. */ +# define OS2 +#endif + +#if !defined WIN32_NATIVE +# if HAVE_LANGINFO_CODESET +# include <langinfo.h> +# else +# if 0 /* see comment below */ +# include <locale.h> +# endif +# endif +# ifdef __CYGWIN__ +# define WIN32_LEAN_AND_MEAN +# include <windows.h> +# endif +#elif defined WIN32_NATIVE +# define WIN32_LEAN_AND_MEAN +# include <windows.h> +#endif +#if defined OS2 +# define INCL_DOS +# include <os2.h> +#endif + +#if ENABLE_RELOCATABLE +# include "relocatable.h" +#else +# define relocate(pathname) (pathname) +#endif + +/* Get LIBDIR. */ +#ifndef LIBDIR +# include "configmake.h" +#endif + +#if defined _WIN32 || defined __WIN32__ || defined __CYGWIN__ || defined __EMX__ || defined __DJGPP__ + /* Win32, Cygwin, OS/2, DOS */ +# define ISSLASH(C) ((C) == '/' || (C) == '\\') +#endif + +#ifndef DIRECTORY_SEPARATOR +# define DIRECTORY_SEPARATOR '/' +#endif + +#ifndef ISSLASH +# define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR) +#endif + +#if HAVE_DECL_GETC_UNLOCKED +# undef getc +# define getc getc_unlocked +#endif + +/* The following static variable is declared 'volatile' to avoid a + possible multithread problem in the function get_charset_aliases. If we + are running in a threaded environment, and if two threads initialize + 'charset_aliases' simultaneously, both will produce the same value, + and everything will be ok if the two assignments to 'charset_aliases' + are atomic. But I don't know what will happen if the two assignments mix. */ +#if __STDC__ != 1 +# define volatile /* empty */ +#endif +/* Pointer to the contents of the charset.alias file, if it has already been + read, else NULL. Its format is: + ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0' */ +static const char * volatile charset_aliases; + +/* Return a pointer to the contents of the charset.alias file. */ +static const char * +get_charset_aliases (void) +{ + const char *cp; + + cp = charset_aliases; + if (cp == NULL) + { +#if !(defined VMS || defined WIN32_NATIVE || defined __CYGWIN__) + FILE *fp; + const char *dir; + const char *base = "charset.alias"; + char *file_name; + + /* Make it possible to override the charset.alias location. This is + necessary for running the testsuite before "make install". */ + dir = getenv ("CHARSETALIASDIR"); + if (dir == NULL || dir[0] == '\0') + dir = relocate (LIBDIR); + + /* Concatenate dir and base into freshly allocated file_name. */ + { + size_t dir_len = strlen (dir); + size_t base_len = strlen (base); + int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1])); + file_name = (char *) malloc (dir_len + add_slash + base_len + 1); + if (file_name != NULL) + { + memcpy (file_name, dir, dir_len); + if (add_slash) + file_name[dir_len] = DIRECTORY_SEPARATOR; + memcpy (file_name + dir_len + add_slash, base, base_len + 1); + } + } + + if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL) + /* Out of memory or file not found, treat it as empty. */ + cp = ""; + else + { + /* Parse the file's contents. */ + char *res_ptr = NULL; + size_t res_size = 0; + + for (;;) + { + int c; + char buf1[50+1]; + char buf2[50+1]; + size_t l1, l2; + char *old_res_ptr; + + c = getc (fp); + if (c == EOF) + break; + if (c == '\n' || c == ' ' || c == '\t') + continue; + if (c == '#') + { + /* Skip comment, to end of line. */ + do + c = getc (fp); + while (!(c == EOF || c == '\n')); + if (c == EOF) + break; + continue; + } + ungetc (c, fp); + if (fscanf (fp, "%50s %50s", buf1, buf2) < 2) + break; + l1 = strlen (buf1); + l2 = strlen (buf2); + old_res_ptr = res_ptr; + if (res_size == 0) + { + res_size = l1 + 1 + l2 + 1; + res_ptr = (char *) malloc (res_size + 1); + } + else + { + res_size += l1 + 1 + l2 + 1; + res_ptr = (char *) realloc (res_ptr, res_size + 1); + } + if (res_ptr == NULL) + { + /* Out of memory. */ + res_size = 0; + if (old_res_ptr != NULL) + free (old_res_ptr); + break; + } + strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1); + strcpy (res_ptr + res_size - (l2 + 1), buf2); + } + fclose (fp); + if (res_size == 0) + cp = ""; + else + { + *(res_ptr + res_size) = '\0'; + cp = res_ptr; + } + } + + if (file_name != NULL) + free (file_name); + +#else + +# if defined VMS + /* To avoid the troubles of an extra file charset.alias_vms in the + sources of many GNU packages, simply inline the aliases here. */ + /* The list of encodings is taken from the OpenVMS 7.3-1 documentation + "Compaq C Run-Time Library Reference Manual for OpenVMS systems" + section 10.7 "Handling Different Character Sets". */ + cp = "ISO8859-1" "\0" "ISO-8859-1" "\0" + "ISO8859-2" "\0" "ISO-8859-2" "\0" + "ISO8859-5" "\0" "ISO-8859-5" "\0" + "ISO8859-7" "\0" "ISO-8859-7" "\0" + "ISO8859-8" "\0" "ISO-8859-8" "\0" + "ISO8859-9" "\0" "ISO-8859-9" "\0" + /* Japanese */ + "eucJP" "\0" "EUC-JP" "\0" + "SJIS" "\0" "SHIFT_JIS" "\0" + "DECKANJI" "\0" "DEC-KANJI" "\0" + "SDECKANJI" "\0" "EUC-JP" "\0" + /* Chinese */ + "eucTW" "\0" "EUC-TW" "\0" + "DECHANYU" "\0" "DEC-HANYU" "\0" + "DECHANZI" "\0" "GB2312" "\0" + /* Korean */ + "DECKOREAN" "\0" "EUC-KR" "\0"; +# endif + +# if defined WIN32_NATIVE || defined __CYGWIN__ + /* To avoid the troubles of installing a separate file in the same + directory as the DLL and of retrieving the DLL's directory at + runtime, simply inline the aliases here. */ + + cp = "CP936" "\0" "GBK" "\0" + "CP1361" "\0" "JOHAB" "\0" + "CP20127" "\0" "ASCII" "\0" + "CP20866" "\0" "KOI8-R" "\0" + "CP20936" "\0" "GB2312" "\0" + "CP21866" "\0" "KOI8-RU" "\0" + "CP28591" "\0" "ISO-8859-1" "\0" + "CP28592" "\0" "ISO-8859-2" "\0" + "CP28593" "\0" "ISO-8859-3" "\0" + "CP28594" "\0" "ISO-8859-4" "\0" + "CP28595" "\0" "ISO-8859-5" "\0" + "CP28596" "\0" "ISO-8859-6" "\0" + "CP28597" "\0" "ISO-8859-7" "\0" + "CP28598" "\0" "ISO-8859-8" "\0" + "CP28599" "\0" "ISO-8859-9" "\0" + "CP28605" "\0" "ISO-8859-15" "\0" + "CP38598" "\0" "ISO-8859-8" "\0" + "CP51932" "\0" "EUC-JP" "\0" + "CP51936" "\0" "GB2312" "\0" + "CP51949" "\0" "EUC-KR" "\0" + "CP51950" "\0" "EUC-TW" "\0" + "CP54936" "\0" "GB18030" "\0" + "CP65001" "\0" "UTF-8" "\0"; +# endif +#endif + + charset_aliases = cp; + } + + return cp; +} + +/* Determine the current locale's character encoding, and canonicalize it + into one of the canonical names listed in config.charset. + The result must not be freed; it is statically allocated. + If the canonical name cannot be determined, the result is a non-canonical + name. */ + +#ifdef STATIC +STATIC +#endif +const char * +locale_charset (void) +{ + const char *codeset; + const char *aliases; + +#if !(defined WIN32_NATIVE || defined OS2) + +# if HAVE_LANGINFO_CODESET + + /* Most systems support nl_langinfo (CODESET) nowadays. */ + codeset = nl_langinfo (CODESET); + +# ifdef __CYGWIN__ + /* Cygwin 2006 does not have locales. nl_langinfo (CODESET) always + returns "US-ASCII". As long as this is not fixed, return the suffix + of the locale name from the environment variables (if present) or + the codepage as a number. */ + if (codeset != NULL && strcmp (codeset, "US-ASCII") == 0) + { + const char *locale; + static char buf[2 + 10 + 1]; + + locale = getenv ("LC_ALL"); + if (locale == NULL || locale[0] == '\0') + { + locale = getenv ("LC_CTYPE"); + if (locale == NULL || locale[0] == '\0') + locale = getenv ("LANG"); + } + if (locale != NULL && locale[0] != '\0') + { + /* If the locale name contains an encoding after the dot, return + it. */ + const char *dot = strchr (locale, '.'); + + if (dot != NULL) + { + const char *modifier; + + dot++; + /* Look for the possible @... trailer and remove it, if any. */ + modifier = strchr (dot, '@'); + if (modifier == NULL) + return dot; + if (modifier - dot < sizeof (buf)) + { + memcpy (buf, dot, modifier - dot); + buf [modifier - dot] = '\0'; + return buf; + } + } + } + + /* Woe32 has a function returning the locale's codepage as a number. */ + sprintf (buf, "CP%u", GetACP ()); + codeset = buf; + } +# endif + +# else + + /* On old systems which lack it, use setlocale or getenv. */ + const char *locale = NULL; + + /* But most old systems don't have a complete set of locales. Some + (like SunOS 4 or DJGPP) have only the C locale. Therefore we don't + use setlocale here; it would return "C" when it doesn't support the + locale name the user has set. */ +# if 0 + locale = setlocale (LC_CTYPE, NULL); +# endif + if (locale == NULL || locale[0] == '\0') + { + locale = getenv ("LC_ALL"); + if (locale == NULL || locale[0] == '\0') + { + locale = getenv ("LC_CTYPE"); + if (locale == NULL || locale[0] == '\0') + locale = getenv ("LANG"); + } + } + + /* On some old systems, one used to set locale = "iso8859_1". On others, + you set it to "language_COUNTRY.charset". In any case, we resolve it + through the charset.alias file. */ + codeset = locale; + +# endif + +#elif defined WIN32_NATIVE + + static char buf[2 + 10 + 1]; + + /* Woe32 has a function returning the locale's codepage as a number. */ + sprintf (buf, "CP%u", GetACP ()); + codeset = buf; + +#elif defined OS2 + + const char *locale; + static char buf[2 + 10 + 1]; + ULONG cp[3]; + ULONG cplen; + + /* Allow user to override the codeset, as set in the operating system, + with standard language environment variables. */ + locale = getenv ("LC_ALL"); + if (locale == NULL || locale[0] == '\0') + { + locale = getenv ("LC_CTYPE"); + if (locale == NULL || locale[0] == '\0') + locale = getenv ("LANG"); + } + if (locale != NULL && locale[0] != '\0') + { + /* If the locale name contains an encoding after the dot, return it. */ + const char *dot = strchr (locale, '.'); + + if (dot != NULL) + { + const char *modifier; + + dot++; + /* Look for the possible @... trailer and remove it, if any. */ + modifier = strchr (dot, '@'); + if (modifier == NULL) + return dot; + if (modifier - dot < sizeof (buf)) + { + memcpy (buf, dot, modifier - dot); + buf [modifier - dot] = '\0'; + return buf; + } + } + + /* Resolve through the charset.alias file. */ + codeset = locale; + } + else + { + /* OS/2 has a function returning the locale's codepage as a number. */ + if (DosQueryCp (sizeof (cp), cp, &cplen)) + codeset = ""; + else + { + sprintf (buf, "CP%u", cp[0]); + codeset = buf; + } + } + +#endif + + if (codeset == NULL) + /* The canonical name cannot be determined. */ + codeset = ""; + + /* Resolve alias. */ + for (aliases = get_charset_aliases (); + *aliases != '\0'; + aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1) + if (strcmp (codeset, aliases) == 0 + || (aliases[0] == '*' && aliases[1] == '\0')) + { + codeset = aliases + strlen (aliases) + 1; + break; + } + + /* Don't return an empty string. GNU libc and GNU libiconv interpret + the empty string as denoting "the locale's character encoding", + thus GNU libiconv would call this function a second time. */ + if (codeset[0] == '\0') + codeset = "ASCII"; + + return codeset; +} diff --git a/lib/localcharset.h b/lib/localcharset.h new file mode 100644 index 0000000..5030210 --- /dev/null +++ b/lib/localcharset.h @@ -0,0 +1,41 @@ +/* Determine a canonical name for the current locale's character encoding. + Copyright (C) 2000-2003 Free Software Foundation, Inc. + This file is part of the GNU CHARSET Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LOCALCHARSET_H +#define _LOCALCHARSET_H + + +#ifdef __cplusplus +extern "C" { +#endif + + +/* Determine the current locale's character encoding, and canonicalize it + into one of the canonical names listed in config.charset. + The result must not be freed; it is statically allocated. + If the canonical name cannot be determined, the result is a non-canonical + name. */ +extern const char * locale_charset (void); + + +#ifdef __cplusplus +} +#endif + + +#endif /* _LOCALCHARSET_H */ diff --git a/lib/lstat.c b/lib/lstat.c new file mode 100644 index 0000000..77dd228 --- /dev/null +++ b/lib/lstat.c @@ -0,0 +1,76 @@ +/* Work around a bug of lstat on some systems + + Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +/* The specification of these functions is in sys_stat.h. But we cannot + include this include file here, because on some systems, a + "#define lstat lstat64" is being used, and sys_stat.h deletes this + definition. */ + +#include <sys/types.h> +#include <sys/stat.h> +#include <string.h> +#include <errno.h> + +/* lstat works differently on Linux and Solaris systems. POSIX (see + `pathname resolution' in the glossary) requires that programs like + `ls' take into consideration the fact that FILE has a trailing slash + when FILE is a symbolic link. On Linux and Solaris 10 systems, the + lstat function already has the desired semantics (in treating + `lstat ("symlink/", sbuf)' just like `lstat ("symlink/.", sbuf)', + but on Solaris 9 and earlier it does not. + + If FILE has a trailing slash and specifies a symbolic link, + then use stat() to get more info on the referent of FILE. + If the referent is a non-directory, then set errno to ENOTDIR + and return -1. Otherwise, return stat's result. */ + +int +rpl_lstat (const char *file, struct stat *sbuf) +{ + size_t len; + int lstat_result = lstat (file, sbuf); + + if (lstat_result != 0 || !S_ISLNK (sbuf->st_mode)) + return lstat_result; + + len = strlen (file); + if (len == 0 || file[len - 1] != '/') + return 0; + + /* FILE refers to a symbolic link and the name ends with a slash. + Call stat() to get info about the link's referent. */ + + /* If stat fails, then we do the same. */ + if (stat (file, sbuf) != 0) + return -1; + + /* If FILE references a directory, return 0. */ + if (S_ISDIR (sbuf->st_mode)) + return 0; + + /* Here, we know stat succeeded and FILE references a non-directory. + But it was specified via a name including a trailing slash. + Fail with errno set to ENOTDIR to indicate the contradiction. */ + errno = ENOTDIR; + return -1; +} diff --git a/lib/lstat.h b/lib/lstat.h new file mode 100644 index 0000000..6a83fbf --- /dev/null +++ b/lib/lstat.h @@ -0,0 +1,24 @@ +/* Retrieving information about files. + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <sys/stat.h> + +#if !LSTAT_FOLLOWS_SLASHED_SYMLINK +extern int rpl_lstat (const char *name, struct stat *buf); +# undef lstat +# define lstat rpl_lstat +#endif diff --git a/lib/malloc.c b/lib/malloc.c new file mode 100644 index 0000000..d4dae3e --- /dev/null +++ b/lib/malloc.c @@ -0,0 +1,35 @@ +/* malloc() function that is glibc compatible. + + Copyright (C) 1997, 1998, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> +#undef malloc + +#include <stdlib.h> + +/* Allocate an N-byte block of memory from the heap. + If N is zero, allocate a 1-byte block. */ + +void * +rpl_malloc (size_t n) +{ + if (n == 0) + n = 1; + return malloc (n); +} diff --git a/lib/mbchar.c b/lib/mbchar.c new file mode 100644 index 0000000..95373f5 --- /dev/null +++ b/lib/mbchar.c @@ -0,0 +1,36 @@ +/* Copyright (C) 2001, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + + +#include <config.h> + +#include <limits.h> + +#include "mbchar.h" + +#if IS_BASIC_ASCII + +/* Bit table of characters in the ISO C "basic character set". */ +const unsigned int is_basic_table [UCHAR_MAX / 32 + 1] = +{ + 0x00001a00, /* '\t' '\v' '\f' */ + 0xffffffef, /* ' '...'#' '%'...'?' */ + 0xfffffffe, /* 'A'...'Z' '[' '\\' ']' '^' '_' */ + 0x7ffffffe /* 'a'...'z' '{' '|' '}' '~' */ + /* The remaining bits are 0. */ +}; + +#endif /* IS_BASIC_ASCII */ diff --git a/lib/mbchar.h b/lib/mbchar.h new file mode 100644 index 0000000..f3e28ef --- /dev/null +++ b/lib/mbchar.h @@ -0,0 +1,353 @@ +/* Multibyte character data type. + Copyright (C) 2001, 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Bruno Haible <bruno@clisp.org>. */ + +/* A multibyte character is a short subsequence of a char* string, + representing a single wide character. + + We use multibyte characters instead of wide characters because of + the following goals: + 1) correct multibyte handling, i.e. operate according to the LC_CTYPE + locale, + 2) ease of maintenance, i.e. the maintainer needs not know all details + of the ISO C 99 standard, + 3) don't fail grossly if the input is not in the encoding set by the + locale, because often different encodings are in use in the same + countries (ISO-8859-1/UTF-8, EUC-JP/Shift_JIS, ...), + 4) fast in the case of ASCII characters, + 5) portability, i.e. don't make unportable assumptions about wchar_t. + + Multibyte characters are only accessed through the mb* macros. + + mb_ptr (mbc) + return a pointer to the beginning of the multibyte sequence. + + mb_len (mbc) + returns the number of bytes occupied by the multibyte sequence. + Always > 0. + + mb_iseq (mbc, sc) + returns true if mbc is the standard ASCII character sc. + + mb_isnul (mbc) + returns true if mbc is the nul character. + + mb_cmp (mbc1, mbc2) + returns a positive, zero, or negative value depending on whether mbc1 + sorts after, same or before mbc2. + + mb_casecmp (mbc1, mbc2) + returns a positive, zero, or negative value depending on whether mbc1 + sorts after, same or before mbc2, modulo upper/lowercase conversion. + + mb_equal (mbc1, mbc2) + returns true if mbc1 and mbc2 are equal. + + mb_caseequal (mbc1, mbc2) + returns true if mbc1 and mbc2 are equal modulo upper/lowercase conversion. + + mb_isalnum (mbc) + returns true if mbc is alphanumeric. + + mb_isalpha (mbc) + returns true if mbc is alphabetic. + + mb_isascii(mbc) + returns true if mbc is plain ASCII. + + mb_isblank (mbc) + returns true if mbc is a blank. + + mb_iscntrl (mbc) + returns true if mbc is a control character. + + mb_isdigit (mbc) + returns true if mbc is a decimal digit. + + mb_isgraph (mbc) + returns true if mbc is a graphic character. + + mb_islower (mbc) + returns true if mbc is lowercase. + + mb_isprint (mbc) + returns true if mbc is a printable character. + + mb_ispunct (mbc) + returns true if mbc is a punctuation character. + + mb_isspace (mbc) + returns true if mbc is a space character. + + mb_isupper (mbc) + returns true if mbc is uppercase. + + mb_isxdigit (mbc) + returns true if mbc is a hexadecimal digit. + + mb_width (mbc) + returns the number of columns on the output device occupied by mbc. + Always >= 0. + + mb_putc (mbc, stream) + outputs mbc on stream, a byte oriented FILE stream opened for output. + + mb_setascii (&mbc, sc) + assigns the standard ASCII character sc to mbc. + + mb_copy (&destmbc, &srcmbc) + copies srcmbc to destmbc. + + Here are the function prototypes of the macros. + + extern const char * mb_ptr (const mbchar_t mbc); + extern size_t mb_len (const mbchar_t mbc); + extern bool mb_iseq (const mbchar_t mbc, char sc); + extern bool mb_isnul (const mbchar_t mbc); + extern int mb_cmp (const mbchar_t mbc1, const mbchar_t mbc2); + extern int mb_casecmp (const mbchar_t mbc1, const mbchar_t mbc2); + extern bool mb_equal (const mbchar_t mbc1, const mbchar_t mbc2); + extern bool mb_caseequal (const mbchar_t mbc1, const mbchar_t mbc2); + extern bool mb_isalnum (const mbchar_t mbc); + extern bool mb_isalpha (const mbchar_t mbc); + extern bool mb_isascii (const mbchar_t mbc); + extern bool mb_isblank (const mbchar_t mbc); + extern bool mb_iscntrl (const mbchar_t mbc); + extern bool mb_isdigit (const mbchar_t mbc); + extern bool mb_isgraph (const mbchar_t mbc); + extern bool mb_islower (const mbchar_t mbc); + extern bool mb_isprint (const mbchar_t mbc); + extern bool mb_ispunct (const mbchar_t mbc); + extern bool mb_isspace (const mbchar_t mbc); + extern bool mb_isupper (const mbchar_t mbc); + extern bool mb_isxdigit (const mbchar_t mbc); + extern int mb_width (const mbchar_t mbc); + extern void mb_putc (const mbchar_t mbc, FILE *stream); + extern void mb_setascii (mbchar_t *new, char sc); + extern void mb_copy (mbchar_t *new, const mbchar_t *old); + */ + +#ifndef _MBCHAR_H +#define _MBCHAR_H 1 + +#include <stdbool.h> +#include <string.h> + +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before + <wchar.h>. */ +#include <stdio.h> +#include <time.h> +#include <wchar.h> +#include <wctype.h> + +#include "wcwidth.h" + +#define MBCHAR_BUF_SIZE 24 + +struct mbchar +{ + const char *ptr; /* pointer to current character */ + size_t bytes; /* number of bytes of current character, > 0 */ + bool wc_valid; /* true if wc is a valid wide character */ + wchar_t wc; /* if wc_valid: the current character */ + char buf[MBCHAR_BUF_SIZE]; /* room for the bytes, used for file input only */ +}; + +/* EOF (not a real character) is represented with bytes = 0 and + wc_valid = false. */ + +typedef struct mbchar mbchar_t; + +/* Access the current character. */ +#define mb_ptr(mbc) ((mbc).ptr) +#define mb_len(mbc) ((mbc).bytes) + +/* Comparison of characters. */ +#define mb_iseq(mbc, sc) ((mbc).wc_valid && (mbc).wc == (sc)) +#define mb_isnul(mbc) ((mbc).wc_valid && (mbc).wc == 0) +#define mb_cmp(mbc1, mbc2) \ + ((mbc1).wc_valid \ + ? ((mbc2).wc_valid \ + ? (int) (mbc1).wc - (int) (mbc2).wc \ + : -1) \ + : ((mbc2).wc_valid \ + ? 1 \ + : (mbc1).bytes == (mbc2).bytes \ + ? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \ + : (mbc1).bytes < (mbc2).bytes \ + ? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \ + : (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1))) +#define mb_casecmp(mbc1, mbc2) \ + ((mbc1).wc_valid \ + ? ((mbc2).wc_valid \ + ? (int) towlower ((mbc1).wc) - (int) towlower ((mbc2).wc) \ + : -1) \ + : ((mbc2).wc_valid \ + ? 1 \ + : (mbc1).bytes == (mbc2).bytes \ + ? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \ + : (mbc1).bytes < (mbc2).bytes \ + ? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \ + : (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1))) +#define mb_equal(mbc1, mbc2) \ + ((mbc1).wc_valid && (mbc2).wc_valid \ + ? (mbc1).wc == (mbc2).wc \ + : (mbc1).bytes == (mbc2).bytes \ + && memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0) +#define mb_caseequal(mbc1, mbc2) \ + ((mbc1).wc_valid && (mbc2).wc_valid \ + ? towlower ((mbc1).wc) == towlower ((mbc2).wc) \ + : (mbc1).bytes == (mbc2).bytes \ + && memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0) + +/* <ctype.h>, <wctype.h> classification. */ +#define mb_isascii(mbc) \ + ((mbc).wc_valid && (mbc).wc >= 0 && (mbc).wc <= 127) +#define mb_isalnum(mbc) ((mbc).wc_valid && iswalnum ((mbc).wc)) +#define mb_isalpha(mbc) ((mbc).wc_valid && iswalpha ((mbc).wc)) +#define mb_isblank(mbc) ((mbc).wc_valid && iswblank ((mbc).wc)) +#define mb_iscntrl(mbc) ((mbc).wc_valid && iswcntrl ((mbc).wc)) +#define mb_isdigit(mbc) ((mbc).wc_valid && iswdigit ((mbc).wc)) +#define mb_isgraph(mbc) ((mbc).wc_valid && iswgraph ((mbc).wc)) +#define mb_islower(mbc) ((mbc).wc_valid && iswlower ((mbc).wc)) +#define mb_isprint(mbc) ((mbc).wc_valid && iswprint ((mbc).wc)) +#define mb_ispunct(mbc) ((mbc).wc_valid && iswpunct ((mbc).wc)) +#define mb_isspace(mbc) ((mbc).wc_valid && iswspace ((mbc).wc)) +#define mb_isupper(mbc) ((mbc).wc_valid && iswupper ((mbc).wc)) +#define mb_isxdigit(mbc) ((mbc).wc_valid && iswxdigit ((mbc).wc)) + +/* Extra <wchar.h> function. */ + +/* Unprintable characters appear as a small box of width 1. */ +#define MB_UNPRINTABLE_WIDTH 1 + +static inline int +mb_width_aux (wint_t wc) +{ + int w = wcwidth (wc); + /* For unprintable characters, arbitrarily return 0 for control characters + and MB_UNPRINTABLE_WIDTH otherwise. */ + return (w >= 0 ? w : iswcntrl (wc) ? 0 : MB_UNPRINTABLE_WIDTH); +} + +#define mb_width(mbc) \ + ((mbc).wc_valid ? mb_width_aux ((mbc).wc) : MB_UNPRINTABLE_WIDTH) + +/* Output. */ +#define mb_putc(mbc, stream) fwrite ((mbc).ptr, 1, (mbc).bytes, (stream)) + +/* Assignment. */ +#define mb_setascii(mbc, sc) \ + ((mbc)->ptr = (mbc)->buf, (mbc)->bytes = 1, (mbc)->wc_valid = 1, \ + (mbc)->wc = (mbc)->buf[0] = (sc)) + +/* Copying a character. */ +static inline void +mb_copy (mbchar_t *new_mbc, const mbchar_t *old_mbc) +{ + if (old_mbc->ptr == &old_mbc->buf[0]) + { + memcpy (&new_mbc->buf[0], &old_mbc->buf[0], old_mbc->bytes); + new_mbc->ptr = &new_mbc->buf[0]; + } + else + new_mbc->ptr = old_mbc->ptr; + new_mbc->bytes = old_mbc->bytes; + if ((new_mbc->wc_valid = old_mbc->wc_valid)) + new_mbc->wc = old_mbc->wc; +} + + +/* is_basic(c) tests whether the single-byte character c is in the + ISO C "basic character set". + This is a convenience function, and is in this file only to share code + between mbiter_multi.h and mbfile_multi.h. */ +#if (' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \ + && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \ + && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \ + && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \ + && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \ + && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \ + && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \ + && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \ + && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \ + && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \ + && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \ + && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \ + && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \ + && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \ + && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \ + && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \ + && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \ + && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \ + && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \ + && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \ + && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \ + && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \ + && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126) +/* The character set is ISO-646, not EBCDIC. */ +# define IS_BASIC_ASCII 1 + +extern const unsigned int is_basic_table[]; + +static inline bool +is_basic (char c) +{ + return (is_basic_table [(unsigned char) c >> 5] >> ((unsigned char) c & 31)) + & 1; +} + +#else + +static inline bool +is_basic (char c) +{ + switch (c) + { + case '\t': case '\v': case '\f': + case ' ': case '!': case '"': case '#': case '%': + case '&': case '\'': case '(': case ')': case '*': + case '+': case ',': case '-': case '.': case '/': + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + case ':': case ';': case '<': case '=': case '>': + case '?': + case 'A': case 'B': case 'C': case 'D': case 'E': + case 'F': case 'G': case 'H': case 'I': case 'J': + case 'K': case 'L': case 'M': case 'N': case 'O': + case 'P': case 'Q': case 'R': case 'S': case 'T': + case 'U': case 'V': case 'W': case 'X': case 'Y': + case 'Z': + case '[': case '\\': case ']': case '^': case '_': + case 'a': case 'b': case 'c': case 'd': case 'e': + case 'f': case 'g': case 'h': case 'i': case 'j': + case 'k': case 'l': case 'm': case 'n': case 'o': + case 'p': case 'q': case 'r': case 's': case 't': + case 'u': case 'v': case 'w': case 'x': case 'y': + case 'z': case '{': case '|': case '}': case '~': + return 1; + default: + return 0; + } +} + +#endif + +#endif /* _MBCHAR_H */ diff --git a/lib/mbscasecmp.c b/lib/mbscasecmp.c new file mode 100644 index 0000000..8a2f434 --- /dev/null +++ b/lib/mbscasecmp.c @@ -0,0 +1,103 @@ +/* Case-insensitive string comparison function. + Copyright (C) 1998-1999, 2005-2007 Free Software Foundation, Inc. + Written by Bruno Haible <bruno@clisp.org>, 2005, + based on earlier glibc code. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <string.h> + +#include <ctype.h> +#include <limits.h> + +#if HAVE_MBRTOWC +# include "mbuiter.h" +#endif + +#define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch)) + +/* Compare the character strings S1 and S2, ignoring case, returning less than, + equal to or greater than zero if S1 is lexicographically less than, equal to + or greater than S2. + Note: This function may, in multibyte locales, return 0 for strings of + different lengths! */ +int +mbscasecmp (const char *s1, const char *s2) +{ + if (s1 == s2) + return 0; + + /* Be careful not to look at the entire extent of s1 or s2 until needed. + This is useful because when two strings differ, the difference is + most often already in the very few first characters. */ +#if HAVE_MBRTOWC + if (MB_CUR_MAX > 1) + { + mbui_iterator_t iter1; + mbui_iterator_t iter2; + + mbui_init (iter1, s1); + mbui_init (iter2, s2); + + while (mbui_avail (iter1) && mbui_avail (iter2)) + { + int cmp = mb_casecmp (mbui_cur (iter1), mbui_cur (iter2)); + + if (cmp != 0) + return cmp; + + mbui_advance (iter1); + mbui_advance (iter2); + } + if (mbui_avail (iter1)) + /* s2 terminated before s1. */ + return 1; + if (mbui_avail (iter2)) + /* s1 terminated before s2. */ + return -1; + return 0; + } + else +#endif + { + const unsigned char *p1 = (const unsigned char *) s1; + const unsigned char *p2 = (const unsigned char *) s2; + unsigned char c1, c2; + + do + { + c1 = TOLOWER (*p1); + c2 = TOLOWER (*p2); + + if (c1 == '\0') + break; + + ++p1; + ++p2; + } + while (c1 == c2); + + if (UCHAR_MAX <= INT_MAX) + return c1 - c2; + else + /* On machines where 'char' and 'int' are types of the same size, the + difference of two 'unsigned char' values - including the sign bit - + doesn't fit in an 'int'. */ + return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0); + } +} diff --git a/lib/mbuiter.h b/lib/mbuiter.h new file mode 100644 index 0000000..e6ad488 --- /dev/null +++ b/lib/mbuiter.h @@ -0,0 +1,222 @@ +/* Iterating through multibyte strings: macros for multi-byte encodings. + Copyright (C) 2001, 2005, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Bruno Haible <bruno@clisp.org>. */ + +/* The macros in this file implement forward iteration through a + multi-byte string, without knowing its length a-priori. + + With these macros, an iteration loop that looks like + + char *iter; + for (iter = buf; *iter != '\0'; iter++) + { + do_something (*iter); + } + + becomes + + mbui_iterator_t iter; + for (mbui_init (iter, buf); mbui_avail (iter); mbui_advance (iter)) + { + do_something (mbui_cur_ptr (iter), mb_len (mbui_cur (iter))); + } + + The benefit of these macros over plain use of mbrtowc is: + - Handling of invalid multibyte sequences is possible without + making the code more complicated, while still preserving the + invalid multibyte sequences. + + Compared to mbiter.h, the macros here don't need to know the string's + length a-priori. The downside is that at each step, the look-ahead + that guards against overrunning the terminating '\0' is more expensive. + The mbui_* macros are therefore suitable when there is a high probability + that only the first few multibyte characters need to be inspected. + Whereas the mbi_* macros are better if usually the iteration runs + through the entire string. + + mbui_iterator_t + is a type usable for variable declarations. + + mbui_init (iter, startptr) + initializes the iterator, starting at startptr. + + mbui_avail (iter) + returns true if there are more multibyte chracters available before + the end of string is reached. In this case, mbui_cur (iter) is + initialized to the next multibyte chracter. + + mbui_advance (iter) + advances the iterator by one multibyte character. + + mbui_cur (iter) + returns the current multibyte character, of type mbchar_t. All the + macros defined in mbchar.h can be used on it. + + mbui_cur_ptr (iter) + return a pointer to the beginning of the current multibyte character. + + mbui_reloc (iter, ptrdiff) + relocates iterator when the string is moved by ptrdiff bytes. + + mbui_copy (&destiter, &srciter) + copies srciter to destiter. + + Here are the function prototypes of the macros. + + extern void mbui_init (mbui_iterator_t iter, const char *startptr); + extern bool mbui_avail (mbui_iterator_t iter); + extern void mbui_advance (mbui_iterator_t iter); + extern mbchar_t mbui_cur (mbui_iterator_t iter); + extern const char * mbui_cur_ptr (mbui_iterator_t iter); + extern void mbui_reloc (mbui_iterator_t iter, ptrdiff_t ptrdiff); + extern void mbui_copy (mbui_iterator_t *new, const mbui_iterator_t *old); + */ + +#ifndef _MBUITER_H +#define _MBUITER_H 1 + +#include <assert.h> +#include <stdbool.h> +#include <stdlib.h> +#include <string.h> + +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before + <wchar.h>. */ +#include <stdio.h> +#include <time.h> +#include <wchar.h> + +#include "mbchar.h" +#include "strnlen1.h" + +struct mbuiter_multi +{ + bool in_shift; /* true if next byte may not be interpreted as ASCII */ + mbstate_t state; /* if in_shift: current shift state */ + bool next_done; /* true if mbui_avail has already filled the following */ + struct mbchar cur; /* the current character: + const char *cur.ptr pointer to current character + The following are only valid after mbui_avail. + size_t cur.bytes number of bytes of current character + bool cur.wc_valid true if wc is a valid wide character + wchar_t cur.wc if wc_valid: the current character + */ +}; + +static inline void +mbuiter_multi_next (struct mbuiter_multi *iter) +{ + if (iter->next_done) + return; + if (iter->in_shift) + goto with_shift; + /* Handle most ASCII characters quickly, without calling mbrtowc(). */ + if (is_basic (*iter->cur.ptr)) + { + /* These characters are part of the basic character set. ISO C 99 + guarantees that their wide character code is identical to their + char code. */ + iter->cur.bytes = 1; + iter->cur.wc = *iter->cur.ptr; + iter->cur.wc_valid = true; + } + else + { + assert (mbsinit (&iter->state)); + iter->in_shift = true; + with_shift: + iter->cur.bytes = mbrtowc (&iter->cur.wc, iter->cur.ptr, + strnlen1 (iter->cur.ptr, MB_CUR_MAX), + &iter->state); + if (iter->cur.bytes == (size_t) -1) + { + /* An invalid multibyte sequence was encountered. */ + iter->cur.bytes = 1; + iter->cur.wc_valid = false; + /* Whether to set iter->in_shift = false and reset iter->state + or not is not very important; the string is bogus anyway. */ + } + else if (iter->cur.bytes == (size_t) -2) + { + /* An incomplete multibyte character at the end. */ + iter->cur.bytes = strlen (iter->cur.ptr); + iter->cur.wc_valid = false; + /* Whether to set iter->in_shift = false and reset iter->state + or not is not important; the string end is reached anyway. */ + } + else + { + if (iter->cur.bytes == 0) + { + /* A null wide character was encountered. */ + iter->cur.bytes = 1; + assert (*iter->cur.ptr == '\0'); + assert (iter->cur.wc == 0); + } + iter->cur.wc_valid = true; + + /* When in the initial state, we can go back treating ASCII + characters more quickly. */ + if (mbsinit (&iter->state)) + iter->in_shift = false; + } + } + iter->next_done = true; +} + +static inline void +mbuiter_multi_reloc (struct mbuiter_multi *iter, ptrdiff_t ptrdiff) +{ + iter->cur.ptr += ptrdiff; +} + +static inline void +mbuiter_multi_copy (struct mbuiter_multi *new_iter, const struct mbuiter_multi *old_iter) +{ + if ((new_iter->in_shift = old_iter->in_shift)) + memcpy (&new_iter->state, &old_iter->state, sizeof (mbstate_t)); + else + memset (&new_iter->state, 0, sizeof (mbstate_t)); + new_iter->next_done = old_iter->next_done; + mb_copy (&new_iter->cur, &old_iter->cur); +} + +/* Iteration macros. */ +typedef struct mbuiter_multi mbui_iterator_t; +#define mbui_init(iter, startptr) \ + ((iter).cur.ptr = (startptr), \ + (iter).in_shift = false, memset (&(iter).state, '\0', sizeof (mbstate_t)), \ + (iter).next_done = false) +#define mbui_avail(iter) \ + (mbuiter_multi_next (&(iter)), !mb_isnul ((iter).cur)) +#define mbui_advance(iter) \ + ((iter).cur.ptr += (iter).cur.bytes, (iter).next_done = false) + +/* Access to the current character. */ +#define mbui_cur(iter) (iter).cur +#define mbui_cur_ptr(iter) (iter).cur.ptr + +/* Relocation. */ +#define mbui_reloc(iter, ptrdiff) mbuiter_multi_reloc (&iter, ptrdiff) + +/* Copying an iterator. */ +#define mbui_copy mbuiter_multi_copy + +#endif /* _MBUITER_H */ diff --git a/lib/memchr.c b/lib/memchr.c new file mode 100644 index 0000000..d44ad6d --- /dev/null +++ b/lib/memchr.c @@ -0,0 +1,201 @@ +/* Copyright (C) 1991, 1993, 1996, 1997, 1999, 2000, 2003, 2004, 2006 Free + Software Foundation, Inc. + + Based on strlen implementation by Torbjorn Granlund (tege@sics.se), + with help from Dan Sahlin (dan@sics.se) and + commentary by Jim Blandy (jimb@ai.mit.edu); + adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu), + and implemented by Roland McGrath (roland@ai.mit.edu). + +NOTE: The canonical source of this file is maintained with the GNU C Library. +Bugs can be reported to bug-glibc@prep.ai.mit.edu. + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 2, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software Foundation, +Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +#endif + +#include <string.h> + +#include <stddef.h> + +#if defined _LIBC +# include <memcopy.h> +#else +# define reg_char char +#endif + +#include <limits.h> + +#if HAVE_BP_SYM_H || defined _LIBC +# include <bp-sym.h> +#else +# define BP_SYM(sym) sym +#endif + +#undef memchr +#undef __memchr + +/* Search no more than N bytes of S for C. */ +void * +__memchr (void const *s, int c_in, size_t n) +{ + const unsigned char *char_ptr; + const unsigned long int *longword_ptr; + unsigned long int longword, magic_bits, charmask; + unsigned reg_char c; + int i; + + c = (unsigned char) c_in; + + /* Handle the first few characters by reading one character at a time. + Do this until CHAR_PTR is aligned on a longword boundary. */ + for (char_ptr = (const unsigned char *) s; + n > 0 && (size_t) char_ptr % sizeof longword != 0; + --n, ++char_ptr) + if (*char_ptr == c) + return (void *) char_ptr; + + /* All these elucidatory comments refer to 4-byte longwords, + but the theory applies equally well to any size longwords. */ + + longword_ptr = (const unsigned long int *) char_ptr; + + /* Bits 31, 24, 16, and 8 of this number are zero. Call these bits + the "holes." Note that there is a hole just to the left of + each byte, with an extra at the end: + + bits: 01111110 11111110 11111110 11111111 + bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD + + The 1-bits make sure that carries propagate to the next 0-bit. + The 0-bits provide holes for carries to fall into. */ + + /* Set MAGIC_BITS to be this pattern of 1 and 0 bits. + Set CHARMASK to be a longword, each of whose bytes is C. */ + + magic_bits = 0xfefefefe; + charmask = c | (c << 8); + charmask |= charmask << 16; +#if 0xffffffffU < ULONG_MAX + magic_bits |= magic_bits << 32; + charmask |= charmask << 32; + if (8 < sizeof longword) + for (i = 64; i < sizeof longword * 8; i *= 2) + { + magic_bits |= magic_bits << i; + charmask |= charmask << i; + } +#endif + magic_bits = (ULONG_MAX >> 1) & (magic_bits | 1); + + /* Instead of the traditional loop which tests each character, + we will test a longword at a time. The tricky part is testing + if *any of the four* bytes in the longword in question are zero. */ + while (n >= sizeof longword) + { + /* We tentatively exit the loop if adding MAGIC_BITS to + LONGWORD fails to change any of the hole bits of LONGWORD. + + 1) Is this safe? Will it catch all the zero bytes? + Suppose there is a byte with all zeros. Any carry bits + propagating from its left will fall into the hole at its + least significant bit and stop. Since there will be no + carry from its most significant bit, the LSB of the + byte to the left will be unchanged, and the zero will be + detected. + + 2) Is this worthwhile? Will it ignore everything except + zero bytes? Suppose every byte of LONGWORD has a bit set + somewhere. There will be a carry into bit 8. If bit 8 + is set, this will carry into bit 16. If bit 8 is clear, + one of bits 9-15 must be set, so there will be a carry + into bit 16. Similarly, there will be a carry into bit + 24. If one of bits 24-30 is set, there will be a carry + into bit 31, so all of the hole bits will be changed. + + The one misfire occurs when bits 24-30 are clear and bit + 31 is set; in this case, the hole at bit 31 is not + changed. If we had access to the processor carry flag, + we could close this loophole by putting the fourth hole + at bit 32! + + So it ignores everything except 128's, when they're aligned + properly. + + 3) But wait! Aren't we looking for C, not zero? + Good point. So what we do is XOR LONGWORD with a longword, + each of whose bytes is C. This turns each byte that is C + into a zero. */ + + longword = *longword_ptr++ ^ charmask; + + /* Add MAGIC_BITS to LONGWORD. */ + if ((((longword + magic_bits) + + /* Set those bits that were unchanged by the addition. */ + ^ ~longword) + + /* Look at only the hole bits. If any of the hole bits + are unchanged, most likely one of the bytes was a + zero. */ + & ~magic_bits) != 0) + { + /* Which of the bytes was C? If none of them were, it was + a misfire; continue the search. */ + + const unsigned char *cp = (const unsigned char *) (longword_ptr - 1); + + if (cp[0] == c) + return (void *) cp; + if (cp[1] == c) + return (void *) &cp[1]; + if (cp[2] == c) + return (void *) &cp[2]; + if (cp[3] == c) + return (void *) &cp[3]; + if (4 < sizeof longword && cp[4] == c) + return (void *) &cp[4]; + if (5 < sizeof longword && cp[5] == c) + return (void *) &cp[5]; + if (6 < sizeof longword && cp[6] == c) + return (void *) &cp[6]; + if (7 < sizeof longword && cp[7] == c) + return (void *) &cp[7]; + if (8 < sizeof longword) + for (i = 8; i < sizeof longword; i++) + if (cp[i] == c) + return (void *) &cp[i]; + } + + n -= sizeof longword; + } + + char_ptr = (const unsigned char *) longword_ptr; + + while (n-- > 0) + { + if (*char_ptr == c) + return (void *) char_ptr; + else + ++char_ptr; + } + + return 0; +} +#ifdef weak_alias +weak_alias (__memchr, BP_SYM (memchr)) +#endif diff --git a/lib/mempcpy.c b/lib/mempcpy.c new file mode 100644 index 0000000..1c702c7 --- /dev/null +++ b/lib/mempcpy.c @@ -0,0 +1,29 @@ +/* Copy memory area and return pointer after last written byte. + Copyright (C) 2003, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <string.h> + +/* Copy N bytes of SRC to DEST, return pointer to bytes after the + last written byte. */ +void * +mempcpy (void *dest, const void *src, size_t n) +{ + return (char *) memcpy (dest, src, n) + n; +} diff --git a/lib/memrchr.c b/lib/memrchr.c new file mode 100644 index 0000000..29fd531 --- /dev/null +++ b/lib/memrchr.c @@ -0,0 +1,190 @@ +/* memrchr -- find the last occurrence of a byte in a memory block + + Copyright (C) 1991, 1993, 1996, 1997, 1999, 2000, 2003, 2004, 2005, + 2006, 2007 Free Software Foundation, Inc. + + Based on strlen implementation by Torbjorn Granlund (tege@sics.se), + with help from Dan Sahlin (dan@sics.se) and + commentary by Jim Blandy (jimb@ai.mit.edu); + adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu), + and implemented by Roland McGrath (roland@ai.mit.edu). + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if defined _LIBC +# include <memcopy.h> +#else +# include <config.h> +# define reg_char char +#endif + +#include <string.h> +#include <limits.h> + +#undef __memrchr +#undef memrchr + +#ifndef weak_alias +# define __memrchr memrchr +#endif + +/* Search no more than N bytes of S for C. */ +void * +__memrchr (void const *s, int c_in, size_t n) +{ + const unsigned char *char_ptr; + const unsigned long int *longword_ptr; + unsigned long int longword, magic_bits, charmask; + unsigned reg_char c; + int i; + + c = (unsigned char) c_in; + + /* Handle the last few characters by reading one character at a time. + Do this until CHAR_PTR is aligned on a longword boundary. */ + for (char_ptr = (const unsigned char *) s + n; + n > 0 && (size_t) char_ptr % sizeof longword != 0; + --n) + if (*--char_ptr == c) + return (void *) char_ptr; + + /* All these elucidatory comments refer to 4-byte longwords, + but the theory applies equally well to any size longwords. */ + + longword_ptr = (const unsigned long int *) char_ptr; + + /* Bits 31, 24, 16, and 8 of this number are zero. Call these bits + the "holes." Note that there is a hole just to the left of + each byte, with an extra at the end: + + bits: 01111110 11111110 11111110 11111111 + bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD + + The 1-bits make sure that carries propagate to the next 0-bit. + The 0-bits provide holes for carries to fall into. */ + + /* Set MAGIC_BITS to be this pattern of 1 and 0 bits. + Set CHARMASK to be a longword, each of whose bytes is C. */ + + magic_bits = 0xfefefefe; + charmask = c | (c << 8); + charmask |= charmask << 16; +#if 0xffffffffU < ULONG_MAX + magic_bits |= magic_bits << 32; + charmask |= charmask << 32; + if (8 < sizeof longword) + for (i = 64; i < sizeof longword * 8; i *= 2) + { + magic_bits |= magic_bits << i; + charmask |= charmask << i; + } +#endif + magic_bits = (ULONG_MAX >> 1) & (magic_bits | 1); + + /* Instead of the traditional loop which tests each character, + we will test a longword at a time. The tricky part is testing + if *any of the four* bytes in the longword in question are zero. */ + while (n >= sizeof longword) + { + /* We tentatively exit the loop if adding MAGIC_BITS to + LONGWORD fails to change any of the hole bits of LONGWORD. + + 1) Is this safe? Will it catch all the zero bytes? + Suppose there is a byte with all zeros. Any carry bits + propagating from its left will fall into the hole at its + least significant bit and stop. Since there will be no + carry from its most significant bit, the LSB of the + byte to the left will be unchanged, and the zero will be + detected. + + 2) Is this worthwhile? Will it ignore everything except + zero bytes? Suppose every byte of LONGWORD has a bit set + somewhere. There will be a carry into bit 8. If bit 8 + is set, this will carry into bit 16. If bit 8 is clear, + one of bits 9-15 must be set, so there will be a carry + into bit 16. Similarly, there will be a carry into bit + 24. If one of bits 24-30 is set, there will be a carry + into bit 31, so all of the hole bits will be changed. + + The one misfire occurs when bits 24-30 are clear and bit + 31 is set; in this case, the hole at bit 31 is not + changed. If we had access to the processor carry flag, + we could close this loophole by putting the fourth hole + at bit 32! + + So it ignores everything except 128's, when they're aligned + properly. + + 3) But wait! Aren't we looking for C, not zero? + Good point. So what we do is XOR LONGWORD with a longword, + each of whose bytes is C. This turns each byte that is C + into a zero. */ + + longword = *--longword_ptr ^ charmask; + + /* Add MAGIC_BITS to LONGWORD. */ + if ((((longword + magic_bits) + + /* Set those bits that were unchanged by the addition. */ + ^ ~longword) + + /* Look at only the hole bits. If any of the hole bits + are unchanged, most likely one of the bytes was a + zero. */ + & ~magic_bits) != 0) + { + /* Which of the bytes was C? If none of them were, it was + a misfire; continue the search. */ + + const unsigned char *cp = (const unsigned char *) longword_ptr; + + if (8 < sizeof longword) + for (i = sizeof longword - 1; 8 <= i; i--) + if (cp[i] == c) + return (void *) &cp[i]; + if (7 < sizeof longword && cp[7] == c) + return (void *) &cp[7]; + if (6 < sizeof longword && cp[6] == c) + return (void *) &cp[6]; + if (5 < sizeof longword && cp[5] == c) + return (void *) &cp[5]; + if (4 < sizeof longword && cp[4] == c) + return (void *) &cp[4]; + if (cp[3] == c) + return (void *) &cp[3]; + if (cp[2] == c) + return (void *) &cp[2]; + if (cp[1] == c) + return (void *) &cp[1]; + if (cp[0] == c) + return (void *) cp; + } + + n -= sizeof longword; + } + + char_ptr = (const unsigned char *) longword_ptr; + + while (n-- > 0) + { + if (*--char_ptr == c) + return (void *) char_ptr; + } + + return 0; +} +#ifdef weak_alias +weak_alias (__memrchr, memrchr) +#endif diff --git a/lib/memset.c b/lib/memset.c new file mode 100644 index 0000000..890cbf1 --- /dev/null +++ b/lib/memset.c @@ -0,0 +1,28 @@ +/* memset.c -- set an area of memory to a given value + Copyright (C) 1991, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +void * +memset (void *str, int c, size_t len) +{ + register char *st = str; + + while (len-- > 0) + *st++ = c; + return str; +} diff --git a/lib/mkdirat.c b/lib/mkdirat.c new file mode 100644 index 0000000..da0b262 --- /dev/null +++ b/lib/mkdirat.c @@ -0,0 +1,43 @@ +/* fd-relative mkdir + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +#include "openat.h" + +#include <unistd.h> + +#include "dirname.h" /* solely for definition of IS_ABSOLUTE_FILE_NAME */ +#include "save-cwd.h" +#include "openat-priv.h" + +/* Solaris 10 has no function like this. + Create a subdirectory, FILE, with mode MODE, in the directory + open on descriptor FD. If possible, do it without changing the + working directory. Otherwise, resort to using save_cwd/fchdir, + then mkdir/restore_cwd. If either the save_cwd or the restore_cwd + fails, then give a diagnostic and exit nonzero. */ + +#define AT_FUNC_NAME mkdirat +#define AT_FUNC_F1 mkdir +#define AT_FUNC_F2 mkdir +#define AT_FUNC_USE_F1_COND 1 +#define AT_FUNC_POST_FILE_PARAM_DECLS , mode_t mode +#define AT_FUNC_POST_FILE_ARGS , mode +#include "at-func.c" diff --git a/lib/mkdtemp.c b/lib/mkdtemp.c new file mode 100644 index 0000000..36b6c75 --- /dev/null +++ b/lib/mkdtemp.c @@ -0,0 +1,39 @@ +/* Copyright (C) 1999, 2001-2003, 2006-2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Extracted from misc/mkdtemp.c. */ + +#include <config.h> + +/* Specification. */ +#include <stdlib.h> + +#include "tempname.h" + +/* Generate a unique temporary directory from TEMPLATE. + The last six characters of TEMPLATE must be "XXXXXX"; + they are replaced with a string that makes the filename unique. + The directory is created, mode 700, and its name is returned. + (This function comes from OpenBSD.) */ +char * +mkdtemp (char *template) +{ + if (gen_tempname (template, GT_DIR)) + return NULL; + else + return template; +} diff --git a/lib/mktime.c b/lib/mktime.c new file mode 100644 index 0000000..a91fb20 --- /dev/null +++ b/lib/mktime.c @@ -0,0 +1,663 @@ +/* Convert a `struct tm' to a time_t value. + Copyright (C) 1993-1999, 2002-2005, 2006, 2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Paul Eggert <eggert@twinsun.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Define this to have a standalone program to test this implementation of + mktime. */ +/* #define DEBUG 1 */ + +#ifndef _LIBC +# include <config.h> +#endif + +/* Assume that leap seconds are possible, unless told otherwise. + If the host has a `zic' command with a `-L leapsecondfilename' option, + then it supports leap seconds; otherwise it probably doesn't. */ +#ifndef LEAP_SECONDS_POSSIBLE +# define LEAP_SECONDS_POSSIBLE 1 +#endif + +#include <time.h> + +#include <limits.h> + +#include <string.h> /* For the real memcpy prototype. */ + +#if DEBUG +# include <stdio.h> +# include <stdlib.h> +/* Make it work even if the system's libc has its own mktime routine. */ +# define mktime my_mktime +#endif /* DEBUG */ + +/* Shift A right by B bits portably, by dividing A by 2**B and + truncating towards minus infinity. A and B should be free of side + effects, and B should be in the range 0 <= B <= INT_BITS - 2, where + INT_BITS is the number of useful bits in an int. GNU code can + assume that INT_BITS is at least 32. + + ISO C99 says that A >> B is implementation-defined if A < 0. Some + implementations (e.g., UNICOS 9.0 on a Cray Y-MP EL) don't shift + right in the usual way when A < 0, so SHR falls back on division if + ordinary A >> B doesn't seem to be the usual signed shift. */ +#define SHR(a, b) \ + (-1 >> 1 == -1 \ + ? (a) >> (b) \ + : (a) / (1 << (b)) - ((a) % (1 << (b)) < 0)) + +/* The extra casts in the following macros work around compiler bugs, + e.g., in Cray C 5.0.3.0. */ + +/* True if the arithmetic type T is an integer type. bool counts as + an integer. */ +#define TYPE_IS_INTEGER(t) ((t) 1.5 == 1) + +/* True if negative values of the signed integer type T use two's + complement, ones' complement, or signed magnitude representation, + respectively. Much GNU code assumes two's complement, but some + people like to be portable to all possible C hosts. */ +#define TYPE_TWOS_COMPLEMENT(t) ((t) ~ (t) 0 == (t) -1) +#define TYPE_ONES_COMPLEMENT(t) ((t) ~ (t) 0 == 0) +#define TYPE_SIGNED_MAGNITUDE(t) ((t) ~ (t) 0 < (t) -1) + +/* True if the arithmetic type T is signed. */ +#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) + +/* The maximum and minimum values for the integer type T. These + macros have undefined behavior if T is signed and has padding bits. + If this is a problem for you, please let us know how to fix it for + your host. */ +#define TYPE_MINIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) 0 \ + : TYPE_SIGNED_MAGNITUDE (t) \ + ? ~ (t) 0 \ + : ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1))) +#define TYPE_MAXIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) -1 \ + : ~ (~ (t) 0 << (sizeof (t) * CHAR_BIT - 1)))) + +#ifndef TIME_T_MIN +# define TIME_T_MIN TYPE_MINIMUM (time_t) +#endif +#ifndef TIME_T_MAX +# define TIME_T_MAX TYPE_MAXIMUM (time_t) +#endif +#define TIME_T_MIDPOINT (SHR (TIME_T_MIN + TIME_T_MAX, 1) + 1) + +/* Verify a requirement at compile-time (unlike assert, which is runtime). */ +#define verify(name, assertion) struct name { char a[(assertion) ? 1 : -1]; } + +verify (time_t_is_integer, TYPE_IS_INTEGER (time_t)); +verify (twos_complement_arithmetic, TYPE_TWOS_COMPLEMENT (int)); +/* The code also assumes that signed integer overflow silently wraps + around, but this assumption can't be stated without causing a + diagnostic on some hosts. */ + +#define EPOCH_YEAR 1970 +#define TM_YEAR_BASE 1900 +verify (base_year_is_a_multiple_of_100, TM_YEAR_BASE % 100 == 0); + +/* Return 1 if YEAR + TM_YEAR_BASE is a leap year. */ +static inline int +leapyear (long int year) +{ + /* Don't add YEAR to TM_YEAR_BASE, as that might overflow. + Also, work even if YEAR is negative. */ + return + ((year & 3) == 0 + && (year % 100 != 0 + || ((year / 100) & 3) == (- (TM_YEAR_BASE / 100) & 3))); +} + +/* How many days come before each month (0-12). */ +#ifndef _LIBC +static +#endif +const unsigned short int __mon_yday[2][13] = + { + /* Normal years. */ + { 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365 }, + /* Leap years. */ + { 0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366 } + }; + + +#ifndef _LIBC +/* Portable standalone applications should supply a <time.h> that + declares a POSIX-compliant localtime_r, for the benefit of older + implementations that lack localtime_r or have a nonstandard one. + See the gnulib time_r module for one way to implement this. */ +# undef __localtime_r +# define __localtime_r localtime_r +# define __mktime_internal mktime_internal +#endif + +/* Return an integer value measuring (YEAR1-YDAY1 HOUR1:MIN1:SEC1) - + (YEAR0-YDAY0 HOUR0:MIN0:SEC0) in seconds, assuming that the clocks + were not adjusted between the time stamps. + + The YEAR values uses the same numbering as TP->tm_year. Values + need not be in the usual range. However, YEAR1 must not be less + than 2 * INT_MIN or greater than 2 * INT_MAX. + + The result may overflow. It is the caller's responsibility to + detect overflow. */ + +static inline time_t +ydhms_diff (long int year1, long int yday1, int hour1, int min1, int sec1, + int year0, int yday0, int hour0, int min0, int sec0) +{ + verify (C99_integer_division, -1 / 2 == 0); + verify (long_int_year_and_yday_are_wide_enough, + INT_MAX <= LONG_MAX / 2 || TIME_T_MAX <= UINT_MAX); + + /* Compute intervening leap days correctly even if year is negative. + Take care to avoid integer overflow here. */ + int a4 = SHR (year1, 2) + SHR (TM_YEAR_BASE, 2) - ! (year1 & 3); + int b4 = SHR (year0, 2) + SHR (TM_YEAR_BASE, 2) - ! (year0 & 3); + int a100 = a4 / 25 - (a4 % 25 < 0); + int b100 = b4 / 25 - (b4 % 25 < 0); + int a400 = SHR (a100, 2); + int b400 = SHR (b100, 2); + int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400); + + /* Compute the desired time in time_t precision. Overflow might + occur here. */ + time_t tyear1 = year1; + time_t years = tyear1 - year0; + time_t days = 365 * years + yday1 - yday0 + intervening_leap_days; + time_t hours = 24 * days + hour1 - hour0; + time_t minutes = 60 * hours + min1 - min0; + time_t seconds = 60 * minutes + sec1 - sec0; + return seconds; +} + + +/* Return a time_t value corresponding to (YEAR-YDAY HOUR:MIN:SEC), + assuming that *T corresponds to *TP and that no clock adjustments + occurred between *TP and the desired time. + If TP is null, return a value not equal to *T; this avoids false matches. + If overflow occurs, yield the minimal or maximal value, except do not + yield a value equal to *T. */ +static time_t +guess_time_tm (long int year, long int yday, int hour, int min, int sec, + const time_t *t, const struct tm *tp) +{ + if (tp) + { + time_t d = ydhms_diff (year, yday, hour, min, sec, + tp->tm_year, tp->tm_yday, + tp->tm_hour, tp->tm_min, tp->tm_sec); + time_t t1 = *t + d; + if ((t1 < *t) == (TYPE_SIGNED (time_t) ? d < 0 : TIME_T_MAX / 2 < d)) + return t1; + } + + /* Overflow occurred one way or another. Return the nearest result + that is actually in range, except don't report a zero difference + if the actual difference is nonzero, as that would cause a false + match; and don't oscillate between two values, as that would + confuse the spring-forward gap detector. */ + return (*t < TIME_T_MIDPOINT + ? (*t <= TIME_T_MIN + 1 ? *t + 1 : TIME_T_MIN) + : (TIME_T_MAX - 1 <= *t ? *t - 1 : TIME_T_MAX)); +} + +/* Use CONVERT to convert *T to a broken down time in *TP. + If *T is out of range for conversion, adjust it so that + it is the nearest in-range value and then convert that. */ +static struct tm * +ranged_convert (struct tm *(*convert) (const time_t *, struct tm *), + time_t *t, struct tm *tp) +{ + struct tm *r = convert (t, tp); + + if (!r && *t) + { + time_t bad = *t; + time_t ok = 0; + + /* BAD is a known unconvertible time_t, and OK is a known good one. + Use binary search to narrow the range between BAD and OK until + they differ by 1. */ + while (bad != ok + (bad < 0 ? -1 : 1)) + { + time_t mid = *t = (bad < 0 + ? bad + ((ok - bad) >> 1) + : ok + ((bad - ok) >> 1)); + r = convert (t, tp); + if (r) + ok = mid; + else + bad = mid; + } + + if (!r && ok) + { + /* The last conversion attempt failed; + revert to the most recent successful attempt. */ + *t = ok; + r = convert (t, tp); + } + } + + return r; +} + + +/* Convert *TP to a time_t value, inverting + the monotonic and mostly-unit-linear conversion function CONVERT. + Use *OFFSET to keep track of a guess at the offset of the result, + compared to what the result would be for UTC without leap seconds. + If *OFFSET's guess is correct, only one CONVERT call is needed. + This function is external because it is used also by timegm.c. */ +time_t +__mktime_internal (struct tm *tp, + struct tm *(*convert) (const time_t *, struct tm *), + time_t *offset) +{ + time_t t, gt, t0, t1, t2; + struct tm tm; + + /* The maximum number of probes (calls to CONVERT) should be enough + to handle any combinations of time zone rule changes, solar time, + leap seconds, and oscillations around a spring-forward gap. + POSIX.1 prohibits leap seconds, but some hosts have them anyway. */ + int remaining_probes = 6; + + /* Time requested. Copy it in case CONVERT modifies *TP; this can + occur if TP is localtime's returned value and CONVERT is localtime. */ + int sec = tp->tm_sec; + int min = tp->tm_min; + int hour = tp->tm_hour; + int mday = tp->tm_mday; + int mon = tp->tm_mon; + int year_requested = tp->tm_year; + int isdst = tp->tm_isdst; + + /* 1 if the previous probe was DST. */ + int dst2; + + /* Ensure that mon is in range, and set year accordingly. */ + int mon_remainder = mon % 12; + int negative_mon_remainder = mon_remainder < 0; + int mon_years = mon / 12 - negative_mon_remainder; + long int lyear_requested = year_requested; + long int year = lyear_requested + mon_years; + + /* The other values need not be in range: + the remaining code handles minor overflows correctly, + assuming int and time_t arithmetic wraps around. + Major overflows are caught at the end. */ + + /* Calculate day of year from year, month, and day of month. + The result need not be in range. */ + int mon_yday = ((__mon_yday[leapyear (year)] + [mon_remainder + 12 * negative_mon_remainder]) + - 1); + long int lmday = mday; + long int yday = mon_yday + lmday; + + time_t guessed_offset = *offset; + + int sec_requested = sec; + + if (LEAP_SECONDS_POSSIBLE) + { + /* Handle out-of-range seconds specially, + since ydhms_tm_diff assumes every minute has 60 seconds. */ + if (sec < 0) + sec = 0; + if (59 < sec) + sec = 59; + } + + /* Invert CONVERT by probing. First assume the same offset as last + time. */ + + t0 = ydhms_diff (year, yday, hour, min, sec, + EPOCH_YEAR - TM_YEAR_BASE, 0, 0, 0, - guessed_offset); + + if (TIME_T_MAX / INT_MAX / 366 / 24 / 60 / 60 < 3) + { + /* time_t isn't large enough to rule out overflows, so check + for major overflows. A gross check suffices, since if t0 + has overflowed, it is off by a multiple of TIME_T_MAX - + TIME_T_MIN + 1. So ignore any component of the difference + that is bounded by a small value. */ + + /* Approximate log base 2 of the number of time units per + biennium. A biennium is 2 years; use this unit instead of + years to avoid integer overflow. For example, 2 average + Gregorian years are 2 * 365.2425 * 24 * 60 * 60 seconds, + which is 63113904 seconds, and rint (log2 (63113904)) is + 26. */ + int ALOG2_SECONDS_PER_BIENNIUM = 26; + int ALOG2_MINUTES_PER_BIENNIUM = 20; + int ALOG2_HOURS_PER_BIENNIUM = 14; + int ALOG2_DAYS_PER_BIENNIUM = 10; + int LOG2_YEARS_PER_BIENNIUM = 1; + + int approx_requested_biennia = + (SHR (year_requested, LOG2_YEARS_PER_BIENNIUM) + - SHR (EPOCH_YEAR - TM_YEAR_BASE, LOG2_YEARS_PER_BIENNIUM) + + SHR (mday, ALOG2_DAYS_PER_BIENNIUM) + + SHR (hour, ALOG2_HOURS_PER_BIENNIUM) + + SHR (min, ALOG2_MINUTES_PER_BIENNIUM) + + (LEAP_SECONDS_POSSIBLE + ? 0 + : SHR (sec, ALOG2_SECONDS_PER_BIENNIUM))); + + int approx_biennia = SHR (t0, ALOG2_SECONDS_PER_BIENNIUM); + int diff = approx_biennia - approx_requested_biennia; + int abs_diff = diff < 0 ? - diff : diff; + + /* IRIX 4.0.5 cc miscaculates TIME_T_MIN / 3: it erroneously + gives a positive value of 715827882. Setting a variable + first then doing math on it seems to work. + (ghazi@caip.rutgers.edu) */ + time_t time_t_max = TIME_T_MAX; + time_t time_t_min = TIME_T_MIN; + time_t overflow_threshold = + (time_t_max / 3 - time_t_min / 3) >> ALOG2_SECONDS_PER_BIENNIUM; + + if (overflow_threshold < abs_diff) + { + /* Overflow occurred. Try repairing it; this might work if + the time zone offset is enough to undo the overflow. */ + time_t repaired_t0 = -1 - t0; + approx_biennia = SHR (repaired_t0, ALOG2_SECONDS_PER_BIENNIUM); + diff = approx_biennia - approx_requested_biennia; + abs_diff = diff < 0 ? - diff : diff; + if (overflow_threshold < abs_diff) + return -1; + guessed_offset += repaired_t0 - t0; + t0 = repaired_t0; + } + } + + /* Repeatedly use the error to improve the guess. */ + + for (t = t1 = t2 = t0, dst2 = 0; + (gt = guess_time_tm (year, yday, hour, min, sec, &t, + ranged_convert (convert, &t, &tm)), + t != gt); + t1 = t2, t2 = t, t = gt, dst2 = tm.tm_isdst != 0) + if (t == t1 && t != t2 + && (tm.tm_isdst < 0 + || (isdst < 0 + ? dst2 <= (tm.tm_isdst != 0) + : (isdst != 0) != (tm.tm_isdst != 0)))) + /* We can't possibly find a match, as we are oscillating + between two values. The requested time probably falls + within a spring-forward gap of size GT - T. Follow the common + practice in this case, which is to return a time that is GT - T + away from the requested time, preferring a time whose + tm_isdst differs from the requested value. (If no tm_isdst + was requested and only one of the two values has a nonzero + tm_isdst, prefer that value.) In practice, this is more + useful than returning -1. */ + goto offset_found; + else if (--remaining_probes == 0) + return -1; + + /* We have a match. Check whether tm.tm_isdst has the requested + value, if any. */ + if (isdst != tm.tm_isdst && 0 <= isdst && 0 <= tm.tm_isdst) + { + /* tm.tm_isdst has the wrong value. Look for a neighboring + time with the right value, and use its UTC offset. + + Heuristic: probe the adjacent timestamps in both directions, + looking for the desired isdst. This should work for all real + time zone histories in the tz database. */ + + /* Distance between probes when looking for a DST boundary. In + tzdata2003a, the shortest period of DST is 601200 seconds + (e.g., America/Recife starting 2000-10-08 01:00), and the + shortest period of non-DST surrounded by DST is 694800 + seconds (Africa/Tunis starting 1943-04-17 01:00). Use the + minimum of these two values, so we don't miss these short + periods when probing. */ + int stride = 601200; + + /* The longest period of DST in tzdata2003a is 536454000 seconds + (e.g., America/Jujuy starting 1946-10-01 01:00). The longest + period of non-DST is much longer, but it makes no real sense + to search for more than a year of non-DST, so use the DST + max. */ + int duration_max = 536454000; + + /* Search in both directions, so the maximum distance is half + the duration; add the stride to avoid off-by-1 problems. */ + int delta_bound = duration_max / 2 + stride; + + int delta, direction; + + for (delta = stride; delta < delta_bound; delta += stride) + for (direction = -1; direction <= 1; direction += 2) + { + time_t ot = t + delta * direction; + if ((ot < t) == (direction < 0)) + { + struct tm otm; + ranged_convert (convert, &ot, &otm); + if (otm.tm_isdst == isdst) + { + /* We found the desired tm_isdst. + Extrapolate back to the desired time. */ + t = guess_time_tm (year, yday, hour, min, sec, &ot, &otm); + ranged_convert (convert, &t, &tm); + goto offset_found; + } + } + } + } + + offset_found: + *offset = guessed_offset + t - t0; + + if (LEAP_SECONDS_POSSIBLE && sec_requested != tm.tm_sec) + { + /* Adjust time to reflect the tm_sec requested, not the normalized value. + Also, repair any damage from a false match due to a leap second. */ + int sec_adjustment = (sec == 0 && tm.tm_sec == 60) - sec; + t1 = t + sec_requested; + t2 = t1 + sec_adjustment; + if (((t1 < t) != (sec_requested < 0)) + | ((t2 < t1) != (sec_adjustment < 0)) + | ! convert (&t2, &tm)) + return -1; + t = t2; + } + + *tp = tm; + return t; +} + + +/* FIXME: This should use a signed type wide enough to hold any UTC + offset in seconds. 'int' should be good enough for GNU code. We + can't fix this unilaterally though, as other modules invoke + __mktime_internal. */ +static time_t localtime_offset; + +/* Convert *TP to a time_t value. */ +time_t +mktime (struct tm *tp) +{ +#ifdef _LIBC + /* POSIX.1 8.1.1 requires that whenever mktime() is called, the + time zone names contained in the external variable `tzname' shall + be set as if the tzset() function had been called. */ + __tzset (); +#endif + + return __mktime_internal (tp, __localtime_r, &localtime_offset); +} + +#ifdef weak_alias +weak_alias (mktime, timelocal) +#endif + +#ifdef _LIBC +libc_hidden_def (mktime) +libc_hidden_weak (timelocal) +#endif + +#if DEBUG + +static int +not_equal_tm (const struct tm *a, const struct tm *b) +{ + return ((a->tm_sec ^ b->tm_sec) + | (a->tm_min ^ b->tm_min) + | (a->tm_hour ^ b->tm_hour) + | (a->tm_mday ^ b->tm_mday) + | (a->tm_mon ^ b->tm_mon) + | (a->tm_year ^ b->tm_year) + | (a->tm_yday ^ b->tm_yday) + | (a->tm_isdst ^ b->tm_isdst)); +} + +static void +print_tm (const struct tm *tp) +{ + if (tp) + printf ("%04d-%02d-%02d %02d:%02d:%02d yday %03d wday %d isdst %d", + tp->tm_year + TM_YEAR_BASE, tp->tm_mon + 1, tp->tm_mday, + tp->tm_hour, tp->tm_min, tp->tm_sec, + tp->tm_yday, tp->tm_wday, tp->tm_isdst); + else + printf ("0"); +} + +static int +check_result (time_t tk, struct tm tmk, time_t tl, const struct tm *lt) +{ + if (tk != tl || !lt || not_equal_tm (&tmk, lt)) + { + printf ("mktime ("); + print_tm (lt); + printf (")\nyields ("); + print_tm (&tmk); + printf (") == %ld, should be %ld\n", (long int) tk, (long int) tl); + return 1; + } + + return 0; +} + +int +main (int argc, char **argv) +{ + int status = 0; + struct tm tm, tmk, tml; + struct tm *lt; + time_t tk, tl, tl1; + char trailer; + + if ((argc == 3 || argc == 4) + && (sscanf (argv[1], "%d-%d-%d%c", + &tm.tm_year, &tm.tm_mon, &tm.tm_mday, &trailer) + == 3) + && (sscanf (argv[2], "%d:%d:%d%c", + &tm.tm_hour, &tm.tm_min, &tm.tm_sec, &trailer) + == 3)) + { + tm.tm_year -= TM_YEAR_BASE; + tm.tm_mon--; + tm.tm_isdst = argc == 3 ? -1 : atoi (argv[3]); + tmk = tm; + tl = mktime (&tmk); + lt = localtime (&tl); + if (lt) + { + tml = *lt; + lt = &tml; + } + printf ("mktime returns %ld == ", (long int) tl); + print_tm (&tmk); + printf ("\n"); + status = check_result (tl, tmk, tl, lt); + } + else if (argc == 4 || (argc == 5 && strcmp (argv[4], "-") == 0)) + { + time_t from = atol (argv[1]); + time_t by = atol (argv[2]); + time_t to = atol (argv[3]); + + if (argc == 4) + for (tl = from; by < 0 ? to <= tl : tl <= to; tl = tl1) + { + lt = localtime (&tl); + if (lt) + { + tmk = tml = *lt; + tk = mktime (&tmk); + status |= check_result (tk, tmk, tl, &tml); + } + else + { + printf ("localtime (%ld) yields 0\n", (long int) tl); + status = 1; + } + tl1 = tl + by; + if ((tl1 < tl) != (by < 0)) + break; + } + else + for (tl = from; by < 0 ? to <= tl : tl <= to; tl = tl1) + { + /* Null benchmark. */ + lt = localtime (&tl); + if (lt) + { + tmk = tml = *lt; + tk = tl; + status |= check_result (tk, tmk, tl, &tml); + } + else + { + printf ("localtime (%ld) yields 0\n", (long int) tl); + status = 1; + } + tl1 = tl + by; + if ((tl1 < tl) != (by < 0)) + break; + } + } + else + printf ("Usage:\ +\t%s YYYY-MM-DD HH:MM:SS [ISDST] # Test given time.\n\ +\t%s FROM BY TO # Test values FROM, FROM+BY, ..., TO.\n\ +\t%s FROM BY TO - # Do not test those values (for benchmark).\n", + argv[0], argv[0], argv[0]); + + return status; +} + +#endif /* DEBUG */ + +/* +Local Variables: +compile-command: "gcc -DDEBUG -Wall -W -O -g mktime.c -o mktime" +End: +*/ diff --git a/lib/modechange.c b/lib/modechange.c new file mode 100644 index 0000000..e1f7ceb --- /dev/null +++ b/lib/modechange.c @@ -0,0 +1,386 @@ +/* modechange.c -- file mode manipulation + + Copyright (C) 1989, 1990, 1997, 1998, 1999, 2001, 2003, 2004, 2005, + 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@ai.mit.edu> */ + +/* The ASCII mode string is compiled into an array of `struct + modechange', which can then be applied to each file to be changed. + We do this instead of re-parsing the ASCII string for each file + because the compiled form requires less computation to use; when + changing the mode of many files, this probably results in a + performance gain. */ + +#include <config.h> + +#include "modechange.h" +#include <sys/stat.h> +#include "stat-macros.h" +#include "xalloc.h" +#include <stdlib.h> + +/* The traditional octal values corresponding to each mode bit. */ +#define SUID 04000 +#define SGID 02000 +#define SVTX 01000 +#define RUSR 00400 +#define WUSR 00200 +#define XUSR 00100 +#define RGRP 00040 +#define WGRP 00020 +#define XGRP 00010 +#define ROTH 00004 +#define WOTH 00002 +#define XOTH 00001 +#define ALLM 07777 /* all octal mode bits */ + +/* Convert OCTAL, which uses one of the traditional octal values, to + an internal mode_t value. */ +static mode_t +octal_to_mode (unsigned int octal) +{ + /* Help the compiler optimize the usual case where mode_t uses + the traditional octal representation. */ + return ((S_ISUID == SUID && S_ISGID == SGID && S_ISVTX == SVTX + && S_IRUSR == RUSR && S_IWUSR == WUSR && S_IXUSR == XUSR + && S_IRGRP == RGRP && S_IWGRP == WGRP && S_IXGRP == XGRP + && S_IROTH == ROTH && S_IWOTH == WOTH && S_IXOTH == XOTH) + ? octal + : (mode_t) ((octal & SUID ? S_ISUID : 0) + | (octal & SGID ? S_ISGID : 0) + | (octal & SVTX ? S_ISVTX : 0) + | (octal & RUSR ? S_IRUSR : 0) + | (octal & WUSR ? S_IWUSR : 0) + | (octal & XUSR ? S_IXUSR : 0) + | (octal & RGRP ? S_IRGRP : 0) + | (octal & WGRP ? S_IWGRP : 0) + | (octal & XGRP ? S_IXGRP : 0) + | (octal & ROTH ? S_IROTH : 0) + | (octal & WOTH ? S_IWOTH : 0) + | (octal & XOTH ? S_IXOTH : 0))); +} + +/* Special operations flags. */ +enum + { + /* For the sentinel at the end of the mode changes array. */ + MODE_DONE, + + /* The typical case. */ + MODE_ORDINARY_CHANGE, + + /* In addition to the typical case, affect the execute bits if at + least one execute bit is set already, or if the file is a + directory. */ + MODE_X_IF_ANY_X, + + /* Instead of the typical case, copy some existing permissions for + u, g, or o onto the other two. Which of u, g, or o is copied + is determined by which bits are set in the `value' field. */ + MODE_COPY_EXISTING + }; + +/* Description of a mode change. */ +struct mode_change +{ + char op; /* One of "=+-". */ + char flag; /* Special operations flag. */ + mode_t affected; /* Set for u, g, o, or a. */ + mode_t value; /* Bits to add/remove. */ + mode_t mentioned; /* Bits explicitly mentioned. */ +}; + +/* Return a mode_change array with the specified `=ddd'-style + mode change operation, where NEW_MODE is `ddd' and MENTIONED + contains the bits explicitly mentioned in the mode are MENTIONED. */ + +static struct mode_change * +make_node_op_equals (mode_t new_mode, mode_t mentioned) +{ + struct mode_change *p = xmalloc (2 * sizeof *p); + p->op = '='; + p->flag = MODE_ORDINARY_CHANGE; + p->affected = CHMOD_MODE_BITS; + p->value = new_mode; + p->mentioned = mentioned; + p[1].flag = MODE_DONE; + return p; +} + +/* Return a pointer to an array of file mode change operations created from + MODE_STRING, an ASCII string that contains either an octal number + specifying an absolute mode, or symbolic mode change operations with + the form: + [ugoa...][[+-=][rwxXstugo...]...][,...] + + Return NULL if `mode_string' does not contain a valid + representation of file mode change operations. */ + +struct mode_change * +mode_compile (char const *mode_string) +{ + /* The array of mode-change directives to be returned. */ + struct mode_change *mc; + size_t used = 0; + + if ('0' <= *mode_string && *mode_string < '8') + { + unsigned int octal_mode = 0; + mode_t mode; + mode_t mentioned; + + do + { + octal_mode = 8 * octal_mode + *mode_string++ - '0'; + if (ALLM < octal_mode) + return NULL; + } + while ('0' <= *mode_string && *mode_string < '8'); + + if (*mode_string) + return NULL; + + mode = octal_to_mode (octal_mode); + mentioned = (mode & (S_ISUID | S_ISGID)) | S_ISVTX | S_IRWXUGO; + return make_node_op_equals (mode, mentioned); + } + + /* Allocate enough space to hold the result. */ + { + size_t needed = 1; + char const *p; + for (p = mode_string; *p; p++) + needed += (*p == '=' || *p == '+' || *p == '-'); + mc = xnmalloc (needed, sizeof *mc); + } + + /* One loop iteration for each `[ugoa]*([-+=]([rwxXst]*|[ugo]))+'. */ + for (;; mode_string++) + { + /* Which bits in the mode are operated on. */ + mode_t affected = 0; + + /* Turn on all the bits in `affected' for each group given. */ + for (;; mode_string++) + switch (*mode_string) + { + default: + goto invalid; + case 'u': + affected |= S_ISUID | S_IRWXU; + break; + case 'g': + affected |= S_ISGID | S_IRWXG; + break; + case 'o': + affected |= S_ISVTX | S_IRWXO; + break; + case 'a': + affected |= CHMOD_MODE_BITS; + break; + case '=': case '+': case '-': + goto no_more_affected; + } + no_more_affected:; + + do + { + char op = *mode_string++; + mode_t value; + char flag = MODE_COPY_EXISTING; + struct mode_change *change; + + switch (*mode_string++) + { + case 'u': + /* Set the affected bits to the value of the `u' bits + on the same file. */ + value = S_IRWXU; + break; + case 'g': + /* Set the affected bits to the value of the `g' bits + on the same file. */ + value = S_IRWXG; + break; + case 'o': + /* Set the affected bits to the value of the `o' bits + on the same file. */ + value = S_IRWXO; + break; + + default: + value = 0; + flag = MODE_ORDINARY_CHANGE; + + for (mode_string--;; mode_string++) + switch (*mode_string) + { + case 'r': + value |= S_IRUSR | S_IRGRP | S_IROTH; + break; + case 'w': + value |= S_IWUSR | S_IWGRP | S_IWOTH; + break; + case 'x': + value |= S_IXUSR | S_IXGRP | S_IXOTH; + break; + case 'X': + flag = MODE_X_IF_ANY_X; + break; + case 's': + /* Set the setuid/gid bits if `u' or `g' is selected. */ + value |= S_ISUID | S_ISGID; + break; + case 't': + /* Set the "save text image" bit if `o' is selected. */ + value |= S_ISVTX; + break; + default: + goto no_more_values; + } + no_more_values:; + } + + change = &mc[used++]; + change->op = op; + change->flag = flag; + change->affected = affected; + change->value = value; + change->mentioned = (affected ? affected & value : value); + } + while (*mode_string == '=' || *mode_string == '+' + || *mode_string == '-'); + + if (*mode_string != ',') + break; + } + + if (*mode_string == 0) + { + mc[used].flag = MODE_DONE; + return mc; + } + +invalid: + free (mc); + return NULL; +} + +/* Return a file mode change operation that sets permissions to match those + of REF_FILE. Return NULL (setting errno) if REF_FILE can't be accessed. */ + +struct mode_change * +mode_create_from_ref (const char *ref_file) +{ + struct stat ref_stats; + + if (stat (ref_file, &ref_stats) != 0) + return NULL; + return make_node_op_equals (ref_stats.st_mode, CHMOD_MODE_BITS); +} + +/* Return the file mode bits of OLDMODE (which is the mode of a + directory if DIR), assuming the umask is UMASK_VALUE, adjusted as + indicated by the list of change operations CHANGES. If DIR, the + type 'X' change affects the returned value even if no execute bits + were set in OLDMODE, and set user and group ID bits are preserved + unless CHANGES mentioned them. If PMODE_BITS is not null, store into + *PMODE_BITS a mask denoting file mode bits that are affected by + CHANGES. + + The returned value and *PMODE_BITS contain only file mode bits. + For example, they have the S_IFMT bits cleared on a standard + Unix-like host. */ + +mode_t +mode_adjust (mode_t oldmode, bool dir, mode_t umask_value, + struct mode_change const *changes, mode_t *pmode_bits) +{ + /* The adjusted mode. */ + mode_t newmode = oldmode & CHMOD_MODE_BITS; + + /* File mode bits that CHANGES cares about. */ + mode_t mode_bits = 0; + + for (; changes->flag != MODE_DONE; changes++) + { + mode_t affected = changes->affected; + mode_t omit_change = + (dir ? S_ISUID | S_ISGID : 0) & ~ changes->mentioned; + mode_t value = changes->value; + + switch (changes->flag) + { + case MODE_ORDINARY_CHANGE: + break; + + case MODE_COPY_EXISTING: + /* Isolate in `value' the bits in `newmode' to copy. */ + value &= newmode; + + /* Copy the isolated bits to the other two parts. */ + value |= ((value & (S_IRUSR | S_IRGRP | S_IROTH) + ? S_IRUSR | S_IRGRP | S_IROTH : 0) + | (value & (S_IWUSR | S_IWGRP | S_IWOTH) + ? S_IWUSR | S_IWGRP | S_IWOTH : 0) + | (value & (S_IXUSR | S_IXGRP | S_IXOTH) + ? S_IXUSR | S_IXGRP | S_IXOTH : 0)); + break; + + case MODE_X_IF_ANY_X: + /* Affect the execute bits if execute bits are already set + or if the file is a directory. */ + if ((newmode & (S_IXUSR | S_IXGRP | S_IXOTH)) | dir) + value |= S_IXUSR | S_IXGRP | S_IXOTH; + break; + } + + /* If WHO was specified, limit the change to the affected bits. + Otherwise, apply the umask. Either way, omit changes as + requested. */ + value &= (affected ? affected : ~umask_value) & ~ omit_change; + + switch (changes->op) + { + case '=': + /* If WHO was specified, preserve the previous values of + bits that are not affected by this change operation. + Otherwise, clear all the bits. */ + { + mode_t preserved = (affected ? ~affected : 0) | omit_change; + mode_bits |= CHMOD_MODE_BITS & ~preserved; + newmode = (newmode & preserved) | value; + break; + } + + case '+': + mode_bits |= value; + newmode |= value; + break; + + case '-': + mode_bits |= value; + newmode &= ~value; + break; + } + } + + if (pmode_bits) + *pmode_bits = mode_bits; + return newmode; +} diff --git a/lib/modechange.h b/lib/modechange.h new file mode 100644 index 0000000..76e0178 --- /dev/null +++ b/lib/modechange.h @@ -0,0 +1,31 @@ +/* modechange.h -- definitions for file mode manipulation + + Copyright (C) 1989, 1990, 1997, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if ! defined MODECHANGE_H_ +# define MODECHANGE_H_ + +# include <stdbool.h> +# include <sys/types.h> + +struct mode_change *mode_compile (const char *); +struct mode_change *mode_create_from_ref (const char *); +mode_t mode_adjust (mode_t, bool, mode_t, struct mode_change const *, + mode_t *); + +#endif diff --git a/lib/obstack.c b/lib/obstack.c new file mode 100644 index 0000000..5cd0b7a --- /dev/null +++ b/lib/obstack.c @@ -0,0 +1,431 @@ +/* obstack.c - subroutines used implicitly by object stack macros + + Copyright (C) 1988, 1989, 1990, 1991, 1992, 1993, 1994, 1996, 1997, + 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef _LIBC +# include <obstack.h> +# include <shlib-compat.h> +#else +# include <config.h> +# include "obstack.h" +#endif + +/* NOTE BEFORE MODIFYING THIS FILE: This version number must be + incremented whenever callers compiled using an old obstack.h can no + longer properly call the functions in this obstack.c. */ +#define OBSTACK_INTERFACE_VERSION 1 + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself, and the installed library + supports the same library interface we do. This code is part of the GNU + C Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object + files, it is simpler to just do this in the source for each such file. */ + +#include <stdio.h> /* Random thing to get __GNU_LIBRARY__. */ +#if !defined _LIBC && defined __GNU_LIBRARY__ && __GNU_LIBRARY__ > 1 +# include <gnu-versions.h> +# if _GNU_OBSTACK_INTERFACE_VERSION == OBSTACK_INTERFACE_VERSION +# define ELIDE_CODE +# endif +#endif + +#include <stddef.h> + +#ifndef ELIDE_CODE + +# include <stdint.h> + +/* Determine default alignment. */ +union fooround +{ + uintmax_t i; + long double d; + void *p; +}; +struct fooalign +{ + char c; + union fooround u; +}; +/* If malloc were really smart, it would round addresses to DEFAULT_ALIGNMENT. + But in fact it might be less smart and round addresses to as much as + DEFAULT_ROUNDING. So we prepare for it to do that. */ +enum + { + DEFAULT_ALIGNMENT = offsetof (struct fooalign, u), + DEFAULT_ROUNDING = sizeof (union fooround) + }; + +/* When we copy a long block of data, this is the unit to do it with. + On some machines, copying successive ints does not work; + in such a case, redefine COPYING_UNIT to `long' (if that works) + or `char' as a last resort. */ +# ifndef COPYING_UNIT +# define COPYING_UNIT int +# endif + + +/* The functions allocating more room by calling `obstack_chunk_alloc' + jump to the handler pointed to by `obstack_alloc_failed_handler'. + This can be set to a user defined function which should either + abort gracefully or use longjump - but shouldn't return. This + variable by default points to the internal function + `print_and_abort'. */ +static void print_and_abort (void); +void (*obstack_alloc_failed_handler) (void) = print_and_abort; + +/* Exit value used when `print_and_abort' is used. */ +# include <stdlib.h> +# ifdef _LIBC +int obstack_exit_failure = EXIT_FAILURE; +# else +# include "exitfail.h" +# define obstack_exit_failure exit_failure +# endif + +# ifdef _LIBC +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3_4) +/* A looong time ago (before 1994, anyway; we're not sure) this global variable + was used by non-GNU-C macros to avoid multiple evaluation. The GNU C + library still exports it because somebody might use it. */ +struct obstack *_obstack_compat; +compat_symbol (libc, _obstack_compat, _obstack, GLIBC_2_0); +# endif +# endif + +/* Define a macro that either calls functions with the traditional malloc/free + calling interface, or calls functions with the mmalloc/mfree interface + (that adds an extra first argument), based on the state of use_extra_arg. + For free, do not use ?:, since some compilers, like the MIPS compilers, + do not allow (expr) ? void : void. */ + +# define CALL_CHUNKFUN(h, size) \ + (((h) -> use_extra_arg) \ + ? (*(h)->chunkfun) ((h)->extra_arg, (size)) \ + : (*(struct _obstack_chunk *(*) (long)) (h)->chunkfun) ((size))) + +# define CALL_FREEFUN(h, old_chunk) \ + do { \ + if ((h) -> use_extra_arg) \ + (*(h)->freefun) ((h)->extra_arg, (old_chunk)); \ + else \ + (*(void (*) (void *)) (h)->freefun) ((old_chunk)); \ + } while (0) + + +/* Initialize an obstack H for use. Specify chunk size SIZE (0 means default). + Objects start on multiples of ALIGNMENT (0 means use default). + CHUNKFUN is the function to use to allocate chunks, + and FREEFUN the function to free them. + + Return nonzero if successful, calls obstack_alloc_failed_handler if + allocation fails. */ + +int +_obstack_begin (struct obstack *h, + int size, int alignment, + void *(*chunkfun) (long), + void (*freefun) (void *)) +{ + register struct _obstack_chunk *chunk; /* points to new chunk */ + + if (alignment == 0) + alignment = DEFAULT_ALIGNMENT; + if (size == 0) + /* Default size is what GNU malloc can fit in a 4096-byte block. */ + { + /* 12 is sizeof (mhead) and 4 is EXTRA from GNU malloc. + Use the values for range checking, because if range checking is off, + the extra bytes won't be missed terribly, but if range checking is on + and we used a larger request, a whole extra 4096 bytes would be + allocated. + + These number are irrelevant to the new GNU malloc. I suspect it is + less sensitive to the size of the request. */ + int extra = ((((12 + DEFAULT_ROUNDING - 1) & ~(DEFAULT_ROUNDING - 1)) + + 4 + DEFAULT_ROUNDING - 1) + & ~(DEFAULT_ROUNDING - 1)); + size = 4096 - extra; + } + + h->chunkfun = (struct _obstack_chunk * (*)(void *, long)) chunkfun; + h->freefun = (void (*) (void *, struct _obstack_chunk *)) freefun; + h->chunk_size = size; + h->alignment_mask = alignment - 1; + h->use_extra_arg = 0; + + chunk = h->chunk = CALL_CHUNKFUN (h, h -> chunk_size); + if (!chunk) + (*obstack_alloc_failed_handler) (); + h->next_free = h->object_base = __PTR_ALIGN ((char *) chunk, chunk->contents, + alignment - 1); + h->chunk_limit = chunk->limit + = (char *) chunk + h->chunk_size; + chunk->prev = 0; + /* The initial chunk now contains no empty object. */ + h->maybe_empty_object = 0; + h->alloc_failed = 0; + return 1; +} + +int +_obstack_begin_1 (struct obstack *h, int size, int alignment, + void *(*chunkfun) (void *, long), + void (*freefun) (void *, void *), + void *arg) +{ + register struct _obstack_chunk *chunk; /* points to new chunk */ + + if (alignment == 0) + alignment = DEFAULT_ALIGNMENT; + if (size == 0) + /* Default size is what GNU malloc can fit in a 4096-byte block. */ + { + /* 12 is sizeof (mhead) and 4 is EXTRA from GNU malloc. + Use the values for range checking, because if range checking is off, + the extra bytes won't be missed terribly, but if range checking is on + and we used a larger request, a whole extra 4096 bytes would be + allocated. + + These number are irrelevant to the new GNU malloc. I suspect it is + less sensitive to the size of the request. */ + int extra = ((((12 + DEFAULT_ROUNDING - 1) & ~(DEFAULT_ROUNDING - 1)) + + 4 + DEFAULT_ROUNDING - 1) + & ~(DEFAULT_ROUNDING - 1)); + size = 4096 - extra; + } + + h->chunkfun = (struct _obstack_chunk * (*)(void *,long)) chunkfun; + h->freefun = (void (*) (void *, struct _obstack_chunk *)) freefun; + h->chunk_size = size; + h->alignment_mask = alignment - 1; + h->extra_arg = arg; + h->use_extra_arg = 1; + + chunk = h->chunk = CALL_CHUNKFUN (h, h -> chunk_size); + if (!chunk) + (*obstack_alloc_failed_handler) (); + h->next_free = h->object_base = __PTR_ALIGN ((char *) chunk, chunk->contents, + alignment - 1); + h->chunk_limit = chunk->limit + = (char *) chunk + h->chunk_size; + chunk->prev = 0; + /* The initial chunk now contains no empty object. */ + h->maybe_empty_object = 0; + h->alloc_failed = 0; + return 1; +} + +/* Allocate a new current chunk for the obstack *H + on the assumption that LENGTH bytes need to be added + to the current object, or a new object of length LENGTH allocated. + Copies any partial object from the end of the old chunk + to the beginning of the new one. */ + +void +_obstack_newchunk (struct obstack *h, int length) +{ + register struct _obstack_chunk *old_chunk = h->chunk; + register struct _obstack_chunk *new_chunk; + register long new_size; + register long obj_size = h->next_free - h->object_base; + register long i; + long already; + char *object_base; + + /* Compute size for new chunk. */ + new_size = (obj_size + length) + (obj_size >> 3) + h->alignment_mask + 100; + if (new_size < h->chunk_size) + new_size = h->chunk_size; + + /* Allocate and initialize the new chunk. */ + new_chunk = CALL_CHUNKFUN (h, new_size); + if (!new_chunk) + (*obstack_alloc_failed_handler) (); + h->chunk = new_chunk; + new_chunk->prev = old_chunk; + new_chunk->limit = h->chunk_limit = (char *) new_chunk + new_size; + + /* Compute an aligned object_base in the new chunk */ + object_base = + __PTR_ALIGN ((char *) new_chunk, new_chunk->contents, h->alignment_mask); + + /* Move the existing object to the new chunk. + Word at a time is fast and is safe if the object + is sufficiently aligned. */ + if (h->alignment_mask + 1 >= DEFAULT_ALIGNMENT) + { + for (i = obj_size / sizeof (COPYING_UNIT) - 1; + i >= 0; i--) + ((COPYING_UNIT *)object_base)[i] + = ((COPYING_UNIT *)h->object_base)[i]; + /* We used to copy the odd few remaining bytes as one extra COPYING_UNIT, + but that can cross a page boundary on a machine + which does not do strict alignment for COPYING_UNITS. */ + already = obj_size / sizeof (COPYING_UNIT) * sizeof (COPYING_UNIT); + } + else + already = 0; + /* Copy remaining bytes one by one. */ + for (i = already; i < obj_size; i++) + object_base[i] = h->object_base[i]; + + /* If the object just copied was the only data in OLD_CHUNK, + free that chunk and remove it from the chain. + But not if that chunk might contain an empty object. */ + if (! h->maybe_empty_object + && (h->object_base + == __PTR_ALIGN ((char *) old_chunk, old_chunk->contents, + h->alignment_mask))) + { + new_chunk->prev = old_chunk->prev; + CALL_FREEFUN (h, old_chunk); + } + + h->object_base = object_base; + h->next_free = h->object_base + obj_size; + /* The new chunk certainly contains no empty object yet. */ + h->maybe_empty_object = 0; +} +# ifdef _LIBC +libc_hidden_def (_obstack_newchunk) +# endif + +/* Return nonzero if object OBJ has been allocated from obstack H. + This is here for debugging. + If you use it in a program, you are probably losing. */ + +/* Suppress -Wmissing-prototypes warning. We don't want to declare this in + obstack.h because it is just for debugging. */ +int _obstack_allocated_p (struct obstack *h, void *obj); + +int +_obstack_allocated_p (struct obstack *h, void *obj) +{ + register struct _obstack_chunk *lp; /* below addr of any objects in this chunk */ + register struct _obstack_chunk *plp; /* point to previous chunk if any */ + + lp = (h)->chunk; + /* We use >= rather than > since the object cannot be exactly at + the beginning of the chunk but might be an empty object exactly + at the end of an adjacent chunk. */ + while (lp != 0 && ((void *) lp >= obj || (void *) (lp)->limit < obj)) + { + plp = lp->prev; + lp = plp; + } + return lp != 0; +} + +/* Free objects in obstack H, including OBJ and everything allocate + more recently than OBJ. If OBJ is zero, free everything in H. */ + +# undef obstack_free + +void +__obstack_free (struct obstack *h, void *obj) +{ + register struct _obstack_chunk *lp; /* below addr of any objects in this chunk */ + register struct _obstack_chunk *plp; /* point to previous chunk if any */ + + lp = h->chunk; + /* We use >= because there cannot be an object at the beginning of a chunk. + But there can be an empty object at that address + at the end of another chunk. */ + while (lp != 0 && ((void *) lp >= obj || (void *) (lp)->limit < obj)) + { + plp = lp->prev; + CALL_FREEFUN (h, lp); + lp = plp; + /* If we switch chunks, we can't tell whether the new current + chunk contains an empty object, so assume that it may. */ + h->maybe_empty_object = 1; + } + if (lp) + { + h->object_base = h->next_free = (char *) (obj); + h->chunk_limit = lp->limit; + h->chunk = lp; + } + else if (obj != 0) + /* obj is not in any of the chunks! */ + abort (); +} + +# ifdef _LIBC +/* Older versions of libc used a function _obstack_free intended to be + called by non-GCC compilers. */ +strong_alias (obstack_free, _obstack_free) +# endif + +int +_obstack_memory_used (struct obstack *h) +{ + register struct _obstack_chunk* lp; + register int nbytes = 0; + + for (lp = h->chunk; lp != 0; lp = lp->prev) + { + nbytes += lp->limit - (char *) lp; + } + return nbytes; +} + +/* Define the error handler. */ +# ifdef _LIBC +# include <libintl.h> +# else +# include "gettext.h" +# endif +# ifndef _ +# define _(msgid) gettext (msgid) +# endif + +# ifdef _LIBC +# include <libio/iolibio.h> +# endif + +# ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) +# define __attribute__(Spec) /* empty */ +# endif +# endif + +static void +__attribute__ ((noreturn)) +print_and_abort (void) +{ + /* Don't change any of these strings. Yes, it would be possible to add + the newline to the string and use fputs or so. But this must not + happen because the "memory exhausted" message appears in other places + like this and the translation should be reused instead of creating + a very similar string which requires a separate translation. */ +# ifdef _LIBC + (void) __fxprintf (NULL, "%s\n", _("memory exhausted")); +# else + fprintf (stderr, "%s\n", _("memory exhausted")); +# endif + exit (obstack_exit_failure); +} + +#endif /* !ELIDE_CODE */ diff --git a/lib/obstack.h b/lib/obstack.h new file mode 100644 index 0000000..3315dfe --- /dev/null +++ b/lib/obstack.h @@ -0,0 +1,513 @@ +/* obstack.h - object stack macros + Copyright (C) 1988-1994,1996-1999,2003,2004,2005,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Summary: + +All the apparent functions defined here are macros. The idea +is that you would use these pre-tested macros to solve a +very specific set of problems, and they would run fast. +Caution: no side-effects in arguments please!! They may be +evaluated MANY times!! + +These macros operate a stack of objects. Each object starts life +small, and may grow to maturity. (Consider building a word syllable +by syllable.) An object can move while it is growing. Once it has +been "finished" it never changes address again. So the "top of the +stack" is typically an immature growing object, while the rest of the +stack is of mature, fixed size and fixed address objects. + +These routines grab large chunks of memory, using a function you +supply, called `obstack_chunk_alloc'. On occasion, they free chunks, +by calling `obstack_chunk_free'. You must define them and declare +them before using any obstack macros. + +Each independent stack is represented by a `struct obstack'. +Each of the obstack macros expects a pointer to such a structure +as the first argument. + +One motivation for this package is the problem of growing char strings +in symbol tables. Unless you are "fascist pig with a read-only mind" +--Gosper's immortal quote from HAKMEM item 154, out of context--you +would not like to put any arbitrary upper limit on the length of your +symbols. + +In practice this often means you will build many short symbols and a +few long symbols. At the time you are reading a symbol you don't know +how long it is. One traditional method is to read a symbol into a +buffer, realloc()ating the buffer every time you try to read a symbol +that is longer than the buffer. This is beaut, but you still will +want to copy the symbol from the buffer to a more permanent +symbol-table entry say about half the time. + +With obstacks, you can work differently. Use one obstack for all symbol +names. As you read a symbol, grow the name in the obstack gradually. +When the name is complete, finalize it. Then, if the symbol exists already, +free the newly read name. + +The way we do this is to take a large chunk, allocating memory from +low addresses. When you want to build a symbol in the chunk you just +add chars above the current "high water mark" in the chunk. When you +have finished adding chars, because you got to the end of the symbol, +you know how long the chars are, and you can create a new object. +Mostly the chars will not burst over the highest address of the chunk, +because you would typically expect a chunk to be (say) 100 times as +long as an average object. + +In case that isn't clear, when we have enough chars to make up +the object, THEY ARE ALREADY CONTIGUOUS IN THE CHUNK (guaranteed) +so we just point to it where it lies. No moving of chars is +needed and this is the second win: potentially long strings need +never be explicitly shuffled. Once an object is formed, it does not +change its address during its lifetime. + +When the chars burst over a chunk boundary, we allocate a larger +chunk, and then copy the partly formed object from the end of the old +chunk to the beginning of the new larger chunk. We then carry on +accreting characters to the end of the object as we normally would. + +A special macro is provided to add a single char at a time to a +growing object. This allows the use of register variables, which +break the ordinary 'growth' macro. + +Summary: + We allocate large chunks. + We carve out one object at a time from the current chunk. + Once carved, an object never moves. + We are free to append data of any size to the currently + growing object. + Exactly one object is growing in an obstack at any one time. + You can run one obstack per control block. + You may have as many control blocks as you dare. + Because of the way we do it, you can `unwind' an obstack + back to a previous state. (You may remove objects much + as you would with a stack.) +*/ + + +/* Don't do the contents of this file more than once. */ + +#ifndef _OBSTACK_H +#define _OBSTACK_H 1 + +#ifdef __cplusplus +extern "C" { +#endif + +/* We need the type of a pointer subtraction. If __PTRDIFF_TYPE__ is + defined, as with GNU C, use that; that way we don't pollute the + namespace with <stddef.h>'s symbols. Otherwise, include <stddef.h> + and use ptrdiff_t. */ + +#ifdef __PTRDIFF_TYPE__ +# define PTR_INT_TYPE __PTRDIFF_TYPE__ +#else +# include <stddef.h> +# define PTR_INT_TYPE ptrdiff_t +#endif + +/* If B is the base of an object addressed by P, return the result of + aligning P to the next multiple of A + 1. B and P must be of type + char *. A + 1 must be a power of 2. */ + +#define __BPTR_ALIGN(B, P, A) ((B) + (((P) - (B) + (A)) & ~(A))) + +/* Similiar to _BPTR_ALIGN (B, P, A), except optimize the common case + where pointers can be converted to integers, aligned as integers, + and converted back again. If PTR_INT_TYPE is narrower than a + pointer (e.g., the AS/400), play it safe and compute the alignment + relative to B. Otherwise, use the faster strategy of computing the + alignment relative to 0. */ + +#define __PTR_ALIGN(B, P, A) \ + __BPTR_ALIGN (sizeof (PTR_INT_TYPE) < sizeof (void *) ? (B) : (char *) 0, \ + P, A) + +#include <string.h> + +struct _obstack_chunk /* Lives at front of each chunk. */ +{ + char *limit; /* 1 past end of this chunk */ + struct _obstack_chunk *prev; /* address of prior chunk or NULL */ + char contents[4]; /* objects begin here */ +}; + +struct obstack /* control current object in current chunk */ +{ + long chunk_size; /* preferred size to allocate chunks in */ + struct _obstack_chunk *chunk; /* address of current struct obstack_chunk */ + char *object_base; /* address of object we are building */ + char *next_free; /* where to add next char to current object */ + char *chunk_limit; /* address of char after current chunk */ + union + { + PTR_INT_TYPE tempint; + void *tempptr; + } temp; /* Temporary for some macros. */ + int alignment_mask; /* Mask of alignment for each object. */ + /* These prototypes vary based on `use_extra_arg', and we use + casts to the prototypeless function type in all assignments, + but having prototypes here quiets -Wstrict-prototypes. */ + struct _obstack_chunk *(*chunkfun) (void *, long); + void (*freefun) (void *, struct _obstack_chunk *); + void *extra_arg; /* first arg for chunk alloc/dealloc funcs */ + unsigned use_extra_arg:1; /* chunk alloc/dealloc funcs take extra arg */ + unsigned maybe_empty_object:1;/* There is a possibility that the current + chunk contains a zero-length object. This + prevents freeing the chunk if we allocate + a bigger chunk to replace it. */ + unsigned alloc_failed:1; /* No longer used, as we now call the failed + handler on error, but retained for binary + compatibility. */ +}; + +/* Declare the external functions we use; they are in obstack.c. */ + +extern void _obstack_newchunk (struct obstack *, int); +extern int _obstack_begin (struct obstack *, int, int, + void *(*) (long), void (*) (void *)); +extern int _obstack_begin_1 (struct obstack *, int, int, + void *(*) (void *, long), + void (*) (void *, void *), void *); +extern int _obstack_memory_used (struct obstack *); + +/* The default name of the function for freeing a chunk is 'obstack_free', + but gnulib users can override this by defining '__obstack_free'. */ +#ifndef __obstack_free +# define __obstack_free obstack_free +#endif +extern void __obstack_free (struct obstack *obstack, void *block); + + +/* Error handler called when `obstack_chunk_alloc' failed to allocate + more memory. This can be set to a user defined function which + should either abort gracefully or use longjump - but shouldn't + return. The default action is to print a message and abort. */ +extern void (*obstack_alloc_failed_handler) (void); + +/* Exit value used when `print_and_abort' is used. */ +extern int obstack_exit_failure; + +/* Pointer to beginning of object being allocated or to be allocated next. + Note that this might not be the final address of the object + because a new chunk might be needed to hold the final size. */ + +#define obstack_base(h) ((void *) (h)->object_base) + +/* Size for allocating ordinary chunks. */ + +#define obstack_chunk_size(h) ((h)->chunk_size) + +/* Pointer to next byte not yet allocated in current chunk. */ + +#define obstack_next_free(h) ((h)->next_free) + +/* Mask specifying low bits that should be clear in address of an object. */ + +#define obstack_alignment_mask(h) ((h)->alignment_mask) + +/* To prevent prototype warnings provide complete argument list. */ +#define obstack_init(h) \ + _obstack_begin ((h), 0, 0, \ + (void *(*) (long)) obstack_chunk_alloc, \ + (void (*) (void *)) obstack_chunk_free) + +#define obstack_begin(h, size) \ + _obstack_begin ((h), (size), 0, \ + (void *(*) (long)) obstack_chunk_alloc, \ + (void (*) (void *)) obstack_chunk_free) + +#define obstack_specify_allocation(h, size, alignment, chunkfun, freefun) \ + _obstack_begin ((h), (size), (alignment), \ + (void *(*) (long)) (chunkfun), \ + (void (*) (void *)) (freefun)) + +#define obstack_specify_allocation_with_arg(h, size, alignment, chunkfun, freefun, arg) \ + _obstack_begin_1 ((h), (size), (alignment), \ + (void *(*) (void *, long)) (chunkfun), \ + (void (*) (void *, void *)) (freefun), (arg)) + +#define obstack_chunkfun(h, newchunkfun) \ + ((h) -> chunkfun = (struct _obstack_chunk *(*)(void *, long)) (newchunkfun)) + +#define obstack_freefun(h, newfreefun) \ + ((h) -> freefun = (void (*)(void *, struct _obstack_chunk *)) (newfreefun)) + +#define obstack_1grow_fast(h,achar) (*((h)->next_free)++ = (achar)) + +#define obstack_blank_fast(h,n) ((h)->next_free += (n)) + +#define obstack_memory_used(h) _obstack_memory_used (h) + +#if defined __GNUC__ && defined __STDC__ && __STDC__ +/* NextStep 2.0 cc is really gcc 1.93 but it defines __GNUC__ = 2 and + does not implement __extension__. But that compiler doesn't define + __GNUC_MINOR__. */ +# if __GNUC__ < 2 || (__NeXT__ && !__GNUC_MINOR__) +# define __extension__ +# endif + +/* For GNU C, if not -traditional, + we can define these macros to compute all args only once + without using a global variable. + Also, we can avoid using the `temp' slot, to make faster code. */ + +# define obstack_object_size(OBSTACK) \ + __extension__ \ + ({ struct obstack const *__o = (OBSTACK); \ + (unsigned) (__o->next_free - __o->object_base); }) + +# define obstack_room(OBSTACK) \ + __extension__ \ + ({ struct obstack const *__o = (OBSTACK); \ + (unsigned) (__o->chunk_limit - __o->next_free); }) + +# define obstack_make_room(OBSTACK,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->chunk_limit - __o->next_free < __len) \ + _obstack_newchunk (__o, __len); \ + (void) 0; }) + +# define obstack_empty_p(OBSTACK) \ + __extension__ \ + ({ struct obstack const *__o = (OBSTACK); \ + (__o->chunk->prev == 0 \ + && __o->next_free == __PTR_ALIGN ((char *) __o->chunk, \ + __o->chunk->contents, \ + __o->alignment_mask)); }) + +# define obstack_grow(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->next_free + __len > __o->chunk_limit) \ + _obstack_newchunk (__o, __len); \ + memcpy (__o->next_free, where, __len); \ + __o->next_free += __len; \ + (void) 0; }) + +# define obstack_grow0(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->next_free + __len + 1 > __o->chunk_limit) \ + _obstack_newchunk (__o, __len + 1); \ + memcpy (__o->next_free, where, __len); \ + __o->next_free += __len; \ + *(__o->next_free)++ = 0; \ + (void) 0; }) + +# define obstack_1grow(OBSTACK,datum) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + if (__o->next_free + 1 > __o->chunk_limit) \ + _obstack_newchunk (__o, 1); \ + obstack_1grow_fast (__o, datum); \ + (void) 0; }) + +/* These assume that the obstack alignment is good enough for pointers + or ints, and that the data added so far to the current object + shares that much alignment. */ + +# define obstack_ptr_grow(OBSTACK,datum) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + if (__o->next_free + sizeof (void *) > __o->chunk_limit) \ + _obstack_newchunk (__o, sizeof (void *)); \ + obstack_ptr_grow_fast (__o, datum); }) \ + +# define obstack_int_grow(OBSTACK,datum) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + if (__o->next_free + sizeof (int) > __o->chunk_limit) \ + _obstack_newchunk (__o, sizeof (int)); \ + obstack_int_grow_fast (__o, datum); }) + +# define obstack_ptr_grow_fast(OBSTACK,aptr) \ +__extension__ \ +({ struct obstack *__o1 = (OBSTACK); \ + *(const void **) __o1->next_free = (aptr); \ + __o1->next_free += sizeof (const void *); \ + (void) 0; }) + +# define obstack_int_grow_fast(OBSTACK,aint) \ +__extension__ \ +({ struct obstack *__o1 = (OBSTACK); \ + *(int *) __o1->next_free = (aint); \ + __o1->next_free += sizeof (int); \ + (void) 0; }) + +# define obstack_blank(OBSTACK,length) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + int __len = (length); \ + if (__o->chunk_limit - __o->next_free < __len) \ + _obstack_newchunk (__o, __len); \ + obstack_blank_fast (__o, __len); \ + (void) 0; }) + +# define obstack_alloc(OBSTACK,length) \ +__extension__ \ +({ struct obstack *__h = (OBSTACK); \ + obstack_blank (__h, (length)); \ + obstack_finish (__h); }) + +# define obstack_copy(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__h = (OBSTACK); \ + obstack_grow (__h, (where), (length)); \ + obstack_finish (__h); }) + +# define obstack_copy0(OBSTACK,where,length) \ +__extension__ \ +({ struct obstack *__h = (OBSTACK); \ + obstack_grow0 (__h, (where), (length)); \ + obstack_finish (__h); }) + +/* The local variable is named __o1 to avoid a name conflict + when obstack_blank is called. */ +# define obstack_finish(OBSTACK) \ +__extension__ \ +({ struct obstack *__o1 = (OBSTACK); \ + void *__value = (void *) __o1->object_base; \ + if (__o1->next_free == __value) \ + __o1->maybe_empty_object = 1; \ + __o1->next_free \ + = __PTR_ALIGN (__o1->object_base, __o1->next_free, \ + __o1->alignment_mask); \ + if (__o1->next_free - (char *)__o1->chunk \ + > __o1->chunk_limit - (char *)__o1->chunk) \ + __o1->next_free = __o1->chunk_limit; \ + __o1->object_base = __o1->next_free; \ + __value; }) + +# define obstack_free(OBSTACK, OBJ) \ +__extension__ \ +({ struct obstack *__o = (OBSTACK); \ + void *__obj = (OBJ); \ + if (__obj > (void *)__o->chunk && __obj < (void *)__o->chunk_limit) \ + __o->next_free = __o->object_base = (char *)__obj; \ + else (__obstack_free) (__o, __obj); }) + +#else /* not __GNUC__ or not __STDC__ */ + +# define obstack_object_size(h) \ + (unsigned) ((h)->next_free - (h)->object_base) + +# define obstack_room(h) \ + (unsigned) ((h)->chunk_limit - (h)->next_free) + +# define obstack_empty_p(h) \ + ((h)->chunk->prev == 0 \ + && (h)->next_free == __PTR_ALIGN ((char *) (h)->chunk, \ + (h)->chunk->contents, \ + (h)->alignment_mask)) + +/* Note that the call to _obstack_newchunk is enclosed in (..., 0) + so that we can avoid having void expressions + in the arms of the conditional expression. + Casting the third operand to void was tried before, + but some compilers won't accept it. */ + +# define obstack_make_room(h,length) \ +( (h)->temp.tempint = (length), \ + (((h)->next_free + (h)->temp.tempint > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), (h)->temp.tempint), 0) : 0)) + +# define obstack_grow(h,where,length) \ +( (h)->temp.tempint = (length), \ + (((h)->next_free + (h)->temp.tempint > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), (h)->temp.tempint), 0) : 0), \ + memcpy ((h)->next_free, where, (h)->temp.tempint), \ + (h)->next_free += (h)->temp.tempint) + +# define obstack_grow0(h,where,length) \ +( (h)->temp.tempint = (length), \ + (((h)->next_free + (h)->temp.tempint + 1 > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), (h)->temp.tempint + 1), 0) : 0), \ + memcpy ((h)->next_free, where, (h)->temp.tempint), \ + (h)->next_free += (h)->temp.tempint, \ + *((h)->next_free)++ = 0) + +# define obstack_1grow(h,datum) \ +( (((h)->next_free + 1 > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), 1), 0) : 0), \ + obstack_1grow_fast (h, datum)) + +# define obstack_ptr_grow(h,datum) \ +( (((h)->next_free + sizeof (char *) > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), sizeof (char *)), 0) : 0), \ + obstack_ptr_grow_fast (h, datum)) + +# define obstack_int_grow(h,datum) \ +( (((h)->next_free + sizeof (int) > (h)->chunk_limit) \ + ? (_obstack_newchunk ((h), sizeof (int)), 0) : 0), \ + obstack_int_grow_fast (h, datum)) + +# define obstack_ptr_grow_fast(h,aptr) \ + (((const void **) ((h)->next_free += sizeof (void *)))[-1] = (aptr)) + +# define obstack_int_grow_fast(h,aint) \ + (((int *) ((h)->next_free += sizeof (int)))[-1] = (aint)) + +# define obstack_blank(h,length) \ +( (h)->temp.tempint = (length), \ + (((h)->chunk_limit - (h)->next_free < (h)->temp.tempint) \ + ? (_obstack_newchunk ((h), (h)->temp.tempint), 0) : 0), \ + obstack_blank_fast (h, (h)->temp.tempint)) + +# define obstack_alloc(h,length) \ + (obstack_blank ((h), (length)), obstack_finish ((h))) + +# define obstack_copy(h,where,length) \ + (obstack_grow ((h), (where), (length)), obstack_finish ((h))) + +# define obstack_copy0(h,where,length) \ + (obstack_grow0 ((h), (where), (length)), obstack_finish ((h))) + +# define obstack_finish(h) \ +( ((h)->next_free == (h)->object_base \ + ? (((h)->maybe_empty_object = 1), 0) \ + : 0), \ + (h)->temp.tempptr = (h)->object_base, \ + (h)->next_free \ + = __PTR_ALIGN ((h)->object_base, (h)->next_free, \ + (h)->alignment_mask), \ + (((h)->next_free - (char *) (h)->chunk \ + > (h)->chunk_limit - (char *) (h)->chunk) \ + ? ((h)->next_free = (h)->chunk_limit) : 0), \ + (h)->object_base = (h)->next_free, \ + (h)->temp.tempptr) + +# define obstack_free(h,obj) \ +( (h)->temp.tempint = (char *) (obj) - (char *) (h)->chunk, \ + ((((h)->temp.tempint > 0 \ + && (h)->temp.tempint < (h)->chunk_limit - (char *) (h)->chunk)) \ + ? (int) ((h)->next_free = (h)->object_base \ + = (h)->temp.tempint + (char *) (h)->chunk) \ + : (((__obstack_free) ((h), (h)->temp.tempint + (char *) (h)->chunk), 0), 0))) + +#endif /* not __GNUC__ or not __STDC__ */ + +#ifdef __cplusplus +} /* C++ */ +#endif + +#endif /* obstack.h */ diff --git a/lib/offtostr.c b/lib/offtostr.c new file mode 100644 index 0000000..45196e2 --- /dev/null +++ b/lib/offtostr.c @@ -0,0 +1,3 @@ +#define inttostr offtostr +#define inttype off_t +#include "inttostr.c" diff --git a/lib/open-safer.c b/lib/open-safer.c new file mode 100644 index 0000000..04a72eb --- /dev/null +++ b/lib/open-safer.c @@ -0,0 +1,50 @@ +/* Invoke open, but avoid some glitches. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "fcntl-safer.h" + +#include <fcntl.h> +#include <stdarg.h> +#include "unistd-safer.h" + +int +open_safer (char const *file, int flags, ...) +{ + mode_t mode = 0; + + if (flags & O_CREAT) + { + va_list ap; + va_start (ap, flags); + + /* Assume mode_t promotes to int if and only if it is smaller. + This assumption isn't guaranteed by the C standard, but we + don't know of any real-world counterexamples. */ + mode = (sizeof (mode_t) < sizeof (int) + ? va_arg (ap, int) + : va_arg (ap, mode_t)); + + va_end (ap); + } + + return fd_safer (open (file, flags, mode)); +} diff --git a/lib/openat-die.c b/lib/openat-die.c new file mode 100644 index 0000000..7a28570 --- /dev/null +++ b/lib/openat-die.c @@ -0,0 +1,51 @@ +/* Report a save- or restore-cwd failure in our openat replacement and then exit. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <stdlib.h> + +#include "error.h" +#include "exitfail.h" + +#include "gettext.h" +#define _(msgid) gettext (msgid) +#define N_(msgid) msgid + +void +openat_save_fail (int errno) +{ + error (exit_failure, errno, + _("unable to record current working directory")); + + /* The `noreturn' attribute cannot be applied to error, since it returns + when its first argument is 0. To help compilers understand that this + function does not return, call abort. Also, the abort is a + safety feature if exit_failure is 0 (which shouldn't happen). */ + abort (); +} + +void +openat_restore_fail (int errno) +{ + error (exit_failure, errno, + _("failed to return to initial working directory")); + + /* As above. */ + abort (); +} diff --git a/lib/openat-priv.h b/lib/openat-priv.h new file mode 100644 index 0000000..2d98821 --- /dev/null +++ b/lib/openat-priv.h @@ -0,0 +1,55 @@ +/* Internals for openat-like functions. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <errno.h> +#include <stdlib.h> + +#define OPENAT_BUFFER_SIZE 512 +char *openat_proc_name (char buf[OPENAT_BUFFER_SIZE], int fd, char const *file); + +/* Some systems don't have ENOSYS. */ +#ifndef ENOSYS +# ifdef ENOTSUP +# define ENOSYS ENOTSUP +# else +/* Some systems don't have ENOTSUP either. */ +# define ENOSYS EINVAL +# endif +#endif + +/* Some systems don't have EOPNOTSUPP. */ +#ifndef EOPNOTSUPP +# ifdef ENOTSUP +# define EOPNOTSUPP ENOTSUP +# else +/* Some systems don't have ENOTSUP either. */ +# define EOPNOTSUPP EINVAL +# endif +#endif + +/* Trying to access a BUILD_PROC_NAME file will fail on systems without + /proc support, and even on systems *with* ProcFS support. Return + nonzero if the failure may be legitimate, e.g., because /proc is not + readable, or the particular .../fd/N directory is not present. */ +#define EXPECTED_ERRNO(Errno) \ + ((Errno) == ENOTDIR || (Errno) == ENOENT \ + || (Errno) == EPERM || (Errno) == EACCES \ + || (Errno) == ENOSYS /* Solaris 8 */ \ + || (Errno) == EOPNOTSUPP /* FreeBSD */) diff --git a/lib/openat-proc.c b/lib/openat-proc.c new file mode 100644 index 0000000..ff2fff0 --- /dev/null +++ b/lib/openat-proc.c @@ -0,0 +1,95 @@ +/* Create /proc/self/fd-related names for subfiles of open directories. + + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include "openat-priv.h" + +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> + +#include <stdio.h> +#include <string.h> + +#include "dirname.h" +#include "intprops.h" +#include "same-inode.h" +#include "xalloc.h" + +/* The results of open() in this file are not used with fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef open +#undef close + +#define PROC_SELF_FD_FORMAT "/proc/self/fd/%d/%s" + +#define PROC_SELF_FD_NAME_SIZE_BOUND(len) \ + (sizeof PROC_SELF_FD_FORMAT - sizeof "%d%s" \ + + INT_STRLEN_BOUND (int) + (len) + 1) + + +/* Set BUF to the expansion of PROC_SELF_FD_FORMAT, using FD and FILE + respectively for %d and %s. If successful, return BUF if the + result fits in BUF, dynamically allocated memory otherwise. But + return NULL if /proc is not reliable. */ +char * +openat_proc_name (char buf[OPENAT_BUFFER_SIZE], int fd, char const *file) +{ + static int proc_status = 0; + + if (! proc_status) + { + /* Set PROC_STATUS to a positive value if /proc/self/fd is + reliable, and a negative value otherwise. Solaris 10 + /proc/self/fd mishandles "..", and any file name might expand + to ".." after symbolic link expansion, so avoid /proc/self/fd + if it mishandles "..". Solaris 10 has openat, but this + problem is exhibited on code that built on Solaris 8 and + running on Solaris 10. */ + + int proc_self_fd = open ("/proc/self/fd", O_RDONLY); + if (proc_self_fd < 0) + proc_status = -1; + else + { + struct stat proc_self_fd_dotdot_st; + struct stat proc_self_st; + char dotdot_buf[PROC_SELF_FD_NAME_SIZE_BOUND (sizeof ".." - 1)]; + sprintf (dotdot_buf, PROC_SELF_FD_FORMAT, proc_self_fd, ".."); + proc_status = + ((stat (dotdot_buf, &proc_self_fd_dotdot_st) == 0 + && stat ("/proc/self", &proc_self_st) == 0 + && SAME_INODE (proc_self_fd_dotdot_st, proc_self_st)) + ? 1 : -1); + close (proc_self_fd); + } + } + + if (proc_status < 0) + return NULL; + else + { + size_t bufsize = PROC_SELF_FD_NAME_SIZE_BOUND (strlen (file)); + char *result = (bufsize < OPENAT_BUFFER_SIZE ? buf : xmalloc (bufsize)); + sprintf (result, PROC_SELF_FD_FORMAT, fd, file); + return result; + } +} diff --git a/lib/openat.c b/lib/openat.c new file mode 100644 index 0000000..cd49654 --- /dev/null +++ b/lib/openat.c @@ -0,0 +1,270 @@ +/* provide a replacement openat function + Copyright (C) 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <config.h> + +#include "openat.h" + +#include <stdarg.h> +#include <stddef.h> + +#include "dirname.h" /* solely for definition of IS_ABSOLUTE_FILE_NAME */ +#include "fcntl--.h" +#include "lchown.h" +#include "lstat.h" +#include "openat-priv.h" +#include "save-cwd.h" + +/* Replacement for Solaris' openat function. + <http://www.google.com/search?q=openat+site:docs.sun.com> + First, try to simulate it via open ("/proc/self/fd/FD/FILE"). + Failing that, simulate it by doing save_cwd/fchdir/open/restore_cwd. + If either the save_cwd or the restore_cwd fails (relatively unlikely), + then give a diagnostic and exit nonzero. + Otherwise, upon failure, set errno and return -1, as openat does. + Upon successful completion, return a file descriptor. */ +int +openat (int fd, char const *file, int flags, ...) +{ + mode_t mode = 0; + + if (flags & O_CREAT) + { + va_list arg; + va_start (arg, flags); + + /* If mode_t is narrower than int, use the promoted type (int), + not mode_t. Use sizeof to guess whether mode_t is narrower; + we don't know of any practical counterexamples. */ + mode = (sizeof (mode_t) < sizeof (int) + ? va_arg (arg, int) + : va_arg (arg, mode_t)); + + va_end (arg); + } + + return openat_permissive (fd, file, flags, mode, NULL); +} + +/* Like openat (FD, FILE, FLAGS, MODE), but if CWD_ERRNO is + nonnull, set *CWD_ERRNO to an errno value if unable to save + or restore the initial working directory. This is needed only + the first time remove.c's remove_dir opens a command-line + directory argument. + + If a previous attempt to restore the current working directory + failed, then we must not even try to access a `.'-relative name. + It is the caller's responsibility not to call this function + in that case. */ + +int +openat_permissive (int fd, char const *file, int flags, mode_t mode, + int *cwd_errno) +{ + struct saved_cwd saved_cwd; + int saved_errno; + int err; + bool save_ok; + + if (fd == AT_FDCWD || IS_ABSOLUTE_FILE_NAME (file)) + return open (file, flags, mode); + + { + char buf[OPENAT_BUFFER_SIZE]; + char *proc_file = openat_proc_name (buf, fd, file); + if (proc_file) + { + int open_result = open (proc_file, flags, mode); + int open_errno = errno; + if (proc_file != buf) + free (proc_file); + /* If the syscall succeeds, or if it fails with an unexpected + errno value, then return right away. Otherwise, fall through + and resort to using save_cwd/restore_cwd. */ + if (0 <= open_result || ! EXPECTED_ERRNO (open_errno)) + { + errno = open_errno; + return open_result; + } + } + } + + save_ok = (save_cwd (&saved_cwd) == 0); + if (! save_ok) + { + if (! cwd_errno) + openat_save_fail (errno); + *cwd_errno = errno; + } + + err = fchdir (fd); + saved_errno = errno; + + if (! err) + { + err = open (file, flags, mode); + saved_errno = errno; + if (save_ok && restore_cwd (&saved_cwd) != 0) + { + if (! cwd_errno) + openat_restore_fail (errno); + *cwd_errno = errno; + } + } + + free_cwd (&saved_cwd); + errno = saved_errno; + return err; +} + +/* Return true if our openat implementation must resort to + using save_cwd and restore_cwd. */ +bool +openat_needs_fchdir (void) +{ + bool needs_fchdir = true; + int fd = open ("/", O_RDONLY); + + if (0 <= fd) + { + char buf[OPENAT_BUFFER_SIZE]; + char *proc_file = openat_proc_name (buf, fd, "."); + if (proc_file) + { + needs_fchdir = false; + if (proc_file != buf) + free (proc_file); + } + close (fd); + } + + return needs_fchdir; +} + +#if !HAVE_FDOPENDIR + +/* Replacement for Solaris' function by the same name. + <http://www.google.com/search?q=fdopendir+site:docs.sun.com> + First, try to simulate it via opendir ("/proc/self/fd/FD"). Failing + that, simulate it by doing save_cwd/fchdir/opendir(".")/restore_cwd. + If either the save_cwd or the restore_cwd fails (relatively unlikely), + then give a diagnostic and exit nonzero. + Otherwise, this function works just like Solaris' fdopendir. + + W A R N I N G: + Unlike the other fd-related functions here, this one + effectively consumes its FD parameter. The caller should not + close or otherwise manipulate FD if this function returns successfully. */ +DIR * +fdopendir (int fd) +{ + struct saved_cwd saved_cwd; + int saved_errno; + DIR *dir; + + char buf[OPENAT_BUFFER_SIZE]; + char *proc_file = openat_proc_name (buf, fd, "."); + if (proc_file) + { + dir = opendir (proc_file); + saved_errno = errno; + } + else + { + dir = NULL; + saved_errno = EOPNOTSUPP; + } + + /* If the syscall fails with an expected errno value, resort to + save_cwd/restore_cwd. */ + if (! dir && EXPECTED_ERRNO (saved_errno)) + { + if (save_cwd (&saved_cwd) != 0) + openat_save_fail (errno); + + if (fchdir (fd) != 0) + { + dir = NULL; + saved_errno = errno; + } + else + { + dir = opendir ("."); + saved_errno = errno; + + if (restore_cwd (&saved_cwd) != 0) + openat_restore_fail (errno); + } + + free_cwd (&saved_cwd); + } + + if (dir) + close (fd); + if (proc_file != buf) + free (proc_file); + errno = saved_errno; + return dir; +} + +#endif + +/* Replacement for Solaris' function by the same name. + <http://www.google.com/search?q=fstatat+site:docs.sun.com> + First, try to simulate it via l?stat ("/proc/self/fd/FD/FILE"). + Failing that, simulate it via save_cwd/fchdir/(stat|lstat)/restore_cwd. + If either the save_cwd or the restore_cwd fails (relatively unlikely), + then give a diagnostic and exit nonzero. + Otherwise, this function works just like Solaris' fstatat. */ + +#define AT_FUNC_NAME fstatat +#define AT_FUNC_F1 lstat +#define AT_FUNC_F2 stat +#define AT_FUNC_USE_F1_COND flag == AT_SYMLINK_NOFOLLOW +#define AT_FUNC_POST_FILE_PARAM_DECLS , struct stat *st, int flag +#define AT_FUNC_POST_FILE_ARGS , st +#include "at-func.c" +#undef AT_FUNC_NAME +#undef AT_FUNC_F1 +#undef AT_FUNC_F2 +#undef AT_FUNC_USE_F1_COND +#undef AT_FUNC_POST_FILE_PARAM_DECLS +#undef AT_FUNC_POST_FILE_ARGS + +/* Replacement for Solaris' function by the same name. + <http://www.google.com/search?q=unlinkat+site:docs.sun.com> + First, try to simulate it via (unlink|rmdir) ("/proc/self/fd/FD/FILE"). + Failing that, simulate it via save_cwd/fchdir/(unlink|rmdir)/restore_cwd. + If either the save_cwd or the restore_cwd fails (relatively unlikely), + then give a diagnostic and exit nonzero. + Otherwise, this function works just like Solaris' unlinkat. */ + +#define AT_FUNC_NAME unlinkat +#define AT_FUNC_F1 rmdir +#define AT_FUNC_F2 unlink +#define AT_FUNC_USE_F1_COND flag == AT_REMOVEDIR +#define AT_FUNC_POST_FILE_PARAM_DECLS , int flag +#define AT_FUNC_POST_FILE_ARGS /* empty */ +#include "at-func.c" +#undef AT_FUNC_NAME +#undef AT_FUNC_F1 +#undef AT_FUNC_F2 +#undef AT_FUNC_USE_F1_COND +#undef AT_FUNC_POST_FILE_PARAM_DECLS +#undef AT_FUNC_POST_FILE_ARGS diff --git a/lib/openat.h b/lib/openat.h new file mode 100644 index 0000000..f8333f0 --- /dev/null +++ b/lib/openat.h @@ -0,0 +1,127 @@ +/* provide a replacement openat function + Copyright (C) 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* written by Jim Meyering */ + +#include <fcntl.h> + +#include <sys/types.h> +#include <sys/stat.h> +#include <dirent.h> +#include <unistd.h> +#include <stdbool.h> + +#ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) /* empty */ +# endif +#endif + +#ifndef ATTRIBUTE_NORETURN +# define ATTRIBUTE_NORETURN __attribute__ ((__noreturn__)) +#endif + +/* Work around a bug in Solaris 9 and 10: AT_FDCWD is positive. Its + value exceeds INT_MAX, so its use as an int doesn't conform to the + C standard, and GCC and Sun C complain in some cases. If the bug + is present, undef AT_FDCWD here, so it can be redefined below. */ +#if 0 < AT_FDCWD && AT_FDCWD == 0xffd19553 +# undef AT_FDCWD +#endif + +/* Use the same bit pattern as Solaris 9, but with the proper + signedness. The bit pattern is important, in case this actually is + Solaris with the above workaround. */ +#ifndef AT_FDCWD +# define AT_FDCWD (-3041965) +#endif + +/* Use the same values as Solaris 9. This shouldn't matter, but + there's no real reason to differ. */ +#ifndef AT_SYMLINK_NOFOLLOW +# define AT_SYMLINK_NOFOLLOW 4096 +# define AT_REMOVEDIR 1 +#endif + +#ifdef __OPENAT_PREFIX + +# undef openat +# define __OPENAT_CONCAT(x, y) x ## y +# define __OPENAT_XCONCAT(x, y) __OPENAT_CONCAT (x, y) +# define __OPENAT_ID(y) __OPENAT_XCONCAT (__OPENAT_PREFIX, y) +# define openat __OPENAT_ID (openat) +int openat (int fd, char const *file, int flags, /* mode_t mode */ ...); +int openat_permissive (int fd, char const *file, int flags, mode_t mode, + int *cwd_errno); +# if ! HAVE_FDOPENDIR +# define fdopendir __OPENAT_ID (fdopendir) +# endif +DIR *fdopendir (int fd); +# define fstatat __OPENAT_ID (fstatat) +int fstatat (int fd, char const *file, struct stat *st, int flag); +# define unlinkat __OPENAT_ID (unlinkat) +int unlinkat (int fd, char const *file, int flag); +bool openat_needs_fchdir (void); + +#else + +# define openat_permissive(Fd, File, Flags, Mode, Cwd_errno) \ + openat (Fd, File, Flags, Mode) +# define openat_needs_fchdir() false + +#endif + +#if HAVE_OPENAT && ! LSTAT_FOLLOWS_SLASHED_SYMLINK +int rpl_fstatat (int fd, char const *file, struct stat *st, int flag); +# if !COMPILING_FSTATAT +# undef fstatat +# define fstatat rpl_fstatat +# endif +#endif + +int mkdirat (int fd, char const *file, mode_t mode); +void openat_restore_fail (int) ATTRIBUTE_NORETURN; +void openat_save_fail (int) ATTRIBUTE_NORETURN; +int fchmodat (int fd, char const *file, mode_t mode, int flag); +int fchownat (int fd, char const *file, uid_t owner, gid_t group, int flag); + +/* Using these function names makes application code + slightly more readable than it would be with + fchownat (..., 0) or fchownat (..., AT_SYMLINK_NOFOLLOW). */ +static inline int +chownat (int fd, char const *file, uid_t owner, gid_t group) +{ + return fchownat (fd, file, owner, group, 0); +} + +static inline int +lchownat (int fd, char const *file, uid_t owner, gid_t group) +{ + return fchownat (fd, file, owner, group, AT_SYMLINK_NOFOLLOW); +} + +static inline int +chmodat (int fd, char const *file, mode_t mode) +{ + return fchmodat (fd, file, mode, 0); +} + +static inline int +lchmodat (int fd, char const *file, mode_t mode) +{ + return fchmodat (fd, file, mode, AT_SYMLINK_NOFOLLOW); +} diff --git a/lib/pathmax.h b/lib/pathmax.h new file mode 100644 index 0000000..6941e45 --- /dev/null +++ b/lib/pathmax.h @@ -0,0 +1,47 @@ +/* Define PATH_MAX somehow. Requires sys/types.h. + Copyright (C) 1992, 1999, 2001, 2003, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _PATHMAX_H +# define _PATHMAX_H + +# include <unistd.h> + +# include <limits.h> + +# ifndef _POSIX_PATH_MAX +# define _POSIX_PATH_MAX 256 +# endif + +# if !defined PATH_MAX && defined _PC_PATH_MAX +# define PATH_MAX (pathconf ("/", _PC_PATH_MAX) < 1 ? 1024 \ + : pathconf ("/", _PC_PATH_MAX)) +# endif + +/* Don't include sys/param.h if it already has been. */ +# if defined HAVE_SYS_PARAM_H && !defined PATH_MAX && !defined MAXPATHLEN +# include <sys/param.h> +# endif + +# if !defined PATH_MAX && defined MAXPATHLEN +# define PATH_MAX MAXPATHLEN +# endif + +# ifndef PATH_MAX +# define PATH_MAX _POSIX_PATH_MAX +# endif + +#endif /* _PATHMAX_H */ diff --git a/lib/paxerror.c b/lib/paxerror.c new file mode 100644 index 0000000..000d9e4 --- /dev/null +++ b/lib/paxerror.c @@ -0,0 +1,365 @@ +/* Miscellaneous error functions + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any later + version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General + Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, Inc., + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <system.h> +#include <paxlib.h> +#include <quote.h> +#include <quotearg.h> + +/* Decode MODE from its binary form in a stat structure, and encode it + into a 9-byte string STRING, terminated with a NUL. */ + +void +pax_decode_mode (mode_t mode, char *string) +{ + *string++ = mode & S_IRUSR ? 'r' : '-'; + *string++ = mode & S_IWUSR ? 'w' : '-'; + *string++ = (mode & S_ISUID + ? (mode & S_IXUSR ? 's' : 'S') + : (mode & S_IXUSR ? 'x' : '-')); + *string++ = mode & S_IRGRP ? 'r' : '-'; + *string++ = mode & S_IWGRP ? 'w' : '-'; + *string++ = (mode & S_ISGID + ? (mode & S_IXGRP ? 's' : 'S') + : (mode & S_IXGRP ? 'x' : '-')); + *string++ = mode & S_IROTH ? 'r' : '-'; + *string++ = mode & S_IWOTH ? 'w' : '-'; + *string++ = (mode & S_ISVTX + ? (mode & S_IXOTH ? 't' : 'T') + : (mode & S_IXOTH ? 'x' : '-')); + *string = '\0'; +} + +/* Report an error associated with the system call CALL and the + optional name NAME. */ +void +call_arg_error (char const *call, char const *name) +{ + int e = errno; + /* TRANSLATORS: %s after `Cannot' is a function name, e.g. `Cannot open'. + Directly translating this to another language will not work, first because + %s itself is not translated. + Translate it as `%s: Function %s failed'. */ + ERROR ((0, e, _("%s: Cannot %s"), quotearg_colon (name), call)); +} + +/* Report a fatal error associated with the system call CALL and + the optional file name NAME. */ +void +call_arg_fatal (char const *call, char const *name) +{ + int e = errno; + /* TRANSLATORS: %s after `Cannot' is a function name, e.g. `Cannot open'. + Directly translating this to another language will not work, first because + %s itself is not translated. + Translate it as `%s: Function %s failed'. */ + FATAL_ERROR ((0, e, _("%s: Cannot %s"), quotearg_colon (name), call)); +} + +/* Report a warning associated with the system call CALL and + the optional file name NAME. */ +void +call_arg_warn (char const *call, char const *name) +{ + int e = errno; + /* TRANSLATORS: %s after `Cannot' is a function name, e.g. `Cannot open'. + Directly translating this to another language will not work, first because + %s itself is not translated. + Translate it as `%s: Function %s failed'. */ + WARN ((0, e, _("%s: Warning: Cannot %s"), quotearg_colon (name), call)); +} + +void +chmod_error_details (char const *name, mode_t mode) +{ + int e = errno; + char buf[10]; + pax_decode_mode (mode, buf); + ERROR ((0, e, _("%s: Cannot change mode to %s"), + quotearg_colon (name), buf)); +} + +void +chown_error_details (char const *name, uid_t uid, gid_t gid) +{ + int e = errno; + ERROR ((0, e, _("%s: Cannot change ownership to uid %lu, gid %lu"), + quotearg_colon (name), (unsigned long) uid, (unsigned long) gid)); +} + +void +close_error (char const *name) +{ + call_arg_error ("close", name); +} + +void +close_warn (char const *name) +{ + call_arg_warn ("close", name); +} + +void +exec_fatal (char const *name) +{ + call_arg_fatal ("exec", name); +} + +void +link_error (char const *target, char const *source) +{ + int e = errno; + ERROR ((0, e, _("%s: Cannot hard link to %s"), + quotearg_colon (source), quote_n (1, target))); +} + +void +mkdir_error (char const *name) +{ + call_arg_error ("mkdir", name); +} + +void +mkfifo_error (char const *name) +{ + call_arg_error ("mkfifo", name); +} + +void +mknod_error (char const *name) +{ + call_arg_error ("mknod", name); +} + +void +open_error (char const *name) +{ + call_arg_error ("open", name); +} + +void +open_fatal (char const *name) +{ + call_arg_fatal ("open", name); +} + +void +open_warn (char const *name) +{ + call_arg_warn ("open", name); +} + +void +read_error (char const *name) +{ + call_arg_error ("read", name); +} + +void +read_error_details (char const *name, off_t offset, size_t size) +{ + char buf[UINTMAX_STRSIZE_BOUND]; + int e = errno; + ERROR ((0, e, + ngettext ("%s: Read error at byte %s, while reading %lu byte", + "%s: Read error at byte %s, while reading %lu bytes", + size), + quotearg_colon (name), STRINGIFY_BIGINT (offset, buf), + (unsigned long) size)); +} + +void +read_warn_details (char const *name, off_t offset, size_t size) +{ + char buf[UINTMAX_STRSIZE_BOUND]; + int e = errno; + WARN ((0, e, + ngettext ("%s: Warning: Read error at byte %s, while reading %lu byte", + "%s: Warning: Read error at byte %s, while reading %lu bytes", + size), + quotearg_colon (name), STRINGIFY_BIGINT (offset, buf), + (unsigned long) size)); +} + +void +read_fatal (char const *name) +{ + call_arg_fatal ("read", name); +} + +void +read_fatal_details (char const *name, off_t offset, size_t size) +{ + char buf[UINTMAX_STRSIZE_BOUND]; + int e = errno; + FATAL_ERROR ((0, e, + ngettext ("%s: Read error at byte %s, while reading %lu byte", + "%s: Read error at byte %s, while reading %lu bytes", + size), + quotearg_colon (name), STRINGIFY_BIGINT (offset, buf), + (unsigned long) size)); +} + +void +readlink_error (char const *name) +{ + call_arg_error ("readlink", name); +} + +void +readlink_warn (char const *name) +{ + call_arg_warn ("readlink", name); +} + +void +rmdir_error (char const *name) +{ + call_arg_error ("rmdir", name); +} + +void +savedir_error (char const *name) +{ + call_arg_error ("savedir", name); +} + +void +savedir_warn (char const *name) +{ + call_arg_warn ("savedir", name); +} + +void +seek_error (char const *name) +{ + call_arg_error ("seek", name); +} + +void +seek_error_details (char const *name, off_t offset) +{ + char buf[UINTMAX_STRSIZE_BOUND]; + int e = errno; + ERROR ((0, e, _("%s: Cannot seek to %s"), + quotearg_colon (name), + STRINGIFY_BIGINT (offset, buf))); +} + +void +seek_warn (char const *name) +{ + call_arg_warn ("seek", name); +} + +void +seek_warn_details (char const *name, off_t offset) +{ + char buf[UINTMAX_STRSIZE_BOUND]; + int e = errno; + WARN ((0, e, _("%s: Warning: Cannot seek to %s"), + quotearg_colon (name), + STRINGIFY_BIGINT (offset, buf))); +} + +void +symlink_error (char const *contents, char const *name) +{ + int e = errno; + ERROR ((0, e, _("%s: Cannot create symlink to %s"), + quotearg_colon (name), quote_n (1, contents))); +} + +void +stat_fatal (char const *name) +{ + call_arg_fatal ("stat", name); +} + +void +stat_error (char const *name) +{ + call_arg_error ("stat", name); +} + +void +stat_warn (char const *name) +{ + call_arg_warn ("stat", name); +} + +void +truncate_error (char const *name) +{ + call_arg_error ("truncate", name); +} + +void +truncate_warn (char const *name) +{ + call_arg_warn ("truncate", name); +} + +void +unlink_error (char const *name) +{ + call_arg_error ("unlink", name); +} + +void +utime_error (char const *name) +{ + call_arg_error ("utime", name); +} + +void +waitpid_error (char const *name) +{ + call_arg_error ("waitpid", name); +} + +void +write_error (char const *name) +{ + call_arg_error ("write", name); +} + +void +write_error_details (char const *name, size_t status, size_t size) +{ + if (status == 0) + write_error (name); + else + ERROR ((0, 0, + ngettext ("%s: Wrote only %lu of %lu byte", + "%s: Wrote only %lu of %lu bytes", + size), + name, (unsigned long int) status, (unsigned long int) size)); +} + +void +write_fatal (char const *name) +{ + call_arg_fatal ("write", name); +} + +void +chdir_fatal (char const *name) +{ + call_arg_fatal ("chdir", name); +} diff --git a/lib/paxexit.c b/lib/paxexit.c new file mode 100644 index 0000000..a511f79 --- /dev/null +++ b/lib/paxexit.c @@ -0,0 +1,28 @@ +/* Miscellaneous error functions + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any later + version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General + Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, Inc., + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <system.h> +#include <paxlib.h> + +int exit_status = PAXEXIT_SUCCESS; + +void +pax_exit () +{ + exit (exit_status); +} diff --git a/lib/paxlib.h b/lib/paxlib.h new file mode 100644 index 0000000..381c4c7 --- /dev/null +++ b/lib/paxlib.h @@ -0,0 +1,115 @@ +/* This file is part of GNU paxutils + + Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003, + 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. +*/ + +#ifndef _paxlib_h_ +#define _paxlib_h_ + +#include <hash.h> +#include <inttostr.h> + +/* Error reporting functions and definitions */ + +/* Exit status for paxutils app. Let's try to keep this list as simple as + possible. tar -d option strongly invites a status different for unequal + comparison and other errors. */ +#define PAXEXIT_SUCCESS 0 +#define PAXEXIT_DIFFERS 1 +#define PAXEXIT_FAILURE 2 + +/* Both WARN and ERROR write a message on stderr and continue processing, + however ERROR manages so tar will exit unsuccessfully. FATAL_ERROR + writes a message on stderr and aborts immediately, with another message + line telling so. USAGE_ERROR works like FATAL_ERROR except that the + other message line suggests trying --help. All four macros accept a + single argument of the form ((0, errno, _("FORMAT"), Args...)). errno + is zero when the error is not being detected by the system. */ + +#define WARN(Args) \ + error Args +#define ERROR(Args) \ + (error Args, exit_status = PAXEXIT_FAILURE) +#define FATAL_ERROR(Args) \ + (error Args, fatal_exit ()) +#define USAGE_ERROR(Args) \ + (error Args, usage (PAXEXIT_FAILURE)) + +extern int exit_status; + +void pax_decode_mode (mode_t mode, char *string); +void call_arg_error (char const *call, char const *name); +void call_arg_fatal (char const *call, char const *name) __attribute__ ((noreturn)); +void call_arg_warn (char const *call, char const *name); +void chmod_error_details (char const *name, mode_t mode); +void chown_error_details (char const *name, uid_t uid, gid_t gid); + +void decode_mode (mode_t, char *); + +void chdir_fatal (char const *) __attribute__ ((noreturn)); +void chmod_error_details (char const *, mode_t); +void chown_error_details (char const *, uid_t, gid_t); +void close_error (char const *); +void close_warn (char const *); +void exec_fatal (char const *) __attribute__ ((noreturn)); +void link_error (char const *, char const *); +void mkdir_error (char const *); +void mkfifo_error (char const *); +void mknod_error (char const *); +void open_error (char const *); +void open_fatal (char const *) __attribute__ ((noreturn)); +void open_warn (char const *); +void read_error (char const *); +void read_error_details (char const *, off_t, size_t); +void read_fatal (char const *) __attribute__ ((noreturn)); +void read_fatal_details (char const *, off_t, size_t) __attribute__ ((noreturn)); +void read_warn_details (char const *, off_t, size_t); +void readlink_error (char const *); +void readlink_warn (char const *); +void rmdir_error (char const *); +void savedir_error (char const *); +void savedir_warn (char const *); +void seek_error (char const *); +void seek_error_details (char const *, off_t); +void seek_warn (char const *); +void seek_warn_details (char const *, off_t); +void stat_fatal (char const *); +void stat_error (char const *); +void stat_warn (char const *); +void symlink_error (char const *, char const *); +void truncate_error (char const *); +void truncate_warn (char const *); +void unlink_error (char const *); +void utime_error (char const *); +void waitpid_error (char const *); +void write_error (char const *); + +void pax_exit (void); +void fatal_exit (void) __attribute__ ((noreturn)); + +#define STRINGIFY_BIGINT(i, b) umaxtostr (i, b) + + +/* Name-related functions */ +bool hash_string_insert (Hash_table **table, char const *string); +bool hash_string_lookup (Hash_table const *table, char const *string); + +bool removed_prefixes_p (void); +char *safer_name_suffix (char const *file_name, bool link_target, bool absolute_names); + +#endif diff --git a/lib/paxnames.c b/lib/paxnames.c new file mode 100644 index 0000000..3ca8bfa --- /dev/null +++ b/lib/paxnames.c @@ -0,0 +1,156 @@ +/* This file is part of GNU paxutils + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any later + version. + + This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General + Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, Inc., + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <system.h> +#include <hash.h> +#include <paxlib.h> + + +/* Hash tables of strings. */ + +/* Calculate the hash of a string. */ +static size_t +hash_string_hasher (void const *name, size_t n_buckets) +{ + return hash_string (name, n_buckets); +} + +/* Compare two strings for equality. */ +static bool +hash_string_compare (void const *name1, void const *name2) +{ + return strcmp (name1, name2) == 0; +} + +/* Return zero if TABLE contains a copy of STRING; otherwise, insert a + copy of STRING to TABLE and return 1. */ +bool +hash_string_insert (Hash_table **table, char const *string) +{ + Hash_table *t = *table; + char *s = xstrdup (string); + char *e; + + if (! ((t + || (*table = t = hash_initialize (0, 0, hash_string_hasher, + hash_string_compare, 0))) + && (e = hash_insert (t, s)))) + xalloc_die (); + + if (e == s) + return 1; + else + { + free (s); + return 0; + } +} + +/* Return 1 if TABLE contains STRING. */ +bool +hash_string_lookup (Hash_table const *table, char const *string) +{ + return table && hash_lookup (table, string); +} + + +static Hash_table *prefix_table[2]; + +/* Return true if file names of some members in the archive were stripped off + their leading components. We could have used + return prefix_table[0] || prefix_table[1] + but the following seems to be safer: */ +bool +removed_prefixes_p (void) +{ + return (prefix_table[0] && hash_get_n_entries (prefix_table[0]) != 0) + || (prefix_table[1] && hash_get_n_entries (prefix_table[1]) != 0); +} + +/* Return a safer suffix of FILE_NAME, or "." if it has no safer + suffix. Check for fully specified file names and other atrocities. + Warn the user if we do not return NAME. If LINK_TARGET is 1, + FILE_NAME is the target of a hard link, not a member name. + If ABSOLUTE_NAMES is 0, strip filesystem prefix from the file name. */ + +char * +safer_name_suffix (char const *file_name, bool link_target, bool absolute_names) +{ + char const *p; + + if (absolute_names) + p = file_name; + else + { + /* Skip file system prefixes, leading file name components that contain + "..", and leading slashes. */ + + size_t prefix_len = FILE_SYSTEM_PREFIX_LEN (file_name); + + for (p = file_name + prefix_len; *p; ) + { + if (p[0] == '.' && p[1] == '.' && (ISSLASH (p[2]) || !p[2])) + prefix_len = p + 2 - file_name; + + do + { + char c = *p++; + if (ISSLASH (c)) + break; + } + while (*p); + } + + for (p = file_name + prefix_len; ISSLASH (*p); p++) + continue; + prefix_len = p - file_name; + + if (prefix_len) + { + char *prefix = alloca (prefix_len + 1); + memcpy (prefix, file_name, prefix_len); + prefix[prefix_len] = '\0'; + + if (hash_string_insert (&prefix_table[link_target], prefix)) + { + static char const *const diagnostic[] = + { + N_("Removing leading `%s' from member names"), + N_("Removing leading `%s' from hard link targets") + }; + WARN ((0, 0, _(diagnostic[link_target]), prefix)); + } + } + } + + if (! *p) + { + if (p == file_name) + { + static char const *const diagnostic[] = + { + N_("Substituting `.' for empty member name"), + N_("Substituting `.' for empty hard link target") + }; + WARN ((0, 0, "%s", _(diagnostic[link_target]))); + } + + p = "."; + } + + return (char *) p; +} diff --git a/lib/pipe-safer.c b/lib/pipe-safer.c new file mode 100644 index 0000000..e4431b3 --- /dev/null +++ b/lib/pipe-safer.c @@ -0,0 +1,57 @@ +/* Invoke pipe, but avoid some glitches. + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "unistd-safer.h" + +#include <unistd.h> +#include <errno.h> + +/* Like pipe, but ensure that neither of the file descriptors is + STDIN_FILENO, STDOUT_FILENO, or STDERR_FILENO. Fail with ENOSYS on + platforms that lack pipe. */ + +int +pipe_safer (int fd[2]) +{ +#if HAVE_PIPE + if (pipe (fd) == 0) + { + int i; + for (i = 0; i < 2; i++) + { + fd[i] = fd_safer (fd[i]); + if (fd[i] < 0) + { + int e = errno; + close (fd[1 - i]); + errno = e; + return -1; + } + } + + return 0; + } +#else + errno = ENOSYS; +#endif + + return -1; +} diff --git a/lib/prepargs.c b/lib/prepargs.c new file mode 100644 index 0000000..1b0d9f5 --- /dev/null +++ b/lib/prepargs.c @@ -0,0 +1,95 @@ +/* Parse arguments from a string and prepend them to an argv. + Copyright 1999, 2000, 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA + 02110-1301, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com>. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif +#include "prepargs.h" +#include <sys/types.h> +#include <xalloc.h> + +#if HAVE_STRING_H +# include <string.h> +#endif + +#include <ctype.h> + +/* IN_CTYPE_DOMAIN (C) is nonzero if the unsigned char C can safely be given + as an argument to <ctype.h> macros like "isspace". */ +#ifdef STDC_HEADERS +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) ((c) <= 0177) +#endif + +#define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c)) + +/* Find the white-space-separated options specified by OPTIONS, and + using BUF to store copies of these options, set ARGV[0], ARGV[1], + etc. to the option copies. Return the number N of options found. + Do not set ARGV[N]. If ARGV is null, do not store ARGV[0] + etc. Backslash can be used to escape whitespace (and backslashes). */ +static int +prepend_args (char const *options, char *buf, char **argv) +{ + char const *o = options; + char *b = buf; + int n = 0; + + for (;;) + { + while (ISSPACE ((unsigned char) *o)) + o++; + if (!*o) + return n; + if (argv) + argv[n] = b; + n++; + + do + if ((*b++ = *o++) == '\\' && *o) + b[-1] = *o++; + while (*o && ! ISSPACE ((unsigned char) *o)); + + *b++ = '\0'; + } +} + +/* Prepend the whitespace-separated options in OPTIONS to the argument + vector of a main program with argument count *PARGC and argument + vector *PARGV. */ +void +prepend_default_options (char const *options, int *pargc, char ***pargv) +{ + if (options) + { + char *buf = xmalloc (strlen (options) + 1); + int prepended = prepend_args (options, buf, (char **) 0); + int argc = *pargc; + char * const *argv = *pargv; + char **pp = (char **) xmalloc ((prepended + argc + 1) * sizeof *pp); + *pargc = prepended + argc; + *pargv = pp; + *pp++ = *argv++; + pp += prepend_args (options, buf, pp); + while ((*pp++ = *argv++)) + continue; + } +} diff --git a/lib/prepargs.h b/lib/prepargs.h new file mode 100644 index 0000000..ce93ea8 --- /dev/null +++ b/lib/prepargs.h @@ -0,0 +1,3 @@ +/* Parse arguments from a string and prepend them to an argv. */ + +void prepend_default_options (char const *, int *, char ***); diff --git a/lib/printf-args.c b/lib/printf-args.c new file mode 100644 index 0000000..f039b2d --- /dev/null +++ b/lib/printf-args.c @@ -0,0 +1,139 @@ +/* Decomposed printf argument list. + Copyright (C) 1999, 2002-2003, 2005-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "printf-args.h" + +#ifdef STATIC +STATIC +#endif +int +printf_fetchargs (va_list args, arguments *a) +{ + size_t i; + argument *ap; + + for (i = 0, ap = &a->arg[0]; i < a->count; i++, ap++) + switch (ap->type) + { + case TYPE_SCHAR: + ap->a.a_schar = va_arg (args, /*signed char*/ int); + break; + case TYPE_UCHAR: + ap->a.a_uchar = va_arg (args, /*unsigned char*/ int); + break; + case TYPE_SHORT: + ap->a.a_short = va_arg (args, /*short*/ int); + break; + case TYPE_USHORT: + ap->a.a_ushort = va_arg (args, /*unsigned short*/ int); + break; + case TYPE_INT: + ap->a.a_int = va_arg (args, int); + break; + case TYPE_UINT: + ap->a.a_uint = va_arg (args, unsigned int); + break; + case TYPE_LONGINT: + ap->a.a_longint = va_arg (args, long int); + break; + case TYPE_ULONGINT: + ap->a.a_ulongint = va_arg (args, unsigned long int); + break; +#if HAVE_LONG_LONG_INT + case TYPE_LONGLONGINT: + ap->a.a_longlongint = va_arg (args, long long int); + break; + case TYPE_ULONGLONGINT: + ap->a.a_ulonglongint = va_arg (args, unsigned long long int); + break; +#endif + case TYPE_DOUBLE: + ap->a.a_double = va_arg (args, double); + break; + case TYPE_LONGDOUBLE: + ap->a.a_longdouble = va_arg (args, long double); + break; + case TYPE_CHAR: + ap->a.a_char = va_arg (args, int); + break; +#if HAVE_WINT_T + case TYPE_WIDE_CHAR: + /* Although ISO C 99 7.24.1.(2) says that wint_t is "unchanged by + default argument promotions", this is not the case in mingw32, + where wint_t is 'unsigned short'. */ + ap->a.a_wide_char = + (sizeof (wint_t) < sizeof (int) + ? va_arg (args, int) + : va_arg (args, wint_t)); + break; +#endif + case TYPE_STRING: + ap->a.a_string = va_arg (args, const char *); + /* A null pointer is an invalid argument for "%s", but in practice + it occurs quite frequently in printf statements that produce + debug output. Use a fallback in this case. */ + if (ap->a.a_string == NULL) + ap->a.a_string = "(NULL)"; + break; +#if HAVE_WCHAR_T + case TYPE_WIDE_STRING: + ap->a.a_wide_string = va_arg (args, const wchar_t *); + /* A null pointer is an invalid argument for "%ls", but in practice + it occurs quite frequently in printf statements that produce + debug output. Use a fallback in this case. */ + if (ap->a.a_wide_string == NULL) + { + static const wchar_t wide_null_string[] = + { + (wchar_t)'(', + (wchar_t)'N', (wchar_t)'U', (wchar_t)'L', (wchar_t)'L', + (wchar_t)')', + (wchar_t)0 + }; + ap->a.a_wide_string = wide_null_string; + } + break; +#endif + case TYPE_POINTER: + ap->a.a_pointer = va_arg (args, void *); + break; + case TYPE_COUNT_SCHAR_POINTER: + ap->a.a_count_schar_pointer = va_arg (args, signed char *); + break; + case TYPE_COUNT_SHORT_POINTER: + ap->a.a_count_short_pointer = va_arg (args, short *); + break; + case TYPE_COUNT_INT_POINTER: + ap->a.a_count_int_pointer = va_arg (args, int *); + break; + case TYPE_COUNT_LONGINT_POINTER: + ap->a.a_count_longint_pointer = va_arg (args, long int *); + break; +#if HAVE_LONG_LONG_INT + case TYPE_COUNT_LONGLONGINT_POINTER: + ap->a.a_count_longlongint_pointer = va_arg (args, long long int *); + break; +#endif + default: + /* Unknown type. */ + return -1; + } + return 0; +} diff --git a/lib/printf-args.h b/lib/printf-args.h new file mode 100644 index 0000000..bcf4470 --- /dev/null +++ b/lib/printf-args.h @@ -0,0 +1,132 @@ +/* Decomposed printf argument list. + Copyright (C) 1999, 2002-2003, 2006-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _PRINTF_ARGS_H +#define _PRINTF_ARGS_H + +/* Get size_t. */ +#include <stddef.h> + +/* Get wchar_t. */ +#if HAVE_WCHAR_T +# include <stddef.h> +#endif + +/* Get wint_t. */ +#if HAVE_WINT_T +# include <wchar.h> +#endif + +/* Get va_list. */ +#include <stdarg.h> + + +/* Argument types */ +typedef enum +{ + TYPE_NONE, + TYPE_SCHAR, + TYPE_UCHAR, + TYPE_SHORT, + TYPE_USHORT, + TYPE_INT, + TYPE_UINT, + TYPE_LONGINT, + TYPE_ULONGINT, +#if HAVE_LONG_LONG_INT + TYPE_LONGLONGINT, + TYPE_ULONGLONGINT, +#endif + TYPE_DOUBLE, + TYPE_LONGDOUBLE, + TYPE_CHAR, +#if HAVE_WINT_T + TYPE_WIDE_CHAR, +#endif + TYPE_STRING, +#if HAVE_WCHAR_T + TYPE_WIDE_STRING, +#endif + TYPE_POINTER, + TYPE_COUNT_SCHAR_POINTER, + TYPE_COUNT_SHORT_POINTER, + TYPE_COUNT_INT_POINTER, + TYPE_COUNT_LONGINT_POINTER +#if HAVE_LONG_LONG_INT +, TYPE_COUNT_LONGLONGINT_POINTER +#endif +} arg_type; + +/* Polymorphic argument */ +typedef struct +{ + arg_type type; + union + { + signed char a_schar; + unsigned char a_uchar; + short a_short; + unsigned short a_ushort; + int a_int; + unsigned int a_uint; + long int a_longint; + unsigned long int a_ulongint; +#if HAVE_LONG_LONG_INT + long long int a_longlongint; + unsigned long long int a_ulonglongint; +#endif + float a_float; + double a_double; + long double a_longdouble; + int a_char; +#if HAVE_WINT_T + wint_t a_wide_char; +#endif + const char* a_string; +#if HAVE_WCHAR_T + const wchar_t* a_wide_string; +#endif + void* a_pointer; + signed char * a_count_schar_pointer; + short * a_count_short_pointer; + int * a_count_int_pointer; + long int * a_count_longint_pointer; +#if HAVE_LONG_LONG_INT + long long int * a_count_longlongint_pointer; +#endif + } + a; +} +argument; + +typedef struct +{ + size_t count; + argument *arg; +} +arguments; + + +/* Fetch the arguments, putting them into a. */ +#ifdef STATIC +STATIC +#else +extern +#endif +int printf_fetchargs (va_list args, arguments *a); + +#endif /* _PRINTF_ARGS_H */ diff --git a/lib/printf-parse.c b/lib/printf-parse.c new file mode 100644 index 0000000..f8e0ced --- /dev/null +++ b/lib/printf-parse.c @@ -0,0 +1,535 @@ +/* Formatted output to strings. + Copyright (C) 1999-2000, 2002-2004, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#if WIDE_CHAR_VERSION +# include "wprintf-parse.h" +#else +# include "printf-parse.h" +#endif + +/* Get size_t, NULL. */ +#include <stddef.h> + +/* Get intmax_t. */ +#if HAVE_STDINT_H_WITH_UINTMAX +# include <stdint.h> +#endif +#if HAVE_INTTYPES_H_WITH_UINTMAX +# include <inttypes.h> +#endif + +/* malloc(), realloc(), free(). */ +#include <stdlib.h> + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +#if WIDE_CHAR_VERSION +# define PRINTF_PARSE wprintf_parse +# define CHAR_T wchar_t +# define DIRECTIVE wchar_t_directive +# define DIRECTIVES wchar_t_directives +#else +# define PRINTF_PARSE printf_parse +# define CHAR_T char +# define DIRECTIVE char_directive +# define DIRECTIVES char_directives +#endif + +#ifdef STATIC +STATIC +#endif +int +PRINTF_PARSE (const CHAR_T *format, DIRECTIVES *d, arguments *a) +{ + const CHAR_T *cp = format; /* pointer into format */ + size_t arg_posn = 0; /* number of regular arguments consumed */ + size_t d_allocated; /* allocated elements of d->dir */ + size_t a_allocated; /* allocated elements of a->arg */ + size_t max_width_length = 0; + size_t max_precision_length = 0; + + d->count = 0; + d_allocated = 1; + d->dir = malloc (d_allocated * sizeof (DIRECTIVE)); + if (d->dir == NULL) + /* Out of memory. */ + return -1; + + a->count = 0; + a_allocated = 0; + a->arg = NULL; + +#define REGISTER_ARG(_index_,_type_) \ + { \ + size_t n = (_index_); \ + if (n >= a_allocated) \ + { \ + size_t memory_size; \ + argument *memory; \ + \ + a_allocated *= 2; \ + if (a_allocated <= n) \ + a_allocated = n + 1; \ + if (SIZE_MAX / sizeof (argument) < a_allocated) \ + /* Overflow, would lead to out of memory. */ \ + goto error; \ + memory_size = a_allocated * sizeof (argument); \ + memory = (a->arg \ + ? realloc (a->arg, memory_size) \ + : malloc (memory_size)); \ + if (memory == NULL) \ + /* Out of memory. */ \ + goto error; \ + a->arg = memory; \ + } \ + while (a->count <= n) \ + a->arg[a->count++].type = TYPE_NONE; \ + if (a->arg[n].type == TYPE_NONE) \ + a->arg[n].type = (_type_); \ + else if (a->arg[n].type != (_type_)) \ + /* Ambiguous type for positional argument. */ \ + goto error; \ + } + + while (*cp != '\0') + { + CHAR_T c = *cp++; + if (c == '%') + { + size_t arg_index = ARG_NONE; + DIRECTIVE *dp = &d->dir[d->count];/* pointer to next directive */ + + /* Initialize the next directive. */ + dp->dir_start = cp - 1; + dp->flags = 0; + dp->width_start = NULL; + dp->width_end = NULL; + dp->width_arg_index = ARG_NONE; + dp->precision_start = NULL; + dp->precision_end = NULL; + dp->precision_arg_index = ARG_NONE; + dp->arg_index = ARG_NONE; + + /* Test for positional argument. */ + if (*cp >= '0' && *cp <= '9') + { + const CHAR_T *np; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + ; + if (*np == '$') + { + size_t n = 0; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + if (n < SIZE_MAX / 10) + n = 10 * n + (*np - '0'); + else + /* n too large for memory. */ + goto error; + if (n == 0) + /* Positional argument 0. */ + goto error; + arg_index = n - 1; + cp = np + 1; + } + } + + /* Read the flags. */ + for (;;) + { + if (*cp == '\'') + { + dp->flags |= FLAG_GROUP; + cp++; + } + else if (*cp == '-') + { + dp->flags |= FLAG_LEFT; + cp++; + } + else if (*cp == '+') + { + dp->flags |= FLAG_SHOWSIGN; + cp++; + } + else if (*cp == ' ') + { + dp->flags |= FLAG_SPACE; + cp++; + } + else if (*cp == '#') + { + dp->flags |= FLAG_ALT; + cp++; + } + else if (*cp == '0') + { + dp->flags |= FLAG_ZERO; + cp++; + } + else + break; + } + + /* Parse the field width. */ + if (*cp == '*') + { + dp->width_start = cp; + cp++; + dp->width_end = cp; + if (max_width_length < 1) + max_width_length = 1; + + /* Test for positional argument. */ + if (*cp >= '0' && *cp <= '9') + { + const CHAR_T *np; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + ; + if (*np == '$') + { + size_t n = 0; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + if (n < SIZE_MAX / 10) + n = 10 * n + (*np - '0'); + else + /* n too large for memory. */ + goto error; + if (n == 0) + /* Positional argument 0. */ + goto error; + dp->width_arg_index = n - 1; + cp = np + 1; + } + } + if (dp->width_arg_index == ARG_NONE) + { + dp->width_arg_index = arg_posn++; + if (dp->width_arg_index == ARG_NONE) + /* arg_posn wrapped around. */ + goto error; + } + REGISTER_ARG (dp->width_arg_index, TYPE_INT); + } + else if (*cp >= '0' && *cp <= '9') + { + size_t width_length; + + dp->width_start = cp; + for (; *cp >= '0' && *cp <= '9'; cp++) + ; + dp->width_end = cp; + width_length = dp->width_end - dp->width_start; + if (max_width_length < width_length) + max_width_length = width_length; + } + + /* Parse the precision. */ + if (*cp == '.') + { + cp++; + if (*cp == '*') + { + dp->precision_start = cp - 1; + cp++; + dp->precision_end = cp; + if (max_precision_length < 2) + max_precision_length = 2; + + /* Test for positional argument. */ + if (*cp >= '0' && *cp <= '9') + { + const CHAR_T *np; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + ; + if (*np == '$') + { + size_t n = 0; + + for (np = cp; *np >= '0' && *np <= '9'; np++) + if (n < SIZE_MAX / 10) + n = 10 * n + (*np - '0'); + else + /* n too large for memory. */ + goto error; + if (n == 0) + /* Positional argument 0. */ + goto error; + dp->precision_arg_index = n - 1; + cp = np + 1; + } + } + if (dp->precision_arg_index == ARG_NONE) + { + dp->precision_arg_index = arg_posn++; + if (dp->precision_arg_index == ARG_NONE) + /* arg_posn wrapped around. */ + goto error; + } + REGISTER_ARG (dp->precision_arg_index, TYPE_INT); + } + else + { + size_t precision_length; + + dp->precision_start = cp - 1; + for (; *cp >= '0' && *cp <= '9'; cp++) + ; + dp->precision_end = cp; + precision_length = dp->precision_end - dp->precision_start; + if (max_precision_length < precision_length) + max_precision_length = precision_length; + } + } + + { + arg_type type; + + /* Parse argument type/size specifiers. */ + { + int flags = 0; + + for (;;) + { + if (*cp == 'h') + { + flags |= (1 << (flags & 1)); + cp++; + } + else if (*cp == 'L') + { + flags |= 4; + cp++; + } + else if (*cp == 'l') + { + flags += 8; + cp++; + } +#ifdef HAVE_INTMAX_T + else if (*cp == 'j') + { + if (sizeof (intmax_t) > sizeof (long)) + { + /* intmax_t = long long */ + flags += 16; + } + else if (sizeof (intmax_t) > sizeof (int)) + { + /* intmax_t = long */ + flags += 8; + } + cp++; + } +#endif + else if (*cp == 'z' || *cp == 'Z') + { + /* 'z' is standardized in ISO C 99, but glibc uses 'Z' + because the warning facility in gcc-2.95.2 understands + only 'Z' (see gcc-2.95.2/gcc/c-common.c:1784). */ + if (sizeof (size_t) > sizeof (long)) + { + /* size_t = long long */ + flags += 16; + } + else if (sizeof (size_t) > sizeof (int)) + { + /* size_t = long */ + flags += 8; + } + cp++; + } + else if (*cp == 't') + { + if (sizeof (ptrdiff_t) > sizeof (long)) + { + /* ptrdiff_t = long long */ + flags += 16; + } + else if (sizeof (ptrdiff_t) > sizeof (int)) + { + /* ptrdiff_t = long */ + flags += 8; + } + cp++; + } + else + break; + } + + /* Read the conversion character. */ + c = *cp++; + switch (c) + { + case 'd': case 'i': +#ifdef HAVE_LONG_LONG + if (flags >= 16 || (flags & 4)) + type = TYPE_LONGLONGINT; + else +#endif + if (flags >= 8) + type = TYPE_LONGINT; + else if (flags & 2) + type = TYPE_SCHAR; + else if (flags & 1) + type = TYPE_SHORT; + else + type = TYPE_INT; + break; + case 'o': case 'u': case 'x': case 'X': +#ifdef HAVE_LONG_LONG + if (flags >= 16 || (flags & 4)) + type = TYPE_ULONGLONGINT; + else +#endif + if (flags >= 8) + type = TYPE_ULONGINT; + else if (flags & 2) + type = TYPE_UCHAR; + else if (flags & 1) + type = TYPE_USHORT; + else + type = TYPE_UINT; + break; + case 'f': case 'F': case 'e': case 'E': case 'g': case 'G': + case 'a': case 'A': +#ifdef HAVE_LONG_DOUBLE + if (flags >= 16 || (flags & 4)) + type = TYPE_LONGDOUBLE; + else +#endif + type = TYPE_DOUBLE; + break; + case 'c': + if (flags >= 8) +#ifdef HAVE_WINT_T + type = TYPE_WIDE_CHAR; +#else + goto error; +#endif + else + type = TYPE_CHAR; + break; +#ifdef HAVE_WINT_T + case 'C': + type = TYPE_WIDE_CHAR; + c = 'c'; + break; +#endif + case 's': + if (flags >= 8) +#ifdef HAVE_WCHAR_T + type = TYPE_WIDE_STRING; +#else + goto error; +#endif + else + type = TYPE_STRING; + break; +#ifdef HAVE_WCHAR_T + case 'S': + type = TYPE_WIDE_STRING; + c = 's'; + break; +#endif + case 'p': + type = TYPE_POINTER; + break; + case 'n': +#ifdef HAVE_LONG_LONG + if (flags >= 16 || (flags & 4)) + type = TYPE_COUNT_LONGLONGINT_POINTER; + else +#endif + if (flags >= 8) + type = TYPE_COUNT_LONGINT_POINTER; + else if (flags & 2) + type = TYPE_COUNT_SCHAR_POINTER; + else if (flags & 1) + type = TYPE_COUNT_SHORT_POINTER; + else + type = TYPE_COUNT_INT_POINTER; + break; + case '%': + type = TYPE_NONE; + break; + default: + /* Unknown conversion character. */ + goto error; + } + } + + if (type != TYPE_NONE) + { + dp->arg_index = arg_index; + if (dp->arg_index == ARG_NONE) + { + dp->arg_index = arg_posn++; + if (dp->arg_index == ARG_NONE) + /* arg_posn wrapped around. */ + goto error; + } + REGISTER_ARG (dp->arg_index, type); + } + dp->conversion = c; + dp->dir_end = cp; + } + + d->count++; + if (d->count >= d_allocated) + { + DIRECTIVE *memory; + + if (SIZE_MAX / (2 * sizeof (DIRECTIVE)) < d_allocated) + /* Overflow, would lead to out of memory. */ + goto error; + d_allocated *= 2; + memory = realloc (d->dir, d_allocated * sizeof (DIRECTIVE)); + if (memory == NULL) + /* Out of memory. */ + goto error; + d->dir = memory; + } + } + } + d->dir[d->count].dir_start = cp; + + d->max_width_length = max_width_length; + d->max_precision_length = max_precision_length; + return 0; + +error: + if (a->arg) + free (a->arg); + if (d->dir) + free (d->dir); + return -1; +} + +#undef DIRECTIVES +#undef DIRECTIVE +#undef CHAR_T +#undef PRINTF_PARSE diff --git a/lib/printf-parse.h b/lib/printf-parse.h new file mode 100644 index 0000000..82a0d37 --- /dev/null +++ b/lib/printf-parse.h @@ -0,0 +1,74 @@ +/* Parse printf format string. + Copyright (C) 1999, 2002-2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _PRINTF_PARSE_H +#define _PRINTF_PARSE_H + +#include "printf-args.h" + + +/* Flags */ +#define FLAG_GROUP 1 /* ' flag */ +#define FLAG_LEFT 2 /* - flag */ +#define FLAG_SHOWSIGN 4 /* + flag */ +#define FLAG_SPACE 8 /* space flag */ +#define FLAG_ALT 16 /* # flag */ +#define FLAG_ZERO 32 + +/* arg_index value indicating that no argument is consumed. */ +#define ARG_NONE (~(size_t)0) + +/* A parsed directive. */ +typedef struct +{ + const char* dir_start; + const char* dir_end; + int flags; + const char* width_start; + const char* width_end; + size_t width_arg_index; + const char* precision_start; + const char* precision_end; + size_t precision_arg_index; + char conversion; /* d i o u x X f e E g G c s p n U % but not C S */ + size_t arg_index; +} +char_directive; + +/* A parsed format string. */ +typedef struct +{ + size_t count; + char_directive *dir; + size_t max_width_length; + size_t max_precision_length; +} +char_directives; + + +/* Parses the format string. Fills in the number N of directives, and fills + in directives[0], ..., directives[N-1], and sets directives[N].dir_start + to the end of the format string. Also fills in the arg_type fields of the + arguments and the needed count of arguments. */ +#ifdef STATIC +STATIC +#else +extern +#endif +int printf_parse (const char *format, char_directives *d, arguments *a); + +#endif /* _PRINTF_PARSE_H */ diff --git a/lib/quote.c b/lib/quote.c new file mode 100644 index 0000000..119be72 --- /dev/null +++ b/lib/quote.c @@ -0,0 +1,41 @@ +/* quote.c - quote arguments for output + + Copyright (C) 1998, 1999, 2000, 2001, 2003, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#include <config.h> + +#include "quotearg.h" +#include "quote.h" + +/* Return an unambiguous printable representation of NAME, + allocated in slot N, suitable for diagnostics. */ +char const * +quote_n (int n, char const *name) +{ + return quotearg_n_style (n, locale_quoting_style, name); +} + +/* Return an unambiguous printable representation of NAME, + suitable for diagnostics. */ +char const * +quote (char const *name) +{ + return quote_n (0, name); +} diff --git a/lib/quote.h b/lib/quote.h new file mode 100644 index 0000000..5400ead --- /dev/null +++ b/lib/quote.h @@ -0,0 +1,22 @@ +/* quote.h - prototypes for quote.c + + Copyright (C) 1998, 1999, 2000, 2001, 2003 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + + +char const *quote_n (int n, char const *name); +char const *quote (char const *name); diff --git a/lib/quotearg.c b/lib/quotearg.c new file mode 100644 index 0000000..f7f326a --- /dev/null +++ b/lib/quotearg.c @@ -0,0 +1,697 @@ +/* quotearg.c - quote arguments for output + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 2004, 2005, 2006, 2007 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#include <config.h> + +#include "quotearg.h" + +#include "xalloc.h" + +#include <ctype.h> +#include <errno.h> +#include <limits.h> +#include <stdbool.h> +#include <stdlib.h> +#include <string.h> +#include <wchar.h> +#include <wctype.h> + +#include "gettext.h" +#define _(msgid) gettext (msgid) +#define N_(msgid) msgid + +#if !HAVE_MBRTOWC +/* Disable multibyte processing entirely. Since MB_CUR_MAX is 1, the + other macros are defined only for documentation and to satisfy C + syntax. */ +# undef MB_CUR_MAX +# define MB_CUR_MAX 1 +# undef mbstate_t +# define mbstate_t int +# define mbrtowc(pwc, s, n, ps) ((*(pwc) = *(s)) != 0) +# define iswprint(wc) isprint ((unsigned char) (wc)) +# undef HAVE_MBSINIT +#endif + +#if !defined mbsinit && !HAVE_MBSINIT +# define mbsinit(ps) 1 +#endif + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +#define INT_BITS (sizeof (int) * CHAR_BIT) + +struct quoting_options +{ + /* Basic quoting style. */ + enum quoting_style style; + + /* Quote the characters indicated by this bit vector even if the + quoting style would not normally require them to be quoted. */ + unsigned int quote_these_too[(UCHAR_MAX / INT_BITS) + 1]; +}; + +/* Names of quoting styles. */ +char const *const quoting_style_args[] = +{ + "literal", + "shell", + "shell-always", + "c", + "escape", + "locale", + "clocale", + 0 +}; + +/* Correspondences to quoting style names. */ +enum quoting_style const quoting_style_vals[] = +{ + literal_quoting_style, + shell_quoting_style, + shell_always_quoting_style, + c_quoting_style, + escape_quoting_style, + locale_quoting_style, + clocale_quoting_style +}; + +/* The default quoting options. */ +static struct quoting_options default_quoting_options; + +/* Allocate a new set of quoting options, with contents initially identical + to O if O is not null, or to the default if O is null. + It is the caller's responsibility to free the result. */ +struct quoting_options * +clone_quoting_options (struct quoting_options *o) +{ + int e = errno; + struct quoting_options *p = xmemdup (o ? o : &default_quoting_options, + sizeof *o); + errno = e; + return p; +} + +/* Get the value of O's quoting style. If O is null, use the default. */ +enum quoting_style +get_quoting_style (struct quoting_options *o) +{ + return (o ? o : &default_quoting_options)->style; +} + +/* In O (or in the default if O is null), + set the value of the quoting style to S. */ +void +set_quoting_style (struct quoting_options *o, enum quoting_style s) +{ + (o ? o : &default_quoting_options)->style = s; +} + +/* In O (or in the default if O is null), + set the value of the quoting options for character C to I. + Return the old value. Currently, the only values defined for I are + 0 (the default) and 1 (which means to quote the character even if + it would not otherwise be quoted). */ +int +set_char_quoting (struct quoting_options *o, char c, int i) +{ + unsigned char uc = c; + unsigned int *p = + (o ? o : &default_quoting_options)->quote_these_too + uc / INT_BITS; + int shift = uc % INT_BITS; + int r = (*p >> shift) & 1; + *p ^= ((i & 1) ^ r) << shift; + return r; +} + +/* MSGID approximates a quotation mark. Return its translation if it + has one; otherwise, return either it or "\"", depending on S. */ +static char const * +gettext_quote (char const *msgid, enum quoting_style s) +{ + char const *translation = _(msgid); + if (translation == msgid && s == clocale_quoting_style) + translation = "\""; + return translation; +} + +/* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of + argument ARG (of size ARGSIZE), using QUOTING_STYLE and the + non-quoting-style part of O to control quoting. + Terminate the output with a null character, and return the written + size of the output, not counting the terminating null. + If BUFFERSIZE is too small to store the output string, return the + value that would have been returned had BUFFERSIZE been large enough. + If ARGSIZE is SIZE_MAX, use the string length of the argument for ARGSIZE. + + This function acts like quotearg_buffer (BUFFER, BUFFERSIZE, ARG, + ARGSIZE, O), except it uses QUOTING_STYLE instead of the quoting + style specified by O, and O may not be null. */ + +static size_t +quotearg_buffer_restyled (char *buffer, size_t buffersize, + char const *arg, size_t argsize, + enum quoting_style quoting_style, + struct quoting_options const *o) +{ + size_t i; + size_t len = 0; + char const *quote_string = 0; + size_t quote_string_len = 0; + bool backslash_escapes = false; + bool unibyte_locale = MB_CUR_MAX == 1; + +#define STORE(c) \ + do \ + { \ + if (len < buffersize) \ + buffer[len] = (c); \ + len++; \ + } \ + while (0) + + switch (quoting_style) + { + case c_quoting_style: + STORE ('"'); + backslash_escapes = true; + quote_string = "\""; + quote_string_len = 1; + break; + + case escape_quoting_style: + backslash_escapes = true; + break; + + case locale_quoting_style: + case clocale_quoting_style: + { + /* TRANSLATORS: + Get translations for open and closing quotation marks. + + The message catalog should translate "`" to a left + quotation mark suitable for the locale, and similarly for + "'". If the catalog has no translation, + locale_quoting_style quotes `like this', and + clocale_quoting_style quotes "like this". + + For example, an American English Unicode locale should + translate "`" to U+201C (LEFT DOUBLE QUOTATION MARK), and + should translate "'" to U+201D (RIGHT DOUBLE QUOTATION + MARK). A British English Unicode locale should instead + translate these to U+2018 (LEFT SINGLE QUOTATION MARK) and + U+2019 (RIGHT SINGLE QUOTATION MARK), respectively. + + If you don't know what to put here, please see + <http://en.wikipedia.org/wiki/Quotation_mark#Glyphs> + and use glyphs suitable for your language. */ + + char const *left = gettext_quote (N_("`"), quoting_style); + char const *right = gettext_quote (N_("'"), quoting_style); + for (quote_string = left; *quote_string; quote_string++) + STORE (*quote_string); + backslash_escapes = true; + quote_string = right; + quote_string_len = strlen (quote_string); + } + break; + + case shell_always_quoting_style: + STORE ('\''); + quote_string = "'"; + quote_string_len = 1; + break; + + default: + break; + } + + for (i = 0; ! (argsize == SIZE_MAX ? arg[i] == '\0' : i == argsize); i++) + { + unsigned char c; + unsigned char esc; + + if (backslash_escapes + && quote_string_len + && i + quote_string_len <= argsize + && memcmp (arg + i, quote_string, quote_string_len) == 0) + STORE ('\\'); + + c = arg[i]; + switch (c) + { + case '\0': + if (backslash_escapes) + { + STORE ('\\'); + STORE ('0'); + STORE ('0'); + c = '0'; + } + break; + + case '?': + switch (quoting_style) + { + case shell_quoting_style: + goto use_shell_always_quoting_style; + + case c_quoting_style: + if (i + 2 < argsize && arg[i + 1] == '?') + switch (arg[i + 2]) + { + case '!': case '\'': + case '(': case ')': case '-': case '/': + case '<': case '=': case '>': + /* Escape the second '?' in what would otherwise be + a trigraph. */ + c = arg[i + 2]; + i += 2; + STORE ('?'); + STORE ('\\'); + STORE ('?'); + break; + + default: + break; + } + break; + + default: + break; + } + break; + + case '\a': esc = 'a'; goto c_escape; + case '\b': esc = 'b'; goto c_escape; + case '\f': esc = 'f'; goto c_escape; + case '\n': esc = 'n'; goto c_and_shell_escape; + case '\r': esc = 'r'; goto c_and_shell_escape; + case '\t': esc = 't'; goto c_and_shell_escape; + case '\v': esc = 'v'; goto c_escape; + case '\\': esc = c; goto c_and_shell_escape; + + c_and_shell_escape: + if (quoting_style == shell_quoting_style) + goto use_shell_always_quoting_style; + c_escape: + if (backslash_escapes) + { + c = esc; + goto store_escape; + } + break; + + case '{': case '}': /* sometimes special if isolated */ + if (! (argsize == SIZE_MAX ? arg[1] == '\0' : argsize == 1)) + break; + /* Fall through. */ + case '#': case '~': + if (i != 0) + break; + /* Fall through. */ + case ' ': + case '!': /* special in bash */ + case '"': case '$': case '&': + case '(': case ')': case '*': case ';': + case '<': + case '=': /* sometimes special in 0th or (with "set -k") later args */ + case '>': case '[': + case '^': /* special in old /bin/sh, e.g. SunOS 4.1.4 */ + case '`': case '|': + /* A shell special character. In theory, '$' and '`' could + be the first bytes of multibyte characters, which means + we should check them with mbrtowc, but in practice this + doesn't happen so it's not worth worrying about. */ + if (quoting_style == shell_quoting_style) + goto use_shell_always_quoting_style; + break; + + case '\'': + switch (quoting_style) + { + case shell_quoting_style: + goto use_shell_always_quoting_style; + + case shell_always_quoting_style: + STORE ('\''); + STORE ('\\'); + STORE ('\''); + break; + + default: + break; + } + break; + + case '%': case '+': case ',': case '-': case '.': case '/': + case '0': case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': case ':': + case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': + case 'G': case 'H': case 'I': case 'J': case 'K': case 'L': + case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R': + case 'S': case 'T': case 'U': case 'V': case 'W': case 'X': + case 'Y': case 'Z': case ']': case '_': case 'a': case 'b': + case 'c': case 'd': case 'e': case 'f': case 'g': case 'h': + case 'i': case 'j': case 'k': case 'l': case 'm': case 'n': + case 'o': case 'p': case 'q': case 'r': case 's': case 't': + case 'u': case 'v': case 'w': case 'x': case 'y': case 'z': + /* These characters don't cause problems, no matter what the + quoting style is. They cannot start multibyte sequences. */ + break; + + default: + /* If we have a multibyte sequence, copy it until we reach + its end, find an error, or come back to the initial shift + state. For C-like styles, if the sequence has + unprintable characters, escape the whole sequence, since + we can't easily escape single characters within it. */ + { + /* Length of multibyte sequence found so far. */ + size_t m; + + bool printable; + + if (unibyte_locale) + { + m = 1; + printable = isprint (c) != 0; + } + else + { + mbstate_t mbstate; + memset (&mbstate, 0, sizeof mbstate); + + m = 0; + printable = true; + if (argsize == SIZE_MAX) + argsize = strlen (arg); + + do + { + wchar_t w; + size_t bytes = mbrtowc (&w, &arg[i + m], + argsize - (i + m), &mbstate); + if (bytes == 0) + break; + else if (bytes == (size_t) -1) + { + printable = false; + break; + } + else if (bytes == (size_t) -2) + { + printable = false; + while (i + m < argsize && arg[i + m]) + m++; + break; + } + else + { + /* Work around a bug with older shells that "see" a '\' + that is really the 2nd byte of a multibyte character. + In practice the problem is limited to ASCII + chars >= '@' that are shell special chars. */ + if ('[' == 0x5b && quoting_style == shell_quoting_style) + { + size_t j; + for (j = 1; j < bytes; j++) + switch (arg[i + m + j]) + { + case '[': case '\\': case '^': + case '`': case '|': + goto use_shell_always_quoting_style; + + default: + break; + } + } + + if (! iswprint (w)) + printable = false; + m += bytes; + } + } + while (! mbsinit (&mbstate)); + } + + if (1 < m || (backslash_escapes && ! printable)) + { + /* Output a multibyte sequence, or an escaped + unprintable unibyte character. */ + size_t ilim = i + m; + + for (;;) + { + if (backslash_escapes && ! printable) + { + STORE ('\\'); + STORE ('0' + (c >> 6)); + STORE ('0' + ((c >> 3) & 7)); + c = '0' + (c & 7); + } + if (ilim <= i + 1) + break; + STORE (c); + c = arg[++i]; + } + + goto store_c; + } + } + } + + if (! (backslash_escapes + && o->quote_these_too[c / INT_BITS] & (1 << (c % INT_BITS)))) + goto store_c; + + store_escape: + STORE ('\\'); + + store_c: + STORE (c); + } + + if (i == 0 && quoting_style == shell_quoting_style) + goto use_shell_always_quoting_style; + + if (quote_string) + for (; *quote_string; quote_string++) + STORE (*quote_string); + + if (len < buffersize) + buffer[len] = '\0'; + return len; + + use_shell_always_quoting_style: + return quotearg_buffer_restyled (buffer, buffersize, arg, argsize, + shell_always_quoting_style, o); +} + +/* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of + argument ARG (of size ARGSIZE), using O to control quoting. + If O is null, use the default. + Terminate the output with a null character, and return the written + size of the output, not counting the terminating null. + If BUFFERSIZE is too small to store the output string, return the + value that would have been returned had BUFFERSIZE been large enough. + If ARGSIZE is SIZE_MAX, use the string length of the argument for + ARGSIZE. */ +size_t +quotearg_buffer (char *buffer, size_t buffersize, + char const *arg, size_t argsize, + struct quoting_options const *o) +{ + struct quoting_options const *p = o ? o : &default_quoting_options; + int e = errno; + size_t r = quotearg_buffer_restyled (buffer, buffersize, arg, argsize, + p->style, p); + errno = e; + return r; +} + +/* Like quotearg_buffer (..., ARG, ARGSIZE, O), except return newly + allocated storage containing the quoted string. */ +char * +quotearg_alloc (char const *arg, size_t argsize, + struct quoting_options const *o) +{ + int e = errno; + size_t bufsize = quotearg_buffer (0, 0, arg, argsize, o) + 1; + char *buf = xcharalloc (bufsize); + quotearg_buffer (buf, bufsize, arg, argsize, o); + errno = e; + return buf; +} + +/* A storage slot with size and pointer to a value. */ +struct slotvec +{ + size_t size; + char *val; +}; + +/* Preallocate a slot 0 buffer, so that the caller can always quote + one small component of a "memory exhausted" message in slot 0. */ +static char slot0[256]; +static unsigned int nslots = 1; +static struct slotvec slotvec0 = {sizeof slot0, slot0}; +static struct slotvec *slotvec = &slotvec0; + +void +quotearg_free (void) +{ + struct slotvec *sv = slotvec; + unsigned int i; + for (i = 1; i < nslots; i++) + free (sv[i].val); + if (sv[0].val != slot0) + { + free (sv[0].val); + slotvec0.size = sizeof slot0; + slotvec0.val = slot0; + } + if (sv != &slotvec0) + { + free (sv); + slotvec = &slotvec0; + } + nslots = 1; +} + +/* Use storage slot N to return a quoted version of argument ARG. + ARG is of size ARGSIZE, but if that is SIZE_MAX, ARG is a + null-terminated string. + OPTIONS specifies the quoting options. + The returned value points to static storage that can be + reused by the next call to this function with the same value of N. + N must be nonnegative. N is deliberately declared with type "int" + to allow for future extensions (using negative values). */ +static char * +quotearg_n_options (int n, char const *arg, size_t argsize, + struct quoting_options const *options) +{ + int e = errno; + + unsigned int n0 = n; + struct slotvec *sv = slotvec; + + if (n < 0) + abort (); + + if (nslots <= n0) + { + /* FIXME: technically, the type of n1 should be `unsigned int', + but that evokes an unsuppressible warning from gcc-4.0.1 and + older. If gcc ever provides an option to suppress that warning, + revert to the original type, so that the test in xalloc_oversized + is once again performed only at compile time. */ + size_t n1 = n0 + 1; + bool preallocated = (sv == &slotvec0); + + if (xalloc_oversized (n1, sizeof *sv)) + xalloc_die (); + + slotvec = sv = xrealloc (preallocated ? NULL : sv, n1 * sizeof *sv); + if (preallocated) + *sv = slotvec0; + memset (sv + nslots, 0, (n1 - nslots) * sizeof *sv); + nslots = n1; + } + + { + size_t size = sv[n].size; + char *val = sv[n].val; + size_t qsize = quotearg_buffer (val, size, arg, argsize, options); + + if (size <= qsize) + { + sv[n].size = size = qsize + 1; + if (val != slot0) + free (val); + sv[n].val = val = xcharalloc (size); + quotearg_buffer (val, size, arg, argsize, options); + } + + errno = e; + return val; + } +} + +char * +quotearg_n (int n, char const *arg) +{ + return quotearg_n_options (n, arg, SIZE_MAX, &default_quoting_options); +} + +char * +quotearg (char const *arg) +{ + return quotearg_n (0, arg); +} + +/* Return quoting options for STYLE, with no extra quoting. */ +static struct quoting_options +quoting_options_from_style (enum quoting_style style) +{ + struct quoting_options o; + o.style = style; + memset (o.quote_these_too, 0, sizeof o.quote_these_too); + return o; +} + +char * +quotearg_n_style (int n, enum quoting_style s, char const *arg) +{ + struct quoting_options const o = quoting_options_from_style (s); + return quotearg_n_options (n, arg, SIZE_MAX, &o); +} + +char * +quotearg_n_style_mem (int n, enum quoting_style s, + char const *arg, size_t argsize) +{ + struct quoting_options const o = quoting_options_from_style (s); + return quotearg_n_options (n, arg, argsize, &o); +} + +char * +quotearg_style (enum quoting_style s, char const *arg) +{ + return quotearg_n_style (0, s, arg); +} + +char * +quotearg_char (char const *arg, char ch) +{ + struct quoting_options options; + options = default_quoting_options; + set_char_quoting (&options, ch, 1); + return quotearg_n_options (0, arg, SIZE_MAX, &options); +} + +char * +quotearg_colon (char const *arg) +{ + return quotearg_char (arg, ':'); +} diff --git a/lib/quotearg.h b/lib/quotearg.h new file mode 100644 index 0000000..4887df3 --- /dev/null +++ b/lib/quotearg.h @@ -0,0 +1,140 @@ +/* quotearg.h - quote arguments for output + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 2004, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert <eggert@twinsun.com> */ + +#ifndef QUOTEARG_H_ +# define QUOTEARG_H_ 1 + +# include <stddef.h> + +/* Basic quoting styles. */ +enum quoting_style + { + /* Output names as-is (ls --quoting-style=literal). */ + literal_quoting_style, + + /* Quote names for the shell if they contain shell metacharacters + or would cause ambiguous output (ls --quoting-style=shell). */ + shell_quoting_style, + + /* Quote names for the shell, even if they would normally not + require quoting (ls --quoting-style=shell-always). */ + shell_always_quoting_style, + + /* Quote names as for a C language string (ls --quoting-style=c). */ + c_quoting_style, + + /* Like c_quoting_style except omit the surrounding double-quote + characters (ls --quoting-style=escape). */ + escape_quoting_style, + + /* Like clocale_quoting_style, but quote `like this' instead of + "like this" in the default C locale (ls --quoting-style=locale). */ + locale_quoting_style, + + /* Like c_quoting_style except use quotation marks appropriate for + the locale (ls --quoting-style=clocale). */ + clocale_quoting_style + }; + +/* For now, --quoting-style=literal is the default, but this may change. */ +# ifndef DEFAULT_QUOTING_STYLE +# define DEFAULT_QUOTING_STYLE literal_quoting_style +# endif + +/* Names of quoting styles and their corresponding values. */ +extern char const *const quoting_style_args[]; +extern enum quoting_style const quoting_style_vals[]; + +struct quoting_options; + +/* The functions listed below set and use a hidden variable + that contains the default quoting style options. */ + +/* Allocate a new set of quoting options, with contents initially identical + to O if O is not null, or to the default if O is null. + It is the caller's responsibility to free the result. */ +struct quoting_options *clone_quoting_options (struct quoting_options *o); + +/* Get the value of O's quoting style. If O is null, use the default. */ +enum quoting_style get_quoting_style (struct quoting_options *o); + +/* In O (or in the default if O is null), + set the value of the quoting style to S. */ +void set_quoting_style (struct quoting_options *o, enum quoting_style s); + +/* In O (or in the default if O is null), + set the value of the quoting options for character C to I. + Return the old value. Currently, the only values defined for I are + 0 (the default) and 1 (which means to quote the character even if + it would not otherwise be quoted). */ +int set_char_quoting (struct quoting_options *o, char c, int i); + +/* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of + argument ARG (of size ARGSIZE), using O to control quoting. + If O is null, use the default. + Terminate the output with a null character, and return the written + size of the output, not counting the terminating null. + If BUFFERSIZE is too small to store the output string, return the + value that would have been returned had BUFFERSIZE been large enough. + If ARGSIZE is -1, use the string length of the argument for ARGSIZE. */ +size_t quotearg_buffer (char *buffer, size_t buffersize, + char const *arg, size_t argsize, + struct quoting_options const *o); + +/* Like quotearg_buffer, except return the result in a newly allocated + buffer. It is the caller's responsibility to free the result. */ +char *quotearg_alloc (char const *arg, size_t argsize, + struct quoting_options const *o); + +/* Use storage slot N to return a quoted version of the string ARG. + Use the default quoting options. + The returned value points to static storage that can be + reused by the next call to this function with the same value of N. + N must be nonnegative. */ +char *quotearg_n (int n, char const *arg); + +/* Equivalent to quotearg_n (0, ARG). */ +char *quotearg (char const *arg); + +/* Use style S and storage slot N to return a quoted version of the string ARG. + This is like quotearg_n (N, ARG), except that it uses S with no other + options to specify the quoting method. */ +char *quotearg_n_style (int n, enum quoting_style s, char const *arg); + +/* Use style S and storage slot N to return a quoted version of the + argument ARG of size ARGSIZE. This is like quotearg_n_style + (N, S, ARG), except it can quote null bytes. */ +char *quotearg_n_style_mem (int n, enum quoting_style s, + char const *arg, size_t argsize); + +/* Equivalent to quotearg_n_style (0, S, ARG). */ +char *quotearg_style (enum quoting_style s, char const *arg); + +/* Like quotearg (ARG), except also quote any instances of CH. */ +char *quotearg_char (char const *arg, char ch); + +/* Equivalent to quotearg_char (ARG, ':'). */ +char *quotearg_colon (char const *arg); + +/* Free any dynamically allocated memory. */ +void quotearg_free (void); + +#endif /* !QUOTEARG_H_ */ diff --git a/lib/readlink.c b/lib/readlink.c new file mode 100644 index 0000000..3cbdc1c --- /dev/null +++ b/lib/readlink.c @@ -0,0 +1,50 @@ +/* Stub for readlink(). + Copyright (C) 2003-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <unistd.h> + +#include <errno.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <stddef.h> + +#if !HAVE_READLINK + +/* readlink() substitute for systems that don't have a readlink() function, + such as DJGPP 2.03 and mingw32. */ + +/* The official POSIX return type of readlink() is ssize_t, but since here + we have no declaration in a public header file, we use 'int' as return + type. */ + +int +readlink (const char *path, char *buf, size_t bufsize) +{ + struct stat statbuf; + + /* In general we should use lstat() here, not stat(). But on platforms + without symbolic links lstat() - if it exists - would be equivalent to + stat(), therefore we can use stat(). This saves us a configure check. */ + if (stat (path, &statbuf) >= 0) + errno = EINVAL; + return -1; +} + +#endif diff --git a/lib/ref-add.sin b/lib/ref-add.sin new file mode 100644 index 0000000..bc5cc79 --- /dev/null +++ b/lib/ref-add.sin @@ -0,0 +1,30 @@ +# Add this package to a list of references stored in a text file. +# +# Copyright (C) 2000 Free Software Foundation, Inc. +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License along +# with this program; if not, write to the Free Software Foundation, +# Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. +# +# Written by Bruno Haible <haible@clisp.cons.org>. +# +/^# Packages using this file: / { + s/# Packages using this file:// + ta + :a + s/ @PACKAGE@ / @PACKAGE@ / + tb + s/ $/ @PACKAGE@ / + :b + s/^/# Packages using this file:/ +} diff --git a/lib/ref-del.sin b/lib/ref-del.sin new file mode 100644 index 0000000..e9301bf --- /dev/null +++ b/lib/ref-del.sin @@ -0,0 +1,25 @@ +# Remove this package from a list of references stored in a text file. +# +# Copyright (C) 2000 Free Software Foundation, Inc. +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License along +# with this program; if not, write to the Free Software Foundation, +# Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. +# +# Written by Bruno Haible <haible@clisp.cons.org>. +# +/^# Packages using this file: / { + s/# Packages using this file:// + s/ @PACKAGE@ / / + s/^/# Packages using this file:/ +} diff --git a/lib/regcomp.c b/lib/regcomp.c new file mode 100644 index 0000000..fe4d243 --- /dev/null +++ b/lib/regcomp.c @@ -0,0 +1,3832 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002,2003,2004,2005,2006,2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +static reg_errcode_t re_compile_internal (regex_t *preg, const char * pattern, + size_t length, reg_syntax_t syntax); +static void re_compile_fastmap_iter (regex_t *bufp, + const re_dfastate_t *init_state, + char *fastmap); +static reg_errcode_t init_dfa (re_dfa_t *dfa, size_t pat_len); +#ifdef RE_ENABLE_I18N +static void free_charset (re_charset_t *cset); +#endif /* RE_ENABLE_I18N */ +static void free_workarea_compile (regex_t *preg); +static reg_errcode_t create_initial_state (re_dfa_t *dfa); +#ifdef RE_ENABLE_I18N +static void optimize_utf8 (re_dfa_t *dfa); +#endif +static reg_errcode_t analyze (regex_t *preg); +static reg_errcode_t preorder (bin_tree_t *root, + reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra); +static reg_errcode_t postorder (bin_tree_t *root, + reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra); +static reg_errcode_t optimize_subexps (void *extra, bin_tree_t *node); +static reg_errcode_t lower_subexps (void *extra, bin_tree_t *node); +static bin_tree_t *lower_subexp (reg_errcode_t *err, regex_t *preg, + bin_tree_t *node); +static reg_errcode_t calc_first (void *extra, bin_tree_t *node); +static reg_errcode_t calc_next (void *extra, bin_tree_t *node); +static reg_errcode_t link_nfa_nodes (void *extra, bin_tree_t *node); +static Idx duplicate_node (re_dfa_t *dfa, Idx org_idx, unsigned int constraint); +static Idx search_duplicated_node (const re_dfa_t *dfa, Idx org_node, + unsigned int constraint); +static reg_errcode_t calc_eclosure (re_dfa_t *dfa); +static reg_errcode_t calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, + Idx node, bool root); +static reg_errcode_t calc_inveclosure (re_dfa_t *dfa); +static Idx fetch_number (re_string_t *input, re_token_t *token, + reg_syntax_t syntax); +static int peek_token (re_token_t *token, re_string_t *input, + reg_syntax_t syntax) internal_function; +static bin_tree_t *parse (re_string_t *regexp, regex_t *preg, + reg_syntax_t syntax, reg_errcode_t *err); +static bin_tree_t *parse_reg_exp (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_branch (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_expression (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_sub_exp (re_string_t *regexp, regex_t *preg, + re_token_t *token, reg_syntax_t syntax, + Idx nest, reg_errcode_t *err); +static bin_tree_t *parse_dup_op (bin_tree_t *dup_elem, re_string_t *regexp, + re_dfa_t *dfa, re_token_t *token, + reg_syntax_t syntax, reg_errcode_t *err); +static bin_tree_t *parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa, + re_token_t *token, reg_syntax_t syntax, + reg_errcode_t *err); +static reg_errcode_t parse_bracket_element (bracket_elem_t *elem, + re_string_t *regexp, + re_token_t *token, int token_len, + re_dfa_t *dfa, + reg_syntax_t syntax, + bool accept_hyphen); +static reg_errcode_t parse_bracket_symbol (bracket_elem_t *elem, + re_string_t *regexp, + re_token_t *token); +#ifdef RE_ENABLE_I18N +static reg_errcode_t build_equiv_class (bitset_t sbcset, + re_charset_t *mbcset, + Idx *equiv_class_alloc, + const unsigned char *name); +static reg_errcode_t build_charclass (RE_TRANSLATE_TYPE trans, + bitset_t sbcset, + re_charset_t *mbcset, + Idx *char_class_alloc, + const unsigned char *class_name, + reg_syntax_t syntax); +#else /* not RE_ENABLE_I18N */ +static reg_errcode_t build_equiv_class (bitset_t sbcset, + const unsigned char *name); +static reg_errcode_t build_charclass (RE_TRANSLATE_TYPE trans, + bitset_t sbcset, + const unsigned char *class_name, + reg_syntax_t syntax); +#endif /* not RE_ENABLE_I18N */ +static bin_tree_t *build_charclass_op (re_dfa_t *dfa, + RE_TRANSLATE_TYPE trans, + const unsigned char *class_name, + const unsigned char *extra, + bool non_match, reg_errcode_t *err); +static bin_tree_t *create_tree (re_dfa_t *dfa, + bin_tree_t *left, bin_tree_t *right, + re_token_type_t type); +static bin_tree_t *create_token_tree (re_dfa_t *dfa, + bin_tree_t *left, bin_tree_t *right, + const re_token_t *token); +static bin_tree_t *duplicate_tree (const bin_tree_t *src, re_dfa_t *dfa); +static void free_token (re_token_t *node); +static reg_errcode_t free_tree (void *extra, bin_tree_t *node); +static reg_errcode_t mark_opt_subexp (void *extra, bin_tree_t *node); + +/* This table gives an error message for each of the error codes listed + in regex.h. Obviously the order here has to be same as there. + POSIX doesn't require that we do anything for REG_NOERROR, + but why not be nice? */ + +static const char __re_error_msgid[] = + { +#define REG_NOERROR_IDX 0 + gettext_noop ("Success") /* REG_NOERROR */ + "\0" +#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success") + gettext_noop ("No match") /* REG_NOMATCH */ + "\0" +#define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match") + gettext_noop ("Invalid regular expression") /* REG_BADPAT */ + "\0" +#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression") + gettext_noop ("Invalid collation character") /* REG_ECOLLATE */ + "\0" +#define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character") + gettext_noop ("Invalid character class name") /* REG_ECTYPE */ + "\0" +#define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name") + gettext_noop ("Trailing backslash") /* REG_EESCAPE */ + "\0" +#define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash") + gettext_noop ("Invalid back reference") /* REG_ESUBREG */ + "\0" +#define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference") + gettext_noop ("Unmatched [ or [^") /* REG_EBRACK */ + "\0" +#define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^") + gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */ + "\0" +#define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(") + gettext_noop ("Unmatched \\{") /* REG_EBRACE */ + "\0" +#define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{") + gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */ + "\0" +#define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}") + gettext_noop ("Invalid range end") /* REG_ERANGE */ + "\0" +#define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end") + gettext_noop ("Memory exhausted") /* REG_ESPACE */ + "\0" +#define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted") + gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */ + "\0" +#define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression") + gettext_noop ("Premature end of regular expression") /* REG_EEND */ + "\0" +#define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression") + gettext_noop ("Regular expression too big") /* REG_ESIZE */ + "\0" +#define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big") + gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */ + }; + +static const size_t __re_error_msgid_idx[] = + { + REG_NOERROR_IDX, + REG_NOMATCH_IDX, + REG_BADPAT_IDX, + REG_ECOLLATE_IDX, + REG_ECTYPE_IDX, + REG_EESCAPE_IDX, + REG_ESUBREG_IDX, + REG_EBRACK_IDX, + REG_EPAREN_IDX, + REG_EBRACE_IDX, + REG_BADBR_IDX, + REG_ERANGE_IDX, + REG_ESPACE_IDX, + REG_BADRPT_IDX, + REG_EEND_IDX, + REG_ESIZE_IDX, + REG_ERPAREN_IDX + }; + +/* Entry points for GNU code. */ + +/* re_compile_pattern is the GNU regular expression compiler: it + compiles PATTERN (of length LENGTH) and puts the result in BUFP. + Returns 0 if the pattern was valid, otherwise an error string. + + Assumes the `allocated' (and perhaps `buffer') and `translate' fields + are set in BUFP on entry. */ + +#ifdef _LIBC +const char * +re_compile_pattern (pattern, length, bufp) + const char *pattern; + size_t length; + struct re_pattern_buffer *bufp; +#else /* size_t might promote */ +const char * +re_compile_pattern (const char *pattern, size_t length, + struct re_pattern_buffer *bufp) +#endif +{ + reg_errcode_t ret; + + /* And GNU code determines whether or not to get register information + by passing null for the REGS argument to re_match, etc., not by + setting no_sub, unless RE_NO_SUB is set. */ + bufp->no_sub = !!(re_syntax_options & RE_NO_SUB); + + /* Match anchors at newline. */ + bufp->newline_anchor = 1; + + ret = re_compile_internal (bufp, pattern, length, re_syntax_options); + + if (!ret) + return NULL; + return gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]); +} +#ifdef _LIBC +weak_alias (__re_compile_pattern, re_compile_pattern) +#endif + +/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can + also be assigned to arbitrarily: each pattern buffer stores its own + syntax, so it can be changed between regex compilations. */ +/* This has no initializer because initialized variables in Emacs + become read-only after dumping. */ +reg_syntax_t re_syntax_options; + + +/* Specify the precise syntax of regexps for compilation. This provides + for compatibility for various utilities which historically have + different, incompatible syntaxes. + + The argument SYNTAX is a bit mask comprised of the various bits + defined in regex.h. We return the old syntax. */ + +reg_syntax_t +re_set_syntax (syntax) + reg_syntax_t syntax; +{ + reg_syntax_t ret = re_syntax_options; + + re_syntax_options = syntax; + return ret; +} +#ifdef _LIBC +weak_alias (__re_set_syntax, re_set_syntax) +#endif + +int +re_compile_fastmap (bufp) + struct re_pattern_buffer *bufp; +{ + re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; + char *fastmap = bufp->fastmap; + + memset (fastmap, '\0', sizeof (char) * SBC_MAX); + re_compile_fastmap_iter (bufp, dfa->init_state, fastmap); + if (dfa->init_state != dfa->init_state_word) + re_compile_fastmap_iter (bufp, dfa->init_state_word, fastmap); + if (dfa->init_state != dfa->init_state_nl) + re_compile_fastmap_iter (bufp, dfa->init_state_nl, fastmap); + if (dfa->init_state != dfa->init_state_begbuf) + re_compile_fastmap_iter (bufp, dfa->init_state_begbuf, fastmap); + bufp->fastmap_accurate = 1; + return 0; +} +#ifdef _LIBC +weak_alias (__re_compile_fastmap, re_compile_fastmap) +#endif + +static inline void +__attribute ((always_inline)) +re_set_fastmap (char *fastmap, bool icase, int ch) +{ + fastmap[ch] = 1; + if (icase) + fastmap[tolower (ch)] = 1; +} + +/* Helper function for re_compile_fastmap. + Compile fastmap for the initial_state INIT_STATE. */ + +static void +re_compile_fastmap_iter (regex_t *bufp, const re_dfastate_t *init_state, + char *fastmap) +{ + re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; + Idx node_cnt; + bool icase = (dfa->mb_cur_max == 1 && (bufp->syntax & RE_ICASE)); + for (node_cnt = 0; node_cnt < init_state->nodes.nelem; ++node_cnt) + { + Idx node = init_state->nodes.elems[node_cnt]; + re_token_type_t type = dfa->nodes[node].type; + + if (type == CHARACTER) + { + re_set_fastmap (fastmap, icase, dfa->nodes[node].opr.c); +#ifdef RE_ENABLE_I18N + if ((bufp->syntax & RE_ICASE) && dfa->mb_cur_max > 1) + { + unsigned char buf[MB_LEN_MAX]; + unsigned char *p; + wchar_t wc; + mbstate_t state; + + p = buf; + *p++ = dfa->nodes[node].opr.c; + while (++node < dfa->nodes_len + && dfa->nodes[node].type == CHARACTER + && dfa->nodes[node].mb_partial) + *p++ = dfa->nodes[node].opr.c; + memset (&state, '\0', sizeof (state)); + if (mbrtowc (&wc, (const char *) buf, p - buf, + &state) == p - buf + && (__wcrtomb ((char *) buf, towlower (wc), &state) + != (size_t) -1)) + re_set_fastmap (fastmap, false, buf[0]); + } +#endif + } + else if (type == SIMPLE_BRACKET) + { + int i, ch; + for (i = 0, ch = 0; i < BITSET_WORDS; ++i) + { + int j; + bitset_word_t w = dfa->nodes[node].opr.sbcset[i]; + for (j = 0; j < BITSET_WORD_BITS; ++j, ++ch) + if (w & ((bitset_word_t) 1 << j)) + re_set_fastmap (fastmap, icase, ch); + } + } +#ifdef RE_ENABLE_I18N + else if (type == COMPLEX_BRACKET) + { + Idx i; + re_charset_t *cset = dfa->nodes[node].opr.mbcset; + if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes + || cset->nranges || cset->nchar_classes) + { +# ifdef _LIBC + if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0) + { + /* In this case we want to catch the bytes which are + the first byte of any collation elements. + e.g. In da_DK, we want to catch 'a' since "aa" + is a valid collation element, and don't catch + 'b' since 'b' is the only collation element + which starts from 'b'. */ + const int32_t *table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + for (i = 0; i < SBC_MAX; ++i) + if (table[i] < 0) + re_set_fastmap (fastmap, icase, i); + } +# else + if (dfa->mb_cur_max > 1) + for (i = 0; i < SBC_MAX; ++i) + if (__btowc (i) == WEOF) + re_set_fastmap (fastmap, icase, i); +# endif /* not _LIBC */ + } + for (i = 0; i < cset->nmbchars; ++i) + { + char buf[256]; + mbstate_t state; + memset (&state, '\0', sizeof (state)); + if (__wcrtomb (buf, cset->mbchars[i], &state) != (size_t) -1) + re_set_fastmap (fastmap, icase, *(unsigned char *) buf); + if ((bufp->syntax & RE_ICASE) && dfa->mb_cur_max > 1) + { + if (__wcrtomb (buf, towlower (cset->mbchars[i]), &state) + != (size_t) -1) + re_set_fastmap (fastmap, false, *(unsigned char *) buf); + } + } + } +#endif /* RE_ENABLE_I18N */ + else if (type == OP_PERIOD +#ifdef RE_ENABLE_I18N + || type == OP_UTF8_PERIOD +#endif /* RE_ENABLE_I18N */ + || type == END_OF_RE) + { + memset (fastmap, '\1', sizeof (char) * SBC_MAX); + if (type == END_OF_RE) + bufp->can_be_null = 1; + return; + } + } +} + +/* Entry point for POSIX code. */ +/* regcomp takes a regular expression as a string and compiles it. + + PREG is a regex_t *. We do not expect any fields to be initialized, + since POSIX says we shouldn't. Thus, we set + + `buffer' to the compiled pattern; + `used' to the length of the compiled pattern; + `syntax' to RE_SYNTAX_POSIX_EXTENDED if the + REG_EXTENDED bit in CFLAGS is set; otherwise, to + RE_SYNTAX_POSIX_BASIC; + `newline_anchor' to REG_NEWLINE being set in CFLAGS; + `fastmap' to an allocated space for the fastmap; + `fastmap_accurate' to zero; + `re_nsub' to the number of subexpressions in PATTERN. + + PATTERN is the address of the pattern string. + + CFLAGS is a series of bits which affect compilation. + + If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we + use POSIX basic syntax. + + If REG_NEWLINE is set, then . and [^...] don't match newline. + Also, regexec will try a match beginning after every newline. + + If REG_ICASE is set, then we considers upper- and lowercase + versions of letters to be equivalent when matching. + + If REG_NOSUB is set, then when PREG is passed to regexec, that + routine will report only success or failure, and nothing about the + registers. + + It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for + the return codes and their meanings.) */ + +int +regcomp (preg, pattern, cflags) + regex_t *_Restrict_ preg; + const char *_Restrict_ pattern; + int cflags; +{ + reg_errcode_t ret; + reg_syntax_t syntax = ((cflags & REG_EXTENDED) ? RE_SYNTAX_POSIX_EXTENDED + : RE_SYNTAX_POSIX_BASIC); + + preg->buffer = NULL; + preg->allocated = 0; + preg->used = 0; + + /* Try to allocate space for the fastmap. */ + preg->fastmap = re_malloc (char, SBC_MAX); + if (BE (preg->fastmap == NULL, 0)) + return REG_ESPACE; + + syntax |= (cflags & REG_ICASE) ? RE_ICASE : 0; + + /* If REG_NEWLINE is set, newlines are treated differently. */ + if (cflags & REG_NEWLINE) + { /* REG_NEWLINE implies neither . nor [^...] match newline. */ + syntax &= ~RE_DOT_NEWLINE; + syntax |= RE_HAT_LISTS_NOT_NEWLINE; + /* It also changes the matching behavior. */ + preg->newline_anchor = 1; + } + else + preg->newline_anchor = 0; + preg->no_sub = !!(cflags & REG_NOSUB); + preg->translate = NULL; + + ret = re_compile_internal (preg, pattern, strlen (pattern), syntax); + + /* POSIX doesn't distinguish between an unmatched open-group and an + unmatched close-group: both are REG_EPAREN. */ + if (ret == REG_ERPAREN) + ret = REG_EPAREN; + + /* We have already checked preg->fastmap != NULL. */ + if (BE (ret == REG_NOERROR, 1)) + /* Compute the fastmap now, since regexec cannot modify the pattern + buffer. This function never fails in this implementation. */ + (void) re_compile_fastmap (preg); + else + { + /* Some error occurred while compiling the expression. */ + re_free (preg->fastmap); + preg->fastmap = NULL; + } + + return (int) ret; +} +#ifdef _LIBC +weak_alias (__regcomp, regcomp) +#endif + +/* Returns a message corresponding to an error code, ERRCODE, returned + from either regcomp or regexec. We don't use PREG here. */ + +#ifdef _LIBC +size_t +regerror (errcode, preg, errbuf, errbuf_size) + int errcode; + const regex_t *_Restrict_ preg; + char *_Restrict_ errbuf; + size_t errbuf_size; +#else /* size_t might promote */ +size_t +regerror (int errcode, const regex_t *_Restrict_ preg, + char *_Restrict_ errbuf, size_t errbuf_size) +#endif +{ + const char *msg; + size_t msg_size; + + if (BE (errcode < 0 + || errcode >= (int) (sizeof (__re_error_msgid_idx) + / sizeof (__re_error_msgid_idx[0])), 0)) + /* Only error codes returned by the rest of the code should be passed + to this routine. If we are given anything else, or if other regex + code generates an invalid error code, then the program has a bug. + Dump core so we can fix it. */ + abort (); + + msg = gettext (__re_error_msgid + __re_error_msgid_idx[errcode]); + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (BE (errbuf_size != 0, 1)) + { + size_t cpy_size = msg_size; + if (BE (msg_size > errbuf_size, 0)) + { + cpy_size = errbuf_size - 1; + errbuf[cpy_size] = '\0'; + } + memcpy (errbuf, msg, cpy_size); + } + + return msg_size; +} +#ifdef _LIBC +weak_alias (__regerror, regerror) +#endif + + +#ifdef RE_ENABLE_I18N +/* This static array is used for the map to single-byte characters when + UTF-8 is used. Otherwise we would allocate memory just to initialize + it the same all the time. UTF-8 is the preferred encoding so this is + a worthwhile optimization. */ +static const bitset_t utf8_sb_map = +{ + /* Set the first 128 bits. */ +# if 4 * BITSET_WORD_BITS < ASCII_CHARS +# error "bitset_word_t is narrower than 32 bits" +# elif 3 * BITSET_WORD_BITS < ASCII_CHARS + BITSET_WORD_MAX, BITSET_WORD_MAX, BITSET_WORD_MAX, +# elif 2 * BITSET_WORD_BITS < ASCII_CHARS + BITSET_WORD_MAX, BITSET_WORD_MAX, +# elif 1 * BITSET_WORD_BITS < ASCII_CHARS + BITSET_WORD_MAX, +# endif + (BITSET_WORD_MAX + >> (SBC_MAX % BITSET_WORD_BITS == 0 + ? 0 + : BITSET_WORD_BITS - SBC_MAX % BITSET_WORD_BITS)) +}; +#endif + + +static void +free_dfa_content (re_dfa_t *dfa) +{ + Idx i, j; + + if (dfa->nodes) + for (i = 0; i < dfa->nodes_len; ++i) + free_token (dfa->nodes + i); + re_free (dfa->nexts); + for (i = 0; i < dfa->nodes_len; ++i) + { + if (dfa->eclosures != NULL) + re_node_set_free (dfa->eclosures + i); + if (dfa->inveclosures != NULL) + re_node_set_free (dfa->inveclosures + i); + if (dfa->edests != NULL) + re_node_set_free (dfa->edests + i); + } + re_free (dfa->edests); + re_free (dfa->eclosures); + re_free (dfa->inveclosures); + re_free (dfa->nodes); + + if (dfa->state_table) + for (i = 0; i <= dfa->state_hash_mask; ++i) + { + struct re_state_table_entry *entry = dfa->state_table + i; + for (j = 0; j < entry->num; ++j) + { + re_dfastate_t *state = entry->array[j]; + free_state (state); + } + re_free (entry->array); + } + re_free (dfa->state_table); +#ifdef RE_ENABLE_I18N + if (dfa->sb_char != utf8_sb_map) + re_free (dfa->sb_char); +#endif + re_free (dfa->subexp_map); +#ifdef DEBUG + re_free (dfa->re_str); +#endif + + re_free (dfa); +} + + +/* Free dynamically allocated space used by PREG. */ + +void +regfree (preg) + regex_t *preg; +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + if (BE (dfa != NULL, 1)) + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + + re_free (preg->fastmap); + preg->fastmap = NULL; + + re_free (preg->translate); + preg->translate = NULL; +} +#ifdef _LIBC +weak_alias (__regfree, regfree) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +# ifdef _LIBC +/* Make these definitions weak in libc, so POSIX programs can redefine + these names if they don't use our functions, and still use + regcomp/regexec above without link errors. */ +weak_function +# endif +re_comp (s) + const char *s; +{ + reg_errcode_t ret; + char *fastmap; + + if (!s) + { + if (!re_comp_buf.buffer) + return gettext ("No previous regular expression"); + return 0; + } + + if (re_comp_buf.buffer) + { + fastmap = re_comp_buf.fastmap; + re_comp_buf.fastmap = NULL; + __regfree (&re_comp_buf); + memset (&re_comp_buf, '\0', sizeof (re_comp_buf)); + re_comp_buf.fastmap = fastmap; + } + + if (re_comp_buf.fastmap == NULL) + { + re_comp_buf.fastmap = (char *) malloc (SBC_MAX); + if (re_comp_buf.fastmap == NULL) + return (char *) gettext (__re_error_msgid + + __re_error_msgid_idx[(int) REG_ESPACE]); + } + + /* Since `re_exec' always passes NULL for the `regs' argument, we + don't need to initialize the pattern buffer fields which affect it. */ + + /* Match anchors at newlines. */ + re_comp_buf.newline_anchor = 1; + + ret = re_compile_internal (&re_comp_buf, s, strlen (s), re_syntax_options); + + if (!ret) + return NULL; + + /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */ + return (char *) gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]); +} + +#ifdef _LIBC +libc_freeres_fn (free_mem) +{ + __regfree (&re_comp_buf); +} +#endif + +#endif /* _REGEX_RE_COMP */ + +/* Internal entry point. + Compile the regular expression PATTERN, whose length is LENGTH. + SYNTAX indicate regular expression's syntax. */ + +static reg_errcode_t +re_compile_internal (regex_t *preg, const char * pattern, size_t length, + reg_syntax_t syntax) +{ + reg_errcode_t err = REG_NOERROR; + re_dfa_t *dfa; + re_string_t regexp; + + /* Initialize the pattern buffer. */ + preg->fastmap_accurate = 0; + preg->syntax = syntax; + preg->not_bol = preg->not_eol = 0; + preg->used = 0; + preg->re_nsub = 0; + preg->can_be_null = 0; + preg->regs_allocated = REGS_UNALLOCATED; + + /* Initialize the dfa. */ + dfa = (re_dfa_t *) preg->buffer; + if (BE (preg->allocated < sizeof (re_dfa_t), 0)) + { + /* If zero allocated, but buffer is non-null, try to realloc + enough space. This loses if buffer's address is bogus, but + that is the user's responsibility. If ->buffer is NULL this + is a simple allocation. */ + dfa = re_realloc (preg->buffer, re_dfa_t, 1); + if (dfa == NULL) + return REG_ESPACE; + preg->allocated = sizeof (re_dfa_t); + preg->buffer = (unsigned char *) dfa; + } + preg->used = sizeof (re_dfa_t); + + err = init_dfa (dfa, length); + if (BE (err != REG_NOERROR, 0)) + { + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + return err; + } +#ifdef DEBUG + /* Note: length+1 will not overflow since it is checked in init_dfa. */ + dfa->re_str = re_malloc (char, length + 1); + strncpy (dfa->re_str, pattern, length + 1); +#endif + + __libc_lock_init (dfa->lock); + + err = re_string_construct (®exp, pattern, length, preg->translate, + syntax & RE_ICASE, dfa); + if (BE (err != REG_NOERROR, 0)) + { + re_compile_internal_free_return: + free_workarea_compile (preg); + re_string_destruct (®exp); + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + return err; + } + + /* Parse the regular expression, and build a structure tree. */ + preg->re_nsub = 0; + dfa->str_tree = parse (®exp, preg, syntax, &err); + if (BE (dfa->str_tree == NULL, 0)) + goto re_compile_internal_free_return; + + /* Analyze the tree and create the nfa. */ + err = analyze (preg); + if (BE (err != REG_NOERROR, 0)) + goto re_compile_internal_free_return; + +#ifdef RE_ENABLE_I18N + /* If possible, do searching in single byte encoding to speed things up. */ + if (dfa->is_utf8 && !(syntax & RE_ICASE) && preg->translate == NULL) + optimize_utf8 (dfa); +#endif + + /* Then create the initial state of the dfa. */ + err = create_initial_state (dfa); + + /* Release work areas. */ + free_workarea_compile (preg); + re_string_destruct (®exp); + + if (BE (err != REG_NOERROR, 0)) + { + free_dfa_content (dfa); + preg->buffer = NULL; + preg->allocated = 0; + } + + return err; +} + +/* Initialize DFA. We use the length of the regular expression PAT_LEN + as the initial length of some arrays. */ + +static reg_errcode_t +init_dfa (re_dfa_t *dfa, size_t pat_len) +{ + __re_size_t table_size; +#ifdef RE_ENABLE_I18N + size_t max_i18n_object_size = MAX (sizeof (wchar_t), sizeof (wctype_t)); +#else + size_t max_i18n_object_size = 0; +#endif + size_t max_object_size = + MAX (sizeof (struct re_state_table_entry), + MAX (sizeof (re_token_t), + MAX (sizeof (re_node_set), + MAX (sizeof (regmatch_t), + max_i18n_object_size)))); + + memset (dfa, '\0', sizeof (re_dfa_t)); + + /* Force allocation of str_tree_storage the first time. */ + dfa->str_tree_storage_idx = BIN_TREE_STORAGE_SIZE; + + /* Avoid overflows. The extra "/ 2" is for the table_size doubling + calculation below, and for similar doubling calculations + elsewhere. And it's <= rather than <, because some of the + doubling calculations add 1 afterwards. */ + if (BE (SIZE_MAX / max_object_size / 2 <= pat_len, 0)) + return REG_ESPACE; + + dfa->nodes_alloc = pat_len + 1; + dfa->nodes = re_malloc (re_token_t, dfa->nodes_alloc); + + /* table_size = 2 ^ ceil(log pat_len) */ + for (table_size = 1; ; table_size <<= 1) + if (table_size > pat_len) + break; + + dfa->state_table = calloc (sizeof (struct re_state_table_entry), table_size); + dfa->state_hash_mask = table_size - 1; + + dfa->mb_cur_max = MB_CUR_MAX; +#ifdef _LIBC + if (dfa->mb_cur_max == 6 + && strcmp (_NL_CURRENT (LC_CTYPE, _NL_CTYPE_CODESET_NAME), "UTF-8") == 0) + dfa->is_utf8 = 1; + dfa->map_notascii = (_NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_MAP_TO_NONASCII) + != 0); +#else + if (strcmp (locale_charset (), "UTF-8") == 0) + dfa->is_utf8 = 1; + + /* We check exhaustively in the loop below if this charset is a + superset of ASCII. */ + dfa->map_notascii = 0; +#endif + +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + if (dfa->is_utf8) + dfa->sb_char = (re_bitset_ptr_t) utf8_sb_map; + else + { + int i, j, ch; + + dfa->sb_char = (re_bitset_ptr_t) calloc (sizeof (bitset_t), 1); + if (BE (dfa->sb_char == NULL, 0)) + return REG_ESPACE; + + /* Set the bits corresponding to single byte chars. */ + for (i = 0, ch = 0; i < BITSET_WORDS; ++i) + for (j = 0; j < BITSET_WORD_BITS; ++j, ++ch) + { + wint_t wch = __btowc (ch); + if (wch != WEOF) + dfa->sb_char[i] |= (bitset_word_t) 1 << j; +# ifndef _LIBC + if (isascii (ch) && wch != ch) + dfa->map_notascii = 1; +# endif + } + } + } +#endif + + if (BE (dfa->nodes == NULL || dfa->state_table == NULL, 0)) + return REG_ESPACE; + return REG_NOERROR; +} + +/* Initialize WORD_CHAR table, which indicate which character is + "word". In this case "word" means that it is the word construction + character used by some operators like "\<", "\>", etc. */ + +static void +internal_function +init_word_char (re_dfa_t *dfa) +{ + int i, j, ch; + dfa->word_ops_used = 1; + for (i = 0, ch = 0; i < BITSET_WORDS; ++i) + for (j = 0; j < BITSET_WORD_BITS; ++j, ++ch) + if (isalnum (ch) || ch == '_') + dfa->word_char[i] |= (bitset_word_t) 1 << j; +} + +/* Free the work area which are only used while compiling. */ + +static void +free_workarea_compile (regex_t *preg) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_storage_t *storage, *next; + for (storage = dfa->str_tree_storage; storage; storage = next) + { + next = storage->next; + re_free (storage); + } + dfa->str_tree_storage = NULL; + dfa->str_tree_storage_idx = BIN_TREE_STORAGE_SIZE; + dfa->str_tree = NULL; + re_free (dfa->org_indices); + dfa->org_indices = NULL; +} + +/* Create initial states for all contexts. */ + +static reg_errcode_t +create_initial_state (re_dfa_t *dfa) +{ + Idx first, i; + reg_errcode_t err; + re_node_set init_nodes; + + /* Initial states have the epsilon closure of the node which is + the first node of the regular expression. */ + first = dfa->str_tree->first->node_idx; + dfa->init_node = first; + err = re_node_set_init_copy (&init_nodes, dfa->eclosures + first); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* The back-references which are in initial states can epsilon transit, + since in this case all of the subexpressions can be null. + Then we add epsilon closures of the nodes which are the next nodes of + the back-references. */ + if (dfa->nbackref > 0) + for (i = 0; i < init_nodes.nelem; ++i) + { + Idx node_idx = init_nodes.elems[i]; + re_token_type_t type = dfa->nodes[node_idx].type; + + Idx clexp_idx; + if (type != OP_BACK_REF) + continue; + for (clexp_idx = 0; clexp_idx < init_nodes.nelem; ++clexp_idx) + { + re_token_t *clexp_node; + clexp_node = dfa->nodes + init_nodes.elems[clexp_idx]; + if (clexp_node->type == OP_CLOSE_SUBEXP + && clexp_node->opr.idx == dfa->nodes[node_idx].opr.idx) + break; + } + if (clexp_idx == init_nodes.nelem) + continue; + + if (type == OP_BACK_REF) + { + Idx dest_idx = dfa->edests[node_idx].elems[0]; + if (!re_node_set_contains (&init_nodes, dest_idx)) + { + re_node_set_merge (&init_nodes, dfa->eclosures + dest_idx); + i = 0; + } + } + } + + /* It must be the first time to invoke acquire_state. */ + dfa->init_state = re_acquire_state_context (&err, dfa, &init_nodes, 0); + /* We don't check ERR here, since the initial state must not be NULL. */ + if (BE (dfa->init_state == NULL, 0)) + return err; + if (dfa->init_state->has_constraint) + { + dfa->init_state_word = re_acquire_state_context (&err, dfa, &init_nodes, + CONTEXT_WORD); + dfa->init_state_nl = re_acquire_state_context (&err, dfa, &init_nodes, + CONTEXT_NEWLINE); + dfa->init_state_begbuf = re_acquire_state_context (&err, dfa, + &init_nodes, + CONTEXT_NEWLINE + | CONTEXT_BEGBUF); + if (BE (dfa->init_state_word == NULL || dfa->init_state_nl == NULL + || dfa->init_state_begbuf == NULL, 0)) + return err; + } + else + dfa->init_state_word = dfa->init_state_nl + = dfa->init_state_begbuf = dfa->init_state; + + re_node_set_free (&init_nodes); + return REG_NOERROR; +} + +#ifdef RE_ENABLE_I18N +/* If it is possible to do searching in single byte encoding instead of UTF-8 + to speed things up, set dfa->mb_cur_max to 1, clear is_utf8 and change + DFA nodes where needed. */ + +static void +optimize_utf8 (re_dfa_t *dfa) +{ + Idx node; + int i; + bool mb_chars = false; + bool has_period = false; + + for (node = 0; node < dfa->nodes_len; ++node) + switch (dfa->nodes[node].type) + { + case CHARACTER: + if (dfa->nodes[node].opr.c >= ASCII_CHARS) + mb_chars = true; + break; + case ANCHOR: + switch (dfa->nodes[node].opr.idx) + { + case LINE_FIRST: + case LINE_LAST: + case BUF_FIRST: + case BUF_LAST: + break; + default: + /* Word anchors etc. cannot be handled. */ + return; + } + break; + case OP_PERIOD: + has_period = true; + break; + case OP_BACK_REF: + case OP_ALT: + case END_OF_RE: + case OP_DUP_ASTERISK: + case OP_OPEN_SUBEXP: + case OP_CLOSE_SUBEXP: + break; + case COMPLEX_BRACKET: + return; + case SIMPLE_BRACKET: + /* Just double check. */ + { + int rshift = (ASCII_CHARS % BITSET_WORD_BITS == 0 + ? 0 + : BITSET_WORD_BITS - ASCII_CHARS % BITSET_WORD_BITS); + for (i = ASCII_CHARS / BITSET_WORD_BITS; i < BITSET_WORDS; ++i) + { + if (dfa->nodes[node].opr.sbcset[i] >> rshift != 0) + return; + rshift = 0; + } + } + break; + default: + abort (); + } + + if (mb_chars || has_period) + for (node = 0; node < dfa->nodes_len; ++node) + { + if (dfa->nodes[node].type == CHARACTER + && dfa->nodes[node].opr.c >= ASCII_CHARS) + dfa->nodes[node].mb_partial = 0; + else if (dfa->nodes[node].type == OP_PERIOD) + dfa->nodes[node].type = OP_UTF8_PERIOD; + } + + /* The search can be in single byte locale. */ + dfa->mb_cur_max = 1; + dfa->is_utf8 = 0; + dfa->has_mb_node = dfa->nbackref > 0 || has_period; +} +#endif + +/* Analyze the structure tree, and calculate "first", "next", "edest", + "eclosure", and "inveclosure". */ + +static reg_errcode_t +analyze (regex_t *preg) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + reg_errcode_t ret; + + /* Allocate arrays. */ + dfa->nexts = re_malloc (Idx, dfa->nodes_alloc); + dfa->org_indices = re_malloc (Idx, dfa->nodes_alloc); + dfa->edests = re_malloc (re_node_set, dfa->nodes_alloc); + dfa->eclosures = re_malloc (re_node_set, dfa->nodes_alloc); + if (BE (dfa->nexts == NULL || dfa->org_indices == NULL || dfa->edests == NULL + || dfa->eclosures == NULL, 0)) + return REG_ESPACE; + + dfa->subexp_map = re_malloc (Idx, preg->re_nsub); + if (dfa->subexp_map != NULL) + { + Idx i; + for (i = 0; i < preg->re_nsub; i++) + dfa->subexp_map[i] = i; + preorder (dfa->str_tree, optimize_subexps, dfa); + for (i = 0; i < preg->re_nsub; i++) + if (dfa->subexp_map[i] != i) + break; + if (i == preg->re_nsub) + { + free (dfa->subexp_map); + dfa->subexp_map = NULL; + } + } + + ret = postorder (dfa->str_tree, lower_subexps, preg); + if (BE (ret != REG_NOERROR, 0)) + return ret; + ret = postorder (dfa->str_tree, calc_first, dfa); + if (BE (ret != REG_NOERROR, 0)) + return ret; + preorder (dfa->str_tree, calc_next, dfa); + ret = preorder (dfa->str_tree, link_nfa_nodes, dfa); + if (BE (ret != REG_NOERROR, 0)) + return ret; + ret = calc_eclosure (dfa); + if (BE (ret != REG_NOERROR, 0)) + return ret; + + /* We only need this during the prune_impossible_nodes pass in regexec.c; + skip it if p_i_n will not run, as calc_inveclosure can be quadratic. */ + if ((!preg->no_sub && preg->re_nsub > 0 && dfa->has_plural_match) + || dfa->nbackref) + { + dfa->inveclosures = re_malloc (re_node_set, dfa->nodes_len); + if (BE (dfa->inveclosures == NULL, 0)) + return REG_ESPACE; + ret = calc_inveclosure (dfa); + } + + return ret; +} + +/* Our parse trees are very unbalanced, so we cannot use a stack to + implement parse tree visits. Instead, we use parent pointers and + some hairy code in these two functions. */ +static reg_errcode_t +postorder (bin_tree_t *root, reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra) +{ + bin_tree_t *node, *prev; + + for (node = root; ; ) + { + /* Descend down the tree, preferably to the left (or to the right + if that's the only child). */ + while (node->left || node->right) + if (node->left) + node = node->left; + else + node = node->right; + + do + { + reg_errcode_t err = fn (extra, node); + if (BE (err != REG_NOERROR, 0)) + return err; + if (node->parent == NULL) + return REG_NOERROR; + prev = node; + node = node->parent; + } + /* Go up while we have a node that is reached from the right. */ + while (node->right == prev || node->right == NULL); + node = node->right; + } +} + +static reg_errcode_t +preorder (bin_tree_t *root, reg_errcode_t (fn (void *, bin_tree_t *)), + void *extra) +{ + bin_tree_t *node; + + for (node = root; ; ) + { + reg_errcode_t err = fn (extra, node); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* Go to the left node, or up and to the right. */ + if (node->left) + node = node->left; + else + { + bin_tree_t *prev = NULL; + while (node->right == prev || node->right == NULL) + { + prev = node; + node = node->parent; + if (!node) + return REG_NOERROR; + } + node = node->right; + } + } +} + +/* Optimization pass: if a SUBEXP is entirely contained, strip it and tell + re_search_internal to map the inner one's opr.idx to this one's. Adjust + backreferences as well. Requires a preorder visit. */ +static reg_errcode_t +optimize_subexps (void *extra, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) extra; + + if (node->token.type == OP_BACK_REF && dfa->subexp_map) + { + int idx = node->token.opr.idx; + node->token.opr.idx = dfa->subexp_map[idx]; + dfa->used_bkref_map |= 1 << node->token.opr.idx; + } + + else if (node->token.type == SUBEXP + && node->left && node->left->token.type == SUBEXP) + { + Idx other_idx = node->left->token.opr.idx; + + node->left = node->left->left; + if (node->left) + node->left->parent = node; + + dfa->subexp_map[other_idx] = dfa->subexp_map[node->token.opr.idx]; + if (other_idx < BITSET_WORD_BITS) + dfa->used_bkref_map &= ~((bitset_word_t) 1 << other_idx); + } + + return REG_NOERROR; +} + +/* Lowering pass: Turn each SUBEXP node into the appropriate concatenation + of OP_OPEN_SUBEXP, the body of the SUBEXP (if any) and OP_CLOSE_SUBEXP. */ +static reg_errcode_t +lower_subexps (void *extra, bin_tree_t *node) +{ + regex_t *preg = (regex_t *) extra; + reg_errcode_t err = REG_NOERROR; + + if (node->left && node->left->token.type == SUBEXP) + { + node->left = lower_subexp (&err, preg, node->left); + if (node->left) + node->left->parent = node; + } + if (node->right && node->right->token.type == SUBEXP) + { + node->right = lower_subexp (&err, preg, node->right); + if (node->right) + node->right->parent = node; + } + + return err; +} + +static bin_tree_t * +lower_subexp (reg_errcode_t *err, regex_t *preg, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *body = node->left; + bin_tree_t *op, *cls, *tree1, *tree; + + if (preg->no_sub + /* We do not optimize empty subexpressions, because otherwise we may + have bad CONCAT nodes with NULL children. This is obviously not + very common, so we do not lose much. An example that triggers + this case is the sed "script" /\(\)/x. */ + && node->left != NULL + && (node->token.opr.idx >= BITSET_WORD_BITS + || !(dfa->used_bkref_map + & ((bitset_word_t) 1 << node->token.opr.idx)))) + return node->left; + + /* Convert the SUBEXP node to the concatenation of an + OP_OPEN_SUBEXP, the contents, and an OP_CLOSE_SUBEXP. */ + op = create_tree (dfa, NULL, NULL, OP_OPEN_SUBEXP); + cls = create_tree (dfa, NULL, NULL, OP_CLOSE_SUBEXP); + tree1 = body ? create_tree (dfa, body, cls, CONCAT) : cls; + tree = create_tree (dfa, op, tree1, CONCAT); + if (BE (tree == NULL || tree1 == NULL || op == NULL || cls == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + + op->token.opr.idx = cls->token.opr.idx = node->token.opr.idx; + op->token.opt_subexp = cls->token.opt_subexp = node->token.opt_subexp; + return tree; +} + +/* Pass 1 in building the NFA: compute FIRST and create unlinked automaton + nodes. Requires a postorder visit. */ +static reg_errcode_t +calc_first (void *extra, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) extra; + if (node->token.type == CONCAT) + { + node->first = node->left->first; + node->node_idx = node->left->node_idx; + } + else + { + node->first = node; + node->node_idx = re_dfa_add_node (dfa, node->token); + if (BE (node->node_idx == REG_MISSING, 0)) + return REG_ESPACE; + } + return REG_NOERROR; +} + +/* Pass 2: compute NEXT on the tree. Preorder visit. */ +static reg_errcode_t +calc_next (void *extra, bin_tree_t *node) +{ + switch (node->token.type) + { + case OP_DUP_ASTERISK: + node->left->next = node; + break; + case CONCAT: + node->left->next = node->right->first; + node->right->next = node->next; + break; + default: + if (node->left) + node->left->next = node->next; + if (node->right) + node->right->next = node->next; + break; + } + return REG_NOERROR; +} + +/* Pass 3: link all DFA nodes to their NEXT node (any order will do). */ +static reg_errcode_t +link_nfa_nodes (void *extra, bin_tree_t *node) +{ + re_dfa_t *dfa = (re_dfa_t *) extra; + Idx idx = node->node_idx; + reg_errcode_t err = REG_NOERROR; + + switch (node->token.type) + { + case CONCAT: + break; + + case END_OF_RE: + assert (node->next == NULL); + break; + + case OP_DUP_ASTERISK: + case OP_ALT: + { + Idx left, right; + dfa->has_plural_match = 1; + if (node->left != NULL) + left = node->left->first->node_idx; + else + left = node->next->node_idx; + if (node->right != NULL) + right = node->right->first->node_idx; + else + right = node->next->node_idx; + assert (REG_VALID_INDEX (left)); + assert (REG_VALID_INDEX (right)); + err = re_node_set_init_2 (dfa->edests + idx, left, right); + } + break; + + case ANCHOR: + case OP_OPEN_SUBEXP: + case OP_CLOSE_SUBEXP: + err = re_node_set_init_1 (dfa->edests + idx, node->next->node_idx); + break; + + case OP_BACK_REF: + dfa->nexts[idx] = node->next->node_idx; + if (node->token.type == OP_BACK_REF) + re_node_set_init_1 (dfa->edests + idx, dfa->nexts[idx]); + break; + + default: + assert (!IS_EPSILON_NODE (node->token.type)); + dfa->nexts[idx] = node->next->node_idx; + break; + } + + return err; +} + +/* Duplicate the epsilon closure of the node ROOT_NODE. + Note that duplicated nodes have constraint INIT_CONSTRAINT in addition + to their own constraint. */ + +static reg_errcode_t +internal_function +duplicate_node_closure (re_dfa_t *dfa, Idx top_org_node, Idx top_clone_node, + Idx root_node, unsigned int init_constraint) +{ + Idx org_node, clone_node; + bool ok; + unsigned int constraint = init_constraint; + for (org_node = top_org_node, clone_node = top_clone_node;;) + { + Idx org_dest, clone_dest; + if (dfa->nodes[org_node].type == OP_BACK_REF) + { + /* If the back reference epsilon-transit, its destination must + also have the constraint. Then duplicate the epsilon closure + of the destination of the back reference, and store it in + edests of the back reference. */ + org_dest = dfa->nexts[org_node]; + re_node_set_empty (dfa->edests + clone_node); + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + dfa->nexts[clone_node] = dfa->nexts[org_node]; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + else if (dfa->edests[org_node].nelem == 0) + { + /* In case of the node can't epsilon-transit, don't duplicate the + destination and store the original destination as the + destination of the node. */ + dfa->nexts[clone_node] = dfa->nexts[org_node]; + break; + } + else if (dfa->edests[org_node].nelem == 1) + { + /* In case of the node can epsilon-transit, and it has only one + destination. */ + org_dest = dfa->edests[org_node].elems[0]; + re_node_set_empty (dfa->edests + clone_node); + if (dfa->nodes[org_node].type == ANCHOR) + { + /* In case of the node has another constraint, append it. */ + if (org_node == root_node && clone_node != org_node) + { + /* ...but if the node is root_node itself, it means the + epsilon closure have a loop, then tie it to the + destination of the root_node. */ + ok = re_node_set_insert (dfa->edests + clone_node, org_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + break; + } + constraint |= dfa->nodes[org_node].opr.ctx_type; + } + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + else /* dfa->edests[org_node].nelem == 2 */ + { + /* In case of the node can epsilon-transit, and it has two + destinations. In the bin_tree_t and DFA, that's '|' and '*'. */ + org_dest = dfa->edests[org_node].elems[0]; + re_node_set_empty (dfa->edests + clone_node); + /* Search for a duplicated node which satisfies the constraint. */ + clone_dest = search_duplicated_node (dfa, org_dest, constraint); + if (clone_dest == REG_MISSING) + { + /* There are no such a duplicated node, create a new one. */ + reg_errcode_t err; + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + err = duplicate_node_closure (dfa, org_dest, clone_dest, + root_node, constraint); + if (BE (err != REG_NOERROR, 0)) + return err; + } + else + { + /* There are a duplicated node which satisfy the constraint, + use it to avoid infinite loop. */ + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + + org_dest = dfa->edests[org_node].elems[1]; + clone_dest = duplicate_node (dfa, org_dest, constraint); + if (BE (clone_dest == REG_MISSING, 0)) + return REG_ESPACE; + ok = re_node_set_insert (dfa->edests + clone_node, clone_dest); + if (BE (! ok, 0)) + return REG_ESPACE; + } + org_node = org_dest; + clone_node = clone_dest; + } + return REG_NOERROR; +} + +/* Search for a node which is duplicated from the node ORG_NODE, and + satisfies the constraint CONSTRAINT. */ + +static Idx +search_duplicated_node (const re_dfa_t *dfa, Idx org_node, + unsigned int constraint) +{ + Idx idx; + for (idx = dfa->nodes_len - 1; dfa->nodes[idx].duplicated && idx > 0; --idx) + { + if (org_node == dfa->org_indices[idx] + && constraint == dfa->nodes[idx].constraint) + return idx; /* Found. */ + } + return REG_MISSING; /* Not found. */ +} + +/* Duplicate the node whose index is ORG_IDX and set the constraint CONSTRAINT. + Return the index of the new node, or REG_MISSING if insufficient storage is + available. */ + +static Idx +duplicate_node (re_dfa_t *dfa, Idx org_idx, unsigned int constraint) +{ + Idx dup_idx = re_dfa_add_node (dfa, dfa->nodes[org_idx]); + if (BE (dup_idx != REG_MISSING, 1)) + { + dfa->nodes[dup_idx].constraint = constraint; + if (dfa->nodes[org_idx].type == ANCHOR) + dfa->nodes[dup_idx].constraint |= dfa->nodes[org_idx].opr.ctx_type; + dfa->nodes[dup_idx].duplicated = 1; + + /* Store the index of the original node. */ + dfa->org_indices[dup_idx] = org_idx; + } + return dup_idx; +} + +static reg_errcode_t +calc_inveclosure (re_dfa_t *dfa) +{ + Idx src, idx; + bool ok; + for (idx = 0; idx < dfa->nodes_len; ++idx) + re_node_set_init_empty (dfa->inveclosures + idx); + + for (src = 0; src < dfa->nodes_len; ++src) + { + Idx *elems = dfa->eclosures[src].elems; + for (idx = 0; idx < dfa->eclosures[src].nelem; ++idx) + { + ok = re_node_set_insert_last (dfa->inveclosures + elems[idx], src); + if (BE (! ok, 0)) + return REG_ESPACE; + } + } + + return REG_NOERROR; +} + +/* Calculate "eclosure" for all the node in DFA. */ + +static reg_errcode_t +calc_eclosure (re_dfa_t *dfa) +{ + Idx node_idx; + bool incomplete; +#ifdef DEBUG + assert (dfa->nodes_len > 0); +#endif + incomplete = false; + /* For each nodes, calculate epsilon closure. */ + for (node_idx = 0; ; ++node_idx) + { + reg_errcode_t err; + re_node_set eclosure_elem; + if (node_idx == dfa->nodes_len) + { + if (!incomplete) + break; + incomplete = false; + node_idx = 0; + } + +#ifdef DEBUG + assert (dfa->eclosures[node_idx].nelem != REG_MISSING); +#endif + + /* If we have already calculated, skip it. */ + if (dfa->eclosures[node_idx].nelem != 0) + continue; + /* Calculate epsilon closure of `node_idx'. */ + err = calc_eclosure_iter (&eclosure_elem, dfa, node_idx, true); + if (BE (err != REG_NOERROR, 0)) + return err; + + if (dfa->eclosures[node_idx].nelem == 0) + { + incomplete = true; + re_node_set_free (&eclosure_elem); + } + } + return REG_NOERROR; +} + +/* Calculate epsilon closure of NODE. */ + +static reg_errcode_t +calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, Idx node, bool root) +{ + reg_errcode_t err; + unsigned int constraint; + Idx i; + bool incomplete; + bool ok; + re_node_set eclosure; + incomplete = false; + err = re_node_set_alloc (&eclosure, dfa->edests[node].nelem + 1); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* This indicates that we are calculating this node now. + We reference this value to avoid infinite loop. */ + dfa->eclosures[node].nelem = REG_MISSING; + + constraint = ((dfa->nodes[node].type == ANCHOR) + ? dfa->nodes[node].opr.ctx_type : 0); + /* If the current node has constraints, duplicate all nodes. + Since they must inherit the constraints. */ + if (constraint + && dfa->edests[node].nelem + && !dfa->nodes[dfa->edests[node].elems[0]].duplicated) + { + err = duplicate_node_closure (dfa, node, node, node, constraint); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + /* Expand each epsilon destination nodes. */ + if (IS_EPSILON_NODE(dfa->nodes[node].type)) + for (i = 0; i < dfa->edests[node].nelem; ++i) + { + re_node_set eclosure_elem; + Idx edest = dfa->edests[node].elems[i]; + /* If calculating the epsilon closure of `edest' is in progress, + return intermediate result. */ + if (dfa->eclosures[edest].nelem == REG_MISSING) + { + incomplete = true; + continue; + } + /* If we haven't calculated the epsilon closure of `edest' yet, + calculate now. Otherwise use calculated epsilon closure. */ + if (dfa->eclosures[edest].nelem == 0) + { + err = calc_eclosure_iter (&eclosure_elem, dfa, edest, false); + if (BE (err != REG_NOERROR, 0)) + return err; + } + else + eclosure_elem = dfa->eclosures[edest]; + /* Merge the epsilon closure of `edest'. */ + re_node_set_merge (&eclosure, &eclosure_elem); + /* If the epsilon closure of `edest' is incomplete, + the epsilon closure of this node is also incomplete. */ + if (dfa->eclosures[edest].nelem == 0) + { + incomplete = true; + re_node_set_free (&eclosure_elem); + } + } + + /* Epsilon closures include itself. */ + ok = re_node_set_insert (&eclosure, node); + if (BE (! ok, 0)) + return REG_ESPACE; + if (incomplete && !root) + dfa->eclosures[node].nelem = 0; + else + dfa->eclosures[node] = eclosure; + *new_set = eclosure; + return REG_NOERROR; +} + +/* Functions for token which are used in the parser. */ + +/* Fetch a token from INPUT. + We must not use this function inside bracket expressions. */ + +static void +internal_function +fetch_token (re_token_t *result, re_string_t *input, reg_syntax_t syntax) +{ + re_string_skip_bytes (input, peek_token (result, input, syntax)); +} + +/* Peek a token from INPUT, and return the length of the token. + We must not use this function inside bracket expressions. */ + +static int +internal_function +peek_token (re_token_t *token, re_string_t *input, reg_syntax_t syntax) +{ + unsigned char c; + + if (re_string_eoi (input)) + { + token->type = END_OF_RE; + return 0; + } + + c = re_string_peek_byte (input, 0); + token->opr.c = c; + + token->word_char = 0; +#ifdef RE_ENABLE_I18N + token->mb_partial = 0; + if (input->mb_cur_max > 1 && + !re_string_first_byte (input, re_string_cur_idx (input))) + { + token->type = CHARACTER; + token->mb_partial = 1; + return 1; + } +#endif + if (c == '\\') + { + unsigned char c2; + if (re_string_cur_idx (input) + 1 >= re_string_length (input)) + { + token->type = BACK_SLASH; + return 1; + } + + c2 = re_string_peek_byte_case (input, 1); + token->opr.c = c2; + token->type = CHARACTER; +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1) + { + wint_t wc = re_string_wchar_at (input, + re_string_cur_idx (input) + 1); + token->word_char = IS_WIDE_WORD_CHAR (wc) != 0; + } + else +#endif + token->word_char = IS_WORD_CHAR (c2) != 0; + + switch (c2) + { + case '|': + if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_NO_BK_VBAR)) + token->type = OP_ALT; + break; + case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': + if (!(syntax & RE_NO_BK_REFS)) + { + token->type = OP_BACK_REF; + token->opr.idx = c2 - '1'; + } + break; + case '<': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = WORD_FIRST; + } + break; + case '>': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = WORD_LAST; + } + break; + case 'b': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = WORD_DELIM; + } + break; + case 'B': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = NOT_WORD_DELIM; + } + break; + case 'w': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_WORD; + break; + case 'W': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_NOTWORD; + break; + case 's': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_SPACE; + break; + case 'S': + if (!(syntax & RE_NO_GNU_OPS)) + token->type = OP_NOTSPACE; + break; + case '`': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = BUF_FIRST; + } + break; + case '\'': + if (!(syntax & RE_NO_GNU_OPS)) + { + token->type = ANCHOR; + token->opr.ctx_type = BUF_LAST; + } + break; + case '(': + if (!(syntax & RE_NO_BK_PARENS)) + token->type = OP_OPEN_SUBEXP; + break; + case ')': + if (!(syntax & RE_NO_BK_PARENS)) + token->type = OP_CLOSE_SUBEXP; + break; + case '+': + if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_PLUS; + break; + case '?': + if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_QUESTION; + break; + case '{': + if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES))) + token->type = OP_OPEN_DUP_NUM; + break; + case '}': + if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES))) + token->type = OP_CLOSE_DUP_NUM; + break; + default: + break; + } + return 2; + } + + token->type = CHARACTER; +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1) + { + wint_t wc = re_string_wchar_at (input, re_string_cur_idx (input)); + token->word_char = IS_WIDE_WORD_CHAR (wc) != 0; + } + else +#endif + token->word_char = IS_WORD_CHAR (token->opr.c); + + switch (c) + { + case '\n': + if (syntax & RE_NEWLINE_ALT) + token->type = OP_ALT; + break; + case '|': + if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_NO_BK_VBAR)) + token->type = OP_ALT; + break; + case '*': + token->type = OP_DUP_ASTERISK; + break; + case '+': + if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_PLUS; + break; + case '?': + if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM)) + token->type = OP_DUP_QUESTION; + break; + case '{': + if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + token->type = OP_OPEN_DUP_NUM; + break; + case '}': + if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + token->type = OP_CLOSE_DUP_NUM; + break; + case '(': + if (syntax & RE_NO_BK_PARENS) + token->type = OP_OPEN_SUBEXP; + break; + case ')': + if (syntax & RE_NO_BK_PARENS) + token->type = OP_CLOSE_SUBEXP; + break; + case '[': + token->type = OP_OPEN_BRACKET; + break; + case '.': + token->type = OP_PERIOD; + break; + case '^': + if (!(syntax & (RE_CONTEXT_INDEP_ANCHORS | RE_CARET_ANCHORS_HERE)) && + re_string_cur_idx (input) != 0) + { + char prev = re_string_peek_byte (input, -1); + if (!(syntax & RE_NEWLINE_ALT) || prev != '\n') + break; + } + token->type = ANCHOR; + token->opr.ctx_type = LINE_FIRST; + break; + case '$': + if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) && + re_string_cur_idx (input) + 1 != re_string_length (input)) + { + re_token_t next; + re_string_skip_bytes (input, 1); + peek_token (&next, input, syntax); + re_string_skip_bytes (input, -1); + if (next.type != OP_ALT && next.type != OP_CLOSE_SUBEXP) + break; + } + token->type = ANCHOR; + token->opr.ctx_type = LINE_LAST; + break; + default: + break; + } + return 1; +} + +/* Peek a token from INPUT, and return the length of the token. + We must not use this function out of bracket expressions. */ + +static int +internal_function +peek_token_bracket (re_token_t *token, re_string_t *input, reg_syntax_t syntax) +{ + unsigned char c; + if (re_string_eoi (input)) + { + token->type = END_OF_RE; + return 0; + } + c = re_string_peek_byte (input, 0); + token->opr.c = c; + +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1 && + !re_string_first_byte (input, re_string_cur_idx (input))) + { + token->type = CHARACTER; + return 1; + } +#endif /* RE_ENABLE_I18N */ + + if (c == '\\' && (syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) + && re_string_cur_idx (input) + 1 < re_string_length (input)) + { + /* In this case, '\' escape a character. */ + unsigned char c2; + re_string_skip_bytes (input, 1); + c2 = re_string_peek_byte (input, 0); + token->opr.c = c2; + token->type = CHARACTER; + return 1; + } + if (c == '[') /* '[' is a special char in a bracket exps. */ + { + unsigned char c2; + int token_len; + if (re_string_cur_idx (input) + 1 < re_string_length (input)) + c2 = re_string_peek_byte (input, 1); + else + c2 = 0; + token->opr.c = c2; + token_len = 2; + switch (c2) + { + case '.': + token->type = OP_OPEN_COLL_ELEM; + break; + case '=': + token->type = OP_OPEN_EQUIV_CLASS; + break; + case ':': + if (syntax & RE_CHAR_CLASSES) + { + token->type = OP_OPEN_CHAR_CLASS; + break; + } + /* else fall through. */ + default: + token->type = CHARACTER; + token->opr.c = c; + token_len = 1; + break; + } + return token_len; + } + switch (c) + { + case '-': + token->type = OP_CHARSET_RANGE; + break; + case ']': + token->type = OP_CLOSE_BRACKET; + break; + case '^': + token->type = OP_NON_MATCH_LIST; + break; + default: + token->type = CHARACTER; + } + return 1; +} + +/* Functions for parser. */ + +/* Entry point of the parser. + Parse the regular expression REGEXP and return the structure tree. + If an error is occured, ERR is set by error code, and return NULL. + This function build the following tree, from regular expression <reg_exp>: + CAT + / \ + / \ + <reg_exp> EOR + + CAT means concatenation. + EOR means end of regular expression. */ + +static bin_tree_t * +parse (re_string_t *regexp, regex_t *preg, reg_syntax_t syntax, + reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree, *eor, *root; + re_token_t current_token; + dfa->syntax = syntax; + fetch_token (¤t_token, regexp, syntax | RE_CARET_ANCHORS_HERE); + tree = parse_reg_exp (regexp, preg, ¤t_token, syntax, 0, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + eor = create_tree (dfa, NULL, NULL, END_OF_RE); + if (tree != NULL) + root = create_tree (dfa, tree, eor, CONCAT); + else + root = eor; + if (BE (eor == NULL || root == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + return root; +} + +/* This function build the following tree, from regular expression + <branch1>|<branch2>: + ALT + / \ + / \ + <branch1> <branch2> + + ALT means alternative, which represents the operator `|'. */ + +static bin_tree_t * +parse_reg_exp (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree, *branch = NULL; + tree = parse_branch (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + + while (token->type == OP_ALT) + { + fetch_token (token, regexp, syntax | RE_CARET_ANCHORS_HERE); + if (token->type != OP_ALT && token->type != END_OF_RE + && (nest == 0 || token->type != OP_CLOSE_SUBEXP)) + { + branch = parse_branch (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && branch == NULL, 0)) + return NULL; + } + else + branch = NULL; + tree = create_tree (dfa, tree, branch, OP_ALT); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + return tree; +} + +/* This function build the following tree, from regular expression + <exp1><exp2>: + CAT + / \ + / \ + <exp1> <exp2> + + CAT means concatenation. */ + +static bin_tree_t * +parse_branch (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + bin_tree_t *tree, *expr; + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + tree = parse_expression (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + + while (token->type != OP_ALT && token->type != END_OF_RE + && (nest == 0 || token->type != OP_CLOSE_SUBEXP)) + { + expr = parse_expression (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && expr == NULL, 0)) + { + return NULL; + } + if (tree != NULL && expr != NULL) + { + tree = create_tree (dfa, tree, expr, CONCAT); + if (tree == NULL) + { + *err = REG_ESPACE; + return NULL; + } + } + else if (tree == NULL) + tree = expr; + /* Otherwise expr == NULL, we don't need to create new tree. */ + } + return tree; +} + +/* This function build the following tree, from regular expression a*: + * + | + a +*/ + +static bin_tree_t * +parse_expression (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree; + switch (token->type) + { + case CHARACTER: + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + while (!re_string_eoi (regexp) + && !re_string_first_byte (regexp, re_string_cur_idx (regexp))) + { + bin_tree_t *mbc_remain; + fetch_token (token, regexp, syntax); + mbc_remain = create_token_tree (dfa, NULL, NULL, token); + tree = create_tree (dfa, tree, mbc_remain, CONCAT); + if (BE (mbc_remain == NULL || tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + } +#endif + break; + case OP_OPEN_SUBEXP: + tree = parse_sub_exp (regexp, preg, token, syntax, nest + 1, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_OPEN_BRACKET: + tree = parse_bracket_exp (regexp, dfa, token, syntax, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_BACK_REF: + if (!BE (dfa->completed_bkref_map & (1 << token->opr.idx), 1)) + { + *err = REG_ESUBREG; + return NULL; + } + dfa->used_bkref_map |= 1 << token->opr.idx; + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + ++dfa->nbackref; + dfa->has_mb_node = 1; + break; + case OP_OPEN_DUP_NUM: + if (syntax & RE_CONTEXT_INVALID_DUP) + { + *err = REG_BADRPT; + return NULL; + } + /* FALLTHROUGH */ + case OP_DUP_ASTERISK: + case OP_DUP_PLUS: + case OP_DUP_QUESTION: + if (syntax & RE_CONTEXT_INVALID_OPS) + { + *err = REG_BADRPT; + return NULL; + } + else if (syntax & RE_CONTEXT_INDEP_OPS) + { + fetch_token (token, regexp, syntax); + return parse_expression (regexp, preg, token, syntax, nest, err); + } + /* else fall through */ + case OP_CLOSE_SUBEXP: + if ((token->type == OP_CLOSE_SUBEXP) && + !(syntax & RE_UNMATCHED_RIGHT_PAREN_ORD)) + { + *err = REG_ERPAREN; + return NULL; + } + /* else fall through */ + case OP_CLOSE_DUP_NUM: + /* We treat it as a normal character. */ + + /* Then we can these characters as normal characters. */ + token->type = CHARACTER; + /* mb_partial and word_char bits should be initialized already + by peek_token. */ + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + break; + case ANCHOR: + if ((token->opr.ctx_type + & (WORD_DELIM | NOT_WORD_DELIM | WORD_FIRST | WORD_LAST)) + && dfa->word_ops_used == 0) + init_word_char (dfa); + if (token->opr.ctx_type == WORD_DELIM + || token->opr.ctx_type == NOT_WORD_DELIM) + { + bin_tree_t *tree_first, *tree_last; + if (token->opr.ctx_type == WORD_DELIM) + { + token->opr.ctx_type = WORD_FIRST; + tree_first = create_token_tree (dfa, NULL, NULL, token); + token->opr.ctx_type = WORD_LAST; + } + else + { + token->opr.ctx_type = INSIDE_WORD; + tree_first = create_token_tree (dfa, NULL, NULL, token); + token->opr.ctx_type = INSIDE_NOTWORD; + } + tree_last = create_token_tree (dfa, NULL, NULL, token); + tree = create_tree (dfa, tree_first, tree_last, OP_ALT); + if (BE (tree_first == NULL || tree_last == NULL || tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + else + { + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + } + /* We must return here, since ANCHORs can't be followed + by repetition operators. + eg. RE"^*" is invalid or "<ANCHOR(^)><CHAR(*)>", + it must not be "<ANCHOR(^)><REPEAT(*)>". */ + fetch_token (token, regexp, syntax); + return tree; + case OP_PERIOD: + tree = create_token_tree (dfa, NULL, NULL, token); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + if (dfa->mb_cur_max > 1) + dfa->has_mb_node = 1; + break; + case OP_WORD: + case OP_NOTWORD: + tree = build_charclass_op (dfa, regexp->trans, + (const unsigned char *) "alnum", + (const unsigned char *) "_", + token->type == OP_NOTWORD, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_SPACE: + case OP_NOTSPACE: + tree = build_charclass_op (dfa, regexp->trans, + (const unsigned char *) "space", + (const unsigned char *) "", + token->type == OP_NOTSPACE, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + break; + case OP_ALT: + case END_OF_RE: + return NULL; + case BACK_SLASH: + *err = REG_EESCAPE; + return NULL; + default: + /* Must not happen? */ +#ifdef DEBUG + assert (0); +#endif + return NULL; + } + fetch_token (token, regexp, syntax); + + while (token->type == OP_DUP_ASTERISK || token->type == OP_DUP_PLUS + || token->type == OP_DUP_QUESTION || token->type == OP_OPEN_DUP_NUM) + { + tree = parse_dup_op (tree, regexp, dfa, token, syntax, err); + if (BE (*err != REG_NOERROR && tree == NULL, 0)) + return NULL; + /* In BRE consecutive duplications are not allowed. */ + if ((syntax & RE_CONTEXT_INVALID_DUP) + && (token->type == OP_DUP_ASTERISK + || token->type == OP_OPEN_DUP_NUM)) + { + *err = REG_BADRPT; + return NULL; + } + } + + return tree; +} + +/* This function build the following tree, from regular expression + (<reg_exp>): + SUBEXP + | + <reg_exp> +*/ + +static bin_tree_t * +parse_sub_exp (re_string_t *regexp, regex_t *preg, re_token_t *token, + reg_syntax_t syntax, Idx nest, reg_errcode_t *err) +{ + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; + bin_tree_t *tree; + size_t cur_nsub; + cur_nsub = preg->re_nsub++; + + fetch_token (token, regexp, syntax | RE_CARET_ANCHORS_HERE); + + /* The subexpression may be a null string. */ + if (token->type == OP_CLOSE_SUBEXP) + tree = NULL; + else + { + tree = parse_reg_exp (regexp, preg, token, syntax, nest, err); + if (BE (*err == REG_NOERROR && token->type != OP_CLOSE_SUBEXP, 0)) + *err = REG_EPAREN; + if (BE (*err != REG_NOERROR, 0)) + return NULL; + } + + if (cur_nsub <= '9' - '1') + dfa->completed_bkref_map |= 1 << cur_nsub; + + tree = create_tree (dfa, tree, NULL, SUBEXP); + if (BE (tree == NULL, 0)) + { + *err = REG_ESPACE; + return NULL; + } + tree->token.opr.idx = cur_nsub; + return tree; +} + +/* This function parse repetition operators like "*", "+", "{1,3}" etc. */ + +static bin_tree_t * +parse_dup_op (bin_tree_t *elem, re_string_t *regexp, re_dfa_t *dfa, + re_token_t *token, reg_syntax_t syntax, reg_errcode_t *err) +{ + bin_tree_t *tree = NULL, *old_tree = NULL; + Idx i, start, end, start_idx = re_string_cur_idx (regexp); + re_token_t start_token = *token; + + if (token->type == OP_OPEN_DUP_NUM) + { + end = 0; + start = fetch_number (regexp, token, syntax); + if (start == REG_MISSING) + { + if (token->type == CHARACTER && token->opr.c == ',') + start = 0; /* We treat "{,m}" as "{0,m}". */ + else + { + *err = REG_BADBR; /* <re>{} is invalid. */ + return NULL; + } + } + if (BE (start != REG_ERROR, 1)) + { + /* We treat "{n}" as "{n,n}". */ + end = ((token->type == OP_CLOSE_DUP_NUM) ? start + : ((token->type == CHARACTER && token->opr.c == ',') + ? fetch_number (regexp, token, syntax) : REG_ERROR)); + } + if (BE (start == REG_ERROR || end == REG_ERROR, 0)) + { + /* Invalid sequence. */ + if (BE (!(syntax & RE_INVALID_INTERVAL_ORD), 0)) + { + if (token->type == END_OF_RE) + *err = REG_EBRACE; + else + *err = REG_BADBR; + + return NULL; + } + + /* If the syntax bit is set, rollback. */ + re_string_set_index (regexp, start_idx); + *token = start_token; + token->type = CHARACTER; + /* mb_partial and word_char bits should be already initialized by + peek_token. */ + return elem; + } + + if (BE (end != REG_MISSING && start > end, 0)) + { + /* First number greater than second. */ + *err = REG_BADBR; + return NULL; + } + } + else + { + start = (token->type == OP_DUP_PLUS) ? 1 : 0; + end = (token->type == OP_DUP_QUESTION) ? 1 : REG_MISSING; + } + + fetch_token (token, regexp, syntax); + + if (BE (elem == NULL, 0)) + return NULL; + if (BE (start == 0 && end == 0, 0)) + { + postorder (elem, free_tree, NULL); + return NULL; + } + + /* Extract "<re>{n,m}" to "<re><re>...<re><re>{0,<m-n>}". */ + if (BE (start > 0, 0)) + { + tree = elem; + for (i = 2; i <= start; ++i) + { + elem = duplicate_tree (elem, dfa); + tree = create_tree (dfa, tree, elem, CONCAT); + if (BE (elem == NULL || tree == NULL, 0)) + goto parse_dup_op_espace; + } + + if (start == end) + return tree; + + /* Duplicate ELEM before it is marked optional. */ + elem = duplicate_tree (elem, dfa); + old_tree = tree; + } + else + old_tree = NULL; + + if (elem->token.type == SUBEXP) + postorder (elem, mark_opt_subexp, (void *) (long) elem->token.opr.idx); + + tree = create_tree (dfa, elem, NULL, + (end == REG_MISSING ? OP_DUP_ASTERISK : OP_ALT)); + if (BE (tree == NULL, 0)) + goto parse_dup_op_espace; + + /* This loop is actually executed only when end != REG_MISSING, + to rewrite <re>{0,n} as (<re>(<re>...<re>?)?)?... We have + already created the start+1-th copy. */ + if ((Idx) -1 < 0 || end != REG_MISSING) + for (i = start + 2; i <= end; ++i) + { + elem = duplicate_tree (elem, dfa); + tree = create_tree (dfa, tree, elem, CONCAT); + if (BE (elem == NULL || tree == NULL, 0)) + goto parse_dup_op_espace; + + tree = create_tree (dfa, tree, NULL, OP_ALT); + if (BE (tree == NULL, 0)) + goto parse_dup_op_espace; + } + + if (old_tree) + tree = create_tree (dfa, old_tree, tree, CONCAT); + + return tree; + + parse_dup_op_espace: + *err = REG_ESPACE; + return NULL; +} + +/* Size of the names for collating symbol/equivalence_class/character_class. + I'm not sure, but maybe enough. */ +#define BRACKET_NAME_BUF_SIZE 32 + +#ifndef _LIBC + /* Local function for parse_bracket_exp only used in case of NOT _LIBC. + Build the range expression which starts from START_ELEM, and ends + at END_ELEM. The result are written to MBCSET and SBCSET. + RANGE_ALLOC is the allocated size of mbcset->range_starts, and + mbcset->range_ends, is a pointer argument sinse we may + update it. */ + +static reg_errcode_t +internal_function +# ifdef RE_ENABLE_I18N +build_range_exp (bitset_t sbcset, re_charset_t *mbcset, Idx *range_alloc, + bracket_elem_t *start_elem, bracket_elem_t *end_elem) +# else /* not RE_ENABLE_I18N */ +build_range_exp (bitset_t sbcset, bracket_elem_t *start_elem, + bracket_elem_t *end_elem) +# endif /* not RE_ENABLE_I18N */ +{ + unsigned int start_ch, end_ch; + /* Equivalence Classes and Character Classes can't be a range start/end. */ + if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS + || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS, + 0)) + return REG_ERANGE; + + /* We can handle no multi character collating elements without libc + support. */ + if (BE ((start_elem->type == COLL_SYM + && strlen ((char *) start_elem->opr.name) > 1) + || (end_elem->type == COLL_SYM + && strlen ((char *) end_elem->opr.name) > 1), 0)) + return REG_ECOLLATE; + +# ifdef RE_ENABLE_I18N + { + wchar_t wc; + wint_t start_wc; + wint_t end_wc; + wchar_t cmp_buf[6] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'}; + + start_ch = ((start_elem->type == SB_CHAR) ? start_elem->opr.ch + : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0] + : 0)); + end_ch = ((end_elem->type == SB_CHAR) ? end_elem->opr.ch + : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0] + : 0)); + start_wc = ((start_elem->type == SB_CHAR || start_elem->type == COLL_SYM) + ? __btowc (start_ch) : start_elem->opr.wch); + end_wc = ((end_elem->type == SB_CHAR || end_elem->type == COLL_SYM) + ? __btowc (end_ch) : end_elem->opr.wch); + if (start_wc == WEOF || end_wc == WEOF) + return REG_ECOLLATE; + cmp_buf[0] = start_wc; + cmp_buf[4] = end_wc; + if (wcscoll (cmp_buf, cmp_buf + 4) > 0) + return REG_ERANGE; + + /* Got valid collation sequence values, add them as a new entry. + However, for !_LIBC we have no collation elements: if the + character set is single byte, the single byte character set + that we build below suffices. parse_bracket_exp passes + no MBCSET if dfa->mb_cur_max == 1. */ + if (mbcset) + { + /* Check the space of the arrays. */ + if (BE (*range_alloc == mbcset->nranges, 0)) + { + /* There is not enough space, need realloc. */ + wchar_t *new_array_start, *new_array_end; + Idx new_nranges; + + /* +1 in case of mbcset->nranges is 0. */ + new_nranges = 2 * mbcset->nranges + 1; + /* Use realloc since mbcset->range_starts and mbcset->range_ends + are NULL if *range_alloc == 0. */ + new_array_start = re_realloc (mbcset->range_starts, wchar_t, + new_nranges); + new_array_end = re_realloc (mbcset->range_ends, wchar_t, + new_nranges); + + if (BE (new_array_start == NULL || new_array_end == NULL, 0)) + return REG_ESPACE; + + mbcset->range_starts = new_array_start; + mbcset->range_ends = new_array_end; + *range_alloc = new_nranges; + } + + mbcset->range_starts[mbcset->nranges] = start_wc; + mbcset->range_ends[mbcset->nranges++] = end_wc; + } + + /* Build the table for single byte characters. */ + for (wc = 0; wc < SBC_MAX; ++wc) + { + cmp_buf[2] = wc; + if (wcscoll (cmp_buf, cmp_buf + 2) <= 0 + && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0) + bitset_set (sbcset, wc); + } + } +# else /* not RE_ENABLE_I18N */ + { + unsigned int ch; + start_ch = ((start_elem->type == SB_CHAR ) ? start_elem->opr.ch + : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0] + : 0)); + end_ch = ((end_elem->type == SB_CHAR ) ? end_elem->opr.ch + : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0] + : 0)); + if (start_ch > end_ch) + return REG_ERANGE; + /* Build the table for single byte characters. */ + for (ch = 0; ch < SBC_MAX; ++ch) + if (start_ch <= ch && ch <= end_ch) + bitset_set (sbcset, ch); + } +# endif /* not RE_ENABLE_I18N */ + return REG_NOERROR; +} +#endif /* not _LIBC */ + +#ifndef _LIBC +/* Helper function for parse_bracket_exp only used in case of NOT _LIBC.. + Build the collating element which is represented by NAME. + The result are written to MBCSET and SBCSET. + COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a + pointer argument since we may update it. */ + +static reg_errcode_t +internal_function +build_collating_symbol (bitset_t sbcset, +# ifdef RE_ENABLE_I18N + re_charset_t *mbcset, Idx *coll_sym_alloc, +# endif + const unsigned char *name) +{ + size_t name_len = strlen ((const char *) name); + if (BE (name_len != 1, 0)) + return REG_ECOLLATE; + else + { + bitset_set (sbcset, name[0]); + return REG_NOERROR; + } +} +#endif /* not _LIBC */ + +/* This function parse bracket expression like "[abc]", "[a-c]", + "[[.a-a.]]" etc. */ + +static bin_tree_t * +parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa, re_token_t *token, + reg_syntax_t syntax, reg_errcode_t *err) +{ +#ifdef _LIBC + const unsigned char *collseqmb; + const char *collseqwc; + uint32_t nrules; + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + + /* Local function for parse_bracket_exp used in _LIBC environement. + Seek the collating symbol entry correspondings to NAME. + Return the index of the symbol in the SYMB_TABLE. */ + + auto inline int32_t + __attribute ((always_inline)) + seek_collating_symbol_entry (name, name_len) + const unsigned char *name; + size_t name_len; + { + int32_t hash = elem_hash ((const char *) name, name_len); + int32_t elem = hash % table_size; + if (symb_table[2 * elem] != 0) + { + int32_t second = hash % (table_size - 2) + 1; + + do + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + /* Compare the length of the name. */ + && name_len == extra[symb_table[2 * elem + 1]] + /* Compare the name. */ + && memcmp (name, &extra[symb_table[2 * elem + 1] + 1], + name_len) == 0) + { + /* Yep, this is the entry. */ + break; + } + + /* Next entry. */ + elem += second; + } + while (symb_table[2 * elem] != 0); + } + return elem; + } + + /* Local function for parse_bracket_exp used in _LIBC environement. + Look up the collation sequence value of BR_ELEM. + Return the value if succeeded, UINT_MAX otherwise. */ + + auto inline unsigned int + __attribute ((always_inline)) + lookup_collation_sequence_value (br_elem) + bracket_elem_t *br_elem; + { + if (br_elem->type == SB_CHAR) + { + /* + if (MB_CUR_MAX == 1) + */ + if (nrules == 0) + return collseqmb[br_elem->opr.ch]; + else + { + wint_t wc = __btowc (br_elem->opr.ch); + return __collseq_table_lookup (collseqwc, wc); + } + } + else if (br_elem->type == MB_CHAR) + { + return __collseq_table_lookup (collseqwc, br_elem->opr.wch); + } + else if (br_elem->type == COLL_SYM) + { + size_t sym_name_len = strlen ((char *) br_elem->opr.name); + if (nrules != 0) + { + int32_t elem, idx; + elem = seek_collating_symbol_entry (br_elem->opr.name, + sym_name_len); + if (symb_table[2 * elem] != 0) + { + /* We found the entry. */ + idx = symb_table[2 * elem + 1]; + /* Skip the name of collating element name. */ + idx += 1 + extra[idx]; + /* Skip the byte sequence of the collating element. */ + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + /* Skip the multibyte collation sequence value. */ + idx += sizeof (unsigned int); + /* Skip the wide char sequence of the collating element. */ + idx += sizeof (unsigned int) * + (1 + *(unsigned int *) (extra + idx)); + /* Return the collation sequence value. */ + return *(unsigned int *) (extra + idx); + } + else if (symb_table[2 * elem] == 0 && sym_name_len == 1) + { + /* No valid character. Match it as a single byte + character. */ + return collseqmb[br_elem->opr.name[0]]; + } + } + else if (sym_name_len == 1) + return collseqmb[br_elem->opr.name[0]]; + } + return UINT_MAX; + } + + /* Local function for parse_bracket_exp used in _LIBC environement. + Build the range expression which starts from START_ELEM, and ends + at END_ELEM. The result are written to MBCSET and SBCSET. + RANGE_ALLOC is the allocated size of mbcset->range_starts, and + mbcset->range_ends, is a pointer argument sinse we may + update it. */ + + auto inline reg_errcode_t + __attribute ((always_inline)) + build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem) + re_charset_t *mbcset; + Idx *range_alloc; + bitset_t sbcset; + bracket_elem_t *start_elem, *end_elem; + { + unsigned int ch; + uint32_t start_collseq; + uint32_t end_collseq; + + /* Equivalence Classes and Character Classes can't be a range + start/end. */ + if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS + || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS, + 0)) + return REG_ERANGE; + + start_collseq = lookup_collation_sequence_value (start_elem); + end_collseq = lookup_collation_sequence_value (end_elem); + /* Check start/end collation sequence values. */ + if (BE (start_collseq == UINT_MAX || end_collseq == UINT_MAX, 0)) + return REG_ECOLLATE; + if (BE ((syntax & RE_NO_EMPTY_RANGES) && start_collseq > end_collseq, 0)) + return REG_ERANGE; + + /* Got valid collation sequence values, add them as a new entry. + However, if we have no collation elements, and the character set + is single byte, the single byte character set that we + build below suffices. */ + if (nrules > 0 || dfa->mb_cur_max > 1) + { + /* Check the space of the arrays. */ + if (BE (*range_alloc == mbcset->nranges, 0)) + { + /* There is not enough space, need realloc. */ + uint32_t *new_array_start; + uint32_t *new_array_end; + Idx new_nranges; + + /* +1 in case of mbcset->nranges is 0. */ + new_nranges = 2 * mbcset->nranges + 1; + new_array_start = re_realloc (mbcset->range_starts, uint32_t, + new_nranges); + new_array_end = re_realloc (mbcset->range_ends, uint32_t, + new_nranges); + + if (BE (new_array_start == NULL || new_array_end == NULL, 0)) + return REG_ESPACE; + + mbcset->range_starts = new_array_start; + mbcset->range_ends = new_array_end; + *range_alloc = new_nranges; + } + + mbcset->range_starts[mbcset->nranges] = start_collseq; + mbcset->range_ends[mbcset->nranges++] = end_collseq; + } + + /* Build the table for single byte characters. */ + for (ch = 0; ch < SBC_MAX; ch++) + { + uint32_t ch_collseq; + /* + if (MB_CUR_MAX == 1) + */ + if (nrules == 0) + ch_collseq = collseqmb[ch]; + else + ch_collseq = __collseq_table_lookup (collseqwc, __btowc (ch)); + if (start_collseq <= ch_collseq && ch_collseq <= end_collseq) + bitset_set (sbcset, ch); + } + return REG_NOERROR; + } + + /* Local function for parse_bracket_exp used in _LIBC environement. + Build the collating element which is represented by NAME. + The result are written to MBCSET and SBCSET. + COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a + pointer argument sinse we may update it. */ + + auto inline reg_errcode_t + __attribute ((always_inline)) + build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name) + re_charset_t *mbcset; + Idx *coll_sym_alloc; + bitset_t sbcset; + const unsigned char *name; + { + int32_t elem, idx; + size_t name_len = strlen ((const char *) name); + if (nrules != 0) + { + elem = seek_collating_symbol_entry (name, name_len); + if (symb_table[2 * elem] != 0) + { + /* We found the entry. */ + idx = symb_table[2 * elem + 1]; + /* Skip the name of collating element name. */ + idx += 1 + extra[idx]; + } + else if (symb_table[2 * elem] == 0 && name_len == 1) + { + /* No valid character, treat it as a normal + character. */ + bitset_set (sbcset, name[0]); + return REG_NOERROR; + } + else + return REG_ECOLLATE; + + /* Got valid collation sequence, add it as a new entry. */ + /* Check the space of the arrays. */ + if (BE (*coll_sym_alloc == mbcset->ncoll_syms, 0)) + { + /* Not enough, realloc it. */ + /* +1 in case of mbcset->ncoll_syms is 0. */ + Idx new_coll_sym_alloc = 2 * mbcset->ncoll_syms + 1; + /* Use realloc since mbcset->coll_syms is NULL + if *alloc == 0. */ + int32_t *new_coll_syms = re_realloc (mbcset->coll_syms, int32_t, + new_coll_sym_alloc); + if (BE (new_coll_syms == NULL, 0)) + return REG_ESPACE; + mbcset->coll_syms = new_coll_syms; + *coll_sym_alloc = new_coll_sym_alloc; + } + mbcset->coll_syms[mbcset->ncoll_syms++] = idx; + return REG_NOERROR; + } + else + { + if (BE (name_len != 1, 0)) + return REG_ECOLLATE; + else + { + bitset_set (sbcset, name[0]); + return REG_NOERROR; + } + } + } +#endif + + re_token_t br_token; + re_bitset_ptr_t sbcset; +#ifdef RE_ENABLE_I18N + re_charset_t *mbcset; + Idx coll_sym_alloc = 0, range_alloc = 0, mbchar_alloc = 0; + Idx equiv_class_alloc = 0, char_class_alloc = 0; +#endif /* not RE_ENABLE_I18N */ + bool non_match = false; + bin_tree_t *work_tree; + int token_len; + bool first_round = true; +#ifdef _LIBC + collseqmb = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB); + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules) + { + /* + if (MB_CUR_MAX > 1) + */ + collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC); + table_size = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + } +#endif + sbcset = (re_bitset_ptr_t) calloc (sizeof (bitset_t), 1); +#ifdef RE_ENABLE_I18N + mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1); +#endif /* RE_ENABLE_I18N */ +#ifdef RE_ENABLE_I18N + if (BE (sbcset == NULL || mbcset == NULL, 0)) +#else + if (BE (sbcset == NULL, 0)) +#endif /* RE_ENABLE_I18N */ + { + *err = REG_ESPACE; + return NULL; + } + + token_len = peek_token_bracket (token, regexp, syntax); + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_BADPAT; + goto parse_bracket_exp_free_return; + } + if (token->type == OP_NON_MATCH_LIST) + { +#ifdef RE_ENABLE_I18N + mbcset->non_match = 1; +#endif /* not RE_ENABLE_I18N */ + non_match = true; + if (syntax & RE_HAT_LISTS_NOT_NEWLINE) + bitset_set (sbcset, '\n'); + re_string_skip_bytes (regexp, token_len); /* Skip a token. */ + token_len = peek_token_bracket (token, regexp, syntax); + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_BADPAT; + goto parse_bracket_exp_free_return; + } + } + + /* We treat the first ']' as a normal character. */ + if (token->type == OP_CLOSE_BRACKET) + token->type = CHARACTER; + + while (1) + { + bracket_elem_t start_elem, end_elem; + unsigned char start_name_buf[BRACKET_NAME_BUF_SIZE]; + unsigned char end_name_buf[BRACKET_NAME_BUF_SIZE]; + reg_errcode_t ret; + int token_len2 = 0; + bool is_range_exp = false; + re_token_t token2; + + start_elem.opr.name = start_name_buf; + ret = parse_bracket_element (&start_elem, regexp, token, token_len, dfa, + syntax, first_round); + if (BE (ret != REG_NOERROR, 0)) + { + *err = ret; + goto parse_bracket_exp_free_return; + } + first_round = false; + + /* Get information about the next token. We need it in any case. */ + token_len = peek_token_bracket (token, regexp, syntax); + + /* Do not check for ranges if we know they are not allowed. */ + if (start_elem.type != CHAR_CLASS && start_elem.type != EQUIV_CLASS) + { + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_EBRACK; + goto parse_bracket_exp_free_return; + } + if (token->type == OP_CHARSET_RANGE) + { + re_string_skip_bytes (regexp, token_len); /* Skip '-'. */ + token_len2 = peek_token_bracket (&token2, regexp, syntax); + if (BE (token2.type == END_OF_RE, 0)) + { + *err = REG_EBRACK; + goto parse_bracket_exp_free_return; + } + if (token2.type == OP_CLOSE_BRACKET) + { + /* We treat the last '-' as a normal character. */ + re_string_skip_bytes (regexp, -token_len); + token->type = CHARACTER; + } + else + is_range_exp = true; + } + } + + if (is_range_exp == true) + { + end_elem.opr.name = end_name_buf; + ret = parse_bracket_element (&end_elem, regexp, &token2, token_len2, + dfa, syntax, true); + if (BE (ret != REG_NOERROR, 0)) + { + *err = ret; + goto parse_bracket_exp_free_return; + } + + token_len = peek_token_bracket (token, regexp, syntax); + +#ifdef _LIBC + *err = build_range_exp (sbcset, mbcset, &range_alloc, + &start_elem, &end_elem); +#else +# ifdef RE_ENABLE_I18N + *err = build_range_exp (sbcset, + dfa->mb_cur_max > 1 ? mbcset : NULL, + &range_alloc, &start_elem, &end_elem); +# else + *err = build_range_exp (sbcset, &start_elem, &end_elem); +# endif +#endif /* RE_ENABLE_I18N */ + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + } + else + { + switch (start_elem.type) + { + case SB_CHAR: + bitset_set (sbcset, start_elem.opr.ch); + break; +#ifdef RE_ENABLE_I18N + case MB_CHAR: + /* Check whether the array has enough space. */ + if (BE (mbchar_alloc == mbcset->nmbchars, 0)) + { + wchar_t *new_mbchars; + /* Not enough, realloc it. */ + /* +1 in case of mbcset->nmbchars is 0. */ + mbchar_alloc = 2 * mbcset->nmbchars + 1; + /* Use realloc since array is NULL if *alloc == 0. */ + new_mbchars = re_realloc (mbcset->mbchars, wchar_t, + mbchar_alloc); + if (BE (new_mbchars == NULL, 0)) + goto parse_bracket_exp_espace; + mbcset->mbchars = new_mbchars; + } + mbcset->mbchars[mbcset->nmbchars++] = start_elem.opr.wch; + break; +#endif /* RE_ENABLE_I18N */ + case EQUIV_CLASS: + *err = build_equiv_class (sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &equiv_class_alloc, +#endif /* RE_ENABLE_I18N */ + start_elem.opr.name); + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + break; + case COLL_SYM: + *err = build_collating_symbol (sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &coll_sym_alloc, +#endif /* RE_ENABLE_I18N */ + start_elem.opr.name); + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + break; + case CHAR_CLASS: + *err = build_charclass (regexp->trans, sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &char_class_alloc, +#endif /* RE_ENABLE_I18N */ + start_elem.opr.name, syntax); + if (BE (*err != REG_NOERROR, 0)) + goto parse_bracket_exp_free_return; + break; + default: + assert (0); + break; + } + } + if (BE (token->type == END_OF_RE, 0)) + { + *err = REG_EBRACK; + goto parse_bracket_exp_free_return; + } + if (token->type == OP_CLOSE_BRACKET) + break; + } + + re_string_skip_bytes (regexp, token_len); /* Skip a token. */ + + /* If it is non-matching list. */ + if (non_match) + bitset_not (sbcset); + +#ifdef RE_ENABLE_I18N + /* Ensure only single byte characters are set. */ + if (dfa->mb_cur_max > 1) + bitset_mask (sbcset, dfa->sb_char); + + if (mbcset->nmbchars || mbcset->ncoll_syms || mbcset->nequiv_classes + || mbcset->nranges || (dfa->mb_cur_max > 1 && (mbcset->nchar_classes + || mbcset->non_match))) + { + bin_tree_t *mbc_tree; + int sbc_idx; + /* Build a tree for complex bracket. */ + dfa->has_mb_node = 1; + br_token.type = COMPLEX_BRACKET; + br_token.opr.mbcset = mbcset; + mbc_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (mbc_tree == NULL, 0)) + goto parse_bracket_exp_espace; + for (sbc_idx = 0; sbc_idx < BITSET_WORDS; ++sbc_idx) + if (sbcset[sbc_idx]) + break; + /* If there are no bits set in sbcset, there is no point + of having both SIMPLE_BRACKET and COMPLEX_BRACKET. */ + if (sbc_idx < BITSET_WORDS) + { + /* Build a tree for simple bracket. */ + br_token.type = SIMPLE_BRACKET; + br_token.opr.sbcset = sbcset; + work_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (work_tree == NULL, 0)) + goto parse_bracket_exp_espace; + + /* Then join them by ALT node. */ + work_tree = create_tree (dfa, work_tree, mbc_tree, OP_ALT); + if (BE (work_tree == NULL, 0)) + goto parse_bracket_exp_espace; + } + else + { + re_free (sbcset); + work_tree = mbc_tree; + } + } + else +#endif /* not RE_ENABLE_I18N */ + { +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif + /* Build a tree for simple bracket. */ + br_token.type = SIMPLE_BRACKET; + br_token.opr.sbcset = sbcset; + work_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (work_tree == NULL, 0)) + goto parse_bracket_exp_espace; + } + return work_tree; + + parse_bracket_exp_espace: + *err = REG_ESPACE; + parse_bracket_exp_free_return: + re_free (sbcset); +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif /* RE_ENABLE_I18N */ + return NULL; +} + +/* Parse an element in the bracket expression. */ + +static reg_errcode_t +parse_bracket_element (bracket_elem_t *elem, re_string_t *regexp, + re_token_t *token, int token_len, re_dfa_t *dfa, + reg_syntax_t syntax, bool accept_hyphen) +{ +#ifdef RE_ENABLE_I18N + int cur_char_size; + cur_char_size = re_string_char_size_at (regexp, re_string_cur_idx (regexp)); + if (cur_char_size > 1) + { + elem->type = MB_CHAR; + elem->opr.wch = re_string_wchar_at (regexp, re_string_cur_idx (regexp)); + re_string_skip_bytes (regexp, cur_char_size); + return REG_NOERROR; + } +#endif /* RE_ENABLE_I18N */ + re_string_skip_bytes (regexp, token_len); /* Skip a token. */ + if (token->type == OP_OPEN_COLL_ELEM || token->type == OP_OPEN_CHAR_CLASS + || token->type == OP_OPEN_EQUIV_CLASS) + return parse_bracket_symbol (elem, regexp, token); + if (BE (token->type == OP_CHARSET_RANGE, 0) && !accept_hyphen) + { + /* A '-' must only appear as anything but a range indicator before + the closing bracket. Everything else is an error. */ + re_token_t token2; + (void) peek_token_bracket (&token2, regexp, syntax); + if (token2.type != OP_CLOSE_BRACKET) + /* The actual error value is not standardized since this whole + case is undefined. But ERANGE makes good sense. */ + return REG_ERANGE; + } + elem->type = SB_CHAR; + elem->opr.ch = token->opr.c; + return REG_NOERROR; +} + +/* Parse a bracket symbol in the bracket expression. Bracket symbols are + such as [:<character_class>:], [.<collating_element>.], and + [=<equivalent_class>=]. */ + +static reg_errcode_t +parse_bracket_symbol (bracket_elem_t *elem, re_string_t *regexp, + re_token_t *token) +{ + unsigned char ch, delim = token->opr.c; + int i = 0; + if (re_string_eoi(regexp)) + return REG_EBRACK; + for (;; ++i) + { + if (i >= BRACKET_NAME_BUF_SIZE) + return REG_EBRACK; + if (token->type == OP_OPEN_CHAR_CLASS) + ch = re_string_fetch_byte_case (regexp); + else + ch = re_string_fetch_byte (regexp); + if (re_string_eoi(regexp)) + return REG_EBRACK; + if (ch == delim && re_string_peek_byte (regexp, 0) == ']') + break; + elem->opr.name[i] = ch; + } + re_string_skip_bytes (regexp, 1); + elem->opr.name[i] = '\0'; + switch (token->type) + { + case OP_OPEN_COLL_ELEM: + elem->type = COLL_SYM; + break; + case OP_OPEN_EQUIV_CLASS: + elem->type = EQUIV_CLASS; + break; + case OP_OPEN_CHAR_CLASS: + elem->type = CHAR_CLASS; + break; + default: + break; + } + return REG_NOERROR; +} + + /* Helper function for parse_bracket_exp. + Build the equivalence class which is represented by NAME. + The result are written to MBCSET and SBCSET. + EQUIV_CLASS_ALLOC is the allocated size of mbcset->equiv_classes, + is a pointer argument sinse we may update it. */ + +static reg_errcode_t +#ifdef RE_ENABLE_I18N +build_equiv_class (bitset_t sbcset, re_charset_t *mbcset, + Idx *equiv_class_alloc, const unsigned char *name) +#else /* not RE_ENABLE_I18N */ +build_equiv_class (bitset_t sbcset, const unsigned char *name) +#endif /* not RE_ENABLE_I18N */ +{ +#ifdef _LIBC + uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules != 0) + { + const int32_t *table, *indirect; + const unsigned char *weights, *extra, *cp; + unsigned char char_buf[2]; + int32_t idx1, idx2; + unsigned int ch; + size_t len; + /* This #include defines a local function! */ +# include <locale/weight.h> + /* Calculate the index for equivalence class. */ + cp = name; + table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_INDIRECTMB); + idx1 = findidx (&cp); + if (BE (idx1 == 0 || cp < name + strlen ((const char *) name), 0)) + /* This isn't a valid character. */ + return REG_ECOLLATE; + + /* Build single byte matcing table for this equivalence class. */ + char_buf[1] = (unsigned char) '\0'; + len = weights[idx1]; + for (ch = 0; ch < SBC_MAX; ++ch) + { + char_buf[0] = ch; + cp = char_buf; + idx2 = findidx (&cp); +/* + idx2 = table[ch]; +*/ + if (idx2 == 0) + /* This isn't a valid character. */ + continue; + if (len == weights[idx2]) + { + int cnt = 0; + while (cnt <= len && + weights[idx1 + 1 + cnt] == weights[idx2 + 1 + cnt]) + ++cnt; + + if (cnt > len) + bitset_set (sbcset, ch); + } + } + /* Check whether the array has enough space. */ + if (BE (*equiv_class_alloc == mbcset->nequiv_classes, 0)) + { + /* Not enough, realloc it. */ + /* +1 in case of mbcset->nequiv_classes is 0. */ + Idx new_equiv_class_alloc = 2 * mbcset->nequiv_classes + 1; + /* Use realloc since the array is NULL if *alloc == 0. */ + int32_t *new_equiv_classes = re_realloc (mbcset->equiv_classes, + int32_t, + new_equiv_class_alloc); + if (BE (new_equiv_classes == NULL, 0)) + return REG_ESPACE; + mbcset->equiv_classes = new_equiv_classes; + *equiv_class_alloc = new_equiv_class_alloc; + } + mbcset->equiv_classes[mbcset->nequiv_classes++] = idx1; + } + else +#endif /* _LIBC */ + { + if (BE (strlen ((const char *) name) != 1, 0)) + return REG_ECOLLATE; + bitset_set (sbcset, *name); + } + return REG_NOERROR; +} + + /* Helper function for parse_bracket_exp. + Build the character class which is represented by NAME. + The result are written to MBCSET and SBCSET. + CHAR_CLASS_ALLOC is the allocated size of mbcset->char_classes, + is a pointer argument sinse we may update it. */ + +static reg_errcode_t +#ifdef RE_ENABLE_I18N +build_charclass (RE_TRANSLATE_TYPE trans, bitset_t sbcset, + re_charset_t *mbcset, Idx *char_class_alloc, + const unsigned char *class_name, reg_syntax_t syntax) +#else /* not RE_ENABLE_I18N */ +build_charclass (RE_TRANSLATE_TYPE trans, bitset_t sbcset, + const unsigned char *class_name, reg_syntax_t syntax) +#endif /* not RE_ENABLE_I18N */ +{ + int i; + const char *name = (const char *) class_name; + + /* In case of REG_ICASE "upper" and "lower" match the both of + upper and lower cases. */ + if ((syntax & RE_ICASE) + && (strcmp (name, "upper") == 0 || strcmp (name, "lower") == 0)) + name = "alpha"; + +#ifdef RE_ENABLE_I18N + /* Check the space of the arrays. */ + if (BE (*char_class_alloc == mbcset->nchar_classes, 0)) + { + /* Not enough, realloc it. */ + /* +1 in case of mbcset->nchar_classes is 0. */ + Idx new_char_class_alloc = 2 * mbcset->nchar_classes + 1; + /* Use realloc since array is NULL if *alloc == 0. */ + wctype_t *new_char_classes = re_realloc (mbcset->char_classes, wctype_t, + new_char_class_alloc); + if (BE (new_char_classes == NULL, 0)) + return REG_ESPACE; + mbcset->char_classes = new_char_classes; + *char_class_alloc = new_char_class_alloc; + } + mbcset->char_classes[mbcset->nchar_classes++] = __wctype (name); +#endif /* RE_ENABLE_I18N */ + +#define BUILD_CHARCLASS_LOOP(ctype_func) \ + do { \ + if (BE (trans != NULL, 0)) \ + { \ + for (i = 0; i < SBC_MAX; ++i) \ + if (ctype_func (i)) \ + bitset_set (sbcset, trans[i]); \ + } \ + else \ + { \ + for (i = 0; i < SBC_MAX; ++i) \ + if (ctype_func (i)) \ + bitset_set (sbcset, i); \ + } \ + } while (0) + + if (strcmp (name, "alnum") == 0) + BUILD_CHARCLASS_LOOP (isalnum); + else if (strcmp (name, "cntrl") == 0) + BUILD_CHARCLASS_LOOP (iscntrl); + else if (strcmp (name, "lower") == 0) + BUILD_CHARCLASS_LOOP (islower); + else if (strcmp (name, "space") == 0) + BUILD_CHARCLASS_LOOP (isspace); + else if (strcmp (name, "alpha") == 0) + BUILD_CHARCLASS_LOOP (isalpha); + else if (strcmp (name, "digit") == 0) + BUILD_CHARCLASS_LOOP (isdigit); + else if (strcmp (name, "print") == 0) + BUILD_CHARCLASS_LOOP (isprint); + else if (strcmp (name, "upper") == 0) + BUILD_CHARCLASS_LOOP (isupper); + else if (strcmp (name, "blank") == 0) + BUILD_CHARCLASS_LOOP (isblank); + else if (strcmp (name, "graph") == 0) + BUILD_CHARCLASS_LOOP (isgraph); + else if (strcmp (name, "punct") == 0) + BUILD_CHARCLASS_LOOP (ispunct); + else if (strcmp (name, "xdigit") == 0) + BUILD_CHARCLASS_LOOP (isxdigit); + else + return REG_ECTYPE; + + return REG_NOERROR; +} + +static bin_tree_t * +build_charclass_op (re_dfa_t *dfa, RE_TRANSLATE_TYPE trans, + const unsigned char *class_name, + const unsigned char *extra, bool non_match, + reg_errcode_t *err) +{ + re_bitset_ptr_t sbcset; +#ifdef RE_ENABLE_I18N + re_charset_t *mbcset; + Idx alloc = 0; +#endif /* not RE_ENABLE_I18N */ + reg_errcode_t ret; + re_token_t br_token; + bin_tree_t *tree; + + sbcset = (re_bitset_ptr_t) calloc (sizeof (bitset_t), 1); +#ifdef RE_ENABLE_I18N + mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1); +#endif /* RE_ENABLE_I18N */ + +#ifdef RE_ENABLE_I18N + if (BE (sbcset == NULL || mbcset == NULL, 0)) +#else /* not RE_ENABLE_I18N */ + if (BE (sbcset == NULL, 0)) +#endif /* not RE_ENABLE_I18N */ + { + *err = REG_ESPACE; + return NULL; + } + + if (non_match) + { +#ifdef RE_ENABLE_I18N + mbcset->non_match = 1; +#endif /* not RE_ENABLE_I18N */ + } + + /* We don't care the syntax in this case. */ + ret = build_charclass (trans, sbcset, +#ifdef RE_ENABLE_I18N + mbcset, &alloc, +#endif /* RE_ENABLE_I18N */ + class_name, 0); + + if (BE (ret != REG_NOERROR, 0)) + { + re_free (sbcset); +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif /* RE_ENABLE_I18N */ + *err = ret; + return NULL; + } + /* \w match '_' also. */ + for (; *extra; extra++) + bitset_set (sbcset, *extra); + + /* If it is non-matching list. */ + if (non_match) + bitset_not (sbcset); + +#ifdef RE_ENABLE_I18N + /* Ensure only single byte characters are set. */ + if (dfa->mb_cur_max > 1) + bitset_mask (sbcset, dfa->sb_char); +#endif + + /* Build a tree for simple bracket. */ + br_token.type = SIMPLE_BRACKET; + br_token.opr.sbcset = sbcset; + tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (tree == NULL, 0)) + goto build_word_op_espace; + +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + bin_tree_t *mbc_tree; + /* Build a tree for complex bracket. */ + br_token.type = COMPLEX_BRACKET; + br_token.opr.mbcset = mbcset; + dfa->has_mb_node = 1; + mbc_tree = create_token_tree (dfa, NULL, NULL, &br_token); + if (BE (mbc_tree == NULL, 0)) + goto build_word_op_espace; + /* Then join them by ALT node. */ + tree = create_tree (dfa, tree, mbc_tree, OP_ALT); + if (BE (mbc_tree != NULL, 1)) + return tree; + } + else + { + free_charset (mbcset); + return tree; + } +#else /* not RE_ENABLE_I18N */ + return tree; +#endif /* not RE_ENABLE_I18N */ + + build_word_op_espace: + re_free (sbcset); +#ifdef RE_ENABLE_I18N + free_charset (mbcset); +#endif /* RE_ENABLE_I18N */ + *err = REG_ESPACE; + return NULL; +} + +/* This is intended for the expressions like "a{1,3}". + Fetch a number from `input', and return the number. + Return REG_MISSING if the number field is empty like "{,1}". + Return REG_ERROR if an error occurred. */ + +static Idx +fetch_number (re_string_t *input, re_token_t *token, reg_syntax_t syntax) +{ + Idx num = REG_MISSING; + unsigned char c; + while (1) + { + fetch_token (token, input, syntax); + c = token->opr.c; + if (BE (token->type == END_OF_RE, 0)) + return REG_ERROR; + if (token->type == OP_CLOSE_DUP_NUM || c == ',') + break; + num = ((token->type != CHARACTER || c < '0' || '9' < c + || num == REG_ERROR) + ? REG_ERROR + : ((num == REG_MISSING) ? c - '0' : num * 10 + c - '0')); + num = (num > RE_DUP_MAX) ? REG_ERROR : num; + } + return num; +} + +#ifdef RE_ENABLE_I18N +static void +free_charset (re_charset_t *cset) +{ + re_free (cset->mbchars); +# ifdef _LIBC + re_free (cset->coll_syms); + re_free (cset->equiv_classes); + re_free (cset->range_starts); + re_free (cset->range_ends); +# endif + re_free (cset->char_classes); + re_free (cset); +} +#endif /* RE_ENABLE_I18N */ + +/* Functions for binary tree operation. */ + +/* Create a tree node. */ + +static bin_tree_t * +create_tree (re_dfa_t *dfa, bin_tree_t *left, bin_tree_t *right, + re_token_type_t type) +{ + re_token_t t; + t.type = type; + return create_token_tree (dfa, left, right, &t); +} + +static bin_tree_t * +create_token_tree (re_dfa_t *dfa, bin_tree_t *left, bin_tree_t *right, + const re_token_t *token) +{ + bin_tree_t *tree; + if (BE (dfa->str_tree_storage_idx == BIN_TREE_STORAGE_SIZE, 0)) + { + bin_tree_storage_t *storage = re_malloc (bin_tree_storage_t, 1); + + if (storage == NULL) + return NULL; + storage->next = dfa->str_tree_storage; + dfa->str_tree_storage = storage; + dfa->str_tree_storage_idx = 0; + } + tree = &dfa->str_tree_storage->data[dfa->str_tree_storage_idx++]; + + tree->parent = NULL; + tree->left = left; + tree->right = right; + tree->token = *token; + tree->token.duplicated = 0; + tree->token.opt_subexp = 0; + tree->first = NULL; + tree->next = NULL; + tree->node_idx = REG_MISSING; + + if (left != NULL) + left->parent = tree; + if (right != NULL) + right->parent = tree; + return tree; +} + +/* Mark the tree SRC as an optional subexpression. + To be called from preorder or postorder. */ + +static reg_errcode_t +mark_opt_subexp (void *extra, bin_tree_t *node) +{ + Idx idx = (Idx) (long) extra; + if (node->token.type == SUBEXP && node->token.opr.idx == idx) + node->token.opt_subexp = 1; + + return REG_NOERROR; +} + +/* Free the allocated memory inside NODE. */ + +static void +free_token (re_token_t *node) +{ +#ifdef RE_ENABLE_I18N + if (node->type == COMPLEX_BRACKET && node->duplicated == 0) + free_charset (node->opr.mbcset); + else +#endif /* RE_ENABLE_I18N */ + if (node->type == SIMPLE_BRACKET && node->duplicated == 0) + re_free (node->opr.sbcset); +} + +/* Worker function for tree walking. Free the allocated memory inside NODE + and its children. */ + +static reg_errcode_t +free_tree (void *extra, bin_tree_t *node) +{ + free_token (&node->token); + return REG_NOERROR; +} + + +/* Duplicate the node SRC, and return new node. This is a preorder + visit similar to the one implemented by the generic visitor, but + we need more infrastructure to maintain two parallel trees --- so, + it's easier to duplicate. */ + +static bin_tree_t * +duplicate_tree (const bin_tree_t *root, re_dfa_t *dfa) +{ + const bin_tree_t *node; + bin_tree_t *dup_root; + bin_tree_t **p_new = &dup_root, *dup_node = root->parent; + + for (node = root; ; ) + { + /* Create a new tree and link it back to the current parent. */ + *p_new = create_token_tree (dfa, NULL, NULL, &node->token); + if (*p_new == NULL) + return NULL; + (*p_new)->parent = dup_node; + (*p_new)->token.duplicated = 1; + dup_node = *p_new; + + /* Go to the left node, or up and to the right. */ + if (node->left) + { + node = node->left; + p_new = &dup_node->left; + } + else + { + const bin_tree_t *prev = NULL; + while (node->right == prev || node->right == NULL) + { + prev = node; + node = node->parent; + dup_node = dup_node->parent; + if (!node) + return dup_root; + } + node = node->right; + p_new = &dup_node->right; + } + } +} diff --git a/lib/regex.c b/lib/regex.c new file mode 100644 index 0000000..d4eb726 --- /dev/null +++ b/lib/regex.c @@ -0,0 +1,71 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2005, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Make sure noone compiles this code with a C++ compiler. */ +#if defined __cplusplus && defined _LIBC +# error "This is C code, use a C compiler" +#endif + +#ifdef _LIBC +/* We have to keep the namespace clean. */ +# define regfree(preg) __regfree (preg) +# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef) +# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags) +# define regerror(errcode, preg, errbuf, errbuf_size) \ + __regerror(errcode, preg, errbuf, errbuf_size) +# define re_set_registers(bu, re, nu, st, en) \ + __re_set_registers (bu, re, nu, st, en) +# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \ + __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) +# define re_match(bufp, string, size, pos, regs) \ + __re_match (bufp, string, size, pos, regs) +# define re_search(bufp, string, size, startpos, range, regs) \ + __re_search (bufp, string, size, startpos, range, regs) +# define re_compile_pattern(pattern, length, bufp) \ + __re_compile_pattern (pattern, length, bufp) +# define re_set_syntax(syntax) __re_set_syntax (syntax) +# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \ + __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop) +# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp) + +# include "../locale/localeinfo.h" +#endif + +/* On some systems, limits.h sets RE_DUP_MAX to a lower value than + GNU regex allows. Include it before <regex.h>, which correctly + #undefs RE_DUP_MAX and sets it to the right value. */ +#include <limits.h> + +#include <regex.h> +#include "regex_internal.h" + +#include "regex_internal.c" +#include "regcomp.c" +#include "regexec.c" + +/* Binary backward compatibility. */ +#if _LIBC +# include <shlib-compat.h> +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3) +link_warning (re_max_failures, "the 're_max_failures' variable is obsolete and will go away.") +int re_max_failures = 2000; +# endif +#endif diff --git a/lib/regex.h b/lib/regex.h new file mode 100644 index 0000000..7a79ca3 --- /dev/null +++ b/lib/regex.h @@ -0,0 +1,675 @@ +/* Definitions for data structures and routines for the regular + expression library. + Copyright (C) 1985,1989-93,1995-98,2000,2001,2002,2003,2005,2006 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _REGEX_H +#define _REGEX_H 1 + +#include <sys/types.h> + +/* Allow the use in C++ code. */ +#ifdef __cplusplus +extern "C" { +#endif + +/* Define __USE_GNU_REGEX to declare GNU extensions that violate the + POSIX name space rules. */ +#undef __USE_GNU_REGEX +#if (defined _GNU_SOURCE \ + || (!defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE \ + && !defined _XOPEN_SOURCE)) +# define __USE_GNU_REGEX 1 +#endif + +#ifdef _REGEX_LARGE_OFFSETS + +/* Use types and values that are wide enough to represent signed and + unsigned byte offsets in memory. This currently works only when + the regex code is used outside of the GNU C library; it is not yet + supported within glibc itself, and glibc users should not define + _REGEX_LARGE_OFFSETS. */ + +/* The type of the offset of a byte within a string. + For historical reasons POSIX 1003.1-2004 requires that regoff_t be + at least as wide as off_t. However, many common POSIX platforms set + regoff_t to the more-sensible ssize_t and the Open Group has + signalled its intention to change the requirement to be that + regoff_t be at least as wide as ptrdiff_t and ssize_t; see XBD ERN + 60 (2005-08-25). We don't know of any hosts where ssize_t or + ptrdiff_t is wider than ssize_t, so ssize_t is safe. */ +typedef ssize_t regoff_t; + +/* The type of nonnegative object indexes. Traditionally, GNU regex + uses 'int' for these. Code that uses __re_idx_t should work + regardless of whether the type is signed. */ +typedef size_t __re_idx_t; + +/* The type of object sizes. */ +typedef size_t __re_size_t; + +/* The type of object sizes, in places where the traditional code + uses unsigned long int. */ +typedef size_t __re_long_size_t; + +#else + +/* Use types that are binary-compatible with the traditional GNU regex + implementation, which mishandles strings longer than INT_MAX. */ + +typedef int regoff_t; +typedef int __re_idx_t; +typedef unsigned int __re_size_t; +typedef unsigned long int __re_long_size_t; + +#endif + +/* The following two types have to be signed and unsigned integer type + wide enough to hold a value of a pointer. For most ANSI compilers + ptrdiff_t and size_t should be likely OK. Still size of these two + types is 2 for Microsoft C. Ugh... */ +typedef long int s_reg_t; +typedef unsigned long int active_reg_t; + +/* The following bits are used to determine the regexp syntax we + recognize. The set/not-set meanings are chosen so that Emacs syntax + remains the value 0. The bits are given in alphabetical order, and + the definitions shifted by one from the previous bit; thus, when we + add or remove a bit, only one other definition need change. */ +typedef unsigned long int reg_syntax_t; + +#ifdef __USE_GNU_REGEX + +/* If this bit is not set, then \ inside a bracket expression is literal. + If set, then such a \ quotes the following character. */ +# define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) + +/* If this bit is not set, then + and ? are operators, and \+ and \? are + literals. + If set, then \+ and \? are operators and + and ? are literals. */ +# define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) + +/* If this bit is set, then character classes are supported. They are: + [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], + [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. + If not set, then character classes are not supported. */ +# define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) + +/* If this bit is set, then ^ and $ are always anchors (outside bracket + expressions, of course). + If this bit is not set, then it depends: + ^ is an anchor if it is at the beginning of a regular + expression or after an open-group or an alternation operator; + $ is an anchor if it is at the end of a regular expression, or + before a close-group or an alternation operator. + + This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because + POSIX draft 11.2 says that * etc. in leading positions is undefined. + We already implemented a previous draft which made those constructs + invalid, though, so we haven't changed the code back. */ +# define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) + +/* If this bit is set, then special characters are always special + regardless of where they are in the pattern. + If this bit is not set, then special characters are special only in + some contexts; otherwise they are ordinary. Specifically, + * + ? and intervals are only special when not after the beginning, + open-group, or alternation operator. */ +# define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) + +/* If this bit is set, then *, +, ?, and { cannot be first in an re or + immediately after an alternation or begin-group operator. */ +# define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) + +/* If this bit is set, then . matches newline. + If not set, then it doesn't. */ +# define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) + +/* If this bit is set, then . doesn't match NUL. + If not set, then it does. */ +# define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) + +/* If this bit is set, nonmatching lists [^...] do not match newline. + If not set, they do. */ +# define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) + +/* If this bit is set, either \{...\} or {...} defines an + interval, depending on RE_NO_BK_BRACES. + If not set, \{, \}, {, and } are literals. */ +# define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) + +/* If this bit is set, +, ? and | aren't recognized as operators. + If not set, they are. */ +# define RE_LIMITED_OPS (RE_INTERVALS << 1) + +/* If this bit is set, newline is an alternation operator. + If not set, newline is literal. */ +# define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) + +/* If this bit is set, then `{...}' defines an interval, and \{ and \} + are literals. + If not set, then `\{...\}' defines an interval. */ +# define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) + +/* If this bit is set, (...) defines a group, and \( and \) are literals. + If not set, \(...\) defines a group, and ( and ) are literals. */ +# define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) + +/* If this bit is set, then \<digit> matches <digit>. + If not set, then \<digit> is a back-reference. */ +# define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) + +/* If this bit is set, then | is an alternation operator, and \| is literal. + If not set, then \| is an alternation operator, and | is literal. */ +# define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) + +/* If this bit is set, then an ending range point collating higher + than the starting range point, as in [z-a], is invalid. + If not set, then when ending range point collates higher than the + starting range point, the range is ignored. */ +# define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) + +/* If this bit is set, then an unmatched ) is ordinary. + If not set, then an unmatched ) is invalid. */ +# define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) + +/* If this bit is set, succeed as soon as we match the whole pattern, + without further backtracking. */ +# define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) + +/* If this bit is set, do not process the GNU regex operators. + If not set, then the GNU regex operators are recognized. */ +# define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) + +/* If this bit is set, turn on internal regex debugging. + If not set, and debugging was on, turn it off. + This only works if regex.c is compiled -DDEBUG. + We define this bit always, so that all that's needed to turn on + debugging is to recompile regex.c; the calling code can always have + this bit set, and it won't affect anything in the normal case. */ +# define RE_DEBUG (RE_NO_GNU_OPS << 1) + +/* If this bit is set, a syntactically invalid interval is treated as + a string of ordinary characters. For example, the ERE 'a{1' is + treated as 'a\{1'. */ +# define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1) + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +# define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) + +/* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only + for ^, because it is difficult to scan the regex backwards to find + whether ^ should be special. */ +# define RE_CARET_ANCHORS_HERE (RE_ICASE << 1) + +/* If this bit is set, then \{ cannot be first in an bre or + immediately after an alternation or begin-group operator. */ +# define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1) + +/* If this bit is set, then no_sub will be set to 1 during + re_compile_pattern. */ +# define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1) + +#endif /* defined __USE_GNU_REGEX */ + +/* This global variable defines the particular regexp syntax to use (for + some interfaces). When a regexp is compiled, the syntax used is + stored in the pattern buffer, so changing this does not affect + already-compiled regexps. */ +extern reg_syntax_t re_syntax_options; + +#ifdef __USE_GNU_REGEX +/* Define combinations of the above bits for the standard possibilities. + (The [[[ comments delimit what gets put into the Texinfo file, so + don't delete them!) */ +/* [[[begin syntaxes]]] */ +# define RE_SYNTAX_EMACS 0 + +# define RE_SYNTAX_AWK \ + (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ + | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ + | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) + +# define RE_SYNTAX_GNU_AWK \ + ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ + & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \ + | RE_CONTEXT_INVALID_OPS )) + +# define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ + | RE_INTERVALS | RE_NO_GNU_OPS) + +# define RE_SYNTAX_GREP \ + (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ + | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ + | RE_NEWLINE_ALT) + +# define RE_SYNTAX_EGREP \ + (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ + | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ + | RE_NO_BK_VBAR) + +# define RE_SYNTAX_POSIX_EGREP \ + (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \ + | RE_INVALID_INTERVAL_ORD) + +/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ +# define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC + +# define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC + +/* Syntax bits common to both basic and extended POSIX regex syntax. */ +# define _RE_SYNTAX_POSIX_COMMON \ + (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ + | RE_INTERVALS | RE_NO_EMPTY_RANGES) + +# define RE_SYNTAX_POSIX_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP) + +/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes + RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this + isn't minimal, since other operators, such as \`, aren't disabled. */ +# define RE_SYNTAX_POSIX_MINIMAL_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) + +# define RE_SYNTAX_POSIX_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ + | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) + +/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is + removed and RE_NO_BK_REFS is added. */ +# define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) +/* [[[end syntaxes]]] */ + +#endif /* defined __USE_GNU_REGEX */ + +#ifdef __USE_GNU_REGEX + +/* Maximum number of duplicates an interval can allow. POSIX-conforming + systems might define this in <limits.h>, but we want our + value, so remove any previous define. */ +# ifdef RE_DUP_MAX +# undef RE_DUP_MAX +# endif + +/* RE_DUP_MAX is 2**15 - 1 because an earlier implementation stored + the counter as a 2-byte signed integer. This is no longer true, so + RE_DUP_MAX could be increased to (INT_MAX / 10 - 1), or to + ((SIZE_MAX - 2) / 10 - 1) if _REGEX_LARGE_OFFSETS is defined. + However, there would be a huge performance problem if someone + actually used a pattern like a\{214748363\}, so RE_DUP_MAX retains + its historical value. */ +# define RE_DUP_MAX (0x7fff) + +#endif /* defined __USE_GNU_REGEX */ + + +/* POSIX `cflags' bits (i.e., information for `regcomp'). */ + +/* If this bit is set, then use extended regular expression syntax. + If not set, then use basic regular expression syntax. */ +#define REG_EXTENDED 1 + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +#define REG_ICASE (1 << 1) + +/* If this bit is set, then anchors do not match at newline + characters in the string. + If not set, then anchors do match at newlines. */ +#define REG_NEWLINE (1 << 2) + +/* If this bit is set, then report only success or fail in regexec. + If not set, then returns differ between not matching and errors. */ +#define REG_NOSUB (1 << 3) + + +/* POSIX `eflags' bits (i.e., information for regexec). */ + +/* If this bit is set, then the beginning-of-line operator doesn't match + the beginning of the string (presumably because it's not the + beginning of a line). + If not set, then the beginning-of-line operator does match the + beginning of the string. */ +#define REG_NOTBOL 1 + +/* Like REG_NOTBOL, except for the end-of-line. */ +#define REG_NOTEOL (1 << 1) + +/* Use PMATCH[0] to delimit the start and end of the search in the + buffer. */ +#define REG_STARTEND (1 << 2) + + +/* If any error codes are removed, changed, or added, update the + `__re_error_msgid' table in regcomp.c. */ + +typedef enum +{ + _REG_ENOSYS = -1, /* This will never happen for this implementation. */ + _REG_NOERROR = 0, /* Success. */ + _REG_NOMATCH, /* Didn't find a match (for regexec). */ + + /* POSIX regcomp return error codes. (In the order listed in the + standard.) */ + _REG_BADPAT, /* Invalid pattern. */ + _REG_ECOLLATE, /* Invalid collating element. */ + _REG_ECTYPE, /* Invalid character class name. */ + _REG_EESCAPE, /* Trailing backslash. */ + _REG_ESUBREG, /* Invalid back reference. */ + _REG_EBRACK, /* Unmatched left bracket. */ + _REG_EPAREN, /* Parenthesis imbalance. */ + _REG_EBRACE, /* Unmatched \{. */ + _REG_BADBR, /* Invalid contents of \{\}. */ + _REG_ERANGE, /* Invalid range end. */ + _REG_ESPACE, /* Ran out of memory. */ + _REG_BADRPT, /* No preceding re for repetition op. */ + + /* Error codes we've added. */ + _REG_EEND, /* Premature end. */ + _REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ + _REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ +} reg_errcode_t; + +#ifdef _XOPEN_SOURCE +# define REG_ENOSYS _REG_ENOSYS +#endif +#define REG_NOERROR _REG_NOERROR +#define REG_NOMATCH _REG_NOMATCH +#define REG_BADPAT _REG_BADPAT +#define REG_ECOLLATE _REG_ECOLLATE +#define REG_ECTYPE _REG_ECTYPE +#define REG_EESCAPE _REG_EESCAPE +#define REG_ESUBREG _REG_ESUBREG +#define REG_EBRACK _REG_EBRACK +#define REG_EPAREN _REG_EPAREN +#define REG_EBRACE _REG_EBRACE +#define REG_BADBR _REG_BADBR +#define REG_ERANGE _REG_ERANGE +#define REG_ESPACE _REG_ESPACE +#define REG_BADRPT _REG_BADRPT +#define REG_EEND _REG_EEND +#define REG_ESIZE _REG_ESIZE +#define REG_ERPAREN _REG_ERPAREN + +/* struct re_pattern_buffer normally uses member names like `buffer' + that POSIX does not allow. In POSIX mode these members have names + with leading `re_' (e.g., `re_buffer'). */ +#ifdef __USE_GNU_REGEX +# define _REG_RE_NAME(id) id +# define _REG_RM_NAME(id) id +#else +# define _REG_RE_NAME(id) re_##id +# define _REG_RM_NAME(id) rm_##id +#endif + +/* The user can specify the type of the re_translate member by + defining the macro RE_TRANSLATE_TYPE, which defaults to unsigned + char *. This pollutes the POSIX name space, so in POSIX mode just + use unsigned char *. */ +#ifdef __USE_GNU_REGEX +# ifndef RE_TRANSLATE_TYPE +# define RE_TRANSLATE_TYPE unsigned char * +# endif +# define REG_TRANSLATE_TYPE RE_TRANSLATE_TYPE +#else +# define REG_TRANSLATE_TYPE unsigned char * +#endif + +/* This data structure represents a compiled pattern. Before calling + the pattern compiler, the fields `buffer', `allocated', `fastmap', + `translate', and `no_sub' can be set. After the pattern has been + compiled, the `re_nsub' field is available. All other fields are + private to the regex routines. */ + +struct re_pattern_buffer +{ + /* Space that holds the compiled pattern. It is declared as + `unsigned char *' because its elements are sometimes used as + array indexes. */ + unsigned char *_REG_RE_NAME (buffer); + + /* Number of bytes to which `buffer' points. */ + __re_long_size_t _REG_RE_NAME (allocated); + + /* Number of bytes actually used in `buffer'. */ + __re_long_size_t _REG_RE_NAME (used); + + /* Syntax setting with which the pattern was compiled. */ + reg_syntax_t _REG_RE_NAME (syntax); + + /* Pointer to a fastmap, if any, otherwise zero. re_search uses the + fastmap, if there is one, to skip over impossible starting points + for matches. */ + char *_REG_RE_NAME (fastmap); + + /* Either a translate table to apply to all characters before + comparing them, or zero for no translation. The translation is + applied to a pattern when it is compiled and to a string when it + is matched. */ + REG_TRANSLATE_TYPE _REG_RE_NAME (translate); + + /* Number of subexpressions found by the compiler. */ + size_t re_nsub; + + /* Zero if this pattern cannot match the empty string, one else. + Well, in truth it's used only in `re_search_2', to see whether or + not we should use the fastmap, so we don't set this absolutely + perfectly; see `re_compile_fastmap' (the `duplicate' case). */ + unsigned int _REG_RE_NAME (can_be_null) : 1; + + /* If REGS_UNALLOCATED, allocate space in the `regs' structure + for `max (RE_NREGS, re_nsub + 1)' groups. + If REGS_REALLOCATE, reallocate space if necessary. + If REGS_FIXED, use what's there. */ +#ifdef __USE_GNU_REGEX +# define REGS_UNALLOCATED 0 +# define REGS_REALLOCATE 1 +# define REGS_FIXED 2 +#endif + unsigned int _REG_RE_NAME (regs_allocated) : 2; + + /* Set to zero when `regex_compile' compiles a pattern; set to one + by `re_compile_fastmap' if it updates the fastmap. */ + unsigned int _REG_RE_NAME (fastmap_accurate) : 1; + + /* If set, `re_match_2' does not return information about + subexpressions. */ + unsigned int _REG_RE_NAME (no_sub) : 1; + + /* If set, a beginning-of-line anchor doesn't match at the beginning + of the string. */ + unsigned int _REG_RE_NAME (not_bol) : 1; + + /* Similarly for an end-of-line anchor. */ + unsigned int _REG_RE_NAME (not_eol) : 1; + + /* If true, an anchor at a newline matches. */ + unsigned int _REG_RE_NAME (newline_anchor) : 1; + +/* [[[end pattern_buffer]]] */ +}; + +typedef struct re_pattern_buffer regex_t; + +/* This is the structure we store register match data in. See + regex.texinfo for a full description of what registers match. */ +struct re_registers +{ + __re_size_t _REG_RM_NAME (num_regs); + regoff_t *_REG_RM_NAME (start); + regoff_t *_REG_RM_NAME (end); +}; + + +/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, + `re_match_2' returns information about at least this many registers + the first time a `regs' structure is passed. */ +#if !defined RE_NREGS && defined __USE_GNU_REGEX +# define RE_NREGS 30 +#endif + + +/* POSIX specification for registers. Aside from the different names than + `re_registers', POSIX uses an array of structures, instead of a + structure of arrays. */ +typedef struct +{ + regoff_t rm_so; /* Byte offset from string's start to substring's start. */ + regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ +} regmatch_t; + +/* Declarations for routines. */ + +/* Sets the current default syntax to SYNTAX, and return the old syntax. + You can also simply assign to the `re_syntax_options' variable. */ +extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax); + +/* Compile the regular expression PATTERN, with length LENGTH + and syntax given by the global `re_syntax_options', into the buffer + BUFFER. Return NULL if successful, and an error string if not. */ +extern const char *re_compile_pattern (const char *__pattern, size_t __length, + struct re_pattern_buffer *__buffer); + + +/* Compile a fastmap for the compiled pattern in BUFFER; used to + accelerate searches. Return 0 if successful and -2 if was an + internal error. */ +extern int re_compile_fastmap (struct re_pattern_buffer *__buffer); + + +/* Search in the string STRING (with length LENGTH) for the pattern + compiled into BUFFER. Start searching at position START, for RANGE + characters. Return the starting position of the match, -1 for no + match, or -2 for an internal error. Also return register + information in REGS (if REGS and BUFFER->no_sub are nonzero). */ +extern regoff_t re_search (struct re_pattern_buffer *__buffer, + const char *__string, __re_idx_t __length, + __re_idx_t __start, regoff_t __range, + struct re_registers *__regs); + + +/* Like `re_search', but search in the concatenation of STRING1 and + STRING2. Also, stop searching at index START + STOP. */ +extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer, + const char *__string1, __re_idx_t __length1, + const char *__string2, __re_idx_t __length2, + __re_idx_t __start, regoff_t __range, + struct re_registers *__regs, + __re_idx_t __stop); + + +/* Like `re_search', but return how many characters in STRING the regexp + in BUFFER matched, starting at position START. */ +extern regoff_t re_match (struct re_pattern_buffer *__buffer, + const char *__string, __re_idx_t __length, + __re_idx_t __start, struct re_registers *__regs); + + +/* Relates to `re_match' as `re_search_2' relates to `re_search'. */ +extern regoff_t re_match_2 (struct re_pattern_buffer *__buffer, + const char *__string1, __re_idx_t __length1, + const char *__string2, __re_idx_t __length2, + __re_idx_t __start, struct re_registers *__regs, + __re_idx_t __stop); + + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using BUFFER and REGS will use this memory + for recording register information. STARTS and ENDS must be + allocated with malloc, and must each be at least `NUM_REGS * sizeof + (regoff_t)' bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ +extern void re_set_registers (struct re_pattern_buffer *__buffer, + struct re_registers *__regs, + __re_size_t __num_regs, + regoff_t *__starts, regoff_t *__ends); + +#if defined _REGEX_RE_COMP || defined _LIBC +# ifndef _CRAY +/* 4.2 bsd compatibility. */ +extern char *re_comp (const char *); +extern int re_exec (const char *); +# endif +#endif + +/* GCC 2.95 and later have "__restrict"; C99 compilers have + "restrict", and "configure" may have defined "restrict". + Other compilers use __restrict, __restrict__, and _Restrict, and + 'configure' might #define 'restrict' to those words, so pick a + different name. */ +#ifndef _Restrict_ +# if 199901L <= __STDC_VERSION__ +# define _Restrict_ restrict +# elif 2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__) +# define _Restrict_ __restrict +# else +# define _Restrict_ +# endif +#endif +/* gcc 3.1 and up support the [restrict] syntax. Don't trust + sys/cdefs.h's definition of __restrict_arr, though, as it + mishandles gcc -ansi -pedantic. */ +#ifndef _Restrict_arr_ +# if ((199901L <= __STDC_VERSION__ \ + || ((3 < __GNUC__ || (3 == __GNUC__ && 1 <= __GNUC_MINOR__)) \ + && !__STRICT_ANSI__)) \ + && !defined __GNUG__) +# define _Restrict_arr_ _Restrict_ +# else +# define _Restrict_arr_ +# endif +#endif + +/* POSIX compatibility. */ +extern int regcomp (regex_t *_Restrict_ __preg, + const char *_Restrict_ __pattern, + int __cflags); + +extern int regexec (const regex_t *_Restrict_ __preg, + const char *_Restrict_ __string, size_t __nmatch, + regmatch_t __pmatch[_Restrict_arr_], + int __eflags); + +extern size_t regerror (int __errcode, const regex_t *_Restrict_ __preg, + char *_Restrict_ __errbuf, size_t __errbuf_size); + +extern void regfree (regex_t *__preg); + + +#ifdef __cplusplus +} +#endif /* C++ */ + +#endif /* regex.h */ diff --git a/lib/regex_internal.c b/lib/regex_internal.c new file mode 100644 index 0000000..2129888 --- /dev/null +++ b/lib/regex_internal.c @@ -0,0 +1,1741 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007 Free Software + Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +static void re_string_construct_common (const char *str, Idx len, + re_string_t *pstr, + RE_TRANSLATE_TYPE trans, bool icase, + const re_dfa_t *dfa) internal_function; +static re_dfastate_t *create_ci_newstate (const re_dfa_t *dfa, + const re_node_set *nodes, + re_hashval_t hash) internal_function; +static re_dfastate_t *create_cd_newstate (const re_dfa_t *dfa, + const re_node_set *nodes, + unsigned int context, + re_hashval_t hash) internal_function; + +/* Functions for string operation. */ + +/* This function allocate the buffers. It is necessary to call + re_string_reconstruct before using the object. */ + +static reg_errcode_t +internal_function +re_string_allocate (re_string_t *pstr, const char *str, Idx len, Idx init_len, + RE_TRANSLATE_TYPE trans, bool icase, const re_dfa_t *dfa) +{ + reg_errcode_t ret; + Idx init_buf_len; + + /* Ensure at least one character fits into the buffers. */ + if (init_len < dfa->mb_cur_max) + init_len = dfa->mb_cur_max; + init_buf_len = (len + 1 < init_len) ? len + 1: init_len; + re_string_construct_common (str, len, pstr, trans, icase, dfa); + + ret = re_string_realloc_buffers (pstr, init_buf_len); + if (BE (ret != REG_NOERROR, 0)) + return ret; + + pstr->word_char = dfa->word_char; + pstr->word_ops_used = dfa->word_ops_used; + pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str; + pstr->valid_len = (pstr->mbs_allocated || dfa->mb_cur_max > 1) ? 0 : len; + pstr->valid_raw_len = pstr->valid_len; + return REG_NOERROR; +} + +/* This function allocate the buffers, and initialize them. */ + +static reg_errcode_t +internal_function +re_string_construct (re_string_t *pstr, const char *str, Idx len, + RE_TRANSLATE_TYPE trans, bool icase, const re_dfa_t *dfa) +{ + reg_errcode_t ret; + memset (pstr, '\0', sizeof (re_string_t)); + re_string_construct_common (str, len, pstr, trans, icase, dfa); + + if (len > 0) + { + ret = re_string_realloc_buffers (pstr, len + 1); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str; + + if (icase) + { +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + { + while (1) + { + ret = build_wcs_upper_buffer (pstr); + if (BE (ret != REG_NOERROR, 0)) + return ret; + if (pstr->valid_raw_len >= len) + break; + if (pstr->bufs_len > pstr->valid_len + dfa->mb_cur_max) + break; + ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + } + else +#endif /* RE_ENABLE_I18N */ + build_upper_buffer (pstr); + } + else + { +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + build_wcs_buffer (pstr); + else +#endif /* RE_ENABLE_I18N */ + { + if (trans != NULL) + re_string_translate_buffer (pstr); + else + { + pstr->valid_len = pstr->bufs_len; + pstr->valid_raw_len = pstr->bufs_len; + } + } + } + + return REG_NOERROR; +} + +/* Helper functions for re_string_allocate, and re_string_construct. */ + +static reg_errcode_t +internal_function +re_string_realloc_buffers (re_string_t *pstr, Idx new_buf_len) +{ +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + wint_t *new_wcs; + + /* Avoid overflow. */ + size_t max_object_size = MAX (sizeof (wint_t), sizeof (Idx)); + if (BE (SIZE_MAX / max_object_size < new_buf_len, 0)) + return REG_ESPACE; + + new_wcs = re_realloc (pstr->wcs, wint_t, new_buf_len); + if (BE (new_wcs == NULL, 0)) + return REG_ESPACE; + pstr->wcs = new_wcs; + if (pstr->offsets != NULL) + { + Idx *new_offsets = re_realloc (pstr->offsets, Idx, new_buf_len); + if (BE (new_offsets == NULL, 0)) + return REG_ESPACE; + pstr->offsets = new_offsets; + } + } +#endif /* RE_ENABLE_I18N */ + if (pstr->mbs_allocated) + { + unsigned char *new_mbs = re_realloc (pstr->mbs, unsigned char, + new_buf_len); + if (BE (new_mbs == NULL, 0)) + return REG_ESPACE; + pstr->mbs = new_mbs; + } + pstr->bufs_len = new_buf_len; + return REG_NOERROR; +} + + +static void +internal_function +re_string_construct_common (const char *str, Idx len, re_string_t *pstr, + RE_TRANSLATE_TYPE trans, bool icase, + const re_dfa_t *dfa) +{ + pstr->raw_mbs = (const unsigned char *) str; + pstr->len = len; + pstr->raw_len = len; + pstr->trans = trans; + pstr->icase = icase; + pstr->mbs_allocated = (trans != NULL || icase); + pstr->mb_cur_max = dfa->mb_cur_max; + pstr->is_utf8 = dfa->is_utf8; + pstr->map_notascii = dfa->map_notascii; + pstr->stop = pstr->len; + pstr->raw_stop = pstr->stop; +} + +#ifdef RE_ENABLE_I18N + +/* Build wide character buffer PSTR->WCS. + If the byte sequence of the string are: + <mb1>(0), <mb1>(1), <mb2>(0), <mb2>(1), <sb3> + Then wide character buffer will be: + <wc1> , WEOF , <wc2> , WEOF , <wc3> + We use WEOF for padding, they indicate that the position isn't + a first byte of a multibyte character. + + Note that this function assumes PSTR->VALID_LEN elements are already + built and starts from PSTR->VALID_LEN. */ + +static void +internal_function +build_wcs_buffer (re_string_t *pstr) +{ +#ifdef _LIBC + unsigned char buf[MB_LEN_MAX]; + assert (MB_LEN_MAX >= pstr->mb_cur_max); +#else + unsigned char buf[64]; +#endif + mbstate_t prev_st; + Idx byte_idx, end_idx, remain_len; + size_t mbclen; + + /* Build the buffers from pstr->valid_len to either pstr->len or + pstr->bufs_len. */ + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + for (byte_idx = pstr->valid_len; byte_idx < end_idx;) + { + wchar_t wc; + const char *p; + + remain_len = end_idx - byte_idx; + prev_st = pstr->cur_state; + /* Apply the translation if we need. */ + if (BE (pstr->trans != NULL, 0)) + { + int i, ch; + + for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i) + { + ch = pstr->raw_mbs [pstr->raw_mbs_idx + byte_idx + i]; + buf[i] = pstr->mbs[byte_idx + i] = pstr->trans[ch]; + } + p = (const char *) buf; + } + else + p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx; + mbclen = mbrtowc (&wc, p, remain_len, &pstr->cur_state); + if (BE (mbclen == (size_t) -2, 0)) + { + /* The buffer doesn't have enough space, finish to build. */ + pstr->cur_state = prev_st; + break; + } + else if (BE (mbclen == (size_t) -1 || mbclen == 0, 0)) + { + /* We treat these cases as a singlebyte character. */ + mbclen = 1; + wc = (wchar_t) pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]; + if (BE (pstr->trans != NULL, 0)) + wc = pstr->trans[wc]; + pstr->cur_state = prev_st; + } + + /* Write wide character and padding. */ + pstr->wcs[byte_idx++] = wc; + /* Write paddings. */ + for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;) + pstr->wcs[byte_idx++] = WEOF; + } + pstr->valid_len = byte_idx; + pstr->valid_raw_len = byte_idx; +} + +/* Build wide character buffer PSTR->WCS like build_wcs_buffer, + but for REG_ICASE. */ + +static reg_errcode_t +internal_function +build_wcs_upper_buffer (re_string_t *pstr) +{ + mbstate_t prev_st; + Idx src_idx, byte_idx, end_idx, remain_len; + size_t mbclen; +#ifdef _LIBC + char buf[MB_LEN_MAX]; + assert (MB_LEN_MAX >= pstr->mb_cur_max); +#else + char buf[64]; +#endif + + byte_idx = pstr->valid_len; + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + + /* The following optimization assumes that ASCII characters can be + mapped to wide characters with a simple cast. */ + if (! pstr->map_notascii && pstr->trans == NULL && !pstr->offsets_needed) + { + while (byte_idx < end_idx) + { + wchar_t wc; + + if (isascii (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]) + && mbsinit (&pstr->cur_state)) + { + /* In case of a singlebyte character. */ + pstr->mbs[byte_idx] + = toupper (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]); + /* The next step uses the assumption that wchar_t is encoded + ASCII-safe: all ASCII values can be converted like this. */ + pstr->wcs[byte_idx] = (wchar_t) pstr->mbs[byte_idx]; + ++byte_idx; + continue; + } + + remain_len = end_idx - byte_idx; + prev_st = pstr->cur_state; + mbclen = mbrtowc (&wc, + ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx + + byte_idx), remain_len, &pstr->cur_state); + if (BE (mbclen < (size_t) -2, 1)) + { + wchar_t wcu = wc; + if (iswlower (wc)) + { + size_t mbcdlen; + + wcu = towupper (wc); + mbcdlen = wcrtomb (buf, wcu, &prev_st); + if (BE (mbclen == mbcdlen, 1)) + memcpy (pstr->mbs + byte_idx, buf, mbclen); + else + { + src_idx = byte_idx; + goto offsets_needed; + } + } + else + memcpy (pstr->mbs + byte_idx, + pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx, mbclen); + pstr->wcs[byte_idx++] = wcu; + /* Write paddings. */ + for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;) + pstr->wcs[byte_idx++] = WEOF; + } + else if (mbclen == (size_t) -1 || mbclen == 0) + { + /* It is an invalid character or '\0'. Just use the byte. */ + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]; + pstr->mbs[byte_idx] = ch; + /* And also cast it to wide char. */ + pstr->wcs[byte_idx++] = (wchar_t) ch; + if (BE (mbclen == (size_t) -1, 0)) + pstr->cur_state = prev_st; + } + else + { + /* The buffer doesn't have enough space, finish to build. */ + pstr->cur_state = prev_st; + break; + } + } + pstr->valid_len = byte_idx; + pstr->valid_raw_len = byte_idx; + return REG_NOERROR; + } + else + for (src_idx = pstr->valid_raw_len; byte_idx < end_idx;) + { + wchar_t wc; + const char *p; + offsets_needed: + remain_len = end_idx - byte_idx; + prev_st = pstr->cur_state; + if (BE (pstr->trans != NULL, 0)) + { + int i, ch; + + for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i) + { + ch = pstr->raw_mbs [pstr->raw_mbs_idx + src_idx + i]; + buf[i] = pstr->trans[ch]; + } + p = (const char *) buf; + } + else + p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + src_idx; + mbclen = mbrtowc (&wc, p, remain_len, &pstr->cur_state); + if (BE (mbclen < (size_t) -2, 1)) + { + wchar_t wcu = wc; + if (iswlower (wc)) + { + size_t mbcdlen; + + wcu = towupper (wc); + mbcdlen = wcrtomb ((char *) buf, wcu, &prev_st); + if (BE (mbclen == mbcdlen, 1)) + memcpy (pstr->mbs + byte_idx, buf, mbclen); + else if (mbcdlen != (size_t) -1) + { + size_t i; + + if (byte_idx + mbcdlen > pstr->bufs_len) + { + pstr->cur_state = prev_st; + break; + } + + if (pstr->offsets == NULL) + { + pstr->offsets = re_malloc (Idx, pstr->bufs_len); + + if (pstr->offsets == NULL) + return REG_ESPACE; + } + if (!pstr->offsets_needed) + { + for (i = 0; i < (size_t) byte_idx; ++i) + pstr->offsets[i] = i; + pstr->offsets_needed = 1; + } + + memcpy (pstr->mbs + byte_idx, buf, mbcdlen); + pstr->wcs[byte_idx] = wcu; + pstr->offsets[byte_idx] = src_idx; + for (i = 1; i < mbcdlen; ++i) + { + pstr->offsets[byte_idx + i] + = src_idx + (i < mbclen ? i : mbclen - 1); + pstr->wcs[byte_idx + i] = WEOF; + } + pstr->len += mbcdlen - mbclen; + if (pstr->raw_stop > src_idx) + pstr->stop += mbcdlen - mbclen; + end_idx = (pstr->bufs_len > pstr->len) + ? pstr->len : pstr->bufs_len; + byte_idx += mbcdlen; + src_idx += mbclen; + continue; + } + else + memcpy (pstr->mbs + byte_idx, p, mbclen); + } + else + memcpy (pstr->mbs + byte_idx, p, mbclen); + + if (BE (pstr->offsets_needed != 0, 0)) + { + size_t i; + for (i = 0; i < mbclen; ++i) + pstr->offsets[byte_idx + i] = src_idx + i; + } + src_idx += mbclen; + + pstr->wcs[byte_idx++] = wcu; + /* Write paddings. */ + for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;) + pstr->wcs[byte_idx++] = WEOF; + } + else if (mbclen == (size_t) -1 || mbclen == 0) + { + /* It is an invalid character or '\0'. Just use the byte. */ + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + src_idx]; + + if (BE (pstr->trans != NULL, 0)) + ch = pstr->trans [ch]; + pstr->mbs[byte_idx] = ch; + + if (BE (pstr->offsets_needed != 0, 0)) + pstr->offsets[byte_idx] = src_idx; + ++src_idx; + + /* And also cast it to wide char. */ + pstr->wcs[byte_idx++] = (wchar_t) ch; + if (BE (mbclen == (size_t) -1, 0)) + pstr->cur_state = prev_st; + } + else + { + /* The buffer doesn't have enough space, finish to build. */ + pstr->cur_state = prev_st; + break; + } + } + pstr->valid_len = byte_idx; + pstr->valid_raw_len = src_idx; + return REG_NOERROR; +} + +/* Skip characters until the index becomes greater than NEW_RAW_IDX. + Return the index. */ + +static Idx +internal_function +re_string_skip_chars (re_string_t *pstr, Idx new_raw_idx, wint_t *last_wc) +{ + mbstate_t prev_st; + Idx rawbuf_idx; + size_t mbclen; + wint_t wc = WEOF; + + /* Skip the characters which are not necessary to check. */ + for (rawbuf_idx = pstr->raw_mbs_idx + pstr->valid_raw_len; + rawbuf_idx < new_raw_idx;) + { + wchar_t wc2; + Idx remain_len; + remain_len = pstr->len - rawbuf_idx; + prev_st = pstr->cur_state; + mbclen = mbrtowc (&wc2, (const char *) pstr->raw_mbs + rawbuf_idx, + remain_len, &pstr->cur_state); + if (BE (mbclen == (size_t) -2 || mbclen == (size_t) -1 || mbclen == 0, 0)) + { + /* We treat these cases as a single byte character. */ + if (mbclen == 0 || remain_len == 0) + wc = L'\0'; + else + wc = *(unsigned char *) (pstr->raw_mbs + rawbuf_idx); + mbclen = 1; + pstr->cur_state = prev_st; + } + else + wc = wc2; + /* Then proceed the next character. */ + rawbuf_idx += mbclen; + } + *last_wc = wc; + return rawbuf_idx; +} +#endif /* RE_ENABLE_I18N */ + +/* Build the buffer PSTR->MBS, and apply the translation if we need. + This function is used in case of REG_ICASE. */ + +static void +internal_function +build_upper_buffer (re_string_t *pstr) +{ + Idx char_idx, end_idx; + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + + for (char_idx = pstr->valid_len; char_idx < end_idx; ++char_idx) + { + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + char_idx]; + if (BE (pstr->trans != NULL, 0)) + ch = pstr->trans[ch]; + if (islower (ch)) + pstr->mbs[char_idx] = toupper (ch); + else + pstr->mbs[char_idx] = ch; + } + pstr->valid_len = char_idx; + pstr->valid_raw_len = char_idx; +} + +/* Apply TRANS to the buffer in PSTR. */ + +static void +internal_function +re_string_translate_buffer (re_string_t *pstr) +{ + Idx buf_idx, end_idx; + end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len; + + for (buf_idx = pstr->valid_len; buf_idx < end_idx; ++buf_idx) + { + int ch = pstr->raw_mbs[pstr->raw_mbs_idx + buf_idx]; + pstr->mbs[buf_idx] = pstr->trans[ch]; + } + + pstr->valid_len = buf_idx; + pstr->valid_raw_len = buf_idx; +} + +/* This function re-construct the buffers. + Concretely, convert to wide character in case of pstr->mb_cur_max > 1, + convert to upper case in case of REG_ICASE, apply translation. */ + +static reg_errcode_t +internal_function +re_string_reconstruct (re_string_t *pstr, Idx idx, int eflags) +{ + Idx offset; + + if (BE (pstr->raw_mbs_idx <= idx, 0)) + offset = idx - pstr->raw_mbs_idx; + else + { + /* Reset buffer. */ +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + memset (&pstr->cur_state, '\0', sizeof (mbstate_t)); +#endif /* RE_ENABLE_I18N */ + pstr->len = pstr->raw_len; + pstr->stop = pstr->raw_stop; + pstr->valid_len = 0; + pstr->raw_mbs_idx = 0; + pstr->valid_raw_len = 0; + pstr->offsets_needed = 0; + pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF + : CONTEXT_NEWLINE | CONTEXT_BEGBUF); + if (!pstr->mbs_allocated) + pstr->mbs = (unsigned char *) pstr->raw_mbs; + offset = idx; + } + + if (BE (offset != 0, 1)) + { + /* Should the already checked characters be kept? */ + if (BE (offset < pstr->valid_raw_len, 1)) + { + /* Yes, move them to the front of the buffer. */ +#ifdef RE_ENABLE_I18N + if (BE (pstr->offsets_needed, 0)) + { + Idx low = 0, high = pstr->valid_len, mid; + do + { + mid = (high + low) / 2; + if (pstr->offsets[mid] > offset) + high = mid; + else if (pstr->offsets[mid] < offset) + low = mid + 1; + else + break; + } + while (low < high); + if (pstr->offsets[mid] < offset) + ++mid; + pstr->tip_context = re_string_context_at (pstr, mid - 1, + eflags); + /* This can be quite complicated, so handle specially + only the common and easy case where the character with + different length representation of lower and upper + case is present at or after offset. */ + if (pstr->valid_len > offset + && mid == offset && pstr->offsets[mid] == offset) + { + memmove (pstr->wcs, pstr->wcs + offset, + (pstr->valid_len - offset) * sizeof (wint_t)); + memmove (pstr->mbs, pstr->mbs + offset, pstr->valid_len - offset); + pstr->valid_len -= offset; + pstr->valid_raw_len -= offset; + for (low = 0; low < pstr->valid_len; low++) + pstr->offsets[low] = pstr->offsets[low + offset] - offset; + } + else + { + /* Otherwise, just find out how long the partial multibyte + character at offset is and fill it with WEOF/255. */ + pstr->len = pstr->raw_len - idx + offset; + pstr->stop = pstr->raw_stop - idx + offset; + pstr->offsets_needed = 0; + while (mid > 0 && pstr->offsets[mid - 1] == offset) + --mid; + while (mid < pstr->valid_len) + if (pstr->wcs[mid] != WEOF) + break; + else + ++mid; + if (mid == pstr->valid_len) + pstr->valid_len = 0; + else + { + pstr->valid_len = pstr->offsets[mid] - offset; + if (pstr->valid_len) + { + for (low = 0; low < pstr->valid_len; ++low) + pstr->wcs[low] = WEOF; + memset (pstr->mbs, 255, pstr->valid_len); + } + } + pstr->valid_raw_len = pstr->valid_len; + } + } + else +#endif + { + pstr->tip_context = re_string_context_at (pstr, offset - 1, + eflags); +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + memmove (pstr->wcs, pstr->wcs + offset, + (pstr->valid_len - offset) * sizeof (wint_t)); +#endif /* RE_ENABLE_I18N */ + if (BE (pstr->mbs_allocated, 0)) + memmove (pstr->mbs, pstr->mbs + offset, + pstr->valid_len - offset); + pstr->valid_len -= offset; + pstr->valid_raw_len -= offset; +#if DEBUG + assert (pstr->valid_len > 0); +#endif + } + } + else + { + /* No, skip all characters until IDX. */ + Idx prev_valid_len = pstr->valid_len; + +#ifdef RE_ENABLE_I18N + if (BE (pstr->offsets_needed, 0)) + { + pstr->len = pstr->raw_len - idx + offset; + pstr->stop = pstr->raw_stop - idx + offset; + pstr->offsets_needed = 0; + } +#endif + pstr->valid_len = 0; +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + Idx wcs_idx; + wint_t wc = WEOF; + + if (pstr->is_utf8) + { + const unsigned char *raw, *p, *end; + + /* Special case UTF-8. Multi-byte chars start with any + byte other than 0x80 - 0xbf. */ + raw = pstr->raw_mbs + pstr->raw_mbs_idx; + end = raw + (offset - pstr->mb_cur_max); + if (end < pstr->raw_mbs) + end = pstr->raw_mbs; + p = raw + offset - 1; +#ifdef _LIBC + /* We know the wchar_t encoding is UCS4, so for the simple + case, ASCII characters, skip the conversion step. */ + if (isascii (*p) && BE (pstr->trans == NULL, 1)) + { + memset (&pstr->cur_state, '\0', sizeof (mbstate_t)); + /* pstr->valid_len = 0; */ + wc = (wchar_t) *p; + } + else +#endif + for (; p >= end; --p) + if ((*p & 0xc0) != 0x80) + { + mbstate_t cur_state; + wchar_t wc2; + Idx mlen = raw + pstr->len - p; + unsigned char buf[6]; + size_t mbclen; + + if (BE (pstr->trans != NULL, 0)) + { + int i = mlen < 6 ? mlen : 6; + while (--i >= 0) + buf[i] = pstr->trans[p[i]]; + } + /* XXX Don't use mbrtowc, we know which conversion + to use (UTF-8 -> UCS4). */ + memset (&cur_state, 0, sizeof (cur_state)); + mbclen = mbrtowc (&wc2, (const char *) p, mlen, + &cur_state); + if (raw + offset - p <= mbclen + && mbclen < (size_t) -2) + { + memset (&pstr->cur_state, '\0', + sizeof (mbstate_t)); + pstr->valid_len = mbclen - (raw + offset - p); + wc = wc2; + } + break; + } + } + + if (wc == WEOF) + pstr->valid_len = re_string_skip_chars (pstr, idx, &wc) - idx; + if (wc == WEOF) + pstr->tip_context + = re_string_context_at (pstr, prev_valid_len - 1, eflags); + else + pstr->tip_context = ((BE (pstr->word_ops_used != 0, 0) + && IS_WIDE_WORD_CHAR (wc)) + ? CONTEXT_WORD + : ((IS_WIDE_NEWLINE (wc) + && pstr->newline_anchor) + ? CONTEXT_NEWLINE : 0)); + if (BE (pstr->valid_len, 0)) + { + for (wcs_idx = 0; wcs_idx < pstr->valid_len; ++wcs_idx) + pstr->wcs[wcs_idx] = WEOF; + if (pstr->mbs_allocated) + memset (pstr->mbs, 255, pstr->valid_len); + } + pstr->valid_raw_len = pstr->valid_len; + } + else +#endif /* RE_ENABLE_I18N */ + { + int c = pstr->raw_mbs[pstr->raw_mbs_idx + offset - 1]; + pstr->valid_raw_len = 0; + if (pstr->trans) + c = pstr->trans[c]; + pstr->tip_context = (bitset_contain (pstr->word_char, c) + ? CONTEXT_WORD + : ((IS_NEWLINE (c) && pstr->newline_anchor) + ? CONTEXT_NEWLINE : 0)); + } + } + if (!BE (pstr->mbs_allocated, 0)) + pstr->mbs += offset; + } + pstr->raw_mbs_idx = idx; + pstr->len -= offset; + pstr->stop -= offset; + + /* Then build the buffers. */ +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + if (pstr->icase) + { + reg_errcode_t ret = build_wcs_upper_buffer (pstr); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + else + build_wcs_buffer (pstr); + } + else +#endif /* RE_ENABLE_I18N */ + if (BE (pstr->mbs_allocated, 0)) + { + if (pstr->icase) + build_upper_buffer (pstr); + else if (pstr->trans != NULL) + re_string_translate_buffer (pstr); + } + else + pstr->valid_len = pstr->len; + + pstr->cur_idx = 0; + return REG_NOERROR; +} + +static unsigned char +internal_function __attribute ((pure)) +re_string_peek_byte_case (const re_string_t *pstr, Idx idx) +{ + int ch; + Idx off; + + /* Handle the common (easiest) cases first. */ + if (BE (!pstr->mbs_allocated, 1)) + return re_string_peek_byte (pstr, idx); + +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1 + && ! re_string_is_single_byte_char (pstr, pstr->cur_idx + idx)) + return re_string_peek_byte (pstr, idx); +#endif + + off = pstr->cur_idx + idx; +#ifdef RE_ENABLE_I18N + if (pstr->offsets_needed) + off = pstr->offsets[off]; +#endif + + ch = pstr->raw_mbs[pstr->raw_mbs_idx + off]; + +#ifdef RE_ENABLE_I18N + /* Ensure that e.g. for tr_TR.UTF-8 BACKSLASH DOTLESS SMALL LETTER I + this function returns CAPITAL LETTER I instead of first byte of + DOTLESS SMALL LETTER I. The latter would confuse the parser, + since peek_byte_case doesn't advance cur_idx in any way. */ + if (pstr->offsets_needed && !isascii (ch)) + return re_string_peek_byte (pstr, idx); +#endif + + return ch; +} + +static unsigned char +internal_function __attribute ((pure)) +re_string_fetch_byte_case (re_string_t *pstr) +{ + if (BE (!pstr->mbs_allocated, 1)) + return re_string_fetch_byte (pstr); + +#ifdef RE_ENABLE_I18N + if (pstr->offsets_needed) + { + Idx off; + int ch; + + /* For tr_TR.UTF-8 [[:islower:]] there is + [[: CAPITAL LETTER I WITH DOT lower:]] in mbs. Skip + in that case the whole multi-byte character and return + the original letter. On the other side, with + [[: DOTLESS SMALL LETTER I return [[:I, as doing + anything else would complicate things too much. */ + + if (!re_string_first_byte (pstr, pstr->cur_idx)) + return re_string_fetch_byte (pstr); + + off = pstr->offsets[pstr->cur_idx]; + ch = pstr->raw_mbs[pstr->raw_mbs_idx + off]; + + if (! isascii (ch)) + return re_string_fetch_byte (pstr); + + re_string_skip_bytes (pstr, + re_string_char_size_at (pstr, pstr->cur_idx)); + return ch; + } +#endif + + return pstr->raw_mbs[pstr->raw_mbs_idx + pstr->cur_idx++]; +} + +static void +internal_function +re_string_destruct (re_string_t *pstr) +{ +#ifdef RE_ENABLE_I18N + re_free (pstr->wcs); + re_free (pstr->offsets); +#endif /* RE_ENABLE_I18N */ + if (pstr->mbs_allocated) + re_free (pstr->mbs); +} + +/* Return the context at IDX in INPUT. */ + +static unsigned int +internal_function +re_string_context_at (const re_string_t *input, Idx idx, int eflags) +{ + int c; + if (BE (! REG_VALID_INDEX (idx), 0)) + /* In this case, we use the value stored in input->tip_context, + since we can't know the character in input->mbs[-1] here. */ + return input->tip_context; + if (BE (idx == input->len, 0)) + return ((eflags & REG_NOTEOL) ? CONTEXT_ENDBUF + : CONTEXT_NEWLINE | CONTEXT_ENDBUF); +#ifdef RE_ENABLE_I18N + if (input->mb_cur_max > 1) + { + wint_t wc; + Idx wc_idx = idx; + while(input->wcs[wc_idx] == WEOF) + { +#ifdef DEBUG + /* It must not happen. */ + assert (REG_VALID_INDEX (wc_idx)); +#endif + --wc_idx; + if (! REG_VALID_INDEX (wc_idx)) + return input->tip_context; + } + wc = input->wcs[wc_idx]; + if (BE (input->word_ops_used != 0, 0) && IS_WIDE_WORD_CHAR (wc)) + return CONTEXT_WORD; + return (IS_WIDE_NEWLINE (wc) && input->newline_anchor + ? CONTEXT_NEWLINE : 0); + } + else +#endif + { + c = re_string_byte_at (input, idx); + if (bitset_contain (input->word_char, c)) + return CONTEXT_WORD; + return IS_NEWLINE (c) && input->newline_anchor ? CONTEXT_NEWLINE : 0; + } +} + +/* Functions for set operation. */ + +static reg_errcode_t +internal_function +re_node_set_alloc (re_node_set *set, Idx size) +{ + set->alloc = size; + set->nelem = 0; + set->elems = re_malloc (Idx, size); + if (BE (set->elems == NULL, 0)) + return REG_ESPACE; + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +re_node_set_init_1 (re_node_set *set, Idx elem) +{ + set->alloc = 1; + set->nelem = 1; + set->elems = re_malloc (Idx, 1); + if (BE (set->elems == NULL, 0)) + { + set->alloc = set->nelem = 0; + return REG_ESPACE; + } + set->elems[0] = elem; + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +re_node_set_init_2 (re_node_set *set, Idx elem1, Idx elem2) +{ + set->alloc = 2; + set->elems = re_malloc (Idx, 2); + if (BE (set->elems == NULL, 0)) + return REG_ESPACE; + if (elem1 == elem2) + { + set->nelem = 1; + set->elems[0] = elem1; + } + else + { + set->nelem = 2; + if (elem1 < elem2) + { + set->elems[0] = elem1; + set->elems[1] = elem2; + } + else + { + set->elems[0] = elem2; + set->elems[1] = elem1; + } + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +re_node_set_init_copy (re_node_set *dest, const re_node_set *src) +{ + dest->nelem = src->nelem; + if (src->nelem > 0) + { + dest->alloc = dest->nelem; + dest->elems = re_malloc (Idx, dest->alloc); + if (BE (dest->elems == NULL, 0)) + { + dest->alloc = dest->nelem = 0; + return REG_ESPACE; + } + memcpy (dest->elems, src->elems, src->nelem * sizeof (Idx)); + } + else + re_node_set_init_empty (dest); + return REG_NOERROR; +} + +/* Calculate the intersection of the sets SRC1 and SRC2. And merge it to + DEST. Return value indicate the error code or REG_NOERROR if succeeded. + Note: We assume dest->elems is NULL, when dest->alloc is 0. */ + +static reg_errcode_t +internal_function +re_node_set_add_intersect (re_node_set *dest, const re_node_set *src1, + const re_node_set *src2) +{ + Idx i1, i2, is, id, delta, sbase; + if (src1->nelem == 0 || src2->nelem == 0) + return REG_NOERROR; + + /* We need dest->nelem + 2 * elems_in_intersection; this is a + conservative estimate. */ + if (src1->nelem + src2->nelem + dest->nelem > dest->alloc) + { + Idx new_alloc = src1->nelem + src2->nelem + dest->alloc; + Idx *new_elems = re_realloc (dest->elems, Idx, new_alloc); + if (BE (new_elems == NULL, 0)) + return REG_ESPACE; + dest->elems = new_elems; + dest->alloc = new_alloc; + } + + /* Find the items in the intersection of SRC1 and SRC2, and copy + into the top of DEST those that are not already in DEST itself. */ + sbase = dest->nelem + src1->nelem + src2->nelem; + i1 = src1->nelem - 1; + i2 = src2->nelem - 1; + id = dest->nelem - 1; + for (;;) + { + if (src1->elems[i1] == src2->elems[i2]) + { + /* Try to find the item in DEST. Maybe we could binary search? */ + while (REG_VALID_INDEX (id) && dest->elems[id] > src1->elems[i1]) + --id; + + if (! REG_VALID_INDEX (id) || dest->elems[id] != src1->elems[i1]) + dest->elems[--sbase] = src1->elems[i1]; + + if (! REG_VALID_INDEX (--i1) || ! REG_VALID_INDEX (--i2)) + break; + } + + /* Lower the highest of the two items. */ + else if (src1->elems[i1] < src2->elems[i2]) + { + if (! REG_VALID_INDEX (--i2)) + break; + } + else + { + if (! REG_VALID_INDEX (--i1)) + break; + } + } + + id = dest->nelem - 1; + is = dest->nelem + src1->nelem + src2->nelem - 1; + delta = is - sbase + 1; + + /* Now copy. When DELTA becomes zero, the remaining + DEST elements are already in place; this is more or + less the same loop that is in re_node_set_merge. */ + dest->nelem += delta; + if (delta > 0 && REG_VALID_INDEX (id)) + for (;;) + { + if (dest->elems[is] > dest->elems[id]) + { + /* Copy from the top. */ + dest->elems[id + delta--] = dest->elems[is--]; + if (delta == 0) + break; + } + else + { + /* Slide from the bottom. */ + dest->elems[id + delta] = dest->elems[id]; + if (! REG_VALID_INDEX (--id)) + break; + } + } + + /* Copy remaining SRC elements. */ + memcpy (dest->elems, dest->elems + sbase, delta * sizeof (Idx)); + + return REG_NOERROR; +} + +/* Calculate the union set of the sets SRC1 and SRC2. And store it to + DEST. Return value indicate the error code or REG_NOERROR if succeeded. */ + +static reg_errcode_t +internal_function +re_node_set_init_union (re_node_set *dest, const re_node_set *src1, + const re_node_set *src2) +{ + Idx i1, i2, id; + if (src1 != NULL && src1->nelem > 0 && src2 != NULL && src2->nelem > 0) + { + dest->alloc = src1->nelem + src2->nelem; + dest->elems = re_malloc (Idx, dest->alloc); + if (BE (dest->elems == NULL, 0)) + return REG_ESPACE; + } + else + { + if (src1 != NULL && src1->nelem > 0) + return re_node_set_init_copy (dest, src1); + else if (src2 != NULL && src2->nelem > 0) + return re_node_set_init_copy (dest, src2); + else + re_node_set_init_empty (dest); + return REG_NOERROR; + } + for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;) + { + if (src1->elems[i1] > src2->elems[i2]) + { + dest->elems[id++] = src2->elems[i2++]; + continue; + } + if (src1->elems[i1] == src2->elems[i2]) + ++i2; + dest->elems[id++] = src1->elems[i1++]; + } + if (i1 < src1->nelem) + { + memcpy (dest->elems + id, src1->elems + i1, + (src1->nelem - i1) * sizeof (Idx)); + id += src1->nelem - i1; + } + else if (i2 < src2->nelem) + { + memcpy (dest->elems + id, src2->elems + i2, + (src2->nelem - i2) * sizeof (Idx)); + id += src2->nelem - i2; + } + dest->nelem = id; + return REG_NOERROR; +} + +/* Calculate the union set of the sets DEST and SRC. And store it to + DEST. Return value indicate the error code or REG_NOERROR if succeeded. */ + +static reg_errcode_t +internal_function +re_node_set_merge (re_node_set *dest, const re_node_set *src) +{ + Idx is, id, sbase, delta; + if (src == NULL || src->nelem == 0) + return REG_NOERROR; + if (dest->alloc < 2 * src->nelem + dest->nelem) + { + Idx new_alloc = 2 * (src->nelem + dest->alloc); + Idx *new_buffer = re_realloc (dest->elems, Idx, new_alloc); + if (BE (new_buffer == NULL, 0)) + return REG_ESPACE; + dest->elems = new_buffer; + dest->alloc = new_alloc; + } + + if (BE (dest->nelem == 0, 0)) + { + dest->nelem = src->nelem; + memcpy (dest->elems, src->elems, src->nelem * sizeof (Idx)); + return REG_NOERROR; + } + + /* Copy into the top of DEST the items of SRC that are not + found in DEST. Maybe we could binary search in DEST? */ + for (sbase = dest->nelem + 2 * src->nelem, + is = src->nelem - 1, id = dest->nelem - 1; + REG_VALID_INDEX (is) && REG_VALID_INDEX (id); ) + { + if (dest->elems[id] == src->elems[is]) + is--, id--; + else if (dest->elems[id] < src->elems[is]) + dest->elems[--sbase] = src->elems[is--]; + else /* if (dest->elems[id] > src->elems[is]) */ + --id; + } + + if (REG_VALID_INDEX (is)) + { + /* If DEST is exhausted, the remaining items of SRC must be unique. */ + sbase -= is + 1; + memcpy (dest->elems + sbase, src->elems, (is + 1) * sizeof (Idx)); + } + + id = dest->nelem - 1; + is = dest->nelem + 2 * src->nelem - 1; + delta = is - sbase + 1; + if (delta == 0) + return REG_NOERROR; + + /* Now copy. When DELTA becomes zero, the remaining + DEST elements are already in place. */ + dest->nelem += delta; + for (;;) + { + if (dest->elems[is] > dest->elems[id]) + { + /* Copy from the top. */ + dest->elems[id + delta--] = dest->elems[is--]; + if (delta == 0) + break; + } + else + { + /* Slide from the bottom. */ + dest->elems[id + delta] = dest->elems[id]; + if (! REG_VALID_INDEX (--id)) + { + /* Copy remaining SRC elements. */ + memcpy (dest->elems, dest->elems + sbase, + delta * sizeof (Idx)); + break; + } + } + } + + return REG_NOERROR; +} + +/* Insert the new element ELEM to the re_node_set* SET. + SET should not already have ELEM. + Return true if successful. */ + +static bool +internal_function +re_node_set_insert (re_node_set *set, Idx elem) +{ + Idx idx; + /* In case the set is empty. */ + if (set->alloc == 0) + return BE (re_node_set_init_1 (set, elem) == REG_NOERROR, 1); + + if (BE (set->nelem, 0) == 0) + { + /* We already guaranteed above that set->alloc != 0. */ + set->elems[0] = elem; + ++set->nelem; + return true; + } + + /* Realloc if we need. */ + if (set->alloc == set->nelem) + { + Idx *new_elems; + set->alloc = set->alloc * 2; + new_elems = re_realloc (set->elems, Idx, set->alloc); + if (BE (new_elems == NULL, 0)) + return false; + set->elems = new_elems; + } + + /* Move the elements which follows the new element. Test the + first element separately to skip a check in the inner loop. */ + if (elem < set->elems[0]) + { + idx = 0; + for (idx = set->nelem; idx > 0; idx--) + set->elems[idx] = set->elems[idx - 1]; + } + else + { + for (idx = set->nelem; set->elems[idx - 1] > elem; idx--) + set->elems[idx] = set->elems[idx - 1]; + } + + /* Insert the new element. */ + set->elems[idx] = elem; + ++set->nelem; + return true; +} + +/* Insert the new element ELEM to the re_node_set* SET. + SET should not already have any element greater than or equal to ELEM. + Return true if successful. */ + +static bool +internal_function +re_node_set_insert_last (re_node_set *set, Idx elem) +{ + /* Realloc if we need. */ + if (set->alloc == set->nelem) + { + Idx *new_elems; + set->alloc = (set->alloc + 1) * 2; + new_elems = re_realloc (set->elems, Idx, set->alloc); + if (BE (new_elems == NULL, 0)) + return false; + set->elems = new_elems; + } + + /* Insert the new element. */ + set->elems[set->nelem++] = elem; + return true; +} + +/* Compare two node sets SET1 and SET2. + Return true if SET1 and SET2 are equivalent. */ + +static bool +internal_function __attribute ((pure)) +re_node_set_compare (const re_node_set *set1, const re_node_set *set2) +{ + Idx i; + if (set1 == NULL || set2 == NULL || set1->nelem != set2->nelem) + return false; + for (i = set1->nelem ; REG_VALID_INDEX (--i) ; ) + if (set1->elems[i] != set2->elems[i]) + return false; + return true; +} + +/* Return (idx + 1) if SET contains the element ELEM, return 0 otherwise. */ + +static Idx +internal_function __attribute ((pure)) +re_node_set_contains (const re_node_set *set, Idx elem) +{ + __re_size_t idx, right, mid; + if (! REG_VALID_NONZERO_INDEX (set->nelem)) + return 0; + + /* Binary search the element. */ + idx = 0; + right = set->nelem - 1; + while (idx < right) + { + mid = (idx + right) / 2; + if (set->elems[mid] < elem) + idx = mid + 1; + else + right = mid; + } + return set->elems[idx] == elem ? idx + 1 : 0; +} + +static void +internal_function +re_node_set_remove_at (re_node_set *set, Idx idx) +{ + if (idx < 0 || idx >= set->nelem) + return; + --set->nelem; + for (; idx < set->nelem; idx++) + set->elems[idx] = set->elems[idx + 1]; +} + + +/* Add the token TOKEN to dfa->nodes, and return the index of the token. + Or return REG_MISSING if an error occurred. */ + +static Idx +internal_function +re_dfa_add_node (re_dfa_t *dfa, re_token_t token) +{ + if (BE (dfa->nodes_len >= dfa->nodes_alloc, 0)) + { + size_t new_nodes_alloc = dfa->nodes_alloc * 2; + Idx *new_nexts, *new_indices; + re_node_set *new_edests, *new_eclosures; + re_token_t *new_nodes; + size_t max_object_size = + MAX (sizeof (re_token_t), + MAX (sizeof (re_node_set), + sizeof (Idx))); + + /* Avoid overflows. */ + if (BE (SIZE_MAX / 2 / max_object_size < dfa->nodes_alloc, 0)) + return REG_MISSING; + + new_nodes = re_realloc (dfa->nodes, re_token_t, new_nodes_alloc); + if (BE (new_nodes == NULL, 0)) + return REG_MISSING; + dfa->nodes = new_nodes; + new_nexts = re_realloc (dfa->nexts, Idx, new_nodes_alloc); + new_indices = re_realloc (dfa->org_indices, Idx, new_nodes_alloc); + new_edests = re_realloc (dfa->edests, re_node_set, new_nodes_alloc); + new_eclosures = re_realloc (dfa->eclosures, re_node_set, new_nodes_alloc); + if (BE (new_nexts == NULL || new_indices == NULL + || new_edests == NULL || new_eclosures == NULL, 0)) + return REG_MISSING; + dfa->nexts = new_nexts; + dfa->org_indices = new_indices; + dfa->edests = new_edests; + dfa->eclosures = new_eclosures; + dfa->nodes_alloc = new_nodes_alloc; + } + dfa->nodes[dfa->nodes_len] = token; + dfa->nodes[dfa->nodes_len].constraint = 0; +#ifdef RE_ENABLE_I18N + { + int type = token.type; + dfa->nodes[dfa->nodes_len].accept_mb = + (type == OP_PERIOD && dfa->mb_cur_max > 1) || type == COMPLEX_BRACKET; + } +#endif + dfa->nexts[dfa->nodes_len] = REG_MISSING; + re_node_set_init_empty (dfa->edests + dfa->nodes_len); + re_node_set_init_empty (dfa->eclosures + dfa->nodes_len); + return dfa->nodes_len++; +} + +static inline re_hashval_t +internal_function +calc_state_hash (const re_node_set *nodes, unsigned int context) +{ + re_hashval_t hash = nodes->nelem + context; + Idx i; + for (i = 0 ; i < nodes->nelem ; i++) + hash += nodes->elems[i]; + return hash; +} + +/* Search for the state whose node_set is equivalent to NODES. + Return the pointer to the state, if we found it in the DFA. + Otherwise create the new one and return it. In case of an error + return NULL and set the error code in ERR. + Note: - We assume NULL as the invalid state, then it is possible that + return value is NULL and ERR is REG_NOERROR. + - We never return non-NULL value in case of any errors, it is for + optimization. */ + +static re_dfastate_t * +internal_function +re_acquire_state (reg_errcode_t *err, const re_dfa_t *dfa, + const re_node_set *nodes) +{ + re_hashval_t hash; + re_dfastate_t *new_state; + struct re_state_table_entry *spot; + Idx i; +#ifdef lint + /* Suppress bogus uninitialized-variable warnings. */ + *err = REG_NOERROR; +#endif + if (BE (nodes->nelem == 0, 0)) + { + *err = REG_NOERROR; + return NULL; + } + hash = calc_state_hash (nodes, 0); + spot = dfa->state_table + (hash & dfa->state_hash_mask); + + for (i = 0 ; i < spot->num ; i++) + { + re_dfastate_t *state = spot->array[i]; + if (hash != state->hash) + continue; + if (re_node_set_compare (&state->nodes, nodes)) + return state; + } + + /* There are no appropriate state in the dfa, create the new one. */ + new_state = create_ci_newstate (dfa, nodes, hash); + if (BE (new_state == NULL, 0)) + *err = REG_ESPACE; + + return new_state; +} + +/* Search for the state whose node_set is equivalent to NODES and + whose context is equivalent to CONTEXT. + Return the pointer to the state, if we found it in the DFA. + Otherwise create the new one and return it. In case of an error + return NULL and set the error code in ERR. + Note: - We assume NULL as the invalid state, then it is possible that + return value is NULL and ERR is REG_NOERROR. + - We never return non-NULL value in case of any errors, it is for + optimization. */ + +static re_dfastate_t * +internal_function +re_acquire_state_context (reg_errcode_t *err, const re_dfa_t *dfa, + const re_node_set *nodes, unsigned int context) +{ + re_hashval_t hash; + re_dfastate_t *new_state; + struct re_state_table_entry *spot; + Idx i; +#ifdef lint + /* Suppress bogus uninitialized-variable warnings. */ + *err = REG_NOERROR; +#endif + if (nodes->nelem == 0) + { + *err = REG_NOERROR; + return NULL; + } + hash = calc_state_hash (nodes, context); + spot = dfa->state_table + (hash & dfa->state_hash_mask); + + for (i = 0 ; i < spot->num ; i++) + { + re_dfastate_t *state = spot->array[i]; + if (state->hash == hash + && state->context == context + && re_node_set_compare (state->entrance_nodes, nodes)) + return state; + } + /* There are no appropriate state in `dfa', create the new one. */ + new_state = create_cd_newstate (dfa, nodes, context, hash); + if (BE (new_state == NULL, 0)) + *err = REG_ESPACE; + + return new_state; +} + +/* Finish initialization of the new state NEWSTATE, and using its hash value + HASH put in the appropriate bucket of DFA's state table. Return value + indicates the error code if failed. */ + +static reg_errcode_t +register_state (const re_dfa_t *dfa, re_dfastate_t *newstate, + re_hashval_t hash) +{ + struct re_state_table_entry *spot; + reg_errcode_t err; + Idx i; + + newstate->hash = hash; + err = re_node_set_alloc (&newstate->non_eps_nodes, newstate->nodes.nelem); + if (BE (err != REG_NOERROR, 0)) + return REG_ESPACE; + for (i = 0; i < newstate->nodes.nelem; i++) + { + Idx elem = newstate->nodes.elems[i]; + if (!IS_EPSILON_NODE (dfa->nodes[elem].type)) + if (BE (! re_node_set_insert_last (&newstate->non_eps_nodes, elem), 0)) + return REG_ESPACE; + } + + spot = dfa->state_table + (hash & dfa->state_hash_mask); + if (BE (spot->alloc <= spot->num, 0)) + { + Idx new_alloc = 2 * spot->num + 2; + re_dfastate_t **new_array = re_realloc (spot->array, re_dfastate_t *, + new_alloc); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + spot->array = new_array; + spot->alloc = new_alloc; + } + spot->array[spot->num++] = newstate; + return REG_NOERROR; +} + +static void +free_state (re_dfastate_t *state) +{ + re_node_set_free (&state->non_eps_nodes); + re_node_set_free (&state->inveclosure); + if (state->entrance_nodes != &state->nodes) + { + re_node_set_free (state->entrance_nodes); + re_free (state->entrance_nodes); + } + re_node_set_free (&state->nodes); + re_free (state->word_trtable); + re_free (state->trtable); + re_free (state); +} + +/* Create the new state which is independ of contexts. + Return the new state if succeeded, otherwise return NULL. */ + +static re_dfastate_t * +internal_function +create_ci_newstate (const re_dfa_t *dfa, const re_node_set *nodes, + re_hashval_t hash) +{ + Idx i; + reg_errcode_t err; + re_dfastate_t *newstate; + + newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); + if (BE (newstate == NULL, 0)) + return NULL; + err = re_node_set_init_copy (&newstate->nodes, nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_free (newstate); + return NULL; + } + + newstate->entrance_nodes = &newstate->nodes; + for (i = 0 ; i < nodes->nelem ; i++) + { + re_token_t *node = dfa->nodes + nodes->elems[i]; + re_token_type_t type = node->type; + if (type == CHARACTER && !node->constraint) + continue; +#ifdef RE_ENABLE_I18N + newstate->accept_mb |= node->accept_mb; +#endif /* RE_ENABLE_I18N */ + + /* If the state has the halt node, the state is a halt state. */ + if (type == END_OF_RE) + newstate->halt = 1; + else if (type == OP_BACK_REF) + newstate->has_backref = 1; + else if (type == ANCHOR || node->constraint) + newstate->has_constraint = 1; + } + err = register_state (dfa, newstate, hash); + if (BE (err != REG_NOERROR, 0)) + { + free_state (newstate); + newstate = NULL; + } + return newstate; +} + +/* Create the new state which is depend on the context CONTEXT. + Return the new state if succeeded, otherwise return NULL. */ + +static re_dfastate_t * +internal_function +create_cd_newstate (const re_dfa_t *dfa, const re_node_set *nodes, + unsigned int context, re_hashval_t hash) +{ + Idx i, nctx_nodes = 0; + reg_errcode_t err; + re_dfastate_t *newstate; + + newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); + if (BE (newstate == NULL, 0)) + return NULL; + err = re_node_set_init_copy (&newstate->nodes, nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_free (newstate); + return NULL; + } + + newstate->context = context; + newstate->entrance_nodes = &newstate->nodes; + + for (i = 0 ; i < nodes->nelem ; i++) + { + unsigned int constraint = 0; + re_token_t *node = dfa->nodes + nodes->elems[i]; + re_token_type_t type = node->type; + if (node->constraint) + constraint = node->constraint; + + if (type == CHARACTER && !constraint) + continue; +#ifdef RE_ENABLE_I18N + newstate->accept_mb |= node->accept_mb; +#endif /* RE_ENABLE_I18N */ + + /* If the state has the halt node, the state is a halt state. */ + if (type == END_OF_RE) + newstate->halt = 1; + else if (type == OP_BACK_REF) + newstate->has_backref = 1; + else if (type == ANCHOR) + constraint = node->opr.ctx_type; + + if (constraint) + { + if (newstate->entrance_nodes == &newstate->nodes) + { + newstate->entrance_nodes = re_malloc (re_node_set, 1); + if (BE (newstate->entrance_nodes == NULL, 0)) + { + free_state (newstate); + return NULL; + } + re_node_set_init_copy (newstate->entrance_nodes, nodes); + nctx_nodes = 0; + newstate->has_constraint = 1; + } + + if (NOT_SATISFY_PREV_CONSTRAINT (constraint,context)) + { + re_node_set_remove_at (&newstate->nodes, i - nctx_nodes); + ++nctx_nodes; + } + } + } + err = register_state (dfa, newstate, hash); + if (BE (err != REG_NOERROR, 0)) + { + free_state (newstate); + newstate = NULL; + } + return newstate; +} diff --git a/lib/regex_internal.h b/lib/regex_internal.h new file mode 100644 index 0000000..9bbc6ac --- /dev/null +++ b/lib/regex_internal.h @@ -0,0 +1,857 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _REGEX_INTERNAL_H +#define _REGEX_INTERNAL_H 1 + +#include <assert.h> +#include <ctype.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#ifdef _LIBC +# include <langinfo.h> +#else +# include "localcharset.h" +#endif +#if defined HAVE_LOCALE_H || defined _LIBC +# include <locale.h> +#endif + +#include <wchar.h> +#include <wctype.h> +#include <stdint.h> +#if defined _LIBC +# include <bits/libc-lock.h> +#else +# define __libc_lock_init(NAME) do { } while (0) +# define __libc_lock_lock(NAME) do { } while (0) +# define __libc_lock_unlock(NAME) do { } while (0) +#endif + +/* In case that the system doesn't have isblank(). */ +#if !defined _LIBC && !HAVE_DECL_ISBLANK && !defined isblank +# define isblank(ch) ((ch) == ' ' || (ch) == '\t') +#endif + +#ifdef _LIBC +# ifndef _RE_DEFINE_LOCALE_FUNCTIONS +# define _RE_DEFINE_LOCALE_FUNCTIONS 1 +# include <locale/localeinfo.h> +# include <locale/elem-hash.h> +# include <locale/coll-lookup.h> +# endif +#endif + +/* This is for other GNU distributions with internationalized messages. */ +#if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC +# include <libintl.h> +# ifdef _LIBC +# undef gettext +# define gettext(msgid) \ + INTUSE(__dcgettext) (_libc_intl_domainname, msgid, LC_MESSAGES) +# endif +#else +# define gettext(msgid) (msgid) +#endif + +#ifndef gettext_noop +/* This define is so xgettext can find the internationalizable + strings. */ +# define gettext_noop(String) String +#endif + +/* For loser systems without the definition. */ +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +#if (defined MB_CUR_MAX && HAVE_LOCALE_H && HAVE_WCTYPE_H && HAVE_ISWCTYPE && HAVE_WCRTOMB && HAVE_MBRTOWC && HAVE_WCSCOLL) || _LIBC +# define RE_ENABLE_I18N +#endif + +#if __GNUC__ >= 3 +# define BE(expr, val) __builtin_expect (expr, val) +#else +# define BE(expr, val) (expr) +# ifdef _LIBC +# define inline +# endif +#endif + +/* Number of ASCII characters. */ +#define ASCII_CHARS 0x80 + +/* Number of single byte characters. */ +#define SBC_MAX (UCHAR_MAX + 1) + +#define COLL_ELEM_LEN_MAX 8 + +/* The character which represents newline. */ +#define NEWLINE_CHAR '\n' +#define WIDE_NEWLINE_CHAR L'\n' + +/* Rename to standard API for using out of glibc. */ +#ifndef _LIBC +# define __wctype wctype +# define __iswctype iswctype +# define __btowc btowc +# define __wcrtomb wcrtomb +# define __regfree regfree +# define attribute_hidden +#endif /* not _LIBC */ + +#if __GNUC__ >= 4 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1) +# define __attribute(arg) __attribute__ (arg) +#else +# define __attribute(arg) +#endif + +typedef __re_idx_t Idx; + +/* Special return value for failure to match. */ +#define REG_MISSING ((Idx) -1) + +/* Special return value for internal error. */ +#define REG_ERROR ((Idx) -2) + +/* Test whether N is a valid index, and is not one of the above. */ +#ifdef _REGEX_LARGE_OFFSETS +# define REG_VALID_INDEX(n) ((Idx) (n) < REG_ERROR) +#else +# define REG_VALID_INDEX(n) (0 <= (n)) +#endif + +/* Test whether N is a valid nonzero index. */ +#ifdef _REGEX_LARGE_OFFSETS +# define REG_VALID_NONZERO_INDEX(n) ((Idx) ((n) - 1) < (Idx) (REG_ERROR - 1)) +#else +# define REG_VALID_NONZERO_INDEX(n) (0 < (n)) +#endif + +/* A hash value, suitable for computing hash tables. */ +typedef __re_size_t re_hashval_t; + +/* An integer used to represent a set of bits. It must be unsigned, + and must be at least as wide as unsigned int. */ +typedef unsigned long int bitset_word_t; +/* All bits set in a bitset_word_t. */ +#define BITSET_WORD_MAX ULONG_MAX + +/* Number of bits in a bitset_word_t. For portability to hosts with + padding bits, do not use '(sizeof (bitset_word_t) * CHAR_BIT)'; + instead, deduce it directly from BITSET_WORD_MAX. Avoid + greater-than-32-bit integers and unconditional shifts by more than + 31 bits, as they're not portable. */ +#if BITSET_WORD_MAX == 0xffffffff +# define BITSET_WORD_BITS 32 +#elif BITSET_WORD_MAX >> 31 >> 5 == 1 +# define BITSET_WORD_BITS 36 +#elif BITSET_WORD_MAX >> 31 >> 16 == 1 +# define BITSET_WORD_BITS 48 +#elif BITSET_WORD_MAX >> 31 >> 28 == 1 +# define BITSET_WORD_BITS 60 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 1 == 1 +# define BITSET_WORD_BITS 64 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 9 == 1 +# define BITSET_WORD_BITS 72 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 3 == 1 +# define BITSET_WORD_BITS 128 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 == 1 +# define BITSET_WORD_BITS 256 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 > 1 +# define BITSET_WORD_BITS 257 /* any value > SBC_MAX will do here */ +# if BITSET_WORD_BITS <= SBC_MAX +# error "Invalid SBC_MAX" +# endif +#elif BITSET_WORD_MAX == (0xffffffff + 2) * 0xffffffff +/* Work around a bug in 64-bit PGC (before version 6.1-2), where the + preprocessor mishandles large unsigned values as if they were signed. */ +# define BITSET_WORD_BITS 64 +#else +# error "Add case for new bitset_word_t size" +#endif + +/* Number of bitset_word_t values in a bitset_t. */ +#define BITSET_WORDS ((SBC_MAX + BITSET_WORD_BITS - 1) / BITSET_WORD_BITS) + +typedef bitset_word_t bitset_t[BITSET_WORDS]; +typedef bitset_word_t *re_bitset_ptr_t; +typedef const bitset_word_t *re_const_bitset_ptr_t; + +#define PREV_WORD_CONSTRAINT 0x0001 +#define PREV_NOTWORD_CONSTRAINT 0x0002 +#define NEXT_WORD_CONSTRAINT 0x0004 +#define NEXT_NOTWORD_CONSTRAINT 0x0008 +#define PREV_NEWLINE_CONSTRAINT 0x0010 +#define NEXT_NEWLINE_CONSTRAINT 0x0020 +#define PREV_BEGBUF_CONSTRAINT 0x0040 +#define NEXT_ENDBUF_CONSTRAINT 0x0080 +#define WORD_DELIM_CONSTRAINT 0x0100 +#define NOT_WORD_DELIM_CONSTRAINT 0x0200 + +typedef enum +{ + INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT, + WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT, + WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT, + INSIDE_NOTWORD = PREV_NOTWORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT, + LINE_FIRST = PREV_NEWLINE_CONSTRAINT, + LINE_LAST = NEXT_NEWLINE_CONSTRAINT, + BUF_FIRST = PREV_BEGBUF_CONSTRAINT, + BUF_LAST = NEXT_ENDBUF_CONSTRAINT, + WORD_DELIM = WORD_DELIM_CONSTRAINT, + NOT_WORD_DELIM = NOT_WORD_DELIM_CONSTRAINT +} re_context_type; + +typedef struct +{ + Idx alloc; + Idx nelem; + Idx *elems; +} re_node_set; + +typedef enum +{ + NON_TYPE = 0, + + /* Node type, These are used by token, node, tree. */ + CHARACTER = 1, + END_OF_RE = 2, + SIMPLE_BRACKET = 3, + OP_BACK_REF = 4, + OP_PERIOD = 5, +#ifdef RE_ENABLE_I18N + COMPLEX_BRACKET = 6, + OP_UTF8_PERIOD = 7, +#endif /* RE_ENABLE_I18N */ + + /* We define EPSILON_BIT as a macro so that OP_OPEN_SUBEXP is used + when the debugger shows values of this enum type. */ +#define EPSILON_BIT 8 + OP_OPEN_SUBEXP = EPSILON_BIT | 0, + OP_CLOSE_SUBEXP = EPSILON_BIT | 1, + OP_ALT = EPSILON_BIT | 2, + OP_DUP_ASTERISK = EPSILON_BIT | 3, + ANCHOR = EPSILON_BIT | 4, + + /* Tree type, these are used only by tree. */ + CONCAT = 16, + SUBEXP = 17, + + /* Token type, these are used only by token. */ + OP_DUP_PLUS = 18, + OP_DUP_QUESTION, + OP_OPEN_BRACKET, + OP_CLOSE_BRACKET, + OP_CHARSET_RANGE, + OP_OPEN_DUP_NUM, + OP_CLOSE_DUP_NUM, + OP_NON_MATCH_LIST, + OP_OPEN_COLL_ELEM, + OP_CLOSE_COLL_ELEM, + OP_OPEN_EQUIV_CLASS, + OP_CLOSE_EQUIV_CLASS, + OP_OPEN_CHAR_CLASS, + OP_CLOSE_CHAR_CLASS, + OP_WORD, + OP_NOTWORD, + OP_SPACE, + OP_NOTSPACE, + BACK_SLASH + +} re_token_type_t; + +#ifdef RE_ENABLE_I18N +typedef struct +{ + /* Multibyte characters. */ + wchar_t *mbchars; + + /* Collating symbols. */ +# ifdef _LIBC + int32_t *coll_syms; +# endif + + /* Equivalence classes. */ +# ifdef _LIBC + int32_t *equiv_classes; +# endif + + /* Range expressions. */ +# ifdef _LIBC + uint32_t *range_starts; + uint32_t *range_ends; +# else /* not _LIBC */ + wchar_t *range_starts; + wchar_t *range_ends; +# endif /* not _LIBC */ + + /* Character classes. */ + wctype_t *char_classes; + + /* If this character set is the non-matching list. */ + unsigned int non_match : 1; + + /* # of multibyte characters. */ + Idx nmbchars; + + /* # of collating symbols. */ + Idx ncoll_syms; + + /* # of equivalence classes. */ + Idx nequiv_classes; + + /* # of range expressions. */ + Idx nranges; + + /* # of character classes. */ + Idx nchar_classes; +} re_charset_t; +#endif /* RE_ENABLE_I18N */ + +typedef struct +{ + union + { + unsigned char c; /* for CHARACTER */ + re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */ +#ifdef RE_ENABLE_I18N + re_charset_t *mbcset; /* for COMPLEX_BRACKET */ +#endif /* RE_ENABLE_I18N */ + Idx idx; /* for BACK_REF */ + re_context_type ctx_type; /* for ANCHOR */ + } opr; +#if __GNUC__ >= 2 && !__STRICT_ANSI__ + re_token_type_t type : 8; +#else + re_token_type_t type; +#endif + unsigned int constraint : 10; /* context constraint */ + unsigned int duplicated : 1; + unsigned int opt_subexp : 1; +#ifdef RE_ENABLE_I18N + unsigned int accept_mb : 1; + /* These 2 bits can be moved into the union if needed (e.g. if running out + of bits; move opr.c to opr.c.c and move the flags to opr.c.flags). */ + unsigned int mb_partial : 1; +#endif + unsigned int word_char : 1; +} re_token_t; + +#define IS_EPSILON_NODE(type) ((type) & EPSILON_BIT) + +struct re_string_t +{ + /* Indicate the raw buffer which is the original string passed as an + argument of regexec(), re_search(), etc.. */ + const unsigned char *raw_mbs; + /* Store the multibyte string. In case of "case insensitive mode" like + REG_ICASE, upper cases of the string are stored, otherwise MBS points + the same address that RAW_MBS points. */ + unsigned char *mbs; +#ifdef RE_ENABLE_I18N + /* Store the wide character string which is corresponding to MBS. */ + wint_t *wcs; + Idx *offsets; + mbstate_t cur_state; +#endif + /* Index in RAW_MBS. Each character mbs[i] corresponds to + raw_mbs[raw_mbs_idx + i]. */ + Idx raw_mbs_idx; + /* The length of the valid characters in the buffers. */ + Idx valid_len; + /* The corresponding number of bytes in raw_mbs array. */ + Idx valid_raw_len; + /* The length of the buffers MBS and WCS. */ + Idx bufs_len; + /* The index in MBS, which is updated by re_string_fetch_byte. */ + Idx cur_idx; + /* length of RAW_MBS array. */ + Idx raw_len; + /* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */ + Idx len; + /* End of the buffer may be shorter than its length in the cases such + as re_match_2, re_search_2. Then, we use STOP for end of the buffer + instead of LEN. */ + Idx raw_stop; + /* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */ + Idx stop; + + /* The context of mbs[0]. We store the context independently, since + the context of mbs[0] may be different from raw_mbs[0], which is + the beginning of the input string. */ + unsigned int tip_context; + /* The translation passed as a part of an argument of re_compile_pattern. */ + RE_TRANSLATE_TYPE trans; + /* Copy of re_dfa_t's word_char. */ + re_const_bitset_ptr_t word_char; + /* true if REG_ICASE. */ + unsigned char icase; + unsigned char is_utf8; + unsigned char map_notascii; + unsigned char mbs_allocated; + unsigned char offsets_needed; + unsigned char newline_anchor; + unsigned char word_ops_used; + int mb_cur_max; +}; +typedef struct re_string_t re_string_t; + + +struct re_dfa_t; +typedef struct re_dfa_t re_dfa_t; + +#ifndef _LIBC +# ifdef __i386__ +# define internal_function __attribute ((regparm (3), stdcall)) +# else +# define internal_function +# endif +#endif + +static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr, + Idx new_buf_len) + internal_function; +#ifdef RE_ENABLE_I18N +static void build_wcs_buffer (re_string_t *pstr) internal_function; +static reg_errcode_t build_wcs_upper_buffer (re_string_t *pstr) + internal_function; +#endif /* RE_ENABLE_I18N */ +static void build_upper_buffer (re_string_t *pstr) internal_function; +static void re_string_translate_buffer (re_string_t *pstr) internal_function; +static unsigned int re_string_context_at (const re_string_t *input, Idx idx, + int eflags) + internal_function __attribute ((pure)); +#define re_string_peek_byte(pstr, offset) \ + ((pstr)->mbs[(pstr)->cur_idx + offset]) +#define re_string_fetch_byte(pstr) \ + ((pstr)->mbs[(pstr)->cur_idx++]) +#define re_string_first_byte(pstr, idx) \ + ((idx) == (pstr)->valid_len || (pstr)->wcs[idx] != WEOF) +#define re_string_is_single_byte_char(pstr, idx) \ + ((pstr)->wcs[idx] != WEOF && ((pstr)->valid_len == (idx) + 1 \ + || (pstr)->wcs[(idx) + 1] != WEOF)) +#define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx) +#define re_string_cur_idx(pstr) ((pstr)->cur_idx) +#define re_string_get_buffer(pstr) ((pstr)->mbs) +#define re_string_length(pstr) ((pstr)->len) +#define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx]) +#define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx)) +#define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx)) + +#include <alloca.h> + +#ifndef _LIBC +# if HAVE_ALLOCA +/* The OS usually guarantees only one guard page at the bottom of the stack, + and a page size can be as small as 4096 bytes. So we cannot safely + allocate anything larger than 4096 bytes. Also care for the possibility + of a few compiler-allocated temporary stack slots. */ +# define __libc_use_alloca(n) ((n) < 4032) +# else +/* alloca is implemented with malloc, so just use malloc. */ +# define __libc_use_alloca(n) 0 +# endif +#endif + +#ifndef MAX +# define MAX(a,b) ((a) < (b) ? (b) : (a)) +#endif + +#define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t))) +#define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t))) +#define re_free(p) free (p) + +struct bin_tree_t +{ + struct bin_tree_t *parent; + struct bin_tree_t *left; + struct bin_tree_t *right; + struct bin_tree_t *first; + struct bin_tree_t *next; + + re_token_t token; + + /* `node_idx' is the index in dfa->nodes, if `type' == 0. + Otherwise `type' indicate the type of this node. */ + Idx node_idx; +}; +typedef struct bin_tree_t bin_tree_t; + +#define BIN_TREE_STORAGE_SIZE \ + ((1024 - sizeof (void *)) / sizeof (bin_tree_t)) + +struct bin_tree_storage_t +{ + struct bin_tree_storage_t *next; + bin_tree_t data[BIN_TREE_STORAGE_SIZE]; +}; +typedef struct bin_tree_storage_t bin_tree_storage_t; + +#define CONTEXT_WORD 1 +#define CONTEXT_NEWLINE (CONTEXT_WORD << 1) +#define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1) +#define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1) + +#define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD) +#define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE) +#define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF) +#define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF) +#define IS_ORDINARY_CONTEXT(c) ((c) == 0) + +#define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_') +#define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR) +#define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_') +#define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR) + +#define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \ + ((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \ + || ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \ + || ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\ + || ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context))) + +#define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \ + ((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \ + || (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \ + || (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \ + || (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context))) + +struct re_dfastate_t +{ + re_hashval_t hash; + re_node_set nodes; + re_node_set non_eps_nodes; + re_node_set inveclosure; + re_node_set *entrance_nodes; + struct re_dfastate_t **trtable, **word_trtable; + unsigned int context : 4; + unsigned int halt : 1; + /* If this state can accept `multi byte'. + Note that we refer to multibyte characters, and multi character + collating elements as `multi byte'. */ + unsigned int accept_mb : 1; + /* If this state has backreference node(s). */ + unsigned int has_backref : 1; + unsigned int has_constraint : 1; +}; +typedef struct re_dfastate_t re_dfastate_t; + +struct re_state_table_entry +{ + Idx num; + Idx alloc; + re_dfastate_t **array; +}; + +/* Array type used in re_sub_match_last_t and re_sub_match_top_t. */ + +typedef struct +{ + Idx next_idx; + Idx alloc; + re_dfastate_t **array; +} state_array_t; + +/* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */ + +typedef struct +{ + Idx node; + Idx str_idx; /* The position NODE match at. */ + state_array_t path; +} re_sub_match_last_t; + +/* Store information about the node NODE whose type is OP_OPEN_SUBEXP. + And information about the node, whose type is OP_CLOSE_SUBEXP, + corresponding to NODE is stored in LASTS. */ + +typedef struct +{ + Idx str_idx; + Idx node; + state_array_t *path; + Idx alasts; /* Allocation size of LASTS. */ + Idx nlasts; /* The number of LASTS. */ + re_sub_match_last_t **lasts; +} re_sub_match_top_t; + +struct re_backref_cache_entry +{ + Idx node; + Idx str_idx; + Idx subexp_from; + Idx subexp_to; + char more; + char unused; + unsigned short int eps_reachable_subexps_map; +}; + +typedef struct +{ + /* The string object corresponding to the input string. */ + re_string_t input; +#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L) + const re_dfa_t *const dfa; +#else + const re_dfa_t *dfa; +#endif + /* EFLAGS of the argument of regexec. */ + int eflags; + /* Where the matching ends. */ + Idx match_last; + Idx last_node; + /* The state log used by the matcher. */ + re_dfastate_t **state_log; + Idx state_log_top; + /* Back reference cache. */ + Idx nbkref_ents; + Idx abkref_ents; + struct re_backref_cache_entry *bkref_ents; + int max_mb_elem_len; + Idx nsub_tops; + Idx asub_tops; + re_sub_match_top_t **sub_tops; +} re_match_context_t; + +typedef struct +{ + re_dfastate_t **sifted_states; + re_dfastate_t **limited_states; + Idx last_node; + Idx last_str_idx; + re_node_set limits; +} re_sift_context_t; + +struct re_fail_stack_ent_t +{ + Idx idx; + Idx node; + regmatch_t *regs; + re_node_set eps_via_nodes; +}; + +struct re_fail_stack_t +{ + Idx num; + Idx alloc; + struct re_fail_stack_ent_t *stack; +}; + +struct re_dfa_t +{ + re_token_t *nodes; + size_t nodes_alloc; + size_t nodes_len; + Idx *nexts; + Idx *org_indices; + re_node_set *edests; + re_node_set *eclosures; + re_node_set *inveclosures; + struct re_state_table_entry *state_table; + re_dfastate_t *init_state; + re_dfastate_t *init_state_word; + re_dfastate_t *init_state_nl; + re_dfastate_t *init_state_begbuf; + bin_tree_t *str_tree; + bin_tree_storage_t *str_tree_storage; + re_bitset_ptr_t sb_char; + int str_tree_storage_idx; + + /* number of subexpressions `re_nsub' is in regex_t. */ + re_hashval_t state_hash_mask; + Idx init_node; + Idx nbackref; /* The number of backreference in this dfa. */ + + /* Bitmap expressing which backreference is used. */ + bitset_word_t used_bkref_map; + bitset_word_t completed_bkref_map; + + unsigned int has_plural_match : 1; + /* If this dfa has "multibyte node", which is a backreference or + a node which can accept multibyte character or multi character + collating element. */ + unsigned int has_mb_node : 1; + unsigned int is_utf8 : 1; + unsigned int map_notascii : 1; + unsigned int word_ops_used : 1; + int mb_cur_max; + bitset_t word_char; + reg_syntax_t syntax; + Idx *subexp_map; +#ifdef DEBUG + char* re_str; +#endif +#ifdef _LIBC + __libc_lock_define (, lock) +#endif +}; + +#define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set)) +#define re_node_set_remove(set,id) \ + (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1)) +#define re_node_set_empty(p) ((p)->nelem = 0) +#define re_node_set_free(set) re_free ((set)->elems) + + +typedef enum +{ + SB_CHAR, + MB_CHAR, + EQUIV_CLASS, + COLL_SYM, + CHAR_CLASS +} bracket_elem_type; + +typedef struct +{ + bracket_elem_type type; + union + { + unsigned char ch; + unsigned char *name; + wchar_t wch; + } opr; +} bracket_elem_t; + + +/* Inline functions for bitset_t operation. */ + +static inline void +bitset_set (bitset_t set, Idx i) +{ + set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS; +} + +static inline void +bitset_clear (bitset_t set, Idx i) +{ + set[i / BITSET_WORD_BITS] &= ~ ((bitset_word_t) 1 << i % BITSET_WORD_BITS); +} + +static inline bool +bitset_contain (const bitset_t set, Idx i) +{ + return (set[i / BITSET_WORD_BITS] >> i % BITSET_WORD_BITS) & 1; +} + +static inline void +bitset_empty (bitset_t set) +{ + memset (set, '\0', sizeof (bitset_t)); +} + +static inline void +bitset_set_all (bitset_t set) +{ + memset (set, -1, sizeof (bitset_word_t) * (SBC_MAX / BITSET_WORD_BITS)); + if (SBC_MAX % BITSET_WORD_BITS != 0) + set[BITSET_WORDS - 1] = + ((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1; +} + +static inline void +bitset_copy (bitset_t dest, const bitset_t src) +{ + memcpy (dest, src, sizeof (bitset_t)); +} + +static inline void +bitset_not (bitset_t set) +{ + int bitset_i; + for (bitset_i = 0; bitset_i < SBC_MAX / BITSET_WORD_BITS; ++bitset_i) + set[bitset_i] = ~set[bitset_i]; + if (SBC_MAX % BITSET_WORD_BITS != 0) + set[BITSET_WORDS - 1] = + ((((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1) + & ~set[BITSET_WORDS - 1]); +} + +static inline void +bitset_merge (bitset_t dest, const bitset_t src) +{ + int bitset_i; + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) + dest[bitset_i] |= src[bitset_i]; +} + +static inline void +bitset_mask (bitset_t dest, const bitset_t src) +{ + int bitset_i; + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) + dest[bitset_i] &= src[bitset_i]; +} + +#ifdef RE_ENABLE_I18N +/* Inline functions for re_string. */ +static inline int +internal_function __attribute ((pure)) +re_string_char_size_at (const re_string_t *pstr, Idx idx) +{ + int byte_idx; + if (pstr->mb_cur_max == 1) + return 1; + for (byte_idx = 1; idx + byte_idx < pstr->valid_len; ++byte_idx) + if (pstr->wcs[idx + byte_idx] != WEOF) + break; + return byte_idx; +} + +static inline wint_t +internal_function __attribute ((pure)) +re_string_wchar_at (const re_string_t *pstr, Idx idx) +{ + if (pstr->mb_cur_max == 1) + return (wint_t) pstr->mbs[idx]; + return (wint_t) pstr->wcs[idx]; +} + +static int +internal_function __attribute ((pure)) +re_string_elem_size_at (const re_string_t *pstr, Idx idx) +{ +# ifdef _LIBC + const unsigned char *p, *extra; + const int32_t *table, *indirect; + int32_t tmp; +# include <locale/weight.h> + uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + + if (nrules != 0) + { + table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_INDIRECTMB); + p = pstr->mbs + idx; + tmp = findidx (&p); + return p - pstr->mbs - idx; + } + else +# endif /* _LIBC */ + return 1; +} +#endif /* RE_ENABLE_I18N */ + +#endif /* _REGEX_INTERNAL_H */ diff --git a/lib/regexec.c b/lib/regexec.c new file mode 100644 index 0000000..b136570 --- /dev/null +++ b/lib/regexec.c @@ -0,0 +1,4399 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007 Free Software Foundation, + Inc. + This file is part of the GNU C Library. + Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +static reg_errcode_t match_ctx_init (re_match_context_t *cache, int eflags, + Idx n) internal_function; +static void match_ctx_clean (re_match_context_t *mctx) internal_function; +static void match_ctx_free (re_match_context_t *cache) internal_function; +static reg_errcode_t match_ctx_add_entry (re_match_context_t *cache, Idx node, + Idx str_idx, Idx from, Idx to) + internal_function; +static Idx search_cur_bkref_entry (const re_match_context_t *mctx, Idx str_idx) + internal_function; +static reg_errcode_t match_ctx_add_subtop (re_match_context_t *mctx, Idx node, + Idx str_idx) internal_function; +static re_sub_match_last_t * match_ctx_add_sublast (re_sub_match_top_t *subtop, + Idx node, Idx str_idx) + internal_function; +static void sift_ctx_init (re_sift_context_t *sctx, re_dfastate_t **sifted_sts, + re_dfastate_t **limited_sts, Idx last_node, + Idx last_str_idx) + internal_function; +static reg_errcode_t re_search_internal (const regex_t *preg, + const char *string, Idx length, + Idx start, Idx last_start, Idx stop, + size_t nmatch, regmatch_t pmatch[], + int eflags) internal_function; +static regoff_t re_search_2_stub (struct re_pattern_buffer *bufp, + const char *string1, Idx length1, + const char *string2, Idx length2, + Idx start, regoff_t range, + struct re_registers *regs, + Idx stop, bool ret_len) internal_function; +static regoff_t re_search_stub (struct re_pattern_buffer *bufp, + const char *string, Idx length, Idx start, + regoff_t range, Idx stop, + struct re_registers *regs, + bool ret_len) internal_function; +static unsigned int re_copy_regs (struct re_registers *regs, regmatch_t *pmatch, + Idx nregs, int regs_allocated) + internal_function; +static reg_errcode_t prune_impossible_nodes (re_match_context_t *mctx) + internal_function; +static Idx check_matching (re_match_context_t *mctx, bool fl_longest_match, + Idx *p_match_first) internal_function; +static Idx check_halt_state_context (const re_match_context_t *mctx, + const re_dfastate_t *state, Idx idx) + internal_function; +static void update_regs (const re_dfa_t *dfa, regmatch_t *pmatch, + regmatch_t *prev_idx_match, Idx cur_node, + Idx cur_idx, Idx nmatch) internal_function; +static reg_errcode_t push_fail_stack (struct re_fail_stack_t *fs, + Idx str_idx, Idx dest_node, Idx nregs, + regmatch_t *regs, + re_node_set *eps_via_nodes) + internal_function; +static reg_errcode_t set_regs (const regex_t *preg, + const re_match_context_t *mctx, + size_t nmatch, regmatch_t *pmatch, + bool fl_backtrack) internal_function; +static reg_errcode_t free_fail_stack_return (struct re_fail_stack_t *fs) + internal_function; + +#ifdef RE_ENABLE_I18N +static int sift_states_iter_mb (const re_match_context_t *mctx, + re_sift_context_t *sctx, + Idx node_idx, Idx str_idx, Idx max_str_idx) + internal_function; +#endif /* RE_ENABLE_I18N */ +static reg_errcode_t sift_states_backward (const re_match_context_t *mctx, + re_sift_context_t *sctx) + internal_function; +static reg_errcode_t build_sifted_states (const re_match_context_t *mctx, + re_sift_context_t *sctx, Idx str_idx, + re_node_set *cur_dest) + internal_function; +static reg_errcode_t update_cur_sifted_state (const re_match_context_t *mctx, + re_sift_context_t *sctx, + Idx str_idx, + re_node_set *dest_nodes) + internal_function; +static reg_errcode_t add_epsilon_src_nodes (const re_dfa_t *dfa, + re_node_set *dest_nodes, + const re_node_set *candidates) + internal_function; +static bool check_dst_limits (const re_match_context_t *mctx, + const re_node_set *limits, + Idx dst_node, Idx dst_idx, Idx src_node, + Idx src_idx) internal_function; +static int check_dst_limits_calc_pos_1 (const re_match_context_t *mctx, + int boundaries, Idx subexp_idx, + Idx from_node, Idx bkref_idx) + internal_function; +static int check_dst_limits_calc_pos (const re_match_context_t *mctx, + Idx limit, Idx subexp_idx, + Idx node, Idx str_idx, + Idx bkref_idx) internal_function; +static reg_errcode_t check_subexp_limits (const re_dfa_t *dfa, + re_node_set *dest_nodes, + const re_node_set *candidates, + re_node_set *limits, + struct re_backref_cache_entry *bkref_ents, + Idx str_idx) internal_function; +static reg_errcode_t sift_states_bkref (const re_match_context_t *mctx, + re_sift_context_t *sctx, + Idx str_idx, const re_node_set *candidates) + internal_function; +static reg_errcode_t merge_state_array (const re_dfa_t *dfa, + re_dfastate_t **dst, + re_dfastate_t **src, Idx num) + internal_function; +static re_dfastate_t *find_recover_state (reg_errcode_t *err, + re_match_context_t *mctx) internal_function; +static re_dfastate_t *transit_state (reg_errcode_t *err, + re_match_context_t *mctx, + re_dfastate_t *state) internal_function; +static re_dfastate_t *merge_state_with_log (reg_errcode_t *err, + re_match_context_t *mctx, + re_dfastate_t *next_state) + internal_function; +static reg_errcode_t check_subexp_matching_top (re_match_context_t *mctx, + re_node_set *cur_nodes, + Idx str_idx) internal_function; +#if 0 +static re_dfastate_t *transit_state_sb (reg_errcode_t *err, + re_match_context_t *mctx, + re_dfastate_t *pstate) + internal_function; +#endif +#ifdef RE_ENABLE_I18N +static reg_errcode_t transit_state_mb (re_match_context_t *mctx, + re_dfastate_t *pstate) + internal_function; +#endif /* RE_ENABLE_I18N */ +static reg_errcode_t transit_state_bkref (re_match_context_t *mctx, + const re_node_set *nodes) + internal_function; +static reg_errcode_t get_subexp (re_match_context_t *mctx, + Idx bkref_node, Idx bkref_str_idx) + internal_function; +static reg_errcode_t get_subexp_sub (re_match_context_t *mctx, + const re_sub_match_top_t *sub_top, + re_sub_match_last_t *sub_last, + Idx bkref_node, Idx bkref_str) + internal_function; +static Idx find_subexp_node (const re_dfa_t *dfa, const re_node_set *nodes, + Idx subexp_idx, int type) internal_function; +static reg_errcode_t check_arrival (re_match_context_t *mctx, + state_array_t *path, Idx top_node, + Idx top_str, Idx last_node, Idx last_str, + int type) internal_function; +static reg_errcode_t check_arrival_add_next_nodes (re_match_context_t *mctx, + Idx str_idx, + re_node_set *cur_nodes, + re_node_set *next_nodes) + internal_function; +static reg_errcode_t check_arrival_expand_ecl (const re_dfa_t *dfa, + re_node_set *cur_nodes, + Idx ex_subexp, int type) + internal_function; +static reg_errcode_t check_arrival_expand_ecl_sub (const re_dfa_t *dfa, + re_node_set *dst_nodes, + Idx target, Idx ex_subexp, + int type) internal_function; +static reg_errcode_t expand_bkref_cache (re_match_context_t *mctx, + re_node_set *cur_nodes, Idx cur_str, + Idx subexp_num, int type) + internal_function; +static bool build_trtable (const re_dfa_t *dfa, + re_dfastate_t *state) internal_function; +#ifdef RE_ENABLE_I18N +static int check_node_accept_bytes (const re_dfa_t *dfa, Idx node_idx, + const re_string_t *input, Idx idx) + internal_function; +# ifdef _LIBC +static unsigned int find_collation_sequence_value (const unsigned char *mbs, + size_t name_len) + internal_function; +# endif /* _LIBC */ +#endif /* RE_ENABLE_I18N */ +static Idx group_nodes_into_DFAstates (const re_dfa_t *dfa, + const re_dfastate_t *state, + re_node_set *states_node, + bitset_t *states_ch) internal_function; +static bool check_node_accept (const re_match_context_t *mctx, + const re_token_t *node, Idx idx) + internal_function; +static reg_errcode_t extend_buffers (re_match_context_t *mctx) + internal_function; + +/* Entry point for POSIX code. */ + +/* regexec searches for a given pattern, specified by PREG, in the + string STRING. + + If NMATCH is zero or REG_NOSUB was set in the cflags argument to + `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at + least NMATCH elements, and we set them to the offsets of the + corresponding matched substrings. + + EFLAGS specifies `execution flags' which affect matching: if + REG_NOTBOL is set, then ^ does not match at the beginning of the + string; if REG_NOTEOL is set, then $ does not match at the end. + + We return 0 if we find a match and REG_NOMATCH if not. */ + +int +regexec (preg, string, nmatch, pmatch, eflags) + const regex_t *_Restrict_ preg; + const char *_Restrict_ string; + size_t nmatch; + regmatch_t pmatch[_Restrict_arr_]; + int eflags; +{ + reg_errcode_t err; + Idx start, length; +#ifdef _LIBC + re_dfa_t *dfa = (re_dfa_t *) preg->buffer; +#endif + + if (eflags & ~(REG_NOTBOL | REG_NOTEOL | REG_STARTEND)) + return REG_BADPAT; + + if (eflags & REG_STARTEND) + { + start = pmatch[0].rm_so; + length = pmatch[0].rm_eo; + } + else + { + start = 0; + length = strlen (string); + } + + __libc_lock_lock (dfa->lock); + if (preg->no_sub) + err = re_search_internal (preg, string, length, start, length, + length, 0, NULL, eflags); + else + err = re_search_internal (preg, string, length, start, length, + length, nmatch, pmatch, eflags); + __libc_lock_unlock (dfa->lock); + return err != REG_NOERROR; +} + +#ifdef _LIBC +# include <shlib-compat.h> +versioned_symbol (libc, __regexec, regexec, GLIBC_2_3_4); + +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3_4) +__typeof__ (__regexec) __compat_regexec; + +int +attribute_compat_text_section +__compat_regexec (const regex_t *_Restrict_ preg, + const char *_Restrict_ string, size_t nmatch, + regmatch_t pmatch[], int eflags) +{ + return regexec (preg, string, nmatch, pmatch, + eflags & (REG_NOTBOL | REG_NOTEOL)); +} +compat_symbol (libc, __compat_regexec, regexec, GLIBC_2_0); +# endif +#endif + +/* Entry points for GNU code. */ + +/* re_match, re_search, re_match_2, re_search_2 + + The former two functions operate on STRING with length LENGTH, + while the later two operate on concatenation of STRING1 and STRING2 + with lengths LENGTH1 and LENGTH2, respectively. + + re_match() matches the compiled pattern in BUFP against the string, + starting at index START. + + re_search() first tries matching at index START, then it tries to match + starting from index START + 1, and so on. The last start position tried + is START + RANGE. (Thus RANGE = 0 forces re_search to operate the same + way as re_match().) + + The parameter STOP of re_{match,search}_2 specifies that no match exceeding + the first STOP characters of the concatenation of the strings should be + concerned. + + If REGS is not NULL, and BUFP->no_sub is not set, the offsets of the match + and all groups is stored in REGS. (For the "_2" variants, the offsets are + computed relative to the concatenation, not relative to the individual + strings.) + + On success, re_match* functions return the length of the match, re_search* + return the position of the start of the match. Return value -1 means no + match was found and -2 indicates an internal error. */ + +regoff_t +re_match (bufp, string, length, start, regs) + struct re_pattern_buffer *bufp; + const char *string; + Idx length, start; + struct re_registers *regs; +{ + return re_search_stub (bufp, string, length, start, 0, length, regs, true); +} +#ifdef _LIBC +weak_alias (__re_match, re_match) +#endif + +regoff_t +re_search (bufp, string, length, start, range, regs) + struct re_pattern_buffer *bufp; + const char *string; + Idx length, start; + regoff_t range; + struct re_registers *regs; +{ + return re_search_stub (bufp, string, length, start, range, length, regs, + false); +} +#ifdef _LIBC +weak_alias (__re_search, re_search) +#endif + +regoff_t +re_match_2 (bufp, string1, length1, string2, length2, start, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + Idx length1, length2, start, stop; + struct re_registers *regs; +{ + return re_search_2_stub (bufp, string1, length1, string2, length2, + start, 0, regs, stop, true); +} +#ifdef _LIBC +weak_alias (__re_match_2, re_match_2) +#endif + +regoff_t +re_search_2 (bufp, string1, length1, string2, length2, start, range, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + Idx length1, length2, start, stop; + regoff_t range; + struct re_registers *regs; +{ + return re_search_2_stub (bufp, string1, length1, string2, length2, + start, range, regs, stop, false); +} +#ifdef _LIBC +weak_alias (__re_search_2, re_search_2) +#endif + +static regoff_t +internal_function +re_search_2_stub (struct re_pattern_buffer *bufp, + const char *string1, Idx length1, + const char *string2, Idx length2, + Idx start, regoff_t range, struct re_registers *regs, + Idx stop, bool ret_len) +{ + const char *str; + regoff_t rval; + Idx len = length1 + length2; + char *s = NULL; + + if (BE (length1 < 0 || length2 < 0 || stop < 0 || len < length1, 0)) + return -2; + + /* Concatenate the strings. */ + if (length2 > 0) + if (length1 > 0) + { + s = re_malloc (char, len); + + if (BE (s == NULL, 0)) + return -2; +#ifdef _LIBC + memcpy (__mempcpy (s, string1, length1), string2, length2); +#else + memcpy (s, string1, length1); + memcpy (s + length1, string2, length2); +#endif + str = s; + } + else + str = string2; + else + str = string1; + + rval = re_search_stub (bufp, str, len, start, range, stop, regs, + ret_len); + re_free (s); + return rval; +} + +/* The parameters have the same meaning as those of re_search. + Additional parameters: + If RET_LEN is true the length of the match is returned (re_match style); + otherwise the position of the match is returned. */ + +static regoff_t +internal_function +re_search_stub (struct re_pattern_buffer *bufp, + const char *string, Idx length, + Idx start, regoff_t range, Idx stop, struct re_registers *regs, + bool ret_len) +{ + reg_errcode_t result; + regmatch_t *pmatch; + Idx nregs; + regoff_t rval; + int eflags = 0; +#ifdef _LIBC + re_dfa_t *dfa = (re_dfa_t *) bufp->buffer; +#endif + Idx last_start = start + range; + + /* Check for out-of-range. */ + if (BE (start < 0 || start > length, 0)) + return -1; + if (BE (length < last_start || (0 <= range && last_start < start), 0)) + last_start = length; + else if (BE (last_start < 0 || (range < 0 && start <= last_start), 0)) + last_start = 0; + + __libc_lock_lock (dfa->lock); + + eflags |= (bufp->not_bol) ? REG_NOTBOL : 0; + eflags |= (bufp->not_eol) ? REG_NOTEOL : 0; + + /* Compile fastmap if we haven't yet. */ + if (start < last_start && bufp->fastmap != NULL && !bufp->fastmap_accurate) + re_compile_fastmap (bufp); + + if (BE (bufp->no_sub, 0)) + regs = NULL; + + /* We need at least 1 register. */ + if (regs == NULL) + nregs = 1; + else if (BE (bufp->regs_allocated == REGS_FIXED + && regs->num_regs <= bufp->re_nsub, 0)) + { + nregs = regs->num_regs; + if (BE (nregs < 1, 0)) + { + /* Nothing can be copied to regs. */ + regs = NULL; + nregs = 1; + } + } + else + nregs = bufp->re_nsub + 1; + pmatch = re_malloc (regmatch_t, nregs); + if (BE (pmatch == NULL, 0)) + { + rval = -2; + goto out; + } + + result = re_search_internal (bufp, string, length, start, last_start, stop, + nregs, pmatch, eflags); + + rval = 0; + + /* I hope we needn't fill ther regs with -1's when no match was found. */ + if (result != REG_NOERROR) + rval = -1; + else if (regs != NULL) + { + /* If caller wants register contents data back, copy them. */ + bufp->regs_allocated = re_copy_regs (regs, pmatch, nregs, + bufp->regs_allocated); + if (BE (bufp->regs_allocated == REGS_UNALLOCATED, 0)) + rval = -2; + } + + if (BE (rval == 0, 1)) + { + if (ret_len) + { + assert (pmatch[0].rm_so == start); + rval = pmatch[0].rm_eo - start; + } + else + rval = pmatch[0].rm_so; + } + re_free (pmatch); + out: + __libc_lock_unlock (dfa->lock); + return rval; +} + +static unsigned int +internal_function +re_copy_regs (struct re_registers *regs, regmatch_t *pmatch, Idx nregs, + int regs_allocated) +{ + int rval = REGS_REALLOCATE; + Idx i; + Idx need_regs = nregs + 1; + /* We need one extra element beyond `num_regs' for the `-1' marker GNU code + uses. */ + + /* Have the register data arrays been allocated? */ + if (regs_allocated == REGS_UNALLOCATED) + { /* No. So allocate them with malloc. */ + regs->start = re_malloc (regoff_t, need_regs); + if (BE (regs->start == NULL, 0)) + return REGS_UNALLOCATED; + regs->end = re_malloc (regoff_t, need_regs); + if (BE (regs->end == NULL, 0)) + { + re_free (regs->start); + return REGS_UNALLOCATED; + } + regs->num_regs = need_regs; + } + else if (regs_allocated == REGS_REALLOCATE) + { /* Yes. If we need more elements than were already + allocated, reallocate them. If we need fewer, just + leave it alone. */ + if (BE (need_regs > regs->num_regs, 0)) + { + regoff_t *new_start = re_realloc (regs->start, regoff_t, need_regs); + regoff_t *new_end; + if (BE (new_start == NULL, 0)) + return REGS_UNALLOCATED; + new_end = re_realloc (regs->end, regoff_t, need_regs); + if (BE (new_end == NULL, 0)) + { + re_free (new_start); + return REGS_UNALLOCATED; + } + regs->start = new_start; + regs->end = new_end; + regs->num_regs = need_regs; + } + } + else + { + assert (regs_allocated == REGS_FIXED); + /* This function may not be called with REGS_FIXED and nregs too big. */ + assert (regs->num_regs >= nregs); + rval = REGS_FIXED; + } + + /* Copy the regs. */ + for (i = 0; i < nregs; ++i) + { + regs->start[i] = pmatch[i].rm_so; + regs->end[i] = pmatch[i].rm_eo; + } + for ( ; i < regs->num_regs; ++i) + regs->start[i] = regs->end[i] = -1; + + return rval; +} + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use + this memory for recording register information. STARTS and ENDS + must be allocated using the malloc library routine, and must each + be at least NUM_REGS * sizeof (regoff_t) bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ + +void +re_set_registers (bufp, regs, num_regs, starts, ends) + struct re_pattern_buffer *bufp; + struct re_registers *regs; + __re_size_t num_regs; + regoff_t *starts, *ends; +{ + if (num_regs) + { + bufp->regs_allocated = REGS_REALLOCATE; + regs->num_regs = num_regs; + regs->start = starts; + regs->end = ends; + } + else + { + bufp->regs_allocated = REGS_UNALLOCATED; + regs->num_regs = 0; + regs->start = regs->end = NULL; + } +} +#ifdef _LIBC +weak_alias (__re_set_registers, re_set_registers) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC +int +# ifdef _LIBC +weak_function +# endif +re_exec (s) + const char *s; +{ + return 0 == regexec (&re_comp_buf, s, 0, NULL, 0); +} +#endif /* _REGEX_RE_COMP */ + +/* Internal entry point. */ + +/* Searches for a compiled pattern PREG in the string STRING, whose + length is LENGTH. NMATCH, PMATCH, and EFLAGS have the same + meaning as with regexec. LAST_START is START + RANGE, where + START and RANGE have the same meaning as with re_search. + Return REG_NOERROR if we find a match, and REG_NOMATCH if not, + otherwise return the error code. + Note: We assume front end functions already check ranges. + (0 <= LAST_START && LAST_START <= LENGTH) */ + +static reg_errcode_t +internal_function +re_search_internal (const regex_t *preg, + const char *string, Idx length, + Idx start, Idx last_start, Idx stop, + size_t nmatch, regmatch_t pmatch[], + int eflags) +{ + reg_errcode_t err; + const re_dfa_t *dfa = (const re_dfa_t *) preg->buffer; + Idx left_lim, right_lim; + int incr; + bool fl_longest_match; + int match_kind; + Idx match_first; + Idx match_last = REG_MISSING; + Idx extra_nmatch; + bool sb; + int ch; +#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L) + re_match_context_t mctx = { .dfa = dfa }; +#else + re_match_context_t mctx; +#endif + char *fastmap = ((preg->fastmap != NULL && preg->fastmap_accurate + && start != last_start && !preg->can_be_null) + ? preg->fastmap : NULL); + RE_TRANSLATE_TYPE t = preg->translate; + +#if !(defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)) + memset (&mctx, '\0', sizeof (re_match_context_t)); + mctx.dfa = dfa; +#endif + + extra_nmatch = (nmatch > preg->re_nsub) ? nmatch - (preg->re_nsub + 1) : 0; + nmatch -= extra_nmatch; + + /* Check if the DFA haven't been compiled. */ + if (BE (preg->used == 0 || dfa->init_state == NULL + || dfa->init_state_word == NULL || dfa->init_state_nl == NULL + || dfa->init_state_begbuf == NULL, 0)) + return REG_NOMATCH; + +#ifdef DEBUG + /* We assume front-end functions already check them. */ + assert (0 <= last_start && last_start <= length); +#endif + + /* If initial states with non-begbuf contexts have no elements, + the regex must be anchored. If preg->newline_anchor is set, + we'll never use init_state_nl, so do not check it. */ + if (dfa->init_state->nodes.nelem == 0 + && dfa->init_state_word->nodes.nelem == 0 + && (dfa->init_state_nl->nodes.nelem == 0 + || !preg->newline_anchor)) + { + if (start != 0 && last_start != 0) + return REG_NOMATCH; + start = last_start = 0; + } + + /* We must check the longest matching, if nmatch > 0. */ + fl_longest_match = (nmatch != 0 || dfa->nbackref); + + err = re_string_allocate (&mctx.input, string, length, dfa->nodes_len + 1, + preg->translate, preg->syntax & RE_ICASE, dfa); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + mctx.input.stop = stop; + mctx.input.raw_stop = stop; + mctx.input.newline_anchor = preg->newline_anchor; + + err = match_ctx_init (&mctx, eflags, dfa->nbackref * 2); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + /* We will log all the DFA states through which the dfa pass, + if nmatch > 1, or this dfa has "multibyte node", which is a + back-reference or a node which can accept multibyte character or + multi character collating element. */ + if (nmatch > 1 || dfa->has_mb_node) + { + /* Avoid overflow. */ + if (BE (SIZE_MAX / sizeof (re_dfastate_t *) <= mctx.input.bufs_len, 0)) + { + err = REG_ESPACE; + goto free_return; + } + + mctx.state_log = re_malloc (re_dfastate_t *, mctx.input.bufs_len + 1); + if (BE (mctx.state_log == NULL, 0)) + { + err = REG_ESPACE; + goto free_return; + } + } + else + mctx.state_log = NULL; + + match_first = start; + mctx.input.tip_context = (eflags & REG_NOTBOL) ? CONTEXT_BEGBUF + : CONTEXT_NEWLINE | CONTEXT_BEGBUF; + + /* Check incrementally whether of not the input string match. */ + incr = (last_start < start) ? -1 : 1; + left_lim = (last_start < start) ? last_start : start; + right_lim = (last_start < start) ? start : last_start; + sb = dfa->mb_cur_max == 1; + match_kind = + (fastmap + ? ((sb || !(preg->syntax & RE_ICASE || t) ? 4 : 0) + | (start <= last_start ? 2 : 0) + | (t != NULL ? 1 : 0)) + : 8); + + for (;; match_first += incr) + { + err = REG_NOMATCH; + if (match_first < left_lim || right_lim < match_first) + goto free_return; + + /* Advance as rapidly as possible through the string, until we + find a plausible place to start matching. This may be done + with varying efficiency, so there are various possibilities: + only the most common of them are specialized, in order to + save on code size. We use a switch statement for speed. */ + switch (match_kind) + { + case 8: + /* No fastmap. */ + break; + + case 7: + /* Fastmap with single-byte translation, match forward. */ + while (BE (match_first < right_lim, 1) + && !fastmap[t[(unsigned char) string[match_first]]]) + ++match_first; + goto forward_match_found_start_or_reached_end; + + case 6: + /* Fastmap without translation, match forward. */ + while (BE (match_first < right_lim, 1) + && !fastmap[(unsigned char) string[match_first]]) + ++match_first; + + forward_match_found_start_or_reached_end: + if (BE (match_first == right_lim, 0)) + { + ch = match_first >= length + ? 0 : (unsigned char) string[match_first]; + if (!fastmap[t ? t[ch] : ch]) + goto free_return; + } + break; + + case 4: + case 5: + /* Fastmap without multi-byte translation, match backwards. */ + while (match_first >= left_lim) + { + ch = match_first >= length + ? 0 : (unsigned char) string[match_first]; + if (fastmap[t ? t[ch] : ch]) + break; + --match_first; + } + if (match_first < left_lim) + goto free_return; + break; + + default: + /* In this case, we can't determine easily the current byte, + since it might be a component byte of a multibyte + character. Then we use the constructed buffer instead. */ + for (;;) + { + /* If MATCH_FIRST is out of the valid range, reconstruct the + buffers. */ + __re_size_t offset = match_first - mctx.input.raw_mbs_idx; + if (BE (offset >= (__re_size_t) mctx.input.valid_raw_len, 0)) + { + err = re_string_reconstruct (&mctx.input, match_first, + eflags); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + offset = match_first - mctx.input.raw_mbs_idx; + } + /* If MATCH_FIRST is out of the buffer, leave it as '\0'. + Note that MATCH_FIRST must not be smaller than 0. */ + ch = (match_first >= length + ? 0 : re_string_byte_at (&mctx.input, offset)); + if (fastmap[ch]) + break; + match_first += incr; + if (match_first < left_lim || match_first > right_lim) + { + err = REG_NOMATCH; + goto free_return; + } + } + break; + } + + /* Reconstruct the buffers so that the matcher can assume that + the matching starts from the beginning of the buffer. */ + err = re_string_reconstruct (&mctx.input, match_first, eflags); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + +#ifdef RE_ENABLE_I18N + /* Don't consider this char as a possible match start if it part, + yet isn't the head, of a multibyte character. */ + if (!sb && !re_string_first_byte (&mctx.input, 0)) + continue; +#endif + + /* It seems to be appropriate one, then use the matcher. */ + /* We assume that the matching starts from 0. */ + mctx.state_log_top = mctx.nbkref_ents = mctx.max_mb_elem_len = 0; + match_last = check_matching (&mctx, fl_longest_match, + start <= last_start ? &match_first : NULL); + if (match_last != REG_MISSING) + { + if (BE (match_last == REG_ERROR, 0)) + { + err = REG_ESPACE; + goto free_return; + } + else + { + mctx.match_last = match_last; + if ((!preg->no_sub && nmatch > 1) || dfa->nbackref) + { + re_dfastate_t *pstate = mctx.state_log[match_last]; + mctx.last_node = check_halt_state_context (&mctx, pstate, + match_last); + } + if ((!preg->no_sub && nmatch > 1 && dfa->has_plural_match) + || dfa->nbackref) + { + err = prune_impossible_nodes (&mctx); + if (err == REG_NOERROR) + break; + if (BE (err != REG_NOMATCH, 0)) + goto free_return; + match_last = REG_MISSING; + } + else + break; /* We found a match. */ + } + } + + match_ctx_clean (&mctx); + } + +#ifdef DEBUG + assert (match_last != REG_MISSING); + assert (err == REG_NOERROR); +#endif + + /* Set pmatch[] if we need. */ + if (nmatch > 0) + { + Idx reg_idx; + + /* Initialize registers. */ + for (reg_idx = 1; reg_idx < nmatch; ++reg_idx) + pmatch[reg_idx].rm_so = pmatch[reg_idx].rm_eo = -1; + + /* Set the points where matching start/end. */ + pmatch[0].rm_so = 0; + pmatch[0].rm_eo = mctx.match_last; + /* FIXME: This function should fail if mctx.match_last exceeds + the maximum possible regoff_t value. We need a new error + code REG_OVERFLOW. */ + + if (!preg->no_sub && nmatch > 1) + { + err = set_regs (preg, &mctx, nmatch, pmatch, + dfa->has_plural_match && dfa->nbackref > 0); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + + /* At last, add the offset to the each registers, since we slided + the buffers so that we could assume that the matching starts + from 0. */ + for (reg_idx = 0; reg_idx < nmatch; ++reg_idx) + if (pmatch[reg_idx].rm_so != -1) + { +#ifdef RE_ENABLE_I18N + if (BE (mctx.input.offsets_needed != 0, 0)) + { + pmatch[reg_idx].rm_so = + (pmatch[reg_idx].rm_so == mctx.input.valid_len + ? mctx.input.valid_raw_len + : mctx.input.offsets[pmatch[reg_idx].rm_so]); + pmatch[reg_idx].rm_eo = + (pmatch[reg_idx].rm_eo == mctx.input.valid_len + ? mctx.input.valid_raw_len + : mctx.input.offsets[pmatch[reg_idx].rm_eo]); + } +#else + assert (mctx.input.offsets_needed == 0); +#endif + pmatch[reg_idx].rm_so += match_first; + pmatch[reg_idx].rm_eo += match_first; + } + for (reg_idx = 0; reg_idx < extra_nmatch; ++reg_idx) + { + pmatch[nmatch + reg_idx].rm_so = -1; + pmatch[nmatch + reg_idx].rm_eo = -1; + } + + if (dfa->subexp_map) + for (reg_idx = 0; reg_idx + 1 < nmatch; reg_idx++) + if (dfa->subexp_map[reg_idx] != reg_idx) + { + pmatch[reg_idx + 1].rm_so + = pmatch[dfa->subexp_map[reg_idx] + 1].rm_so; + pmatch[reg_idx + 1].rm_eo + = pmatch[dfa->subexp_map[reg_idx] + 1].rm_eo; + } + } + + free_return: + re_free (mctx.state_log); + if (dfa->nbackref) + match_ctx_free (&mctx); + re_string_destruct (&mctx.input); + return err; +} + +static reg_errcode_t +internal_function +prune_impossible_nodes (re_match_context_t *mctx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx halt_node, match_last; + reg_errcode_t ret; + re_dfastate_t **sifted_states; + re_dfastate_t **lim_states = NULL; + re_sift_context_t sctx; +#ifdef DEBUG + assert (mctx->state_log != NULL); +#endif + match_last = mctx->match_last; + halt_node = mctx->last_node; + + /* Avoid overflow. */ + if (BE (SIZE_MAX / sizeof (re_dfastate_t *) <= match_last, 0)) + return REG_ESPACE; + + sifted_states = re_malloc (re_dfastate_t *, match_last + 1); + if (BE (sifted_states == NULL, 0)) + { + ret = REG_ESPACE; + goto free_return; + } + if (dfa->nbackref) + { + lim_states = re_malloc (re_dfastate_t *, match_last + 1); + if (BE (lim_states == NULL, 0)) + { + ret = REG_ESPACE; + goto free_return; + } + while (1) + { + memset (lim_states, '\0', + sizeof (re_dfastate_t *) * (match_last + 1)); + sift_ctx_init (&sctx, sifted_states, lim_states, halt_node, + match_last); + ret = sift_states_backward (mctx, &sctx); + re_node_set_free (&sctx.limits); + if (BE (ret != REG_NOERROR, 0)) + goto free_return; + if (sifted_states[0] != NULL || lim_states[0] != NULL) + break; + do + { + --match_last; + if (! REG_VALID_INDEX (match_last)) + { + ret = REG_NOMATCH; + goto free_return; + } + } while (mctx->state_log[match_last] == NULL + || !mctx->state_log[match_last]->halt); + halt_node = check_halt_state_context (mctx, + mctx->state_log[match_last], + match_last); + } + ret = merge_state_array (dfa, sifted_states, lim_states, + match_last + 1); + re_free (lim_states); + lim_states = NULL; + if (BE (ret != REG_NOERROR, 0)) + goto free_return; + } + else + { + sift_ctx_init (&sctx, sifted_states, lim_states, halt_node, match_last); + ret = sift_states_backward (mctx, &sctx); + re_node_set_free (&sctx.limits); + if (BE (ret != REG_NOERROR, 0)) + goto free_return; + } + re_free (mctx->state_log); + mctx->state_log = sifted_states; + sifted_states = NULL; + mctx->last_node = halt_node; + mctx->match_last = match_last; + ret = REG_NOERROR; + free_return: + re_free (sifted_states); + re_free (lim_states); + return ret; +} + +/* Acquire an initial state and return it. + We must select appropriate initial state depending on the context, + since initial states may have constraints like "\<", "^", etc.. */ + +static inline re_dfastate_t * +__attribute ((always_inline)) internal_function +acquire_init_state_context (reg_errcode_t *err, const re_match_context_t *mctx, + Idx idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + if (dfa->init_state->has_constraint) + { + unsigned int context; + context = re_string_context_at (&mctx->input, idx - 1, mctx->eflags); + if (IS_WORD_CONTEXT (context)) + return dfa->init_state_word; + else if (IS_ORDINARY_CONTEXT (context)) + return dfa->init_state; + else if (IS_BEGBUF_CONTEXT (context) && IS_NEWLINE_CONTEXT (context)) + return dfa->init_state_begbuf; + else if (IS_NEWLINE_CONTEXT (context)) + return dfa->init_state_nl; + else if (IS_BEGBUF_CONTEXT (context)) + { + /* It is relatively rare case, then calculate on demand. */ + return re_acquire_state_context (err, dfa, + dfa->init_state->entrance_nodes, + context); + } + else + /* Must not happen? */ + return dfa->init_state; + } + else + return dfa->init_state; +} + +/* Check whether the regular expression match input string INPUT or not, + and return the index where the matching end. Return REG_MISSING if + there is no match, and return REG_ERROR in case of an error. + FL_LONGEST_MATCH means we want the POSIX longest matching. + If P_MATCH_FIRST is not NULL, and the match fails, it is set to the + next place where we may want to try matching. + Note that the matcher assume that the maching starts from the current + index of the buffer. */ + +static Idx +internal_function +check_matching (re_match_context_t *mctx, bool fl_longest_match, + Idx *p_match_first) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx match = 0; + Idx match_last = REG_MISSING; + Idx cur_str_idx = re_string_cur_idx (&mctx->input); + re_dfastate_t *cur_state; + bool at_init_state = p_match_first != NULL; + Idx next_start_idx = cur_str_idx; + + err = REG_NOERROR; + cur_state = acquire_init_state_context (&err, mctx, cur_str_idx); + /* An initial state must not be NULL (invalid). */ + if (BE (cur_state == NULL, 0)) + { + assert (err == REG_ESPACE); + return REG_ERROR; + } + + if (mctx->state_log != NULL) + { + mctx->state_log[cur_str_idx] = cur_state; + + /* Check OP_OPEN_SUBEXP in the initial state in case that we use them + later. E.g. Processing back references. */ + if (BE (dfa->nbackref, 0)) + { + at_init_state = false; + err = check_subexp_matching_top (mctx, &cur_state->nodes, 0); + if (BE (err != REG_NOERROR, 0)) + return err; + + if (cur_state->has_backref) + { + err = transit_state_bkref (mctx, &cur_state->nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + } + + /* If the RE accepts NULL string. */ + if (BE (cur_state->halt, 0)) + { + if (!cur_state->has_constraint + || check_halt_state_context (mctx, cur_state, cur_str_idx)) + { + if (!fl_longest_match) + return cur_str_idx; + else + { + match_last = cur_str_idx; + match = 1; + } + } + } + + while (!re_string_eoi (&mctx->input)) + { + re_dfastate_t *old_state = cur_state; + Idx next_char_idx = re_string_cur_idx (&mctx->input) + 1; + + if (BE (next_char_idx >= mctx->input.bufs_len, 0) + || (BE (next_char_idx >= mctx->input.valid_len, 0) + && mctx->input.valid_len < mctx->input.len)) + { + err = extend_buffers (mctx); + if (BE (err != REG_NOERROR, 0)) + { + assert (err == REG_ESPACE); + return REG_ERROR; + } + } + + cur_state = transit_state (&err, mctx, cur_state); + if (mctx->state_log != NULL) + cur_state = merge_state_with_log (&err, mctx, cur_state); + + if (cur_state == NULL) + { + /* Reached the invalid state or an error. Try to recover a valid + state using the state log, if available and if we have not + already found a valid (even if not the longest) match. */ + if (BE (err != REG_NOERROR, 0)) + return REG_ERROR; + + if (mctx->state_log == NULL + || (match && !fl_longest_match) + || (cur_state = find_recover_state (&err, mctx)) == NULL) + break; + } + + if (BE (at_init_state, 0)) + { + if (old_state == cur_state) + next_start_idx = next_char_idx; + else + at_init_state = false; + } + + if (cur_state->halt) + { + /* Reached a halt state. + Check the halt state can satisfy the current context. */ + if (!cur_state->has_constraint + || check_halt_state_context (mctx, cur_state, + re_string_cur_idx (&mctx->input))) + { + /* We found an appropriate halt state. */ + match_last = re_string_cur_idx (&mctx->input); + match = 1; + + /* We found a match, do not modify match_first below. */ + p_match_first = NULL; + if (!fl_longest_match) + break; + } + } + } + + if (p_match_first) + *p_match_first += next_start_idx; + + return match_last; +} + +/* Check NODE match the current context. */ + +static bool +internal_function +check_halt_node_context (const re_dfa_t *dfa, Idx node, unsigned int context) +{ + re_token_type_t type = dfa->nodes[node].type; + unsigned int constraint = dfa->nodes[node].constraint; + if (type != END_OF_RE) + return false; + if (!constraint) + return true; + if (NOT_SATISFY_NEXT_CONSTRAINT (constraint, context)) + return false; + return true; +} + +/* Check the halt state STATE match the current context. + Return 0 if not match, if the node, STATE has, is a halt node and + match the context, return the node. */ + +static Idx +internal_function +check_halt_state_context (const re_match_context_t *mctx, + const re_dfastate_t *state, Idx idx) +{ + Idx i; + unsigned int context; +#ifdef DEBUG + assert (state->halt); +#endif + context = re_string_context_at (&mctx->input, idx, mctx->eflags); + for (i = 0; i < state->nodes.nelem; ++i) + if (check_halt_node_context (mctx->dfa, state->nodes.elems[i], context)) + return state->nodes.elems[i]; + return 0; +} + +/* Compute the next node to which "NFA" transit from NODE("NFA" is a NFA + corresponding to the DFA). + Return the destination node, and update EPS_VIA_NODES; + return REG_MISSING in case of errors. */ + +static Idx +internal_function +proceed_next_node (const re_match_context_t *mctx, Idx nregs, regmatch_t *regs, + Idx *pidx, Idx node, re_node_set *eps_via_nodes, + struct re_fail_stack_t *fs) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx i; + bool ok; + if (IS_EPSILON_NODE (dfa->nodes[node].type)) + { + re_node_set *cur_nodes = &mctx->state_log[*pidx]->nodes; + re_node_set *edests = &dfa->edests[node]; + Idx dest_node; + ok = re_node_set_insert (eps_via_nodes, node); + if (BE (! ok, 0)) + return REG_ERROR; + /* Pick up a valid destination, or return REG_MISSING if none + is found. */ + for (dest_node = REG_MISSING, i = 0; i < edests->nelem; ++i) + { + Idx candidate = edests->elems[i]; + if (!re_node_set_contains (cur_nodes, candidate)) + continue; + if (dest_node == REG_MISSING) + dest_node = candidate; + + else + { + /* In order to avoid infinite loop like "(a*)*", return the second + epsilon-transition if the first was already considered. */ + if (re_node_set_contains (eps_via_nodes, dest_node)) + return candidate; + + /* Otherwise, push the second epsilon-transition on the fail stack. */ + else if (fs != NULL + && push_fail_stack (fs, *pidx, candidate, nregs, regs, + eps_via_nodes)) + return REG_ERROR; + + /* We know we are going to exit. */ + break; + } + } + return dest_node; + } + else + { + Idx naccepted = 0; + re_token_type_t type = dfa->nodes[node].type; + +#ifdef RE_ENABLE_I18N + if (dfa->nodes[node].accept_mb) + naccepted = check_node_accept_bytes (dfa, node, &mctx->input, *pidx); + else +#endif /* RE_ENABLE_I18N */ + if (type == OP_BACK_REF) + { + Idx subexp_idx = dfa->nodes[node].opr.idx + 1; + naccepted = regs[subexp_idx].rm_eo - regs[subexp_idx].rm_so; + if (fs != NULL) + { + if (regs[subexp_idx].rm_so == -1 || regs[subexp_idx].rm_eo == -1) + return REG_MISSING; + else if (naccepted) + { + char *buf = (char *) re_string_get_buffer (&mctx->input); + if (memcmp (buf + regs[subexp_idx].rm_so, buf + *pidx, + naccepted) != 0) + return REG_MISSING; + } + } + + if (naccepted == 0) + { + Idx dest_node; + ok = re_node_set_insert (eps_via_nodes, node); + if (BE (! ok, 0)) + return REG_ERROR; + dest_node = dfa->edests[node].elems[0]; + if (re_node_set_contains (&mctx->state_log[*pidx]->nodes, + dest_node)) + return dest_node; + } + } + + if (naccepted != 0 + || check_node_accept (mctx, dfa->nodes + node, *pidx)) + { + Idx dest_node = dfa->nexts[node]; + *pidx = (naccepted == 0) ? *pidx + 1 : *pidx + naccepted; + if (fs && (*pidx > mctx->match_last || mctx->state_log[*pidx] == NULL + || !re_node_set_contains (&mctx->state_log[*pidx]->nodes, + dest_node))) + return REG_MISSING; + re_node_set_empty (eps_via_nodes); + return dest_node; + } + } + return REG_MISSING; +} + +static reg_errcode_t +internal_function +push_fail_stack (struct re_fail_stack_t *fs, Idx str_idx, Idx dest_node, + Idx nregs, regmatch_t *regs, re_node_set *eps_via_nodes) +{ + reg_errcode_t err; + Idx num = fs->num++; + if (fs->num == fs->alloc) + { + struct re_fail_stack_ent_t *new_array; + new_array = realloc (fs->stack, (sizeof (struct re_fail_stack_ent_t) + * fs->alloc * 2)); + if (new_array == NULL) + return REG_ESPACE; + fs->alloc *= 2; + fs->stack = new_array; + } + fs->stack[num].idx = str_idx; + fs->stack[num].node = dest_node; + fs->stack[num].regs = re_malloc (regmatch_t, nregs); + if (fs->stack[num].regs == NULL) + return REG_ESPACE; + memcpy (fs->stack[num].regs, regs, sizeof (regmatch_t) * nregs); + err = re_node_set_init_copy (&fs->stack[num].eps_via_nodes, eps_via_nodes); + return err; +} + +static Idx +internal_function +pop_fail_stack (struct re_fail_stack_t *fs, Idx *pidx, Idx nregs, + regmatch_t *regs, re_node_set *eps_via_nodes) +{ + Idx num = --fs->num; + assert (REG_VALID_INDEX (num)); + *pidx = fs->stack[num].idx; + memcpy (regs, fs->stack[num].regs, sizeof (regmatch_t) * nregs); + re_node_set_free (eps_via_nodes); + re_free (fs->stack[num].regs); + *eps_via_nodes = fs->stack[num].eps_via_nodes; + return fs->stack[num].node; +} + +/* Set the positions where the subexpressions are starts/ends to registers + PMATCH. + Note: We assume that pmatch[0] is already set, and + pmatch[i].rm_so == pmatch[i].rm_eo == -1 for 0 < i < nmatch. */ + +static reg_errcode_t +internal_function +set_regs (const regex_t *preg, const re_match_context_t *mctx, size_t nmatch, + regmatch_t *pmatch, bool fl_backtrack) +{ + const re_dfa_t *dfa = (const re_dfa_t *) preg->buffer; + Idx idx, cur_node; + re_node_set eps_via_nodes; + struct re_fail_stack_t *fs; + struct re_fail_stack_t fs_body = { 0, 2, NULL }; + regmatch_t *prev_idx_match; + bool prev_idx_match_malloced = false; + +#ifdef DEBUG + assert (nmatch > 1); + assert (mctx->state_log != NULL); +#endif + if (fl_backtrack) + { + fs = &fs_body; + fs->stack = re_malloc (struct re_fail_stack_ent_t, fs->alloc); + if (fs->stack == NULL) + return REG_ESPACE; + } + else + fs = NULL; + + cur_node = dfa->init_node; + re_node_set_init_empty (&eps_via_nodes); + + if (__libc_use_alloca (nmatch * sizeof (regmatch_t))) + prev_idx_match = (regmatch_t *) alloca (nmatch * sizeof (regmatch_t)); + else + { + prev_idx_match = re_malloc (regmatch_t, nmatch); + if (prev_idx_match == NULL) + { + free_fail_stack_return (fs); + return REG_ESPACE; + } + prev_idx_match_malloced = true; + } + memcpy (prev_idx_match, pmatch, sizeof (regmatch_t) * nmatch); + + for (idx = pmatch[0].rm_so; idx <= pmatch[0].rm_eo ;) + { + update_regs (dfa, pmatch, prev_idx_match, cur_node, idx, nmatch); + + if (idx == pmatch[0].rm_eo && cur_node == mctx->last_node) + { + Idx reg_idx; + if (fs) + { + for (reg_idx = 0; reg_idx < nmatch; ++reg_idx) + if (pmatch[reg_idx].rm_so > -1 && pmatch[reg_idx].rm_eo == -1) + break; + if (reg_idx == nmatch) + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return free_fail_stack_return (fs); + } + cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch, + &eps_via_nodes); + } + else + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return REG_NOERROR; + } + } + + /* Proceed to next node. */ + cur_node = proceed_next_node (mctx, nmatch, pmatch, &idx, cur_node, + &eps_via_nodes, fs); + + if (BE (! REG_VALID_INDEX (cur_node), 0)) + { + if (BE (cur_node == REG_ERROR, 0)) + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + free_fail_stack_return (fs); + return REG_ESPACE; + } + if (fs) + cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch, + &eps_via_nodes); + else + { + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return REG_NOMATCH; + } + } + } + re_node_set_free (&eps_via_nodes); + if (prev_idx_match_malloced) + re_free (prev_idx_match); + return free_fail_stack_return (fs); +} + +static reg_errcode_t +internal_function +free_fail_stack_return (struct re_fail_stack_t *fs) +{ + if (fs) + { + Idx fs_idx; + for (fs_idx = 0; fs_idx < fs->num; ++fs_idx) + { + re_node_set_free (&fs->stack[fs_idx].eps_via_nodes); + re_free (fs->stack[fs_idx].regs); + } + re_free (fs->stack); + } + return REG_NOERROR; +} + +static void +internal_function +update_regs (const re_dfa_t *dfa, regmatch_t *pmatch, + regmatch_t *prev_idx_match, Idx cur_node, Idx cur_idx, Idx nmatch) +{ + int type = dfa->nodes[cur_node].type; + if (type == OP_OPEN_SUBEXP) + { + Idx reg_num = dfa->nodes[cur_node].opr.idx + 1; + + /* We are at the first node of this sub expression. */ + if (reg_num < nmatch) + { + pmatch[reg_num].rm_so = cur_idx; + pmatch[reg_num].rm_eo = -1; + } + } + else if (type == OP_CLOSE_SUBEXP) + { + Idx reg_num = dfa->nodes[cur_node].opr.idx + 1; + if (reg_num < nmatch) + { + /* We are at the last node of this sub expression. */ + if (pmatch[reg_num].rm_so < cur_idx) + { + pmatch[reg_num].rm_eo = cur_idx; + /* This is a non-empty match or we are not inside an optional + subexpression. Accept this right away. */ + memcpy (prev_idx_match, pmatch, sizeof (regmatch_t) * nmatch); + } + else + { + if (dfa->nodes[cur_node].opt_subexp + && prev_idx_match[reg_num].rm_so != -1) + /* We transited through an empty match for an optional + subexpression, like (a?)*, and this is not the subexp's + first match. Copy back the old content of the registers + so that matches of an inner subexpression are undone as + well, like in ((a?))*. */ + memcpy (pmatch, prev_idx_match, sizeof (regmatch_t) * nmatch); + else + /* We completed a subexpression, but it may be part of + an optional one, so do not update PREV_IDX_MATCH. */ + pmatch[reg_num].rm_eo = cur_idx; + } + } + } +} + +/* This function checks the STATE_LOG from the SCTX->last_str_idx to 0 + and sift the nodes in each states according to the following rules. + Updated state_log will be wrote to STATE_LOG. + + Rules: We throw away the Node `a' in the STATE_LOG[STR_IDX] if... + 1. When STR_IDX == MATCH_LAST(the last index in the state_log): + If `a' isn't the LAST_NODE and `a' can't epsilon transit to + the LAST_NODE, we throw away the node `a'. + 2. When 0 <= STR_IDX < MATCH_LAST and `a' accepts + string `s' and transit to `b': + i. If 'b' isn't in the STATE_LOG[STR_IDX+strlen('s')], we throw + away the node `a'. + ii. If 'b' is in the STATE_LOG[STR_IDX+strlen('s')] but 'b' is + thrown away, we throw away the node `a'. + 3. When 0 <= STR_IDX < MATCH_LAST and 'a' epsilon transit to 'b': + i. If 'b' isn't in the STATE_LOG[STR_IDX], we throw away the + node `a'. + ii. If 'b' is in the STATE_LOG[STR_IDX] but 'b' is thrown away, + we throw away the node `a'. */ + +#define STATE_NODE_CONTAINS(state,node) \ + ((state) != NULL && re_node_set_contains (&(state)->nodes, node)) + +static reg_errcode_t +internal_function +sift_states_backward (const re_match_context_t *mctx, re_sift_context_t *sctx) +{ + reg_errcode_t err; + int null_cnt = 0; + Idx str_idx = sctx->last_str_idx; + re_node_set cur_dest; + +#ifdef DEBUG + assert (mctx->state_log != NULL && mctx->state_log[str_idx] != NULL); +#endif + + /* Build sifted state_log[str_idx]. It has the nodes which can epsilon + transit to the last_node and the last_node itself. */ + err = re_node_set_init_1 (&cur_dest, sctx->last_node); + if (BE (err != REG_NOERROR, 0)) + return err; + err = update_cur_sifted_state (mctx, sctx, str_idx, &cur_dest); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + /* Then check each states in the state_log. */ + while (str_idx > 0) + { + /* Update counters. */ + null_cnt = (sctx->sifted_states[str_idx] == NULL) ? null_cnt + 1 : 0; + if (null_cnt > mctx->max_mb_elem_len) + { + memset (sctx->sifted_states, '\0', + sizeof (re_dfastate_t *) * str_idx); + re_node_set_free (&cur_dest); + return REG_NOERROR; + } + re_node_set_empty (&cur_dest); + --str_idx; + + if (mctx->state_log[str_idx]) + { + err = build_sifted_states (mctx, sctx, str_idx, &cur_dest); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + + /* Add all the nodes which satisfy the following conditions: + - It can epsilon transit to a node in CUR_DEST. + - It is in CUR_SRC. + And update state_log. */ + err = update_cur_sifted_state (mctx, sctx, str_idx, &cur_dest); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + err = REG_NOERROR; + free_return: + re_node_set_free (&cur_dest); + return err; +} + +static reg_errcode_t +internal_function +build_sifted_states (const re_match_context_t *mctx, re_sift_context_t *sctx, + Idx str_idx, re_node_set *cur_dest) +{ + const re_dfa_t *const dfa = mctx->dfa; + const re_node_set *cur_src = &mctx->state_log[str_idx]->non_eps_nodes; + Idx i; + + /* Then build the next sifted state. + We build the next sifted state on `cur_dest', and update + `sifted_states[str_idx]' with `cur_dest'. + Note: + `cur_dest' is the sifted state from `state_log[str_idx + 1]'. + `cur_src' points the node_set of the old `state_log[str_idx]' + (with the epsilon nodes pre-filtered out). */ + for (i = 0; i < cur_src->nelem; i++) + { + Idx prev_node = cur_src->elems[i]; + int naccepted = 0; + bool ok; + +#ifdef DEBUG + re_token_type_t type = dfa->nodes[prev_node].type; + assert (!IS_EPSILON_NODE (type)); +#endif +#ifdef RE_ENABLE_I18N + /* If the node may accept `multi byte'. */ + if (dfa->nodes[prev_node].accept_mb) + naccepted = sift_states_iter_mb (mctx, sctx, prev_node, + str_idx, sctx->last_str_idx); +#endif /* RE_ENABLE_I18N */ + + /* We don't check backreferences here. + See update_cur_sifted_state(). */ + if (!naccepted + && check_node_accept (mctx, dfa->nodes + prev_node, str_idx) + && STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + 1], + dfa->nexts[prev_node])) + naccepted = 1; + + if (naccepted == 0) + continue; + + if (sctx->limits.nelem) + { + Idx to_idx = str_idx + naccepted; + if (check_dst_limits (mctx, &sctx->limits, + dfa->nexts[prev_node], to_idx, + prev_node, str_idx)) + continue; + } + ok = re_node_set_insert (cur_dest, prev_node); + if (BE (! ok, 0)) + return REG_ESPACE; + } + + return REG_NOERROR; +} + +/* Helper functions. */ + +static reg_errcode_t +internal_function +clean_state_log_if_needed (re_match_context_t *mctx, Idx next_state_log_idx) +{ + Idx top = mctx->state_log_top; + + if (next_state_log_idx >= mctx->input.bufs_len + || (next_state_log_idx >= mctx->input.valid_len + && mctx->input.valid_len < mctx->input.len)) + { + reg_errcode_t err; + err = extend_buffers (mctx); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + if (top < next_state_log_idx) + { + memset (mctx->state_log + top + 1, '\0', + sizeof (re_dfastate_t *) * (next_state_log_idx - top)); + mctx->state_log_top = next_state_log_idx; + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +merge_state_array (const re_dfa_t *dfa, re_dfastate_t **dst, + re_dfastate_t **src, Idx num) +{ + Idx st_idx; + reg_errcode_t err; + for (st_idx = 0; st_idx < num; ++st_idx) + { + if (dst[st_idx] == NULL) + dst[st_idx] = src[st_idx]; + else if (src[st_idx] != NULL) + { + re_node_set merged_set; + err = re_node_set_init_union (&merged_set, &dst[st_idx]->nodes, + &src[st_idx]->nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + dst[st_idx] = re_acquire_state (&err, dfa, &merged_set); + re_node_set_free (&merged_set); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +update_cur_sifted_state (const re_match_context_t *mctx, + re_sift_context_t *sctx, Idx str_idx, + re_node_set *dest_nodes) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err = REG_NOERROR; + const re_node_set *candidates; + candidates = ((mctx->state_log[str_idx] == NULL) ? NULL + : &mctx->state_log[str_idx]->nodes); + + if (dest_nodes->nelem == 0) + sctx->sifted_states[str_idx] = NULL; + else + { + if (candidates) + { + /* At first, add the nodes which can epsilon transit to a node in + DEST_NODE. */ + err = add_epsilon_src_nodes (dfa, dest_nodes, candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + + /* Then, check the limitations in the current sift_context. */ + if (sctx->limits.nelem) + { + err = check_subexp_limits (dfa, dest_nodes, candidates, &sctx->limits, + mctx->bkref_ents, str_idx); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + + sctx->sifted_states[str_idx] = re_acquire_state (&err, dfa, dest_nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + if (candidates && mctx->state_log[str_idx]->has_backref) + { + err = sift_states_bkref (mctx, sctx, str_idx, candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +add_epsilon_src_nodes (const re_dfa_t *dfa, re_node_set *dest_nodes, + const re_node_set *candidates) +{ + reg_errcode_t err = REG_NOERROR; + Idx i; + + re_dfastate_t *state = re_acquire_state (&err, dfa, dest_nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + + if (!state->inveclosure.alloc) + { + err = re_node_set_alloc (&state->inveclosure, dest_nodes->nelem); + if (BE (err != REG_NOERROR, 0)) + return REG_ESPACE; + for (i = 0; i < dest_nodes->nelem; i++) + re_node_set_merge (&state->inveclosure, + dfa->inveclosures + dest_nodes->elems[i]); + } + return re_node_set_add_intersect (dest_nodes, candidates, + &state->inveclosure); +} + +static reg_errcode_t +internal_function +sub_epsilon_src_nodes (const re_dfa_t *dfa, Idx node, re_node_set *dest_nodes, + const re_node_set *candidates) +{ + Idx ecl_idx; + reg_errcode_t err; + re_node_set *inv_eclosure = dfa->inveclosures + node; + re_node_set except_nodes; + re_node_set_init_empty (&except_nodes); + for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx) + { + Idx cur_node = inv_eclosure->elems[ecl_idx]; + if (cur_node == node) + continue; + if (IS_EPSILON_NODE (dfa->nodes[cur_node].type)) + { + Idx edst1 = dfa->edests[cur_node].elems[0]; + Idx edst2 = ((dfa->edests[cur_node].nelem > 1) + ? dfa->edests[cur_node].elems[1] : REG_MISSING); + if ((!re_node_set_contains (inv_eclosure, edst1) + && re_node_set_contains (dest_nodes, edst1)) + || (REG_VALID_NONZERO_INDEX (edst2) + && !re_node_set_contains (inv_eclosure, edst2) + && re_node_set_contains (dest_nodes, edst2))) + { + err = re_node_set_add_intersect (&except_nodes, candidates, + dfa->inveclosures + cur_node); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&except_nodes); + return err; + } + } + } + } + for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx) + { + Idx cur_node = inv_eclosure->elems[ecl_idx]; + if (!re_node_set_contains (&except_nodes, cur_node)) + { + Idx idx = re_node_set_contains (dest_nodes, cur_node) - 1; + re_node_set_remove_at (dest_nodes, idx); + } + } + re_node_set_free (&except_nodes); + return REG_NOERROR; +} + +static bool +internal_function +check_dst_limits (const re_match_context_t *mctx, const re_node_set *limits, + Idx dst_node, Idx dst_idx, Idx src_node, Idx src_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx lim_idx, src_pos, dst_pos; + + Idx dst_bkref_idx = search_cur_bkref_entry (mctx, dst_idx); + Idx src_bkref_idx = search_cur_bkref_entry (mctx, src_idx); + for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx) + { + Idx subexp_idx; + struct re_backref_cache_entry *ent; + ent = mctx->bkref_ents + limits->elems[lim_idx]; + subexp_idx = dfa->nodes[ent->node].opr.idx; + + dst_pos = check_dst_limits_calc_pos (mctx, limits->elems[lim_idx], + subexp_idx, dst_node, dst_idx, + dst_bkref_idx); + src_pos = check_dst_limits_calc_pos (mctx, limits->elems[lim_idx], + subexp_idx, src_node, src_idx, + src_bkref_idx); + + /* In case of: + <src> <dst> ( <subexp> ) + ( <subexp> ) <src> <dst> + ( <subexp1> <src> <subexp2> <dst> <subexp3> ) */ + if (src_pos == dst_pos) + continue; /* This is unrelated limitation. */ + else + return true; + } + return false; +} + +static int +internal_function +check_dst_limits_calc_pos_1 (const re_match_context_t *mctx, int boundaries, + Idx subexp_idx, Idx from_node, Idx bkref_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + const re_node_set *eclosures = dfa->eclosures + from_node; + Idx node_idx; + + /* Else, we are on the boundary: examine the nodes on the epsilon + closure. */ + for (node_idx = 0; node_idx < eclosures->nelem; ++node_idx) + { + Idx node = eclosures->elems[node_idx]; + switch (dfa->nodes[node].type) + { + case OP_BACK_REF: + if (bkref_idx != REG_MISSING) + { + struct re_backref_cache_entry *ent = mctx->bkref_ents + bkref_idx; + do + { + Idx dst; + int cpos; + + if (ent->node != node) + continue; + + if (subexp_idx < BITSET_WORD_BITS + && !(ent->eps_reachable_subexps_map + & ((bitset_word_t) 1 << subexp_idx))) + continue; + + /* Recurse trying to reach the OP_OPEN_SUBEXP and + OP_CLOSE_SUBEXP cases below. But, if the + destination node is the same node as the source + node, don't recurse because it would cause an + infinite loop: a regex that exhibits this behavior + is ()\1*\1* */ + dst = dfa->edests[node].elems[0]; + if (dst == from_node) + { + if (boundaries & 1) + return -1; + else /* if (boundaries & 2) */ + return 0; + } + + cpos = + check_dst_limits_calc_pos_1 (mctx, boundaries, subexp_idx, + dst, bkref_idx); + if (cpos == -1 /* && (boundaries & 1) */) + return -1; + if (cpos == 0 && (boundaries & 2)) + return 0; + + if (subexp_idx < BITSET_WORD_BITS) + ent->eps_reachable_subexps_map + &= ~((bitset_word_t) 1 << subexp_idx); + } + while (ent++->more); + } + break; + + case OP_OPEN_SUBEXP: + if ((boundaries & 1) && subexp_idx == dfa->nodes[node].opr.idx) + return -1; + break; + + case OP_CLOSE_SUBEXP: + if ((boundaries & 2) && subexp_idx == dfa->nodes[node].opr.idx) + return 0; + break; + + default: + break; + } + } + + return (boundaries & 2) ? 1 : 0; +} + +static int +internal_function +check_dst_limits_calc_pos (const re_match_context_t *mctx, Idx limit, + Idx subexp_idx, Idx from_node, Idx str_idx, + Idx bkref_idx) +{ + struct re_backref_cache_entry *lim = mctx->bkref_ents + limit; + int boundaries; + + /* If we are outside the range of the subexpression, return -1 or 1. */ + if (str_idx < lim->subexp_from) + return -1; + + if (lim->subexp_to < str_idx) + return 1; + + /* If we are within the subexpression, return 0. */ + boundaries = (str_idx == lim->subexp_from); + boundaries |= (str_idx == lim->subexp_to) << 1; + if (boundaries == 0) + return 0; + + /* Else, examine epsilon closure. */ + return check_dst_limits_calc_pos_1 (mctx, boundaries, subexp_idx, + from_node, bkref_idx); +} + +/* Check the limitations of sub expressions LIMITS, and remove the nodes + which are against limitations from DEST_NODES. */ + +static reg_errcode_t +internal_function +check_subexp_limits (const re_dfa_t *dfa, re_node_set *dest_nodes, + const re_node_set *candidates, re_node_set *limits, + struct re_backref_cache_entry *bkref_ents, Idx str_idx) +{ + reg_errcode_t err; + Idx node_idx, lim_idx; + + for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx) + { + Idx subexp_idx; + struct re_backref_cache_entry *ent; + ent = bkref_ents + limits->elems[lim_idx]; + + if (str_idx <= ent->subexp_from || ent->str_idx < str_idx) + continue; /* This is unrelated limitation. */ + + subexp_idx = dfa->nodes[ent->node].opr.idx; + if (ent->subexp_to == str_idx) + { + Idx ops_node = REG_MISSING; + Idx cls_node = REG_MISSING; + for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx) + { + Idx node = dest_nodes->elems[node_idx]; + re_token_type_t type = dfa->nodes[node].type; + if (type == OP_OPEN_SUBEXP + && subexp_idx == dfa->nodes[node].opr.idx) + ops_node = node; + else if (type == OP_CLOSE_SUBEXP + && subexp_idx == dfa->nodes[node].opr.idx) + cls_node = node; + } + + /* Check the limitation of the open subexpression. */ + /* Note that (ent->subexp_to = str_idx != ent->subexp_from). */ + if (REG_VALID_INDEX (ops_node)) + { + err = sub_epsilon_src_nodes (dfa, ops_node, dest_nodes, + candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + } + + /* Check the limitation of the close subexpression. */ + if (REG_VALID_INDEX (cls_node)) + for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx) + { + Idx node = dest_nodes->elems[node_idx]; + if (!re_node_set_contains (dfa->inveclosures + node, + cls_node) + && !re_node_set_contains (dfa->eclosures + node, + cls_node)) + { + /* It is against this limitation. + Remove it form the current sifted state. */ + err = sub_epsilon_src_nodes (dfa, node, dest_nodes, + candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + --node_idx; + } + } + } + else /* (ent->subexp_to != str_idx) */ + { + for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx) + { + Idx node = dest_nodes->elems[node_idx]; + re_token_type_t type = dfa->nodes[node].type; + if (type == OP_CLOSE_SUBEXP || type == OP_OPEN_SUBEXP) + { + if (subexp_idx != dfa->nodes[node].opr.idx) + continue; + /* It is against this limitation. + Remove it form the current sifted state. */ + err = sub_epsilon_src_nodes (dfa, node, dest_nodes, + candidates); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + } + } + return REG_NOERROR; +} + +static reg_errcode_t +internal_function +sift_states_bkref (const re_match_context_t *mctx, re_sift_context_t *sctx, + Idx str_idx, const re_node_set *candidates) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx node_idx, node; + re_sift_context_t local_sctx; + Idx first_idx = search_cur_bkref_entry (mctx, str_idx); + + if (first_idx == REG_MISSING) + return REG_NOERROR; + + local_sctx.sifted_states = NULL; /* Mark that it hasn't been initialized. */ + + for (node_idx = 0; node_idx < candidates->nelem; ++node_idx) + { + Idx enabled_idx; + re_token_type_t type; + struct re_backref_cache_entry *entry; + node = candidates->elems[node_idx]; + type = dfa->nodes[node].type; + /* Avoid infinite loop for the REs like "()\1+". */ + if (node == sctx->last_node && str_idx == sctx->last_str_idx) + continue; + if (type != OP_BACK_REF) + continue; + + entry = mctx->bkref_ents + first_idx; + enabled_idx = first_idx; + do + { + Idx subexp_len; + Idx to_idx; + Idx dst_node; + bool ok; + re_dfastate_t *cur_state; + + if (entry->node != node) + continue; + subexp_len = entry->subexp_to - entry->subexp_from; + to_idx = str_idx + subexp_len; + dst_node = (subexp_len ? dfa->nexts[node] + : dfa->edests[node].elems[0]); + + if (to_idx > sctx->last_str_idx + || sctx->sifted_states[to_idx] == NULL + || !STATE_NODE_CONTAINS (sctx->sifted_states[to_idx], dst_node) + || check_dst_limits (mctx, &sctx->limits, node, + str_idx, dst_node, to_idx)) + continue; + + if (local_sctx.sifted_states == NULL) + { + local_sctx = *sctx; + err = re_node_set_init_copy (&local_sctx.limits, &sctx->limits); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + local_sctx.last_node = node; + local_sctx.last_str_idx = str_idx; + ok = re_node_set_insert (&local_sctx.limits, enabled_idx); + if (BE (! ok, 0)) + { + err = REG_ESPACE; + goto free_return; + } + cur_state = local_sctx.sifted_states[str_idx]; + err = sift_states_backward (mctx, &local_sctx); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + if (sctx->limited_states != NULL) + { + err = merge_state_array (dfa, sctx->limited_states, + local_sctx.sifted_states, + str_idx + 1); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + local_sctx.sifted_states[str_idx] = cur_state; + re_node_set_remove (&local_sctx.limits, enabled_idx); + + /* mctx->bkref_ents may have changed, reload the pointer. */ + entry = mctx->bkref_ents + enabled_idx; + } + while (enabled_idx++, entry++->more); + } + err = REG_NOERROR; + free_return: + if (local_sctx.sifted_states != NULL) + { + re_node_set_free (&local_sctx.limits); + } + + return err; +} + + +#ifdef RE_ENABLE_I18N +static int +internal_function +sift_states_iter_mb (const re_match_context_t *mctx, re_sift_context_t *sctx, + Idx node_idx, Idx str_idx, Idx max_str_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + int naccepted; + /* Check the node can accept `multi byte'. */ + naccepted = check_node_accept_bytes (dfa, node_idx, &mctx->input, str_idx); + if (naccepted > 0 && str_idx + naccepted <= max_str_idx && + !STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + naccepted], + dfa->nexts[node_idx])) + /* The node can't accept the `multi byte', or the + destination was already thrown away, then the node + could't accept the current input `multi byte'. */ + naccepted = 0; + /* Otherwise, it is sure that the node could accept + `naccepted' bytes input. */ + return naccepted; +} +#endif /* RE_ENABLE_I18N */ + + +/* Functions for state transition. */ + +/* Return the next state to which the current state STATE will transit by + accepting the current input byte, and update STATE_LOG if necessary. + If STATE can accept a multibyte char/collating element/back reference + update the destination of STATE_LOG. */ + +static re_dfastate_t * +internal_function +transit_state (reg_errcode_t *err, re_match_context_t *mctx, + re_dfastate_t *state) +{ + re_dfastate_t **trtable; + unsigned char ch; + +#ifdef RE_ENABLE_I18N + /* If the current state can accept multibyte. */ + if (BE (state->accept_mb, 0)) + { + *err = transit_state_mb (mctx, state); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + } +#endif /* RE_ENABLE_I18N */ + + /* Then decide the next state with the single byte. */ +#if 0 + if (0) + /* don't use transition table */ + return transit_state_sb (err, mctx, state); +#endif + + /* Use transition table */ + ch = re_string_fetch_byte (&mctx->input); + for (;;) + { + trtable = state->trtable; + if (BE (trtable != NULL, 1)) + return trtable[ch]; + + trtable = state->word_trtable; + if (BE (trtable != NULL, 1)) + { + unsigned int context; + context + = re_string_context_at (&mctx->input, + re_string_cur_idx (&mctx->input) - 1, + mctx->eflags); + if (IS_WORD_CONTEXT (context)) + return trtable[ch + SBC_MAX]; + else + return trtable[ch]; + } + + if (!build_trtable (mctx->dfa, state)) + { + *err = REG_ESPACE; + return NULL; + } + + /* Retry, we now have a transition table. */ + } +} + +/* Update the state_log if we need */ +static re_dfastate_t * +internal_function +merge_state_with_log (reg_errcode_t *err, re_match_context_t *mctx, + re_dfastate_t *next_state) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx cur_idx = re_string_cur_idx (&mctx->input); + + if (cur_idx > mctx->state_log_top) + { + mctx->state_log[cur_idx] = next_state; + mctx->state_log_top = cur_idx; + } + else if (mctx->state_log[cur_idx] == 0) + { + mctx->state_log[cur_idx] = next_state; + } + else + { + re_dfastate_t *pstate; + unsigned int context; + re_node_set next_nodes, *log_nodes, *table_nodes = NULL; + /* If (state_log[cur_idx] != 0), it implies that cur_idx is + the destination of a multibyte char/collating element/ + back reference. Then the next state is the union set of + these destinations and the results of the transition table. */ + pstate = mctx->state_log[cur_idx]; + log_nodes = pstate->entrance_nodes; + if (next_state != NULL) + { + table_nodes = next_state->entrance_nodes; + *err = re_node_set_init_union (&next_nodes, table_nodes, + log_nodes); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + } + else + next_nodes = *log_nodes; + /* Note: We already add the nodes of the initial state, + then we don't need to add them here. */ + + context = re_string_context_at (&mctx->input, + re_string_cur_idx (&mctx->input) - 1, + mctx->eflags); + next_state = mctx->state_log[cur_idx] + = re_acquire_state_context (err, dfa, &next_nodes, context); + /* We don't need to check errors here, since the return value of + this function is next_state and ERR is already set. */ + + if (table_nodes != NULL) + re_node_set_free (&next_nodes); + } + + if (BE (dfa->nbackref, 0) && next_state != NULL) + { + /* Check OP_OPEN_SUBEXP in the current state in case that we use them + later. We must check them here, since the back references in the + next state might use them. */ + *err = check_subexp_matching_top (mctx, &next_state->nodes, + cur_idx); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + + /* If the next state has back references. */ + if (next_state->has_backref) + { + *err = transit_state_bkref (mctx, &next_state->nodes); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + next_state = mctx->state_log[cur_idx]; + } + } + + return next_state; +} + +/* Skip bytes in the input that correspond to part of a + multi-byte match, then look in the log for a state + from which to restart matching. */ +static re_dfastate_t * +internal_function +find_recover_state (reg_errcode_t *err, re_match_context_t *mctx) +{ + re_dfastate_t *cur_state; + do + { + Idx max = mctx->state_log_top; + Idx cur_str_idx = re_string_cur_idx (&mctx->input); + + do + { + if (++cur_str_idx > max) + return NULL; + re_string_skip_bytes (&mctx->input, 1); + } + while (mctx->state_log[cur_str_idx] == NULL); + + cur_state = merge_state_with_log (err, mctx, NULL); + } + while (*err == REG_NOERROR && cur_state == NULL); + return cur_state; +} + +/* Helper functions for transit_state. */ + +/* From the node set CUR_NODES, pick up the nodes whose types are + OP_OPEN_SUBEXP and which have corresponding back references in the regular + expression. And register them to use them later for evaluating the + correspoding back references. */ + +static reg_errcode_t +internal_function +check_subexp_matching_top (re_match_context_t *mctx, re_node_set *cur_nodes, + Idx str_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx node_idx; + reg_errcode_t err; + + /* TODO: This isn't efficient. + Because there might be more than one nodes whose types are + OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all + nodes. + E.g. RE: (a){2} */ + for (node_idx = 0; node_idx < cur_nodes->nelem; ++node_idx) + { + Idx node = cur_nodes->elems[node_idx]; + if (dfa->nodes[node].type == OP_OPEN_SUBEXP + && dfa->nodes[node].opr.idx < BITSET_WORD_BITS + && (dfa->used_bkref_map + & ((bitset_word_t) 1 << dfa->nodes[node].opr.idx))) + { + err = match_ctx_add_subtop (mctx, node, str_idx); + if (BE (err != REG_NOERROR, 0)) + return err; + } + } + return REG_NOERROR; +} + +#if 0 +/* Return the next state to which the current state STATE will transit by + accepting the current input byte. */ + +static re_dfastate_t * +transit_state_sb (reg_errcode_t *err, re_match_context_t *mctx, + re_dfastate_t *state) +{ + const re_dfa_t *const dfa = mctx->dfa; + re_node_set next_nodes; + re_dfastate_t *next_state; + Idx node_cnt, cur_str_idx = re_string_cur_idx (&mctx->input); + unsigned int context; + + *err = re_node_set_alloc (&next_nodes, state->nodes.nelem + 1); + if (BE (*err != REG_NOERROR, 0)) + return NULL; + for (node_cnt = 0; node_cnt < state->nodes.nelem; ++node_cnt) + { + Idx cur_node = state->nodes.elems[node_cnt]; + if (check_node_accept (mctx, dfa->nodes + cur_node, cur_str_idx)) + { + *err = re_node_set_merge (&next_nodes, + dfa->eclosures + dfa->nexts[cur_node]); + if (BE (*err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return NULL; + } + } + } + context = re_string_context_at (&mctx->input, cur_str_idx, mctx->eflags); + next_state = re_acquire_state_context (err, dfa, &next_nodes, context); + /* We don't need to check errors here, since the return value of + this function is next_state and ERR is already set. */ + + re_node_set_free (&next_nodes); + re_string_skip_bytes (&mctx->input, 1); + return next_state; +} +#endif + +#ifdef RE_ENABLE_I18N +static reg_errcode_t +internal_function +transit_state_mb (re_match_context_t *mctx, re_dfastate_t *pstate) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx i; + + for (i = 0; i < pstate->nodes.nelem; ++i) + { + re_node_set dest_nodes, *new_nodes; + Idx cur_node_idx = pstate->nodes.elems[i]; + int naccepted; + Idx dest_idx; + unsigned int context; + re_dfastate_t *dest_state; + + if (!dfa->nodes[cur_node_idx].accept_mb) + continue; + + if (dfa->nodes[cur_node_idx].constraint) + { + context = re_string_context_at (&mctx->input, + re_string_cur_idx (&mctx->input), + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (dfa->nodes[cur_node_idx].constraint, + context)) + continue; + } + + /* How many bytes the node can accept? */ + naccepted = check_node_accept_bytes (dfa, cur_node_idx, &mctx->input, + re_string_cur_idx (&mctx->input)); + if (naccepted == 0) + continue; + + /* The node can accepts `naccepted' bytes. */ + dest_idx = re_string_cur_idx (&mctx->input) + naccepted; + mctx->max_mb_elem_len = ((mctx->max_mb_elem_len < naccepted) ? naccepted + : mctx->max_mb_elem_len); + err = clean_state_log_if_needed (mctx, dest_idx); + if (BE (err != REG_NOERROR, 0)) + return err; +#ifdef DEBUG + assert (dfa->nexts[cur_node_idx] != REG_MISSING); +#endif + new_nodes = dfa->eclosures + dfa->nexts[cur_node_idx]; + + dest_state = mctx->state_log[dest_idx]; + if (dest_state == NULL) + dest_nodes = *new_nodes; + else + { + err = re_node_set_init_union (&dest_nodes, + dest_state->entrance_nodes, new_nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + context = re_string_context_at (&mctx->input, dest_idx - 1, + mctx->eflags); + mctx->state_log[dest_idx] + = re_acquire_state_context (&err, dfa, &dest_nodes, context); + if (dest_state != NULL) + re_node_set_free (&dest_nodes); + if (BE (mctx->state_log[dest_idx] == NULL && err != REG_NOERROR, 0)) + return err; + } + return REG_NOERROR; +} +#endif /* RE_ENABLE_I18N */ + +static reg_errcode_t +internal_function +transit_state_bkref (re_match_context_t *mctx, const re_node_set *nodes) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx i; + Idx cur_str_idx = re_string_cur_idx (&mctx->input); + + for (i = 0; i < nodes->nelem; ++i) + { + Idx dest_str_idx, prev_nelem, bkc_idx; + Idx node_idx = nodes->elems[i]; + unsigned int context; + const re_token_t *node = dfa->nodes + node_idx; + re_node_set *new_dest_nodes; + + /* Check whether `node' is a backreference or not. */ + if (node->type != OP_BACK_REF) + continue; + + if (node->constraint) + { + context = re_string_context_at (&mctx->input, cur_str_idx, + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context)) + continue; + } + + /* `node' is a backreference. + Check the substring which the substring matched. */ + bkc_idx = mctx->nbkref_ents; + err = get_subexp (mctx, node_idx, cur_str_idx); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + + /* And add the epsilon closures (which is `new_dest_nodes') of + the backreference to appropriate state_log. */ +#ifdef DEBUG + assert (dfa->nexts[node_idx] != REG_MISSING); +#endif + for (; bkc_idx < mctx->nbkref_ents; ++bkc_idx) + { + Idx subexp_len; + re_dfastate_t *dest_state; + struct re_backref_cache_entry *bkref_ent; + bkref_ent = mctx->bkref_ents + bkc_idx; + if (bkref_ent->node != node_idx || bkref_ent->str_idx != cur_str_idx) + continue; + subexp_len = bkref_ent->subexp_to - bkref_ent->subexp_from; + new_dest_nodes = (subexp_len == 0 + ? dfa->eclosures + dfa->edests[node_idx].elems[0] + : dfa->eclosures + dfa->nexts[node_idx]); + dest_str_idx = (cur_str_idx + bkref_ent->subexp_to + - bkref_ent->subexp_from); + context = re_string_context_at (&mctx->input, dest_str_idx - 1, + mctx->eflags); + dest_state = mctx->state_log[dest_str_idx]; + prev_nelem = ((mctx->state_log[cur_str_idx] == NULL) ? 0 + : mctx->state_log[cur_str_idx]->nodes.nelem); + /* Add `new_dest_node' to state_log. */ + if (dest_state == NULL) + { + mctx->state_log[dest_str_idx] + = re_acquire_state_context (&err, dfa, new_dest_nodes, + context); + if (BE (mctx->state_log[dest_str_idx] == NULL + && err != REG_NOERROR, 0)) + goto free_return; + } + else + { + re_node_set dest_nodes; + err = re_node_set_init_union (&dest_nodes, + dest_state->entrance_nodes, + new_dest_nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&dest_nodes); + goto free_return; + } + mctx->state_log[dest_str_idx] + = re_acquire_state_context (&err, dfa, &dest_nodes, context); + re_node_set_free (&dest_nodes); + if (BE (mctx->state_log[dest_str_idx] == NULL + && err != REG_NOERROR, 0)) + goto free_return; + } + /* We need to check recursively if the backreference can epsilon + transit. */ + if (subexp_len == 0 + && mctx->state_log[cur_str_idx]->nodes.nelem > prev_nelem) + { + err = check_subexp_matching_top (mctx, new_dest_nodes, + cur_str_idx); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + err = transit_state_bkref (mctx, new_dest_nodes); + if (BE (err != REG_NOERROR, 0)) + goto free_return; + } + } + } + err = REG_NOERROR; + free_return: + return err; +} + +/* Enumerate all the candidates which the backreference BKREF_NODE can match + at BKREF_STR_IDX, and register them by match_ctx_add_entry(). + Note that we might collect inappropriate candidates here. + However, the cost of checking them strictly here is too high, then we + delay these checking for prune_impossible_nodes(). */ + +static reg_errcode_t +internal_function +get_subexp (re_match_context_t *mctx, Idx bkref_node, Idx bkref_str_idx) +{ + const re_dfa_t *const dfa = mctx->dfa; + Idx subexp_num, sub_top_idx; + const char *buf = (const char *) re_string_get_buffer (&mctx->input); + /* Return if we have already checked BKREF_NODE at BKREF_STR_IDX. */ + Idx cache_idx = search_cur_bkref_entry (mctx, bkref_str_idx); + if (cache_idx != REG_MISSING) + { + const struct re_backref_cache_entry *entry + = mctx->bkref_ents + cache_idx; + do + if (entry->node == bkref_node) + return REG_NOERROR; /* We already checked it. */ + while (entry++->more); + } + + subexp_num = dfa->nodes[bkref_node].opr.idx; + + /* For each sub expression */ + for (sub_top_idx = 0; sub_top_idx < mctx->nsub_tops; ++sub_top_idx) + { + reg_errcode_t err; + re_sub_match_top_t *sub_top = mctx->sub_tops[sub_top_idx]; + re_sub_match_last_t *sub_last; + Idx sub_last_idx, sl_str, bkref_str_off; + + if (dfa->nodes[sub_top->node].opr.idx != subexp_num) + continue; /* It isn't related. */ + + sl_str = sub_top->str_idx; + bkref_str_off = bkref_str_idx; + /* At first, check the last node of sub expressions we already + evaluated. */ + for (sub_last_idx = 0; sub_last_idx < sub_top->nlasts; ++sub_last_idx) + { + regoff_t sl_str_diff; + sub_last = sub_top->lasts[sub_last_idx]; + sl_str_diff = sub_last->str_idx - sl_str; + /* The matched string by the sub expression match with the substring + at the back reference? */ + if (sl_str_diff > 0) + { + if (BE (bkref_str_off + sl_str_diff > mctx->input.valid_len, 0)) + { + /* Not enough chars for a successful match. */ + if (bkref_str_off + sl_str_diff > mctx->input.len) + break; + + err = clean_state_log_if_needed (mctx, + bkref_str_off + + sl_str_diff); + if (BE (err != REG_NOERROR, 0)) + return err; + buf = (const char *) re_string_get_buffer (&mctx->input); + } + if (memcmp (buf + bkref_str_off, buf + sl_str, sl_str_diff) != 0) + /* We don't need to search this sub expression any more. */ + break; + } + bkref_str_off += sl_str_diff; + sl_str += sl_str_diff; + err = get_subexp_sub (mctx, sub_top, sub_last, bkref_node, + bkref_str_idx); + + /* Reload buf, since the preceding call might have reallocated + the buffer. */ + buf = (const char *) re_string_get_buffer (&mctx->input); + + if (err == REG_NOMATCH) + continue; + if (BE (err != REG_NOERROR, 0)) + return err; + } + + if (sub_last_idx < sub_top->nlasts) + continue; + if (sub_last_idx > 0) + ++sl_str; + /* Then, search for the other last nodes of the sub expression. */ + for (; sl_str <= bkref_str_idx; ++sl_str) + { + Idx cls_node; + regoff_t sl_str_off; + const re_node_set *nodes; + sl_str_off = sl_str - sub_top->str_idx; + /* The matched string by the sub expression match with the substring + at the back reference? */ + if (sl_str_off > 0) + { + if (BE (bkref_str_off >= mctx->input.valid_len, 0)) + { + /* If we are at the end of the input, we cannot match. */ + if (bkref_str_off >= mctx->input.len) + break; + + err = extend_buffers (mctx); + if (BE (err != REG_NOERROR, 0)) + return err; + + buf = (const char *) re_string_get_buffer (&mctx->input); + } + if (buf [bkref_str_off++] != buf[sl_str - 1]) + break; /* We don't need to search this sub expression + any more. */ + } + if (mctx->state_log[sl_str] == NULL) + continue; + /* Does this state have a ')' of the sub expression? */ + nodes = &mctx->state_log[sl_str]->nodes; + cls_node = find_subexp_node (dfa, nodes, subexp_num, + OP_CLOSE_SUBEXP); + if (cls_node == REG_MISSING) + continue; /* No. */ + if (sub_top->path == NULL) + { + sub_top->path = calloc (sizeof (state_array_t), + sl_str - sub_top->str_idx + 1); + if (sub_top->path == NULL) + return REG_ESPACE; + } + /* Can the OP_OPEN_SUBEXP node arrive the OP_CLOSE_SUBEXP node + in the current context? */ + err = check_arrival (mctx, sub_top->path, sub_top->node, + sub_top->str_idx, cls_node, sl_str, + OP_CLOSE_SUBEXP); + if (err == REG_NOMATCH) + continue; + if (BE (err != REG_NOERROR, 0)) + return err; + sub_last = match_ctx_add_sublast (sub_top, cls_node, sl_str); + if (BE (sub_last == NULL, 0)) + return REG_ESPACE; + err = get_subexp_sub (mctx, sub_top, sub_last, bkref_node, + bkref_str_idx); + if (err == REG_NOMATCH) + continue; + } + } + return REG_NOERROR; +} + +/* Helper functions for get_subexp(). */ + +/* Check SUB_LAST can arrive to the back reference BKREF_NODE at BKREF_STR. + If it can arrive, register the sub expression expressed with SUB_TOP + and SUB_LAST. */ + +static reg_errcode_t +internal_function +get_subexp_sub (re_match_context_t *mctx, const re_sub_match_top_t *sub_top, + re_sub_match_last_t *sub_last, Idx bkref_node, Idx bkref_str) +{ + reg_errcode_t err; + Idx to_idx; + /* Can the subexpression arrive the back reference? */ + err = check_arrival (mctx, &sub_last->path, sub_last->node, + sub_last->str_idx, bkref_node, bkref_str, + OP_OPEN_SUBEXP); + if (err != REG_NOERROR) + return err; + err = match_ctx_add_entry (mctx, bkref_node, bkref_str, sub_top->str_idx, + sub_last->str_idx); + if (BE (err != REG_NOERROR, 0)) + return err; + to_idx = bkref_str + sub_last->str_idx - sub_top->str_idx; + return clean_state_log_if_needed (mctx, to_idx); +} + +/* Find the first node which is '(' or ')' and whose index is SUBEXP_IDX. + Search '(' if FL_OPEN, or search ')' otherwise. + TODO: This function isn't efficient... + Because there might be more than one nodes whose types are + OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all + nodes. + E.g. RE: (a){2} */ + +static Idx +internal_function +find_subexp_node (const re_dfa_t *dfa, const re_node_set *nodes, + Idx subexp_idx, int type) +{ + Idx cls_idx; + for (cls_idx = 0; cls_idx < nodes->nelem; ++cls_idx) + { + Idx cls_node = nodes->elems[cls_idx]; + const re_token_t *node = dfa->nodes + cls_node; + if (node->type == type + && node->opr.idx == subexp_idx) + return cls_node; + } + return REG_MISSING; +} + +/* Check whether the node TOP_NODE at TOP_STR can arrive to the node + LAST_NODE at LAST_STR. We record the path onto PATH since it will be + heavily reused. + Return REG_NOERROR if it can arrive, or REG_NOMATCH otherwise. */ + +static reg_errcode_t +internal_function +check_arrival (re_match_context_t *mctx, state_array_t *path, Idx top_node, + Idx top_str, Idx last_node, Idx last_str, int type) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err = REG_NOERROR; + Idx subexp_num, backup_cur_idx, str_idx, null_cnt; + re_dfastate_t *cur_state = NULL; + re_node_set *cur_nodes, next_nodes; + re_dfastate_t **backup_state_log; + unsigned int context; + + subexp_num = dfa->nodes[top_node].opr.idx; + /* Extend the buffer if we need. */ + if (BE (path->alloc < last_str + mctx->max_mb_elem_len + 1, 0)) + { + re_dfastate_t **new_array; + Idx old_alloc = path->alloc; + Idx new_alloc = old_alloc + last_str + mctx->max_mb_elem_len + 1; + if (BE (new_alloc < old_alloc, 0) + || BE (SIZE_MAX / sizeof (re_dfastate_t *) < new_alloc, 0)) + return REG_ESPACE; + new_array = re_realloc (path->array, re_dfastate_t *, new_alloc); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + path->array = new_array; + path->alloc = new_alloc; + memset (new_array + old_alloc, '\0', + sizeof (re_dfastate_t *) * (path->alloc - old_alloc)); + } + + str_idx = path->next_idx ? path->next_idx : top_str; + + /* Temporary modify MCTX. */ + backup_state_log = mctx->state_log; + backup_cur_idx = mctx->input.cur_idx; + mctx->state_log = path->array; + mctx->input.cur_idx = str_idx; + + /* Setup initial node set. */ + context = re_string_context_at (&mctx->input, str_idx - 1, mctx->eflags); + if (str_idx == top_str) + { + err = re_node_set_init_1 (&next_nodes, top_node); + if (BE (err != REG_NOERROR, 0)) + return err; + err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + else + { + cur_state = mctx->state_log[str_idx]; + if (cur_state && cur_state->has_backref) + { + err = re_node_set_init_copy (&next_nodes, &cur_state->nodes); + if (BE (err != REG_NOERROR, 0)) + return err; + } + else + re_node_set_init_empty (&next_nodes); + } + if (str_idx == top_str || (cur_state && cur_state->has_backref)) + { + if (next_nodes.nelem) + { + err = expand_bkref_cache (mctx, &next_nodes, str_idx, + subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context); + if (BE (cur_state == NULL && err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + mctx->state_log[str_idx] = cur_state; + } + + for (null_cnt = 0; str_idx < last_str && null_cnt <= mctx->max_mb_elem_len;) + { + re_node_set_empty (&next_nodes); + if (mctx->state_log[str_idx + 1]) + { + err = re_node_set_merge (&next_nodes, + &mctx->state_log[str_idx + 1]->nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + if (cur_state) + { + err = check_arrival_add_next_nodes (mctx, str_idx, + &cur_state->non_eps_nodes, + &next_nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + ++str_idx; + if (next_nodes.nelem) + { + err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + err = expand_bkref_cache (mctx, &next_nodes, str_idx, + subexp_num, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + } + context = re_string_context_at (&mctx->input, str_idx - 1, mctx->eflags); + cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context); + if (BE (cur_state == NULL && err != REG_NOERROR, 0)) + { + re_node_set_free (&next_nodes); + return err; + } + mctx->state_log[str_idx] = cur_state; + null_cnt = cur_state == NULL ? null_cnt + 1 : 0; + } + re_node_set_free (&next_nodes); + cur_nodes = (mctx->state_log[last_str] == NULL ? NULL + : &mctx->state_log[last_str]->nodes); + path->next_idx = str_idx; + + /* Fix MCTX. */ + mctx->state_log = backup_state_log; + mctx->input.cur_idx = backup_cur_idx; + + /* Then check the current node set has the node LAST_NODE. */ + if (cur_nodes != NULL && re_node_set_contains (cur_nodes, last_node)) + return REG_NOERROR; + + return REG_NOMATCH; +} + +/* Helper functions for check_arrival. */ + +/* Calculate the destination nodes of CUR_NODES at STR_IDX, and append them + to NEXT_NODES. + TODO: This function is similar to the functions transit_state*(), + however this function has many additional works. + Can't we unify them? */ + +static reg_errcode_t +internal_function +check_arrival_add_next_nodes (re_match_context_t *mctx, Idx str_idx, + re_node_set *cur_nodes, re_node_set *next_nodes) +{ + const re_dfa_t *const dfa = mctx->dfa; + bool ok; + Idx cur_idx; + reg_errcode_t err = REG_NOERROR; + re_node_set union_set; + re_node_set_init_empty (&union_set); + for (cur_idx = 0; cur_idx < cur_nodes->nelem; ++cur_idx) + { + int naccepted = 0; + Idx cur_node = cur_nodes->elems[cur_idx]; +#ifdef DEBUG + re_token_type_t type = dfa->nodes[cur_node].type; + assert (!IS_EPSILON_NODE (type)); +#endif +#ifdef RE_ENABLE_I18N + /* If the node may accept `multi byte'. */ + if (dfa->nodes[cur_node].accept_mb) + { + naccepted = check_node_accept_bytes (dfa, cur_node, &mctx->input, + str_idx); + if (naccepted > 1) + { + re_dfastate_t *dest_state; + Idx next_node = dfa->nexts[cur_node]; + Idx next_idx = str_idx + naccepted; + dest_state = mctx->state_log[next_idx]; + re_node_set_empty (&union_set); + if (dest_state) + { + err = re_node_set_merge (&union_set, &dest_state->nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&union_set); + return err; + } + } + ok = re_node_set_insert (&union_set, next_node); + if (BE (! ok, 0)) + { + re_node_set_free (&union_set); + return REG_ESPACE; + } + mctx->state_log[next_idx] = re_acquire_state (&err, dfa, + &union_set); + if (BE (mctx->state_log[next_idx] == NULL + && err != REG_NOERROR, 0)) + { + re_node_set_free (&union_set); + return err; + } + } + } +#endif /* RE_ENABLE_I18N */ + if (naccepted + || check_node_accept (mctx, dfa->nodes + cur_node, str_idx)) + { + ok = re_node_set_insert (next_nodes, dfa->nexts[cur_node]); + if (BE (! ok, 0)) + { + re_node_set_free (&union_set); + return REG_ESPACE; + } + } + } + re_node_set_free (&union_set); + return REG_NOERROR; +} + +/* For all the nodes in CUR_NODES, add the epsilon closures of them to + CUR_NODES, however exclude the nodes which are: + - inside the sub expression whose number is EX_SUBEXP, if FL_OPEN. + - out of the sub expression whose number is EX_SUBEXP, if !FL_OPEN. +*/ + +static reg_errcode_t +internal_function +check_arrival_expand_ecl (const re_dfa_t *dfa, re_node_set *cur_nodes, + Idx ex_subexp, int type) +{ + reg_errcode_t err; + Idx idx, outside_node; + re_node_set new_nodes; +#ifdef DEBUG + assert (cur_nodes->nelem); +#endif + err = re_node_set_alloc (&new_nodes, cur_nodes->nelem); + if (BE (err != REG_NOERROR, 0)) + return err; + /* Create a new node set NEW_NODES with the nodes which are epsilon + closures of the node in CUR_NODES. */ + + for (idx = 0; idx < cur_nodes->nelem; ++idx) + { + Idx cur_node = cur_nodes->elems[idx]; + const re_node_set *eclosure = dfa->eclosures + cur_node; + outside_node = find_subexp_node (dfa, eclosure, ex_subexp, type); + if (outside_node == REG_MISSING) + { + /* There are no problematic nodes, just merge them. */ + err = re_node_set_merge (&new_nodes, eclosure); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&new_nodes); + return err; + } + } + else + { + /* There are problematic nodes, re-calculate incrementally. */ + err = check_arrival_expand_ecl_sub (dfa, &new_nodes, cur_node, + ex_subexp, type); + if (BE (err != REG_NOERROR, 0)) + { + re_node_set_free (&new_nodes); + return err; + } + } + } + re_node_set_free (cur_nodes); + *cur_nodes = new_nodes; + return REG_NOERROR; +} + +/* Helper function for check_arrival_expand_ecl. + Check incrementally the epsilon closure of TARGET, and if it isn't + problematic append it to DST_NODES. */ + +static reg_errcode_t +internal_function +check_arrival_expand_ecl_sub (const re_dfa_t *dfa, re_node_set *dst_nodes, + Idx target, Idx ex_subexp, int type) +{ + Idx cur_node; + for (cur_node = target; !re_node_set_contains (dst_nodes, cur_node);) + { + bool ok; + + if (dfa->nodes[cur_node].type == type + && dfa->nodes[cur_node].opr.idx == ex_subexp) + { + if (type == OP_CLOSE_SUBEXP) + { + ok = re_node_set_insert (dst_nodes, cur_node); + if (BE (! ok, 0)) + return REG_ESPACE; + } + break; + } + ok = re_node_set_insert (dst_nodes, cur_node); + if (BE (! ok, 0)) + return REG_ESPACE; + if (dfa->edests[cur_node].nelem == 0) + break; + if (dfa->edests[cur_node].nelem == 2) + { + reg_errcode_t err; + err = check_arrival_expand_ecl_sub (dfa, dst_nodes, + dfa->edests[cur_node].elems[1], + ex_subexp, type); + if (BE (err != REG_NOERROR, 0)) + return err; + } + cur_node = dfa->edests[cur_node].elems[0]; + } + return REG_NOERROR; +} + + +/* For all the back references in the current state, calculate the + destination of the back references by the appropriate entry + in MCTX->BKREF_ENTS. */ + +static reg_errcode_t +internal_function +expand_bkref_cache (re_match_context_t *mctx, re_node_set *cur_nodes, + Idx cur_str, Idx subexp_num, int type) +{ + const re_dfa_t *const dfa = mctx->dfa; + reg_errcode_t err; + Idx cache_idx_start = search_cur_bkref_entry (mctx, cur_str); + struct re_backref_cache_entry *ent; + + if (cache_idx_start == REG_MISSING) + return REG_NOERROR; + + restart: + ent = mctx->bkref_ents + cache_idx_start; + do + { + Idx to_idx, next_node; + + /* Is this entry ENT is appropriate? */ + if (!re_node_set_contains (cur_nodes, ent->node)) + continue; /* No. */ + + to_idx = cur_str + ent->subexp_to - ent->subexp_from; + /* Calculate the destination of the back reference, and append it + to MCTX->STATE_LOG. */ + if (to_idx == cur_str) + { + /* The backreference did epsilon transit, we must re-check all the + node in the current state. */ + re_node_set new_dests; + reg_errcode_t err2, err3; + next_node = dfa->edests[ent->node].elems[0]; + if (re_node_set_contains (cur_nodes, next_node)) + continue; + err = re_node_set_init_1 (&new_dests, next_node); + err2 = check_arrival_expand_ecl (dfa, &new_dests, subexp_num, type); + err3 = re_node_set_merge (cur_nodes, &new_dests); + re_node_set_free (&new_dests); + if (BE (err != REG_NOERROR || err2 != REG_NOERROR + || err3 != REG_NOERROR, 0)) + { + err = (err != REG_NOERROR ? err + : (err2 != REG_NOERROR ? err2 : err3)); + return err; + } + /* TODO: It is still inefficient... */ + goto restart; + } + else + { + re_node_set union_set; + next_node = dfa->nexts[ent->node]; + if (mctx->state_log[to_idx]) + { + bool ok; + if (re_node_set_contains (&mctx->state_log[to_idx]->nodes, + next_node)) + continue; + err = re_node_set_init_copy (&union_set, + &mctx->state_log[to_idx]->nodes); + ok = re_node_set_insert (&union_set, next_node); + if (BE (err != REG_NOERROR || ! ok, 0)) + { + re_node_set_free (&union_set); + err = err != REG_NOERROR ? err : REG_ESPACE; + return err; + } + } + else + { + err = re_node_set_init_1 (&union_set, next_node); + if (BE (err != REG_NOERROR, 0)) + return err; + } + mctx->state_log[to_idx] = re_acquire_state (&err, dfa, &union_set); + re_node_set_free (&union_set); + if (BE (mctx->state_log[to_idx] == NULL + && err != REG_NOERROR, 0)) + return err; + } + } + while (ent++->more); + return REG_NOERROR; +} + +/* Build transition table for the state. + Return true if successful. */ + +static bool +internal_function +build_trtable (const re_dfa_t *dfa, re_dfastate_t *state) +{ + reg_errcode_t err; + Idx i, j; + int ch; + bool need_word_trtable = false; + bitset_word_t elem, mask; + bool dests_node_malloced = false; + bool dest_states_malloced = false; + Idx ndests; /* Number of the destination states from `state'. */ + re_dfastate_t **trtable; + re_dfastate_t **dest_states = NULL, **dest_states_word, **dest_states_nl; + re_node_set follows, *dests_node; + bitset_t *dests_ch; + bitset_t acceptable; + + struct dests_alloc + { + re_node_set dests_node[SBC_MAX]; + bitset_t dests_ch[SBC_MAX]; + } *dests_alloc; + + /* We build DFA states which corresponds to the destination nodes + from `state'. `dests_node[i]' represents the nodes which i-th + destination state contains, and `dests_ch[i]' represents the + characters which i-th destination state accepts. */ + if (__libc_use_alloca (sizeof (struct dests_alloc))) + dests_alloc = (struct dests_alloc *) alloca (sizeof (struct dests_alloc)); + else + { + dests_alloc = re_malloc (struct dests_alloc, 1); + if (BE (dests_alloc == NULL, 0)) + return false; + dests_node_malloced = true; + } + dests_node = dests_alloc->dests_node; + dests_ch = dests_alloc->dests_ch; + + /* Initialize transiton table. */ + state->word_trtable = state->trtable = NULL; + + /* At first, group all nodes belonging to `state' into several + destinations. */ + ndests = group_nodes_into_DFAstates (dfa, state, dests_node, dests_ch); + if (BE (! REG_VALID_NONZERO_INDEX (ndests), 0)) + { + if (dests_node_malloced) + free (dests_alloc); + if (ndests == 0) + { + state->trtable = (re_dfastate_t **) + calloc (sizeof (re_dfastate_t *), SBC_MAX); + return true; + } + return false; + } + + err = re_node_set_alloc (&follows, ndests + 1); + if (BE (err != REG_NOERROR, 0)) + goto out_free; + + /* Avoid arithmetic overflow in size calculation. */ + if (BE ((((SIZE_MAX - (sizeof (re_node_set) + sizeof (bitset_t)) * SBC_MAX) + / (3 * sizeof (re_dfastate_t *))) + < ndests), + 0)) + goto out_free; + + if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset_t)) * SBC_MAX + + ndests * 3 * sizeof (re_dfastate_t *))) + dest_states = (re_dfastate_t **) + alloca (ndests * 3 * sizeof (re_dfastate_t *)); + else + { + dest_states = (re_dfastate_t **) + malloc (ndests * 3 * sizeof (re_dfastate_t *)); + if (BE (dest_states == NULL, 0)) + { +out_free: + if (dest_states_malloced) + free (dest_states); + re_node_set_free (&follows); + for (i = 0; i < ndests; ++i) + re_node_set_free (dests_node + i); + if (dests_node_malloced) + free (dests_alloc); + return false; + } + dest_states_malloced = true; + } + dest_states_word = dest_states + ndests; + dest_states_nl = dest_states_word + ndests; + bitset_empty (acceptable); + + /* Then build the states for all destinations. */ + for (i = 0; i < ndests; ++i) + { + Idx next_node; + re_node_set_empty (&follows); + /* Merge the follows of this destination states. */ + for (j = 0; j < dests_node[i].nelem; ++j) + { + next_node = dfa->nexts[dests_node[i].elems[j]]; + if (next_node != REG_MISSING) + { + err = re_node_set_merge (&follows, dfa->eclosures + next_node); + if (BE (err != REG_NOERROR, 0)) + goto out_free; + } + } + dest_states[i] = re_acquire_state_context (&err, dfa, &follows, 0); + if (BE (dest_states[i] == NULL && err != REG_NOERROR, 0)) + goto out_free; + /* If the new state has context constraint, + build appropriate states for these contexts. */ + if (dest_states[i]->has_constraint) + { + dest_states_word[i] = re_acquire_state_context (&err, dfa, &follows, + CONTEXT_WORD); + if (BE (dest_states_word[i] == NULL && err != REG_NOERROR, 0)) + goto out_free; + + if (dest_states[i] != dest_states_word[i] && dfa->mb_cur_max > 1) + need_word_trtable = true; + + dest_states_nl[i] = re_acquire_state_context (&err, dfa, &follows, + CONTEXT_NEWLINE); + if (BE (dest_states_nl[i] == NULL && err != REG_NOERROR, 0)) + goto out_free; + } + else + { + dest_states_word[i] = dest_states[i]; + dest_states_nl[i] = dest_states[i]; + } + bitset_merge (acceptable, dests_ch[i]); + } + + if (!BE (need_word_trtable, 0)) + { + /* We don't care about whether the following character is a word + character, or we are in a single-byte character set so we can + discern by looking at the character code: allocate a + 256-entry transition table. */ + trtable = state->trtable = + (re_dfastate_t **) calloc (sizeof (re_dfastate_t *), SBC_MAX); + if (BE (trtable == NULL, 0)) + goto out_free; + + /* For all characters ch...: */ + for (i = 0; i < BITSET_WORDS; ++i) + for (ch = i * BITSET_WORD_BITS, elem = acceptable[i], mask = 1; + elem; + mask <<= 1, elem >>= 1, ++ch) + if (BE (elem & 1, 0)) + { + /* There must be exactly one destination which accepts + character ch. See group_nodes_into_DFAstates. */ + for (j = 0; (dests_ch[j][i] & mask) == 0; ++j) + ; + + /* j-th destination accepts the word character ch. */ + if (dfa->word_char[i] & mask) + trtable[ch] = dest_states_word[j]; + else + trtable[ch] = dest_states[j]; + } + } + else + { + /* We care about whether the following character is a word + character, and we are in a multi-byte character set: discern + by looking at the character code: build two 256-entry + transition tables, one starting at trtable[0] and one + starting at trtable[SBC_MAX]. */ + trtable = state->word_trtable = + (re_dfastate_t **) calloc (sizeof (re_dfastate_t *), 2 * SBC_MAX); + if (BE (trtable == NULL, 0)) + goto out_free; + + /* For all characters ch...: */ + for (i = 0; i < BITSET_WORDS; ++i) + for (ch = i * BITSET_WORD_BITS, elem = acceptable[i], mask = 1; + elem; + mask <<= 1, elem >>= 1, ++ch) + if (BE (elem & 1, 0)) + { + /* There must be exactly one destination which accepts + character ch. See group_nodes_into_DFAstates. */ + for (j = 0; (dests_ch[j][i] & mask) == 0; ++j) + ; + + /* j-th destination accepts the word character ch. */ + trtable[ch] = dest_states[j]; + trtable[ch + SBC_MAX] = dest_states_word[j]; + } + } + + /* new line */ + if (bitset_contain (acceptable, NEWLINE_CHAR)) + { + /* The current state accepts newline character. */ + for (j = 0; j < ndests; ++j) + if (bitset_contain (dests_ch[j], NEWLINE_CHAR)) + { + /* k-th destination accepts newline character. */ + trtable[NEWLINE_CHAR] = dest_states_nl[j]; + if (need_word_trtable) + trtable[NEWLINE_CHAR + SBC_MAX] = dest_states_nl[j]; + /* There must be only one destination which accepts + newline. See group_nodes_into_DFAstates. */ + break; + } + } + + if (dest_states_malloced) + free (dest_states); + + re_node_set_free (&follows); + for (i = 0; i < ndests; ++i) + re_node_set_free (dests_node + i); + + if (dests_node_malloced) + free (dests_alloc); + + return true; +} + +/* Group all nodes belonging to STATE into several destinations. + Then for all destinations, set the nodes belonging to the destination + to DESTS_NODE[i] and set the characters accepted by the destination + to DEST_CH[i]. This function return the number of destinations. */ + +static Idx +internal_function +group_nodes_into_DFAstates (const re_dfa_t *dfa, const re_dfastate_t *state, + re_node_set *dests_node, bitset_t *dests_ch) +{ + reg_errcode_t err; + bool ok; + Idx i, j, k; + Idx ndests; /* Number of the destinations from `state'. */ + bitset_t accepts; /* Characters a node can accept. */ + const re_node_set *cur_nodes = &state->nodes; + bitset_empty (accepts); + ndests = 0; + + /* For all the nodes belonging to `state', */ + for (i = 0; i < cur_nodes->nelem; ++i) + { + re_token_t *node = &dfa->nodes[cur_nodes->elems[i]]; + re_token_type_t type = node->type; + unsigned int constraint = node->constraint; + + /* Enumerate all single byte character this node can accept. */ + if (type == CHARACTER) + bitset_set (accepts, node->opr.c); + else if (type == SIMPLE_BRACKET) + { + bitset_merge (accepts, node->opr.sbcset); + } + else if (type == OP_PERIOD) + { +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + bitset_merge (accepts, dfa->sb_char); + else +#endif + bitset_set_all (accepts); + if (!(dfa->syntax & RE_DOT_NEWLINE)) + bitset_clear (accepts, '\n'); + if (dfa->syntax & RE_DOT_NOT_NULL) + bitset_clear (accepts, '\0'); + } +#ifdef RE_ENABLE_I18N + else if (type == OP_UTF8_PERIOD) + { + if (ASCII_CHARS % BITSET_WORD_BITS == 0) + memset (accepts, -1, ASCII_CHARS / CHAR_BIT); + else + bitset_merge (accepts, utf8_sb_map); + if (!(dfa->syntax & RE_DOT_NEWLINE)) + bitset_clear (accepts, '\n'); + if (dfa->syntax & RE_DOT_NOT_NULL) + bitset_clear (accepts, '\0'); + } +#endif + else + continue; + + /* Check the `accepts' and sift the characters which are not + match it the context. */ + if (constraint) + { + if (constraint & NEXT_NEWLINE_CONSTRAINT) + { + bool accepts_newline = bitset_contain (accepts, NEWLINE_CHAR); + bitset_empty (accepts); + if (accepts_newline) + bitset_set (accepts, NEWLINE_CHAR); + else + continue; + } + if (constraint & NEXT_ENDBUF_CONSTRAINT) + { + bitset_empty (accepts); + continue; + } + + if (constraint & NEXT_WORD_CONSTRAINT) + { + bitset_word_t any_set = 0; + if (type == CHARACTER && !node->word_char) + { + bitset_empty (accepts); + continue; + } +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= (dfa->word_char[j] | ~dfa->sb_char[j])); + else +#endif + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= dfa->word_char[j]); + if (!any_set) + continue; + } + if (constraint & NEXT_NOTWORD_CONSTRAINT) + { + bitset_word_t any_set = 0; + if (type == CHARACTER && node->word_char) + { + bitset_empty (accepts); + continue; + } +#ifdef RE_ENABLE_I18N + if (dfa->mb_cur_max > 1) + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= ~(dfa->word_char[j] & dfa->sb_char[j])); + else +#endif + for (j = 0; j < BITSET_WORDS; ++j) + any_set |= (accepts[j] &= ~dfa->word_char[j]); + if (!any_set) + continue; + } + } + + /* Then divide `accepts' into DFA states, or create a new + state. Above, we make sure that accepts is not empty. */ + for (j = 0; j < ndests; ++j) + { + bitset_t intersec; /* Intersection sets, see below. */ + bitset_t remains; + /* Flags, see below. */ + bitset_word_t has_intersec, not_subset, not_consumed; + + /* Optimization, skip if this state doesn't accept the character. */ + if (type == CHARACTER && !bitset_contain (dests_ch[j], node->opr.c)) + continue; + + /* Enumerate the intersection set of this state and `accepts'. */ + has_intersec = 0; + for (k = 0; k < BITSET_WORDS; ++k) + has_intersec |= intersec[k] = accepts[k] & dests_ch[j][k]; + /* And skip if the intersection set is empty. */ + if (!has_intersec) + continue; + + /* Then check if this state is a subset of `accepts'. */ + not_subset = not_consumed = 0; + for (k = 0; k < BITSET_WORDS; ++k) + { + not_subset |= remains[k] = ~accepts[k] & dests_ch[j][k]; + not_consumed |= accepts[k] = accepts[k] & ~dests_ch[j][k]; + } + + /* If this state isn't a subset of `accepts', create a + new group state, which has the `remains'. */ + if (not_subset) + { + bitset_copy (dests_ch[ndests], remains); + bitset_copy (dests_ch[j], intersec); + err = re_node_set_init_copy (dests_node + ndests, &dests_node[j]); + if (BE (err != REG_NOERROR, 0)) + goto error_return; + ++ndests; + } + + /* Put the position in the current group. */ + ok = re_node_set_insert (&dests_node[j], cur_nodes->elems[i]); + if (BE (! ok, 0)) + goto error_return; + + /* If all characters are consumed, go to next node. */ + if (!not_consumed) + break; + } + /* Some characters remain, create a new group. */ + if (j == ndests) + { + bitset_copy (dests_ch[ndests], accepts); + err = re_node_set_init_1 (dests_node + ndests, cur_nodes->elems[i]); + if (BE (err != REG_NOERROR, 0)) + goto error_return; + ++ndests; + bitset_empty (accepts); + } + } + return ndests; + error_return: + for (j = 0; j < ndests; ++j) + re_node_set_free (dests_node + j); + return REG_MISSING; +} + +#ifdef RE_ENABLE_I18N +/* Check how many bytes the node `dfa->nodes[node_idx]' accepts. + Return the number of the bytes the node accepts. + STR_IDX is the current index of the input string. + + This function handles the nodes which can accept one character, or + one collating element like '.', '[a-z]', opposite to the other nodes + can only accept one byte. */ + +static int +internal_function +check_node_accept_bytes (const re_dfa_t *dfa, Idx node_idx, + const re_string_t *input, Idx str_idx) +{ + const re_token_t *node = dfa->nodes + node_idx; + int char_len, elem_len; + Idx i; + + if (BE (node->type == OP_UTF8_PERIOD, 0)) + { + unsigned char c = re_string_byte_at (input, str_idx), d; + if (BE (c < 0xc2, 1)) + return 0; + + if (str_idx + 2 > input->len) + return 0; + + d = re_string_byte_at (input, str_idx + 1); + if (c < 0xe0) + return (d < 0x80 || d > 0xbf) ? 0 : 2; + else if (c < 0xf0) + { + char_len = 3; + if (c == 0xe0 && d < 0xa0) + return 0; + } + else if (c < 0xf8) + { + char_len = 4; + if (c == 0xf0 && d < 0x90) + return 0; + } + else if (c < 0xfc) + { + char_len = 5; + if (c == 0xf8 && d < 0x88) + return 0; + } + else if (c < 0xfe) + { + char_len = 6; + if (c == 0xfc && d < 0x84) + return 0; + } + else + return 0; + + if (str_idx + char_len > input->len) + return 0; + + for (i = 1; i < char_len; ++i) + { + d = re_string_byte_at (input, str_idx + i); + if (d < 0x80 || d > 0xbf) + return 0; + } + return char_len; + } + + char_len = re_string_char_size_at (input, str_idx); + if (node->type == OP_PERIOD) + { + if (char_len <= 1) + return 0; + /* FIXME: I don't think this if is needed, as both '\n' + and '\0' are char_len == 1. */ + /* '.' accepts any one character except the following two cases. */ + if ((!(dfa->syntax & RE_DOT_NEWLINE) && + re_string_byte_at (input, str_idx) == '\n') || + ((dfa->syntax & RE_DOT_NOT_NULL) && + re_string_byte_at (input, str_idx) == '\0')) + return 0; + return char_len; + } + + elem_len = re_string_elem_size_at (input, str_idx); + if ((elem_len <= 1 && char_len <= 1) || char_len == 0) + return 0; + + if (node->type == COMPLEX_BRACKET) + { + const re_charset_t *cset = node->opr.mbcset; +# ifdef _LIBC + const unsigned char *pin + = ((const unsigned char *) re_string_get_buffer (input) + str_idx); + Idx j; + uint32_t nrules; +# endif /* _LIBC */ + int match_len = 0; + wchar_t wc = ((cset->nranges || cset->nchar_classes || cset->nmbchars) + ? re_string_wchar_at (input, str_idx) : 0); + + /* match with multibyte character? */ + for (i = 0; i < cset->nmbchars; ++i) + if (wc == cset->mbchars[i]) + { + match_len = char_len; + goto check_node_accept_bytes_match; + } + /* match with character_class? */ + for (i = 0; i < cset->nchar_classes; ++i) + { + wctype_t wt = cset->char_classes[i]; + if (__iswctype (wc, wt)) + { + match_len = char_len; + goto check_node_accept_bytes_match; + } + } + +# ifdef _LIBC + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules != 0) + { + unsigned int in_collseq = 0; + const int32_t *table, *indirect; + const unsigned char *weights, *extra; + const char *collseqwc; + int32_t idx; + /* This #include defines a local function! */ +# include <locale/weight.h> + + /* match with collating_symbol? */ + if (cset->ncoll_syms) + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + for (i = 0; i < cset->ncoll_syms; ++i) + { + const unsigned char *coll_sym = extra + cset->coll_syms[i]; + /* Compare the length of input collating element and + the length of current collating element. */ + if (*coll_sym != elem_len) + continue; + /* Compare each bytes. */ + for (j = 0; j < *coll_sym; j++) + if (pin[j] != coll_sym[1 + j]) + break; + if (j == *coll_sym) + { + /* Match if every bytes is equal. */ + match_len = j; + goto check_node_accept_bytes_match; + } + } + + if (cset->nranges) + { + if (elem_len <= char_len) + { + collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC); + in_collseq = __collseq_table_lookup (collseqwc, wc); + } + else + in_collseq = find_collation_sequence_value (pin, elem_len); + } + /* match with range expression? */ + for (i = 0; i < cset->nranges; ++i) + if (cset->range_starts[i] <= in_collseq + && in_collseq <= cset->range_ends[i]) + { + match_len = elem_len; + goto check_node_accept_bytes_match; + } + + /* match with equivalence_class? */ + if (cset->nequiv_classes) + { + const unsigned char *cp = pin; + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); + idx = findidx (&cp); + if (idx > 0) + for (i = 0; i < cset->nequiv_classes; ++i) + { + int32_t equiv_class_idx = cset->equiv_classes[i]; + size_t weight_len = weights[idx]; + if (weight_len == weights[equiv_class_idx]) + { + Idx cnt = 0; + while (cnt <= weight_len + && (weights[equiv_class_idx + 1 + cnt] + == weights[idx + 1 + cnt])) + ++cnt; + if (cnt > weight_len) + { + match_len = elem_len; + goto check_node_accept_bytes_match; + } + } + } + } + } + else +# endif /* _LIBC */ + { + /* match with range expression? */ +#if __GNUC__ >= 2 && ! (__STDC_VERSION__ < 199901L && __STRICT_ANSI__) + wchar_t cmp_buf[] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'}; +#else + wchar_t cmp_buf[] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'}; + cmp_buf[2] = wc; +#endif + for (i = 0; i < cset->nranges; ++i) + { + cmp_buf[0] = cset->range_starts[i]; + cmp_buf[4] = cset->range_ends[i]; + if (wcscoll (cmp_buf, cmp_buf + 2) <= 0 + && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0) + { + match_len = char_len; + goto check_node_accept_bytes_match; + } + } + } + check_node_accept_bytes_match: + if (!cset->non_match) + return match_len; + else + { + if (match_len > 0) + return 0; + else + return (elem_len > char_len) ? elem_len : char_len; + } + } + return 0; +} + +# ifdef _LIBC +static unsigned int +internal_function +find_collation_sequence_value (const unsigned char *mbs, size_t mbs_len) +{ + uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules == 0) + { + if (mbs_len == 1) + { + /* No valid character. Match it as a single byte character. */ + const unsigned char *collseq = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB); + return collseq[mbs[0]]; + } + return UINT_MAX; + } + else + { + int32_t idx; + const unsigned char *extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + int32_t extrasize = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB + 1) - extra; + + for (idx = 0; idx < extrasize;) + { + int mbs_cnt; + bool found = false; + int32_t elem_mbs_len; + /* Skip the name of collating element name. */ + idx = idx + extra[idx] + 1; + elem_mbs_len = extra[idx++]; + if (mbs_len == elem_mbs_len) + { + for (mbs_cnt = 0; mbs_cnt < elem_mbs_len; ++mbs_cnt) + if (extra[idx + mbs_cnt] != mbs[mbs_cnt]) + break; + if (mbs_cnt == elem_mbs_len) + /* Found the entry. */ + found = true; + } + /* Skip the byte sequence of the collating element. */ + idx += elem_mbs_len; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + /* Skip the collation sequence value. */ + idx += sizeof (uint32_t); + /* Skip the wide char sequence of the collating element. */ + idx = idx + sizeof (uint32_t) * (extra[idx] + 1); + /* If we found the entry, return the sequence value. */ + if (found) + return *(uint32_t *) (extra + idx); + /* Skip the collation sequence value. */ + idx += sizeof (uint32_t); + } + return UINT_MAX; + } +} +# endif /* _LIBC */ +#endif /* RE_ENABLE_I18N */ + +/* Check whether the node accepts the byte which is IDX-th + byte of the INPUT. */ + +static bool +internal_function +check_node_accept (const re_match_context_t *mctx, const re_token_t *node, + Idx idx) +{ + unsigned char ch; + ch = re_string_byte_at (&mctx->input, idx); + switch (node->type) + { + case CHARACTER: + if (node->opr.c != ch) + return false; + break; + + case SIMPLE_BRACKET: + if (!bitset_contain (node->opr.sbcset, ch)) + return false; + break; + +#ifdef RE_ENABLE_I18N + case OP_UTF8_PERIOD: + if (ch >= ASCII_CHARS) + return false; + /* FALLTHROUGH */ +#endif + case OP_PERIOD: + if ((ch == '\n' && !(mctx->dfa->syntax & RE_DOT_NEWLINE)) + || (ch == '\0' && (mctx->dfa->syntax & RE_DOT_NOT_NULL))) + return false; + break; + + default: + return false; + } + + if (node->constraint) + { + /* The node has constraints. Check whether the current context + satisfies the constraints. */ + unsigned int context = re_string_context_at (&mctx->input, idx, + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context)) + return false; + } + + return true; +} + +/* Extend the buffers, if the buffers have run out. */ + +static reg_errcode_t +internal_function +extend_buffers (re_match_context_t *mctx) +{ + reg_errcode_t ret; + re_string_t *pstr = &mctx->input; + + /* Avoid overflow. */ + if (BE (SIZE_MAX / 2 / sizeof (re_dfastate_t *) <= pstr->bufs_len, 0)) + return REG_ESPACE; + + /* Double the lengthes of the buffers. */ + ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2); + if (BE (ret != REG_NOERROR, 0)) + return ret; + + if (mctx->state_log != NULL) + { + /* And double the length of state_log. */ + /* XXX We have no indication of the size of this buffer. If this + allocation fail we have no indication that the state_log array + does not have the right size. */ + re_dfastate_t **new_array = re_realloc (mctx->state_log, re_dfastate_t *, + pstr->bufs_len + 1); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + mctx->state_log = new_array; + } + + /* Then reconstruct the buffers. */ + if (pstr->icase) + { +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + { + ret = build_wcs_upper_buffer (pstr); + if (BE (ret != REG_NOERROR, 0)) + return ret; + } + else +#endif /* RE_ENABLE_I18N */ + build_upper_buffer (pstr); + } + else + { +#ifdef RE_ENABLE_I18N + if (pstr->mb_cur_max > 1) + build_wcs_buffer (pstr); + else +#endif /* RE_ENABLE_I18N */ + { + if (pstr->trans != NULL) + re_string_translate_buffer (pstr); + } + } + return REG_NOERROR; +} + + +/* Functions for matching context. */ + +/* Initialize MCTX. */ + +static reg_errcode_t +internal_function +match_ctx_init (re_match_context_t *mctx, int eflags, Idx n) +{ + mctx->eflags = eflags; + mctx->match_last = REG_MISSING; + if (n > 0) + { + /* Avoid overflow. */ + size_t max_object_size = + MAX (sizeof (struct re_backref_cache_entry), + sizeof (re_sub_match_top_t *)); + if (BE (SIZE_MAX / max_object_size < n, 0)) + return REG_ESPACE; + + mctx->bkref_ents = re_malloc (struct re_backref_cache_entry, n); + mctx->sub_tops = re_malloc (re_sub_match_top_t *, n); + if (BE (mctx->bkref_ents == NULL || mctx->sub_tops == NULL, 0)) + return REG_ESPACE; + } + /* Already zero-ed by the caller. + else + mctx->bkref_ents = NULL; + mctx->nbkref_ents = 0; + mctx->nsub_tops = 0; */ + mctx->abkref_ents = n; + mctx->max_mb_elem_len = 1; + mctx->asub_tops = n; + return REG_NOERROR; +} + +/* Clean the entries which depend on the current input in MCTX. + This function must be invoked when the matcher changes the start index + of the input, or changes the input string. */ + +static void +internal_function +match_ctx_clean (re_match_context_t *mctx) +{ + Idx st_idx; + for (st_idx = 0; st_idx < mctx->nsub_tops; ++st_idx) + { + Idx sl_idx; + re_sub_match_top_t *top = mctx->sub_tops[st_idx]; + for (sl_idx = 0; sl_idx < top->nlasts; ++sl_idx) + { + re_sub_match_last_t *last = top->lasts[sl_idx]; + re_free (last->path.array); + re_free (last); + } + re_free (top->lasts); + if (top->path) + { + re_free (top->path->array); + re_free (top->path); + } + free (top); + } + + mctx->nsub_tops = 0; + mctx->nbkref_ents = 0; +} + +/* Free all the memory associated with MCTX. */ + +static void +internal_function +match_ctx_free (re_match_context_t *mctx) +{ + /* First, free all the memory associated with MCTX->SUB_TOPS. */ + match_ctx_clean (mctx); + re_free (mctx->sub_tops); + re_free (mctx->bkref_ents); +} + +/* Add a new backreference entry to MCTX. + Note that we assume that caller never call this function with duplicate + entry, and call with STR_IDX which isn't smaller than any existing entry. +*/ + +static reg_errcode_t +internal_function +match_ctx_add_entry (re_match_context_t *mctx, Idx node, Idx str_idx, Idx from, + Idx to) +{ + if (mctx->nbkref_ents >= mctx->abkref_ents) + { + struct re_backref_cache_entry* new_entry; + new_entry = re_realloc (mctx->bkref_ents, struct re_backref_cache_entry, + mctx->abkref_ents * 2); + if (BE (new_entry == NULL, 0)) + { + re_free (mctx->bkref_ents); + return REG_ESPACE; + } + mctx->bkref_ents = new_entry; + memset (mctx->bkref_ents + mctx->nbkref_ents, '\0', + sizeof (struct re_backref_cache_entry) * mctx->abkref_ents); + mctx->abkref_ents *= 2; + } + if (mctx->nbkref_ents > 0 + && mctx->bkref_ents[mctx->nbkref_ents - 1].str_idx == str_idx) + mctx->bkref_ents[mctx->nbkref_ents - 1].more = 1; + + mctx->bkref_ents[mctx->nbkref_ents].node = node; + mctx->bkref_ents[mctx->nbkref_ents].str_idx = str_idx; + mctx->bkref_ents[mctx->nbkref_ents].subexp_from = from; + mctx->bkref_ents[mctx->nbkref_ents].subexp_to = to; + + /* This is a cache that saves negative results of check_dst_limits_calc_pos. + If bit N is clear, means that this entry won't epsilon-transition to + an OP_OPEN_SUBEXP or OP_CLOSE_SUBEXP for the N+1-th subexpression. If + it is set, check_dst_limits_calc_pos_1 will recurse and try to find one + such node. + + A backreference does not epsilon-transition unless it is empty, so set + to all zeros if FROM != TO. */ + mctx->bkref_ents[mctx->nbkref_ents].eps_reachable_subexps_map + = (from == to ? -1 : 0); + + mctx->bkref_ents[mctx->nbkref_ents++].more = 0; + if (mctx->max_mb_elem_len < to - from) + mctx->max_mb_elem_len = to - from; + return REG_NOERROR; +} + +/* Return the first entry with the same str_idx, or REG_MISSING if none is + found. Note that MCTX->BKREF_ENTS is already sorted by MCTX->STR_IDX. */ + +static Idx +internal_function +search_cur_bkref_entry (const re_match_context_t *mctx, Idx str_idx) +{ + Idx left, right, mid, last; + last = right = mctx->nbkref_ents; + for (left = 0; left < right;) + { + mid = (left + right) / 2; + if (mctx->bkref_ents[mid].str_idx < str_idx) + left = mid + 1; + else + right = mid; + } + if (left < last && mctx->bkref_ents[left].str_idx == str_idx) + return left; + else + return REG_MISSING; +} + +/* Register the node NODE, whose type is OP_OPEN_SUBEXP, and which matches + at STR_IDX. */ + +static reg_errcode_t +internal_function +match_ctx_add_subtop (re_match_context_t *mctx, Idx node, Idx str_idx) +{ +#ifdef DEBUG + assert (mctx->sub_tops != NULL); + assert (mctx->asub_tops > 0); +#endif + if (BE (mctx->nsub_tops == mctx->asub_tops, 0)) + { + Idx new_asub_tops = mctx->asub_tops * 2; + re_sub_match_top_t **new_array = re_realloc (mctx->sub_tops, + re_sub_match_top_t *, + new_asub_tops); + if (BE (new_array == NULL, 0)) + return REG_ESPACE; + mctx->sub_tops = new_array; + mctx->asub_tops = new_asub_tops; + } + mctx->sub_tops[mctx->nsub_tops] = calloc (1, sizeof (re_sub_match_top_t)); + if (BE (mctx->sub_tops[mctx->nsub_tops] == NULL, 0)) + return REG_ESPACE; + mctx->sub_tops[mctx->nsub_tops]->node = node; + mctx->sub_tops[mctx->nsub_tops++]->str_idx = str_idx; + return REG_NOERROR; +} + +/* Register the node NODE, whose type is OP_CLOSE_SUBEXP, and which matches + at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP. */ + +static re_sub_match_last_t * +internal_function +match_ctx_add_sublast (re_sub_match_top_t *subtop, Idx node, Idx str_idx) +{ + re_sub_match_last_t *new_entry; + if (BE (subtop->nlasts == subtop->alasts, 0)) + { + Idx new_alasts = 2 * subtop->alasts + 1; + re_sub_match_last_t **new_array = re_realloc (subtop->lasts, + re_sub_match_last_t *, + new_alasts); + if (BE (new_array == NULL, 0)) + return NULL; + subtop->lasts = new_array; + subtop->alasts = new_alasts; + } + new_entry = calloc (1, sizeof (re_sub_match_last_t)); + if (BE (new_entry != NULL, 1)) + { + subtop->lasts[subtop->nlasts] = new_entry; + new_entry->node = node; + new_entry->str_idx = str_idx; + ++subtop->nlasts; + } + return new_entry; +} + +static void +internal_function +sift_ctx_init (re_sift_context_t *sctx, re_dfastate_t **sifted_sts, + re_dfastate_t **limited_sts, Idx last_node, Idx last_str_idx) +{ + sctx->sifted_states = sifted_sts; + sctx->limited_states = limited_sts; + sctx->last_node = last_node; + sctx->last_str_idx = last_str_idx; + re_node_set_init_empty (&sctx->limits); +} diff --git a/lib/rmdir.c b/lib/rmdir.c new file mode 100644 index 0000000..1221f4f --- /dev/null +++ b/lib/rmdir.c @@ -0,0 +1,74 @@ +/* BSD compatible remove directory function for System V + + Copyright (C) 1988, 1990, 1999, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <sys/types.h> +#include <sys/stat.h> +#include <errno.h> + +/* rmdir adapted from GNU tar. */ + +/* Remove directory DIR. + Return 0 if successful, -1 if not. */ + +int +rmdir (char const *dir) +{ + pid_t cpid; + int status; + struct stat statbuf; + + if (stat (dir, &statbuf) != 0) + return -1; /* errno already set */ + + if (!S_ISDIR (statbuf.st_mode)) + { + errno = ENOTDIR; + return -1; + } + + cpid = fork (); + switch (cpid) + { + case -1: /* cannot fork */ + return -1; /* errno already set */ + + case 0: /* child process */ + execl ("/bin/rmdir", "rmdir", dir, (char *) 0); + _exit (1); + + default: /* parent process */ + + /* Wait for kid to finish. */ + + while (wait (&status) != cpid) + /* Do nothing. */ ; + + if (status) + { + + /* /bin/rmdir failed. */ + + errno = EIO; + return -1; + } + return 0; + } +} diff --git a/lib/rmt.h b/lib/rmt.h new file mode 100644 index 0000000..9f96cdb --- /dev/null +++ b/lib/rmt.h @@ -0,0 +1,99 @@ +/* Definitions for communicating with a remote tape drive. + + Copyright (C) 1988, 1992, 1996, 1997, 2001, 2003, 2004 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +extern char *rmt_command; +extern char *rmt_dev_name__; + +int rmt_open__ (const char *, int, int, const char *); +int rmt_close__ (int); +size_t rmt_read__ (int, char *, size_t); +size_t rmt_write__ (int, char *, size_t); +off_t rmt_lseek__ (int, off_t, int); +int rmt_ioctl__ (int, int, char *); + +extern bool force_local_option; + +/* A filename is remote if it contains a colon not preceded by a slash, + to take care of `/:/' which is a shorthand for `/.../<CELL-NAME>/fs' + on machines running OSF's Distributing Computing Environment (DCE) and + Distributed File System (DFS). However, when --force-local, a + filename is never remote. */ + +#define _remdev(dev_name) \ + (!force_local_option && (rmt_dev_name__ = strchr (dev_name, ':')) \ + && rmt_dev_name__ > (dev_name) \ + && ! memchr (dev_name, '/', rmt_dev_name__ - (dev_name))) + +#define _isrmt(fd) \ + ((fd) >= __REM_BIAS) + +#define __REM_BIAS (1 << 30) + +#ifndef O_CREAT +# define O_CREAT 01000 +#endif + +#define rmtopen(dev_name, oflag, mode, command) \ + (_remdev (dev_name) ? rmt_open__ (dev_name, oflag, __REM_BIAS, command) \ + : open (dev_name, oflag, mode)) + +#define rmtaccess(dev_name, amode) \ + (_remdev (dev_name) ? 0 : access (dev_name, amode)) + +#define rmtstat(dev_name, buffer) \ + (_remdev (dev_name) ? (errno = EOPNOTSUPP), -1 : stat (dev_name, buffer)) + +#define rmtcreat(dev_name, mode, command) \ + (_remdev (dev_name) \ + ? rmt_open__ (dev_name, 1 | O_CREAT, __REM_BIAS, command) \ + : creat (dev_name, mode)) + +#define rmtlstat(dev_name, muffer) \ + (_remdev (dev_name) ? (errno = EOPNOTSUPP), -1 : lstat (dev_name, buffer)) + +#define rmtread(fd, buffer, length) \ + (_isrmt (fd) ? rmt_read__ (fd - __REM_BIAS, buffer, length) \ + : safe_read (fd, buffer, length)) + +#define rmtwrite(fd, buffer, length) \ + (_isrmt (fd) ? rmt_write__ (fd - __REM_BIAS, buffer, length) \ + : full_write (fd, buffer, length)) + +#define rmtlseek(fd, offset, where) \ + (_isrmt (fd) ? rmt_lseek__ (fd - __REM_BIAS, offset, where) \ + : lseek (fd, offset, where)) + +#define rmtclose(fd) \ + (_isrmt (fd) ? rmt_close__ (fd - __REM_BIAS) : close (fd)) + +#define rmtioctl(fd, request, argument) \ + (_isrmt (fd) ? rmt_ioctl__ (fd - __REM_BIAS, request, argument) \ + : ioctl (fd, request, argument)) + +#define rmtdup(fd) \ + (_isrmt (fd) ? (errno = EOPNOTSUPP), -1 : dup (fd)) + +#define rmtfstat(fd, buffer) \ + (_isrmt (fd) ? (errno = EOPNOTSUPP), -1 : fstat (fd, buffer)) + +#define rmtfcntl(cd, command, argument) \ + (_isrmt (fd) ? (errno = EOPNOTSUPP), -1 : fcntl (fd, command, argument)) + +#define rmtisatty(fd) \ + (_isrmt (fd) ? 0 : isatty (fd)) diff --git a/lib/rpmatch.c b/lib/rpmatch.c new file mode 100644 index 0000000..e5f79f8 --- /dev/null +++ b/lib/rpmatch.c @@ -0,0 +1,79 @@ +/* Determine whether string value is affirmation or negative response + according to current locale's data. + + Copyright (C) 1996, 1998, 2000, 2002, 2003, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <stddef.h> +#include <stdlib.h> + +#if ENABLE_NLS +# include <sys/types.h> +# include <limits.h> +# include <regex.h> +# include "gettext.h" +# define _(msgid) gettext (msgid) + +static int +try (const char *response, const char *pattern, const int match, + const int nomatch, const char **lastp, regex_t *re) +{ + if (pattern != *lastp) + { + /* The pattern has changed. */ + if (*lastp) + { + /* Free the old compiled pattern. */ + regfree (re); + *lastp = NULL; + } + /* Compile the pattern and cache it for future runs. */ + if (regcomp (re, pattern, REG_EXTENDED) != 0) + return -1; + *lastp = pattern; + } + + /* See if the regular expression matches RESPONSE. */ + return regexec (re, response, 0, NULL, 0) == 0 ? match : nomatch; +} +#endif + + +int +rpmatch (const char *response) +{ +#if ENABLE_NLS + /* Match against one of the response patterns, compiling the pattern + first if necessary. */ + + /* We cache the response patterns and compiled regexps here. */ + static const char *yesexpr, *noexpr; + static regex_t yesre, nore; + int result; + + return ((result = try (response, _("^[yY]"), 1, 0, + &yesexpr, &yesre)) + ? result + : try (response, _("^[nN]"), 0, -1, &noexpr, &nore)); +#else + /* Test against "^[yY]" and "^[nN]", hardcoded to avoid requiring regex */ + return (*response == 'y' || *response == 'Y' ? 1 + : *response == 'n' || *response == 'N' ? 0 : -1); +#endif +} diff --git a/lib/rtapelib.c b/lib/rtapelib.c new file mode 100644 index 0000000..af19b04 --- /dev/null +++ b/lib/rtapelib.c @@ -0,0 +1,740 @@ +/* Functions for communicating with a remote tape drive. + + Copyright (C) 1988, 1992, 1994, 1996, 1997, 1999, 2000, 2001, 2004, + 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* The man page rmt(8) for /etc/rmt documents the remote mag tape protocol + which rdump and rrestore use. Unfortunately, the man page is *WRONG*. + The author of the routines I'm including originally wrote his code just + based on the man page, and it didn't work, so he went to the rdump source + to figure out why. The only thing he had to change was to check for the + 'F' return code in addition to the 'E', and to separate the various + arguments with \n instead of a space. I personally don't think that this + is much of a problem, but I wanted to point it out. -- Arnold Robbins + + Originally written by Jeff Lee, modified some by Arnold Robbins. Redone + as a library that can replace open, read, write, etc., by Fred Fish, with + some additional work by Arnold Robbins. Modified to make all rmt* calls + into macros for speed by Jay Fenlason. Use -DWITH_REXEC for rexec + code, courtesy of Dan Kegel. */ + +#include "system.h" +#include "system-ioctl.h" + +#include <safe-read.h> +#include <full-write.h> + +/* Try hard to get EOPNOTSUPP defined. 486/ISC has it in net/errno.h, + 3B2/SVR3 has it in sys/inet.h. Otherwise, like on MSDOS, use EINVAL. */ + +#ifndef EOPNOTSUPP +# if HAVE_NET_ERRNO_H +# include <net/errno.h> +# endif +# if HAVE_SYS_INET_H +# include <sys/inet.h> +# endif +# ifndef EOPNOTSUPP +# define EOPNOTSUPP EINVAL +# endif +#endif + +#include <signal.h> + +#if HAVE_NETDB_H +# include <netdb.h> +#endif + +#include <rmt.h> +#include <rmt-command.h> + +/* Exit status if exec errors. */ +#define EXIT_ON_EXEC_ERROR 128 + +/* FIXME: Size of buffers for reading and writing commands to rmt. */ +#define COMMAND_BUFFER_SIZE 64 + +#ifndef RETSIGTYPE +# define RETSIGTYPE void +#endif + +/* FIXME: Maximum number of simultaneous remote tape connections. */ +#define MAXUNIT 4 + +#define PREAD 0 /* read file descriptor from pipe() */ +#define PWRITE 1 /* write file descriptor from pipe() */ + +/* Return the parent's read side of remote tape connection Fd. */ +#define READ_SIDE(Fd) (from_remote[Fd][PREAD]) + +/* Return the parent's write side of remote tape connection Fd. */ +#define WRITE_SIDE(Fd) (to_remote[Fd][PWRITE]) + +/* The pipes for receiving data from remote tape drives. */ +static int from_remote[MAXUNIT][2] = {{-1, -1}, {-1, -1}, {-1, -1}, {-1, -1}}; + +/* The pipes for sending data to remote tape drives. */ +static int to_remote[MAXUNIT][2] = {{-1, -1}, {-1, -1}, {-1, -1}, {-1, -1}}; + +char *rmt_command = DEFAULT_RMT_COMMAND; + +/* Temporary variable used by macros in rmt.h. */ +char *rmt_dev_name__; + +/* If true, always consider file names to be local, even if they contain + colons */ +bool force_local_option; + + + +/* Close remote tape connection HANDLE, and reset errno to ERRNO_VALUE. */ +static void +_rmt_shutdown (int handle, int errno_value) +{ + close (READ_SIDE (handle)); + close (WRITE_SIDE (handle)); + READ_SIDE (handle) = -1; + WRITE_SIDE (handle) = -1; + errno = errno_value; +} + +/* Attempt to perform the remote tape command specified in BUFFER on + remote tape connection HANDLE. Return 0 if successful, -1 on + error. */ +static int +do_command (int handle, const char *buffer) +{ + /* Save the current pipe handler and try to make the request. */ + + size_t length = strlen (buffer); + RETSIGTYPE (*pipe_handler) () = signal (SIGPIPE, SIG_IGN); + ssize_t written = full_write (WRITE_SIDE (handle), buffer, length); + signal (SIGPIPE, pipe_handler); + + if (written == length) + return 0; + + /* Something went wrong. Close down and go home. */ + + _rmt_shutdown (handle, EIO); + return -1; +} + +static char * +get_status_string (int handle, char *command_buffer) +{ + char *cursor; + int counter; + + /* Read the reply command line. */ + + for (counter = 0, cursor = command_buffer; + counter < COMMAND_BUFFER_SIZE; + counter++, cursor++) + { + if (safe_read (READ_SIDE (handle), cursor, 1) != 1) + { + _rmt_shutdown (handle, EIO); + return 0; + } + if (*cursor == '\n') + { + *cursor = '\0'; + break; + } + } + + if (counter == COMMAND_BUFFER_SIZE) + { + _rmt_shutdown (handle, EIO); + return 0; + } + + /* Check the return status. */ + + for (cursor = command_buffer; *cursor; cursor++) + if (*cursor != ' ') + break; + + if (*cursor == 'E' || *cursor == 'F') + { + /* Skip the error message line. */ + + /* FIXME: there is better to do than merely ignoring error messages + coming from the remote end. Translate them, too... */ + + { + char character; + + while (safe_read (READ_SIDE (handle), &character, 1) == 1) + if (character == '\n') + break; + } + + errno = atoi (cursor + 1); + + if (*cursor == 'F') + _rmt_shutdown (handle, errno); + + return 0; + } + + /* Check for mis-synced pipes. */ + + if (*cursor != 'A') + { + _rmt_shutdown (handle, EIO); + return 0; + } + + /* Got an `A' (success) response. */ + + return cursor + 1; +} + +/* Read and return the status from remote tape connection HANDLE. If + an error occurred, return -1 and set errno. */ +static long int +get_status (int handle) +{ + char command_buffer[COMMAND_BUFFER_SIZE]; + const char *status = get_status_string (handle, command_buffer); + if (status) + { + long int result = atol (status); + if (0 <= result) + return result; + errno = EIO; + } + return -1; +} + +static off_t +get_status_off (int handle) +{ + char command_buffer[COMMAND_BUFFER_SIZE]; + const char *status = get_status_string (handle, command_buffer); + + if (! status) + return -1; + else + { + /* Parse status, taking care to check for overflow. + We can't use standard functions, + since off_t might be longer than long. */ + + off_t count = 0; + int negative; + + for (; *status == ' ' || *status == '\t'; status++) + continue; + + negative = *status == '-'; + status += negative || *status == '+'; + + for (;;) + { + int digit = *status++ - '0'; + if (9 < (unsigned) digit) + break; + else + { + off_t c10 = 10 * count; + off_t nc = negative ? c10 - digit : c10 + digit; + if (c10 / 10 != count || (negative ? c10 < nc : nc < c10)) + return -1; + count = nc; + } + } + + return count; + } +} + +#if WITH_REXEC + +/* Execute /etc/rmt as user USER on remote system HOST using rexec. + Return a file descriptor of a bidirectional socket for stdin and + stdout. If USER is zero, use the current username. + + By default, this code is not used, since it requires that the user + have a .netrc file in his/her home directory, or that the + application designer be willing to have rexec prompt for login and + password info. This may be unacceptable, and .rhosts files for use + with rsh are much more common on BSD systems. */ +static int +_rmt_rexec (char *host, char *user) +{ + int saved_stdin = dup (STDIN_FILENO); + int saved_stdout = dup (STDOUT_FILENO); + struct servent *rexecserv; + int result; + + /* When using cpio -o < filename, stdin is no longer the tty. But the + rexec subroutine reads the login and the passwd on stdin, to allow + remote execution of the command. So, reopen stdin and stdout on + /dev/tty before the rexec and give them back their original value + after. */ + + if (! freopen ("/dev/tty", "r", stdin)) + freopen ("/dev/null", "r", stdin); + if (! freopen ("/dev/tty", "w", stdout)) + freopen ("/dev/null", "w", stdout); + + if (rexecserv = getservbyname ("exec", "tcp"), !rexecserv) + error (EXIT_ON_EXEC_ERROR, 0, _("exec/tcp: Service not available")); + + result = rexec (&host, rexecserv->s_port, user, 0, rmt_command, 0); + if (fclose (stdin) == EOF) + error (0, errno, _("stdin")); + fdopen (saved_stdin, "r"); + if (fclose (stdout) == EOF) + error (0, errno, _("stdout")); + fdopen (saved_stdout, "w"); + + return result; +} + +#endif /* WITH_REXEC */ + +/* Place into BUF a string representing OFLAG, which must be suitable + as argument 2 of `open'. BUF must be large enough to hold the + result. This function should generate a string that decode_oflag + can parse. */ +static void +encode_oflag (char *buf, int oflag) +{ + sprintf (buf, "%d ", oflag); + + switch (oflag & O_ACCMODE) + { + case O_RDONLY: strcat (buf, "O_RDONLY"); break; + case O_RDWR: strcat (buf, "O_RDWR"); break; + case O_WRONLY: strcat (buf, "O_WRONLY"); break; + default: abort (); + } + +#ifdef O_APPEND + if (oflag & O_APPEND) strcat (buf, "|O_APPEND"); +#endif + if (oflag & O_CREAT) strcat (buf, "|O_CREAT"); +#ifdef O_DSYNC + if (oflag & O_DSYNC) strcat (buf, "|O_DSYNC"); +#endif + if (oflag & O_EXCL) strcat (buf, "|O_EXCL"); +#ifdef O_LARGEFILE + if (oflag & O_LARGEFILE) strcat (buf, "|O_LARGEFILE"); +#endif +#ifdef O_NOCTTY + if (oflag & O_NOCTTY) strcat (buf, "|O_NOCTTY"); +#endif + if (oflag & O_NONBLOCK) strcat (buf, "|O_NONBLOCK"); +#ifdef O_RSYNC + if (oflag & O_RSYNC) strcat (buf, "|O_RSYNC"); +#endif +#ifdef O_SYNC + if (oflag & O_SYNC) strcat (buf, "|O_SYNC"); +#endif + if (oflag & O_TRUNC) strcat (buf, "|O_TRUNC"); +} + +/* Open a file (a magnetic tape device?) on the system specified in + FILE_NAME, as the given user. FILE_NAME has the form `[USER@]HOST:FILE'. + OPEN_MODE is O_RDONLY, O_WRONLY, etc. If successful, return the + remote pipe number plus BIAS. REMOTE_SHELL may be overridden. On + error, return -1. */ +int +rmt_open__ (const char *file_name, int open_mode, int bias, + const char *remote_shell) +{ + int remote_pipe_number; /* pseudo, biased file descriptor */ + char *file_name_copy; /* copy of file_name string */ + char *remote_host; /* remote host name */ + char *remote_file; /* remote file name (often a device) */ + char *remote_user; /* remote user name */ + + /* Find an unused pair of file descriptors. */ + + for (remote_pipe_number = 0; + remote_pipe_number < MAXUNIT; + remote_pipe_number++) + if (READ_SIDE (remote_pipe_number) == -1 + && WRITE_SIDE (remote_pipe_number) == -1) + break; + + if (remote_pipe_number == MAXUNIT) + { + errno = EMFILE; + return -1; + } + + /* Pull apart the system and device, and optional user. */ + + { + char *cursor; + + file_name_copy = xstrdup (file_name); + remote_host = file_name_copy; + remote_user = 0; + remote_file = 0; + + for (cursor = file_name_copy; *cursor; cursor++) + switch (*cursor) + { + default: + break; + + case '\n': + /* Do not allow newlines in the file_name, since the protocol + uses newline delimiters. */ + free (file_name_copy); + errno = ENOENT; + return -1; + + case '@': + if (!remote_user) + { + remote_user = remote_host; + *cursor = '\0'; + remote_host = cursor + 1; + } + break; + + case ':': + if (!remote_file) + { + *cursor = '\0'; + remote_file = cursor + 1; + } + break; + } + } + + /* FIXME: Should somewhat validate the decoding, here. */ + + if (remote_user && *remote_user == '\0') + remote_user = 0; + +#if WITH_REXEC + + /* Execute the remote command using rexec. */ + + READ_SIDE (remote_pipe_number) = _rmt_rexec (remote_host, remote_user); + if (READ_SIDE (remote_pipe_number) < 0) + { + int e = errno; + free (file_name_copy); + errno = e; + return -1; + } + + WRITE_SIDE (remote_pipe_number) = READ_SIDE (remote_pipe_number); + +#else /* not WITH_REXEC */ + { + const char *remote_shell_basename; + pid_t status; + + /* Identify the remote command to be executed. */ + + if (!remote_shell) + { +#ifdef REMOTE_SHELL + remote_shell = REMOTE_SHELL; +#else + free (file_name_copy); + errno = EIO; + return -1; +#endif + } + remote_shell_basename = base_name (remote_shell); + + /* Set up the pipes for the `rsh' command, and fork. */ + + if (pipe (to_remote[remote_pipe_number]) == -1 + || pipe (from_remote[remote_pipe_number]) == -1) + { + int e = errno; + free (file_name_copy); + errno = e; + return -1; + } + + status = fork (); + if (status == -1) + { + int e = errno; + free (file_name_copy); + errno = e; + return -1; + } + + if (status == 0) + { + /* Child. */ + + close (STDIN_FILENO); + dup (to_remote[remote_pipe_number][PREAD]); + close (to_remote[remote_pipe_number][PREAD]); + close (to_remote[remote_pipe_number][PWRITE]); + + close (STDOUT_FILENO); + dup (from_remote[remote_pipe_number][PWRITE]); + close (from_remote[remote_pipe_number][PREAD]); + close (from_remote[remote_pipe_number][PWRITE]); + + sys_reset_uid_gid (); + + if (remote_user) + execl (remote_shell, remote_shell_basename, remote_host, + "-l", remote_user, rmt_command, (char *) 0); + else + execl (remote_shell, remote_shell_basename, remote_host, + rmt_command, (char *) 0); + + /* Bad problems if we get here. */ + + /* In a previous version, _exit was used here instead of exit. */ + error (EXIT_ON_EXEC_ERROR, errno, _("Cannot execute remote shell")); + } + + /* Parent. */ + + close (from_remote[remote_pipe_number][PWRITE]); + close (to_remote[remote_pipe_number][PREAD]); + } +#endif /* not WITH_REXEC */ + + /* Attempt to open the tape device. */ + + { + size_t remote_file_len = strlen (remote_file); + char *command_buffer = xmalloc (remote_file_len + 1000); + sprintf (command_buffer, "O%s\n", remote_file); + encode_oflag (command_buffer + remote_file_len + 2, open_mode); + strcat (command_buffer, "\n"); + if (do_command (remote_pipe_number, command_buffer) == -1 + || get_status (remote_pipe_number) == -1) + { + int e = errno; + free (command_buffer); + free (file_name_copy); + _rmt_shutdown (remote_pipe_number, e); + return -1; + } + free (command_buffer); + } + + free (file_name_copy); + return remote_pipe_number + bias; +} + +/* Close remote tape connection HANDLE and shut down. Return 0 if + successful, -1 on error. */ +int +rmt_close__ (int handle) +{ + long int status; + + if (do_command (handle, "C\n") == -1) + return -1; + + status = get_status (handle); + _rmt_shutdown (handle, errno); + return status; +} + +/* Read up to LENGTH bytes into BUFFER from remote tape connection HANDLE. + Return the number of bytes read on success, SAFE_READ_ERROR on error. */ +size_t +rmt_read__ (int handle, char *buffer, size_t length) +{ + char command_buffer[COMMAND_BUFFER_SIZE]; + size_t status; + size_t rlen; + size_t counter; + + sprintf (command_buffer, "R%lu\n", (unsigned long) length); + if (do_command (handle, command_buffer) == -1 + || (status = get_status (handle)) == SAFE_READ_ERROR) + return SAFE_READ_ERROR; + + for (counter = 0; counter < status; counter += rlen, buffer += rlen) + { + rlen = safe_read (READ_SIDE (handle), buffer, status - counter); + if (rlen == SAFE_READ_ERROR || rlen == 0) + { + _rmt_shutdown (handle, EIO); + return SAFE_READ_ERROR; + } + } + + return status; +} + +/* Write LENGTH bytes from BUFFER to remote tape connection HANDLE. + Return the number of bytes written. */ +size_t +rmt_write__ (int handle, char *buffer, size_t length) +{ + char command_buffer[COMMAND_BUFFER_SIZE]; + RETSIGTYPE (*pipe_handler) (); + size_t written; + + sprintf (command_buffer, "W%lu\n", (unsigned long) length); + if (do_command (handle, command_buffer) == -1) + return 0; + + pipe_handler = signal (SIGPIPE, SIG_IGN); + written = full_write (WRITE_SIDE (handle), buffer, length); + signal (SIGPIPE, pipe_handler); + if (written == length) + { + long int r = get_status (handle); + if (r < 0) + return 0; + if (r == length) + return length; + written = r; + } + + /* Write error. */ + + _rmt_shutdown (handle, EIO); + return written; +} + +/* Perform an imitation lseek operation on remote tape connection + HANDLE. Return the new file offset if successful, -1 if on error. */ +off_t +rmt_lseek__ (int handle, off_t offset, int whence) +{ + char command_buffer[COMMAND_BUFFER_SIZE]; + char operand_buffer[UINTMAX_STRSIZE_BOUND]; + uintmax_t u = offset < 0 ? - (uintmax_t) offset : (uintmax_t) offset; + char *p = operand_buffer + sizeof operand_buffer; + + *--p = 0; + do + *--p = '0' + (int) (u % 10); + while ((u /= 10) != 0); + if (offset < 0) + *--p = '-'; + + switch (whence) + { + case SEEK_SET: whence = 0; break; + case SEEK_CUR: whence = 1; break; + case SEEK_END: whence = 2; break; + default: abort (); + } + + sprintf (command_buffer, "L%s\n%d\n", p, whence); + + if (do_command (handle, command_buffer) == -1) + return -1; + + return get_status_off (handle); +} + +/* Perform a raw tape operation on remote tape connection HANDLE. + Return the results of the ioctl, or -1 on error. */ +int +rmt_ioctl__ (int handle, int operation, char *argument) +{ + switch (operation) + { + default: + errno = EOPNOTSUPP; + return -1; + +#ifdef MTIOCTOP + case MTIOCTOP: + { + char command_buffer[COMMAND_BUFFER_SIZE]; + char operand_buffer[UINTMAX_STRSIZE_BOUND]; + uintmax_t u = (((struct mtop *) argument)->mt_count < 0 + ? - (uintmax_t) ((struct mtop *) argument)->mt_count + : (uintmax_t) ((struct mtop *) argument)->mt_count); + char *p = operand_buffer + sizeof operand_buffer; + + *--p = 0; + do + *--p = '0' + (int) (u % 10); + while ((u /= 10) != 0); + if (((struct mtop *) argument)->mt_count < 0) + *--p = '-'; + + /* MTIOCTOP is the easy one. Nothing is transferred in binary. */ + + sprintf (command_buffer, "I%d\n%s\n", + ((struct mtop *) argument)->mt_op, p); + if (do_command (handle, command_buffer) == -1) + return -1; + + return get_status (handle); + } +#endif /* MTIOCTOP */ + +#ifdef MTIOCGET + case MTIOCGET: + { + ssize_t status; + size_t counter; + + /* Grab the status and read it directly into the structure. This + assumes that the status buffer is not padded and that 2 shorts + fit in a long without any word alignment problems; i.e., the + whole struct is contiguous. NOTE - this is probably NOT a good + assumption. */ + + if (do_command (handle, "S") == -1 + || (status = get_status (handle), status == -1)) + return -1; + + for (; status > 0; status -= counter, argument += counter) + { + counter = safe_read (READ_SIDE (handle), argument, status); + if (counter == SAFE_READ_ERROR || counter == 0) + { + _rmt_shutdown (handle, EIO); + return -1; + } + } + + /* Check for byte position. mt_type (or mt_model) is a small integer + field (normally) so we will check its magnitude. If it is larger + than 256, we will assume that the bytes are swapped and go through + and reverse all the bytes. */ + + if (((struct mtget *) argument)->MTIO_CHECK_FIELD < 256) + return 0; + + for (counter = 0; counter < status; counter += 2) + { + char copy = argument[counter]; + + argument[counter] = argument[counter + 1]; + argument[counter + 1] = copy; + } + + return 0; + } +#endif /* MTIOCGET */ + + } +} diff --git a/lib/safe-read.c b/lib/safe-read.c new file mode 100644 index 0000000..b7bf1d5 --- /dev/null +++ b/lib/safe-read.c @@ -0,0 +1,78 @@ +/* An interface to read and write that retries after interrupts. + + Copyright (C) 1993, 1994, 1998, 2002, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#ifdef SAFE_WRITE +# include "safe-write.h" +#else +# include "safe-read.h" +#endif + +/* Get ssize_t. */ +#include <sys/types.h> +#include <unistd.h> + +#include <errno.h> + +#ifdef EINTR +# define IS_EINTR(x) ((x) == EINTR) +#else +# define IS_EINTR(x) 0 +#endif + +#include <limits.h> + +#ifdef SAFE_WRITE +# define safe_rw safe_write +# define rw write +#else +# define safe_rw safe_read +# define rw read +# undef const +# define const /* empty */ +#endif + +/* Read(write) up to COUNT bytes at BUF from(to) descriptor FD, retrying if + interrupted. Return the actual number of bytes read(written), zero for EOF, + or SAFE_READ_ERROR(SAFE_WRITE_ERROR) upon error. */ +size_t +safe_rw (int fd, void const *buf, size_t count) +{ + /* Work around a bug in Tru64 5.1. Attempting to read more than + INT_MAX bytes fails with errno == EINVAL. See + <http://lists.gnu.org/archive/html/bug-gnu-utils/2002-04/msg00010.html>. + When decreasing COUNT, keep it block-aligned. */ + enum { BUGGY_READ_MAXIMUM = INT_MAX & ~8191 }; + + for (;;) + { + ssize_t result = rw (fd, buf, count); + + if (0 <= result) + return result; + else if (IS_EINTR (errno)) + continue; + else if (errno == EINVAL && BUGGY_READ_MAXIMUM < count) + count = BUGGY_READ_MAXIMUM; + else + return result; + } +} diff --git a/lib/safe-read.h b/lib/safe-read.h new file mode 100644 index 0000000..3451955 --- /dev/null +++ b/lib/safe-read.h @@ -0,0 +1,35 @@ +/* An interface to read() that retries after interrupts. + Copyright (C) 2002, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +#ifdef __cplusplus +extern "C" { +#endif + + +#define SAFE_READ_ERROR ((size_t) -1) + +/* Read up to COUNT bytes at BUF from descriptor FD, retrying if interrupted. + Return the actual number of bytes read, zero for EOF, or SAFE_READ_ERROR + upon error. */ +extern size_t safe_read (int fd, void *buf, size_t count); + + +#ifdef __cplusplus +} +#endif diff --git a/lib/safe-write.c b/lib/safe-write.c new file mode 100644 index 0000000..4c375a6 --- /dev/null +++ b/lib/safe-write.c @@ -0,0 +1,19 @@ +/* An interface to write that retries after interrupts. + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#define SAFE_WRITE +#include "safe-read.c" diff --git a/lib/safe-write.h b/lib/safe-write.h new file mode 100644 index 0000000..c194636 --- /dev/null +++ b/lib/safe-write.h @@ -0,0 +1,25 @@ +/* An interface to write() that retries after interrupts. + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +#define SAFE_WRITE_ERROR ((size_t) -1) + +/* Write up to COUNT bytes at BUF to descriptor FD, retrying if interrupted. + Return the actual number of bytes written, zero for EOF, or SAFE_WRITE_ERROR + upon error. */ +extern size_t safe_write (int fd, const void *buf, size_t count); diff --git a/lib/same-inode.h b/lib/same-inode.h new file mode 100644 index 0000000..deca87b --- /dev/null +++ b/lib/same-inode.h @@ -0,0 +1,26 @@ +/* Determine whether two stat buffers refer to the same file. + + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef SAME_INODE_H +# define SAME_INODE_H 1 + +# define SAME_INODE(Stat_buf_1, Stat_buf_2) \ + ((Stat_buf_1).st_ino == (Stat_buf_2).st_ino \ + && (Stat_buf_1).st_dev == (Stat_buf_2).st_dev) + +#endif diff --git a/lib/save-cwd.c b/lib/save-cwd.c new file mode 100644 index 0000000..f07973a --- /dev/null +++ b/lib/save-cwd.c @@ -0,0 +1,103 @@ +/* save-cwd.c -- Save and restore current working directory. + + Copyright (C) 1995, 1997, 1998, 2003, 2004, 2005, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "save-cwd.h" + +#include <errno.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> + +#include "chdir-long.h" +#include "fcntl--.h" +#include "xgetcwd.h" + +/* On systems without the fchdir function (WOE), pretend that open + always returns -1 so that save_cwd resorts to using xgetcwd. + Since chdir_long requires fchdir, use chdir instead. */ +#if !HAVE_FCHDIR +# undef open +# define open(File, Flags) (-1) +# undef fchdir +# define fchdir(Fd) (abort (), -1) +# undef chdir_long +# define chdir_long(Dir) chdir (Dir) +#endif + +/* Record the location of the current working directory in CWD so that + the program may change to other directories and later use restore_cwd + to return to the recorded location. This function may allocate + space using malloc (via xgetcwd) or leave a file descriptor open; + use free_cwd to perform the necessary free or close. Upon failure, + no memory is allocated, any locally opened file descriptors are + closed; return non-zero -- in that case, free_cwd need not be + called, but doing so is ok. Otherwise, return zero. + + The `raison d'etre' for this interface is that the working directory + is sometimes inaccessible, and getcwd is not robust or as efficient. + So, we prefer to use the open/fchdir approach, but fall back on + getcwd if necessary. + + Some systems lack fchdir altogether: e.g., OS/2, pre-2001 Cygwin, + SCO Xenix. Also, SunOS 4 and Irix 5.3 provide the function, yet it + doesn't work for partitions on which auditing is enabled. If + you're still using an obsolete system with these problems, please + send email to the maintainer of this code. */ + +int +save_cwd (struct saved_cwd *cwd) +{ + cwd->name = NULL; + + cwd->desc = open (".", O_RDONLY); + if (cwd->desc < 0) + { + cwd->name = xgetcwd (); + return cwd->name ? 0 : -1; + } + + return 0; +} + +/* Change to recorded location, CWD, in directory hierarchy. + Upon failure, return -1 (errno is set by chdir or fchdir). + Upon success, return zero. */ + +int +restore_cwd (const struct saved_cwd *cwd) +{ + if (0 <= cwd->desc) + return fchdir (cwd->desc); + else + return chdir_long (cwd->name); +} + +void +free_cwd (struct saved_cwd *cwd) +{ + if (cwd->desc >= 0) + close (cwd->desc); + if (cwd->name) + free (cwd->name); +} diff --git a/lib/save-cwd.h b/lib/save-cwd.h new file mode 100644 index 0000000..d646b55 --- /dev/null +++ b/lib/save-cwd.h @@ -0,0 +1,34 @@ +/* Save and restore current working directory. + + Copyright (C) 1995, 1997, 1998, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#ifndef SAVE_CWD_H +# define SAVE_CWD_H 1 + +struct saved_cwd + { + int desc; + char *name; + }; + +int save_cwd (struct saved_cwd *cwd); +int restore_cwd (const struct saved_cwd *cwd); +void free_cwd (struct saved_cwd *cwd); + +#endif /* SAVE_CWD_H */ diff --git a/lib/savedir.c b/lib/savedir.c new file mode 100644 index 0000000..d930fb4 --- /dev/null +++ b/lib/savedir.c @@ -0,0 +1,137 @@ +/* savedir.c -- save the list of files in a directory in a string + + Copyright (C) 1990, 1997, 1998, 1999, 2000, 2001, 2003, 2004, 2005, + 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#include <config.h> + +#include "savedir.h" + +#include <sys/types.h> + +#include <errno.h> + +#include <dirent.h> +#ifndef _D_EXACT_NAMLEN +# define _D_EXACT_NAMLEN(dp) strlen ((dp)->d_name) +#endif + +#include <stddef.h> +#include <stdlib.h> +#include <string.h> + +#include "openat.h" +#include "xalloc.h" + +#ifndef NAME_SIZE_DEFAULT +# define NAME_SIZE_DEFAULT 512 +#endif + +/* The results of opendir() in this file are not used with dirfd and fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef opendir +#undef closedir + +/* Return a freshly allocated string containing the file names + in directory DIRP, separated by '\0' characters; + the end is marked by two '\0' characters in a row. + Return NULL (setting errno) if DIRP cannot be read or closed. + If DIRP is NULL, return NULL without affecting errno. */ + +static char * +savedirstream (DIR *dirp) +{ + char *name_space; + size_t allocated = NAME_SIZE_DEFAULT; + size_t used = 0; + int save_errno; + + if (dirp == NULL) + return NULL; + + name_space = xmalloc (allocated); + + for (;;) + { + struct dirent const *dp; + char const *entry; + + errno = 0; + dp = readdir (dirp); + if (! dp) + break; + + /* Skip "", ".", and "..". "" is returned by at least one buggy + implementation: Solaris 2.4 readdir on NFS file systems. */ + entry = dp->d_name; + if (entry[entry[0] != '.' ? 0 : entry[1] != '.' ? 1 : 2] != '\0') + { + size_t entry_size = _D_EXACT_NAMLEN (dp) + 1; + if (used + entry_size < used) + xalloc_die (); + if (allocated <= used + entry_size) + { + do + { + if (2 * allocated < allocated) + xalloc_die (); + allocated *= 2; + } + while (allocated <= used + entry_size); + + name_space = xrealloc (name_space, allocated); + } + memcpy (name_space + used, entry, entry_size); + used += entry_size; + } + } + name_space[used] = '\0'; + save_errno = errno; + if (closedir (dirp) != 0) + save_errno = errno; + if (save_errno != 0) + { + free (name_space); + errno = save_errno; + return NULL; + } + return name_space; +} + +/* Return a freshly allocated string containing the file names + in directory DIR, separated by '\0' characters; + the end is marked by two '\0' characters in a row. + Return NULL (setting errno) if DIR cannot be opened, read, or closed. */ + +char * +savedir (char const *dir) +{ + return savedirstream (opendir (dir)); +} + +/* Return a freshly allocated string containing the file names + in directory FD, separated by '\0' characters; + the end is marked by two '\0' characters in a row. + Return NULL (setting errno) if FD cannot be read or closed. */ + +char * +fdsavedir (int fd) +{ + return savedirstream (fdopendir (fd)); +} diff --git a/lib/savedir.h b/lib/savedir.h new file mode 100644 index 0000000..5b7bef9 --- /dev/null +++ b/lib/savedir.h @@ -0,0 +1,27 @@ +/* Save the list of files in a directory in a string. + + Copyright (C) 1997, 1999, 2001, 2003, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#if !defined SAVEDIR_H_ +# define SAVEDIR_H_ + +char *savedir (char const *dir); +char *fdsavedir (int fd); + +#endif diff --git a/lib/setenv.c b/lib/setenv.c new file mode 100644 index 0000000..c54c28d --- /dev/null +++ b/lib/setenv.c @@ -0,0 +1,332 @@ +/* Copyright (C) 1992,1995-1999,2000-2003,2005-2007 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if !_LIBC +# include <config.h> +#endif +#include <alloca.h> + +#include <errno.h> +#ifndef __set_errno +# define __set_errno(ev) ((errno) = (ev)) +#endif + +#include <stdlib.h> +#include <string.h> +#if _LIBC || HAVE_UNISTD_H +# include <unistd.h> +#endif + +#if _LIBC || !HAVE_SETENV + +#if !_LIBC +# include "allocsa.h" +#endif + +#if !_LIBC +# define __environ environ +# ifndef HAVE_ENVIRON_DECL +extern char **environ; +# endif +#endif + +#if _LIBC +/* This lock protects against simultaneous modifications of `environ'. */ +# include <bits/libc-lock.h> +__libc_lock_define_initialized (static, envlock) +# define LOCK __libc_lock_lock (envlock) +# define UNLOCK __libc_lock_unlock (envlock) +#else +# define LOCK +# define UNLOCK +#endif + +/* In the GNU C library we must keep the namespace clean. */ +#ifdef _LIBC +# define setenv __setenv +# define clearenv __clearenv +# define tfind __tfind +# define tsearch __tsearch +#endif + +/* In the GNU C library implementation we try to be more clever and + allow arbitrarily many changes of the environment given that the used + values are from a small set. Outside glibc this will eat up all + memory after a while. */ +#if defined _LIBC || (defined HAVE_SEARCH_H && defined HAVE_TSEARCH \ + && defined __GNUC__) +# define USE_TSEARCH 1 +# include <search.h> +typedef int (*compar_fn_t) (const void *, const void *); + +/* This is a pointer to the root of the search tree with the known + values. */ +static void *known_values; + +# define KNOWN_VALUE(Str) \ + ({ \ + void *value = tfind (Str, &known_values, (compar_fn_t) strcmp); \ + value != NULL ? *(char **) value : NULL; \ + }) +# define STORE_VALUE(Str) \ + tsearch (Str, &known_values, (compar_fn_t) strcmp) + +#else +# undef USE_TSEARCH + +# define KNOWN_VALUE(Str) NULL +# define STORE_VALUE(Str) do { } while (0) + +#endif + + +/* If this variable is not a null pointer we allocated the current + environment. */ +static char **last_environ; + + +/* This function is used by `setenv' and `putenv'. The difference between + the two functions is that for the former must create a new string which + is then placed in the environment, while the argument of `putenv' + must be used directly. This is all complicated by the fact that we try + to reuse values once generated for a `setenv' call since we can never + free the strings. */ +int +__add_to_environ (const char *name, const char *value, const char *combined, + int replace) +{ + register char **ep; + register size_t size; + const size_t namelen = strlen (name); + const size_t vallen = value != NULL ? strlen (value) + 1 : 0; + + LOCK; + + /* We have to get the pointer now that we have the lock and not earlier + since another thread might have created a new environment. */ + ep = __environ; + + size = 0; + if (ep != NULL) + { + for (; *ep != NULL; ++ep) + if (!strncmp (*ep, name, namelen) && (*ep)[namelen] == '=') + break; + else + ++size; + } + + if (ep == NULL || *ep == NULL) + { + char **new_environ; +#ifdef USE_TSEARCH + char *new_value; +#endif + + /* We allocated this space; we can extend it. */ + new_environ = + (char **) (last_environ == NULL + ? malloc ((size + 2) * sizeof (char *)) + : realloc (last_environ, (size + 2) * sizeof (char *))); + if (new_environ == NULL) + { + UNLOCK; + return -1; + } + + /* If the whole entry is given add it. */ + if (combined != NULL) + /* We must not add the string to the search tree since it belongs + to the user. */ + new_environ[size] = (char *) combined; + else + { + /* See whether the value is already known. */ +#ifdef USE_TSEARCH +# ifdef _LIBC + new_value = (char *) alloca (namelen + 1 + vallen); + __mempcpy (__mempcpy (__mempcpy (new_value, name, namelen), "=", 1), + value, vallen); +# else + new_value = (char *) allocsa (namelen + 1 + vallen); + if (new_value == NULL) + { + __set_errno (ENOMEM); + UNLOCK; + return -1; + } + memcpy (new_value, name, namelen); + new_value[namelen] = '='; + memcpy (&new_value[namelen + 1], value, vallen); +# endif + + new_environ[size] = KNOWN_VALUE (new_value); + if (new_environ[size] == NULL) +#endif + { + new_environ[size] = (char *) malloc (namelen + 1 + vallen); + if (new_environ[size] == NULL) + { +#if defined USE_TSEARCH && !defined _LIBC + freesa (new_value); +#endif + __set_errno (ENOMEM); + UNLOCK; + return -1; + } + +#ifdef USE_TSEARCH + memcpy (new_environ[size], new_value, namelen + 1 + vallen); +#else + memcpy (new_environ[size], name, namelen); + new_environ[size][namelen] = '='; + memcpy (&new_environ[size][namelen + 1], value, vallen); +#endif + /* And save the value now. We cannot do this when we remove + the string since then we cannot decide whether it is a + user string or not. */ + STORE_VALUE (new_environ[size]); + } +#if defined USE_TSEARCH && !defined _LIBC + freesa (new_value); +#endif + } + + if (__environ != last_environ) + memcpy ((char *) new_environ, (char *) __environ, + size * sizeof (char *)); + + new_environ[size + 1] = NULL; + + last_environ = __environ = new_environ; + } + else if (replace) + { + char *np; + + /* Use the user string if given. */ + if (combined != NULL) + np = (char *) combined; + else + { +#ifdef USE_TSEARCH + char *new_value; +# ifdef _LIBC + new_value = alloca (namelen + 1 + vallen); + __mempcpy (__mempcpy (__mempcpy (new_value, name, namelen), "=", 1), + value, vallen); +# else + new_value = allocsa (namelen + 1 + vallen); + if (new_value == NULL) + { + __set_errno (ENOMEM); + UNLOCK; + return -1; + } + memcpy (new_value, name, namelen); + new_value[namelen] = '='; + memcpy (&new_value[namelen + 1], value, vallen); +# endif + + np = KNOWN_VALUE (new_value); + if (np == NULL) +#endif + { + np = malloc (namelen + 1 + vallen); + if (np == NULL) + { +#if defined USE_TSEARCH && !defined _LIBC + freesa (new_value); +#endif + __set_errno (ENOMEM); + UNLOCK; + return -1; + } + +#ifdef USE_TSEARCH + memcpy (np, new_value, namelen + 1 + vallen); +#else + memcpy (np, name, namelen); + np[namelen] = '='; + memcpy (&np[namelen + 1], value, vallen); +#endif + /* And remember the value. */ + STORE_VALUE (np); + } +#if defined USE_TSEARCH && !defined _LIBC + freesa (new_value); +#endif + } + + *ep = np; + } + + UNLOCK; + + return 0; +} + +int +setenv (const char *name, const char *value, int replace) +{ + return __add_to_environ (name, value, NULL, replace); +} + +/* The `clearenv' was planned to be added to POSIX.1 but probably + never made it. Nevertheless the POSIX.9 standard (POSIX bindings + for Fortran 77) requires this function. */ +int +clearenv (void) +{ + LOCK; + + if (__environ == last_environ && __environ != NULL) + { + /* We allocated this environment so we can free it. */ + free (__environ); + last_environ = NULL; + } + + /* Clear the environment pointer removes the whole environment. */ + __environ = NULL; + + UNLOCK; + + return 0; +} + +#ifdef _LIBC +static void +free_mem (void) +{ + /* Remove all traces. */ + clearenv (); + + /* Now remove the search tree. */ + __tdestroy (known_values, free); + known_values = NULL; +} +text_set_element (__libc_subfreeres, free_mem); + + +# undef setenv +# undef clearenv +weak_alias (__setenv, setenv) +weak_alias (__clearenv, clearenv) +#endif + +#endif /* _LIBC || !HAVE_SETENV */ diff --git a/lib/setenv.h b/lib/setenv.h new file mode 100644 index 0000000..92e7bba --- /dev/null +++ b/lib/setenv.h @@ -0,0 +1,54 @@ +/* Setting environment variables. + Copyright (C) 2001-2004, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if HAVE_SETENV || HAVE_UNSETENV + +/* Get setenv(), unsetenv() declarations. */ +# include <stdlib.h> + +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +#if !HAVE_SETENV + +/* Set NAME to VALUE in the environment. + If REPLACE is nonzero, overwrite an existing value. */ +extern int setenv (const char *name, const char *value, int replace); + +#endif + +#if HAVE_UNSETENV + +# if VOID_UNSETENV +/* On some systems, unsetenv() returns void. + This is the case for FreeBSD 4.8, NetBSD 1.6, OpenBSD 3.4. */ +# define unsetenv(name) ((unsetenv)(name), 0) +# endif + +#else + +/* Remove the variable NAME from the environment. */ +extern int unsetenv (const char *name); + +#endif + +#ifdef __cplusplus +} +#endif diff --git a/lib/sleep.c b/lib/sleep.c new file mode 100644 index 0000000..037559b --- /dev/null +++ b/lib/sleep.c @@ -0,0 +1,47 @@ +/* Pausing execution of the current thread. + Copyright (C) 2007 Free Software Foundation, Inc. + Written by Bruno Haible <bruno@clisp.org>, 2007. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <unistd.h> + +#if (defined _WIN32 || defined __WIN32__) && ! defined __CYGWIN__ + +# define WIN32_LEAN_AND_MEAN /* avoid including junk */ +# include <windows.h> + +unsigned int +sleep (unsigned int seconds) +{ + unsigned int remaining; + + /* Sleep for 1 second many times, because + 1. Sleep is not interruptiple by Ctrl-C, + 2. we want to avoid arithmetic overflow while multiplying with 1000. */ + for (remaining = seconds; remaining > 0; remaining--) + Sleep (1000); + + return remaining; +} + +#else + + #error "Please port gnulib sleep.c to your platform, possibly using usleep() or select(), then report this to bug-gnulib." + +#endif diff --git a/lib/stat-macros.h b/lib/stat-macros.h new file mode 100644 index 0000000..690216c --- /dev/null +++ b/lib/stat-macros.h @@ -0,0 +1,3 @@ +/* All the mode bits that can be affected by chmod. */ +#define CHMOD_MODE_BITS \ + (S_ISUID | S_ISGID | S_ISVTX | S_IRWXU | S_IRWXG | S_IRWXO) diff --git a/lib/stat-time.h b/lib/stat-time.h new file mode 100644 index 0000000..649c848 --- /dev/null +++ b/lib/stat-time.h @@ -0,0 +1,184 @@ +/* stat-related time functions. + + Copyright (C) 2005, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#ifndef STAT_TIME_H +#define STAT_TIME_H 1 + +#include <sys/stat.h> +#include <time.h> + +/* STAT_TIMESPEC (ST, ST_XTIM) is the ST_XTIM member for *ST of type + struct timespec, if available. If not, then STAT_TIMESPEC_NS (ST, + ST_XTIM) is the nanosecond component of the ST_XTIM member for *ST, + if available. ST_XTIM can be st_atim, st_ctim, st_mtim, or st_birthtim + for access, status change, data modification, or birth (creation) + time respectively. + + These macros are private to stat-time.h. */ +#if defined HAVE_STRUCT_STAT_ST_ATIM_TV_NSEC +# ifdef TYPEOF_STRUCT_STAT_ST_ATIM_IS_STRUCT_TIMESPEC +# define STAT_TIMESPEC(st, st_xtim) ((st)->st_xtim) +# else +# define STAT_TIMESPEC_NS(st, st_xtim) ((st)->st_xtim.tv_nsec) +# endif +#elif defined HAVE_STRUCT_STAT_ST_ATIMESPEC_TV_NSEC +# define STAT_TIMESPEC(st, st_xtim) ((st)->st_xtim##espec) +#elif defined HAVE_STRUCT_STAT_ST_ATIMENSEC +# define STAT_TIMESPEC_NS(st, st_xtim) ((st)->st_xtim##ensec) +#elif defined HAVE_STRUCT_STAT_ST_ATIM_ST__TIM_TV_NSEC +# define STAT_TIMESPEC_NS(st, st_xtim) ((st)->st_xtim.st__tim.tv_nsec) +#endif + +/* Return the nanosecond component of *ST's access time. */ +static inline long int +get_stat_atime_ns (struct stat const *st) +{ +# if defined STAT_TIMESPEC + return STAT_TIMESPEC (st, st_atim).tv_nsec; +# elif defined STAT_TIMESPEC_NS + return STAT_TIMESPEC_NS (st, st_atim); +# else + return 0; +# endif +} + +/* Return the nanosecond component of *ST's status change time. */ +static inline long int +get_stat_ctime_ns (struct stat const *st) +{ +# if defined STAT_TIMESPEC + return STAT_TIMESPEC (st, st_ctim).tv_nsec; +# elif defined STAT_TIMESPEC_NS + return STAT_TIMESPEC_NS (st, st_ctim); +# else + return 0; +# endif +} + +/* Return the nanosecond component of *ST's data modification time. */ +static inline long int +get_stat_mtime_ns (struct stat const *st) +{ +# if defined STAT_TIMESPEC + return STAT_TIMESPEC (st, st_mtim).tv_nsec; +# elif defined STAT_TIMESPEC_NS + return STAT_TIMESPEC_NS (st, st_mtim); +# else + return 0; +# endif +} + +/* Return the nanosecond component of *ST's birth time. */ +static inline long int +get_stat_birthtime_ns (struct stat const *st) +{ +# if defined HAVE_STRUCT_STAT_ST_BIRTHTIMESPEC_TV_NSEC + return STAT_TIMESPEC (st, st_birthtim).tv_nsec; +# elif defined HAVE_STRUCT_STAT_ST_BIRTHTIMENSEC + return STAT_TIMESPEC_NS (st, st_birthtim); +# else + return 0; +# endif +} + +/* Return *ST's access time. */ +static inline struct timespec +get_stat_atime (struct stat const *st) +{ +#ifdef STAT_TIMESPEC + return STAT_TIMESPEC (st, st_atim); +#else + struct timespec t; + t.tv_sec = st->st_atime; + t.tv_nsec = get_stat_atime_ns (st); + return t; +#endif +} + +/* Return *ST's status change time. */ +static inline struct timespec +get_stat_ctime (struct stat const *st) +{ +#ifdef STAT_TIMESPEC + return STAT_TIMESPEC (st, st_ctim); +#else + struct timespec t; + t.tv_sec = st->st_ctime; + t.tv_nsec = get_stat_ctime_ns (st); + return t; +#endif +} + +/* Return *ST's data modification time. */ +static inline struct timespec +get_stat_mtime (struct stat const *st) +{ +#ifdef STAT_TIMESPEC + return STAT_TIMESPEC (st, st_mtim); +#else + struct timespec t; + t.tv_sec = st->st_mtime; + t.tv_nsec = get_stat_mtime_ns (st); + return t; +#endif +} + +/* Return *ST's birth time, if available; otherwise return a value + with negative tv_nsec. */ +static inline struct timespec +get_stat_birthtime (struct stat const *st) +{ + struct timespec t; + +#if (defined HAVE_STRUCT_STAT_ST_BIRTHTIMESPEC_TV_NSEC \ + || defined HAVE_STRUCT_STAT_ST_BIRTHTIM_TV_NSEC) + t = STAT_TIMESPEC (st, st_birthtim); +#elif defined HAVE_STRUCT_STAT_ST_BIRTHTIMENSEC + t.tv_sec = st->st_birthtime; + t.tv_nsec = st->st_birthtimensec; +#elif (defined _WIN32 || defined __WIN32__) && ! defined __CYGWIN__ + /* Woe32 native platforms (but not Cygwin) put the "file creation + time" in st_ctime (!). See + <http://msdn2.microsoft.com/de-de/library/14h5k7ff(VS.80).aspx>. */ + t.tv_sec = st->st_ctime; + t.tv_nsec = 0; +#else + /* Birth time is not supported. Set tv_sec to avoid undefined behavior. */ + t.tv_sec = -1; + t.tv_nsec = -1; +#endif + +#if (defined HAVE_STRUCT_STAT_ST_BIRTHTIMESPEC_TV_NSEC \ + || defined HAVE_STRUCT_STAT_ST_BIRTHTIM_TV_NSEC \ + || defined HAVE_STRUCT_STAT_ST_BIRTHTIMENSEC) + /* FreeBSD and NetBSD sometimes signal the absence of knowledge by + using zero. Attempt to work around this problem. Alas, this can + report failure even for valid time stamps. Also, NetBSD + sometimes returns junk in the birth time fields; work around this + bug if it it is detected. There's no need to detect negative + tv_nsec junk as negative tv_nsec already indicates an error. */ + if (t.tv_sec == 0 || 1000000000 <= t.tv_nsec) + t.tv_nsec = -1; +#endif + + return t; +} + +#endif diff --git a/lib/stdbool_.h b/lib/stdbool_.h new file mode 100644 index 0000000..150a010 --- /dev/null +++ b/lib/stdbool_.h @@ -0,0 +1,118 @@ +/* Copyright (C) 2001, 2002, 2003, 2006, 2007 Free Software Foundation, Inc. + Written by Bruno Haible <haible@clisp.cons.org>, 2001. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_STDBOOL_H +#define _GL_STDBOOL_H + +/* ISO C 99 <stdbool.h> for platforms that lack it. */ + +/* Usage suggestions: + + Programs that use <stdbool.h> should be aware of some limitations + and standards compliance issues. + + Standards compliance: + + - <stdbool.h> must be #included before 'bool', 'false', 'true' + can be used. + + - You cannot assume that sizeof (bool) == 1. + + - Programs should not undefine the macros bool, true, and false, + as C99 lists that as an "obsolescent feature". + + Limitations of this substitute, when used in a C89 environment: + + - <stdbool.h> must be #included before the '_Bool' type can be used. + + - You cannot assume that _Bool is a typedef; it might be a macro. + + - Bit-fields of type 'bool' are not supported. Portable code + should use 'unsigned int foo : 1;' rather than 'bool foo : 1;'. + + - In C99, casts and automatic conversions to '_Bool' or 'bool' are + performed in such a way that every nonzero value gets converted + to 'true', and zero gets converted to 'false'. This doesn't work + with this substitute. With this substitute, only the values 0 and 1 + give the expected result when converted to _Bool' or 'bool'. + + Also, it is suggested that programs use 'bool' rather than '_Bool'; + this isn't required, but 'bool' is more common. */ + + +/* 7.16. Boolean type and values */ + +/* BeOS <sys/socket.h> already #defines false 0, true 1. We use the same + definitions below, but temporarily we have to #undef them. */ +#ifdef __BEOS__ +# include <OS.h> /* defines bool but not _Bool */ +# undef false +# undef true +#endif + +/* For the sake of symbolic names in gdb, we define true and false as + enum constants, not only as macros. + It is tempting to write + typedef enum { false = 0, true = 1 } _Bool; + so that gdb prints values of type 'bool' symbolically. But if we do + this, values of type '_Bool' may promote to 'int' or 'unsigned int' + (see ISO C 99 6.7.2.2.(4)); however, '_Bool' must promote to 'int' + (see ISO C 99 6.3.1.1.(2)). So we add a negative value to the + enum; this ensures that '_Bool' promotes to 'int'. */ +#if defined __cplusplus || defined __BEOS__ + /* A compiler known to have 'bool'. */ + /* If the compiler already has both 'bool' and '_Bool', we can assume they + are the same types. */ +# if !@HAVE__BOOL@ +typedef bool _Bool; +# endif +#else +# if !defined __GNUC__ + /* If @HAVE__BOOL@: + Some HP-UX cc and AIX IBM C compiler versions have compiler bugs when + the built-in _Bool type is used. See + http://gcc.gnu.org/ml/gcc-patches/2003-12/msg02303.html + http://lists.gnu.org/archive/html/bug-coreutils/2005-11/msg00161.html + http://lists.gnu.org/archive/html/bug-coreutils/2005-10/msg00086.html + Similar bugs are likely with other compilers as well; this file + wouldn't be used if <stdbool.h> was working. + So we override the _Bool type. + If !@HAVE__BOOL@: + Need to define _Bool ourselves. As 'signed char' or as an enum type? + Use of a typedef, with SunPRO C, leads to a stupid + "warning: _Bool is a keyword in ISO C99". + Use of an enum type, with IRIX cc, leads to a stupid + "warning(1185): enumerated type mixed with another type". + The only benefit of the enum type, debuggability, is not important + with these compilers. So use 'signed char' and no typedef. */ +# define _Bool signed char +enum { false = 0, true = 1 }; +# else + /* With this compiler, trust the _Bool type if the compiler has it. */ +# if !@HAVE__BOOL@ +typedef enum { _Bool_must_promote_to_int = -1, false = 0, true = 1 } _Bool; +# endif +# endif +#endif +#define bool _Bool + +/* The other macros must be usable in preprocessor directives. */ +#define false 0 +#define true 1 +#define __bool_true_false_are_defined 1 + +#endif /* _GL_STDBOOL_H */ diff --git a/lib/stdint_.h b/lib/stdint_.h new file mode 100644 index 0000000..d374535 --- /dev/null +++ b/lib/stdint_.h @@ -0,0 +1,512 @@ +/* Copyright (C) 2001-2002, 2004-2007 Free Software Foundation, Inc. + Written by Paul Eggert, Bruno Haible, Sam Steingold, Peter Burwood. + This file is part of gnulib. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* + * ISO C 99 <stdint.h> for platforms that lack it. + * <http://www.opengroup.org/susv3xbd/stdint.h.html> + */ + +#ifndef _GL_STDINT_H + +/* Get those types that are already defined in other system include + files, so that we can "#define int8_t signed char" below without + worrying about a later system include file containing a "typedef + signed char int8_t;" that will get messed up by our macro. Our + macros should all be consistent with the system versions, except + for the "fast" types and macros, which we recommend against using + in public interfaces due to compiler differences. */ + +#if @HAVE_STDINT_H@ +# if defined __sgi && ! defined __c99 + /* Bypass IRIX's <stdint.h> if in C89 mode, since it merely annoys users + with "This header file is to be used only for c99 mode compilations" + diagnostics. */ +# define __STDINT_H__ +# endif + /* Other systems may have an incomplete or buggy <stdint.h>. + Include it before <inttypes.h>, since any "#include <stdint.h>" + in <inttypes.h> would reinclude us, skipping our contents because + _GL_STDINT_H is defined. + The include_next requires a split double-inclusion guard. */ +# if @HAVE_INCLUDE_NEXT@ +# include_next <stdint.h> +# else +# include @ABSOLUTE_STDINT_H@ +# endif +#endif + +#ifndef _GL_STDINT_H +#define _GL_STDINT_H + +/* <sys/types.h> defines some of the stdint.h types as well, on glibc, + IRIX 6.5, and OpenBSD 3.8 (via <machine/types.h>). + AIX 5.2 <sys/types.h> isn't needed and causes troubles. + MacOS X 10.4.6 <sys/types.h> includes <stdint.h> (which is us), but + relies on the system <stdint.h> definitions, so include + <sys/types.h> after @ABSOLUTE_STDINT_H@. */ +#if @HAVE_SYS_TYPES_H@ && ! defined _AIX +# include <sys/types.h> +#endif + +/* Get LONG_MIN, LONG_MAX, ULONG_MAX. */ +#include <limits.h> + +#if @HAVE_INTTYPES_H@ + /* In OpenBSD 3.8, <inttypes.h> includes <machine/types.h>, which defines + int{8,16,32,64}_t, uint{8,16,32,64}_t and __BIT_TYPES_DEFINED__. + <inttypes.h> also defines intptr_t and uintptr_t. */ +# define _GL_JUST_INCLUDE_ABSOLUTE_INTTYPES_H +# include <inttypes.h> +# undef _GL_JUST_INCLUDE_ABSOLUTE_INTTYPES_H +#elif @HAVE_SYS_INTTYPES_H@ + /* Solaris 7 <sys/inttypes.h> has the types except the *_fast*_t types, and + the macros except for *_FAST*_*, INTPTR_MIN, PTRDIFF_MIN, PTRDIFF_MAX. */ +# include <sys/inttypes.h> +#endif + +#if @HAVE_SYS_BITYPES_H@ && ! defined __BIT_TYPES_DEFINED__ + /* Linux libc4 >= 4.6.7 and libc5 have a <sys/bitypes.h> that defines + int{8,16,32,64}_t and __BIT_TYPES_DEFINED__. In libc5 >= 5.2.2 it is + included by <sys/types.h>. */ +# include <sys/bitypes.h> +#endif + +#if ! defined __cplusplus || defined __STDC_CONSTANT_MACROS + +/* Get WCHAR_MIN, WCHAR_MAX. */ +# if ! (defined WCHAR_MIN && defined WCHAR_MAX) +# include <wchar.h> +# endif + +#endif + +/* Minimum and maximum values for a integer type under the usual assumption. + Return an unspecified value if BITS == 0, adding a check to pacify + picky compilers. */ + +#define _STDINT_MIN(signed, bits, zero) \ + ((signed) ? (- ((zero) + 1) << ((bits) ? (bits) - 1 : 0)) : (zero)) + +#define _STDINT_MAX(signed, bits, zero) \ + ((signed) \ + ? ~ _STDINT_MIN (signed, bits, zero) \ + : ((((zero) + 1) << ((bits) ? (bits) - 1 : 0)) - 1) * 2 + 1) + +/* 7.18.1.1. Exact-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. */ + +#undef int8_t +#undef uint8_t +#define int8_t signed char +#define uint8_t unsigned char + +#undef int16_t +#undef uint16_t +#define int16_t short int +#define uint16_t unsigned short int + +#undef int32_t +#undef uint32_t +#define int32_t int +#define uint32_t unsigned int + +/* Do not undefine int64_t if gnulib is not being used with 64-bit + types, since otherwise it breaks platforms like Tandem/NSK. */ +#if LONG_MAX >> 31 >> 31 == 1 +# undef int64_t +# define int64_t long int +# define GL_INT64_T +#elif defined _MSC_VER +# undef int64_t +# define int64_t __int64 +# define GL_INT64_T +#elif @HAVE_LONG_LONG_INT@ +# undef int64_t +# define int64_t long long int +# define GL_INT64_T +#endif + +#if ULONG_MAX >> 31 >> 31 >> 1 == 1 +# undef uint64_t +# define uint64_t unsigned long int +# define GL_UINT64_T +#elif defined _MSC_VER +# undef uint64_t +# define uint64_t unsigned __int64 +# define GL_UINT64_T +#elif @HAVE_UNSIGNED_LONG_LONG_INT@ +# undef uint64_t +# define uint64_t unsigned long long int +# define GL_UINT64_T +#endif + +/* Avoid collision with Solaris 2.5.1 <pthread.h> etc. */ +#define _UINT8_T +#define _UINT32_T +#define _UINT64_T + + +/* 7.18.1.2. Minimum-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the leastN_t types + are the same as the corresponding N_t types. */ + +#undef int_least8_t +#undef uint_least8_t +#undef int_least16_t +#undef uint_least16_t +#undef int_least32_t +#undef uint_least32_t +#undef int_least64_t +#undef uint_least64_t +#define int_least8_t int8_t +#define uint_least8_t uint8_t +#define int_least16_t int16_t +#define uint_least16_t uint16_t +#define int_least32_t int32_t +#define uint_least32_t uint32_t +#ifdef GL_INT64_T +# define int_least64_t int64_t +#endif +#ifdef GL_UINT64_T +# define uint_least64_t uint64_t +#endif + +/* 7.18.1.3. Fastest minimum-width integer types */ + +/* Note: Other <stdint.h> substitutes may define these types differently. + It is not recommended to use these types in public header files. */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the fastN_t types + are taken from the same list of types. Assume that 'long int' + is fast enough for all narrower integers. */ + +#undef int_fast8_t +#undef uint_fast8_t +#undef int_fast16_t +#undef uint_fast16_t +#undef int_fast32_t +#undef uint_fast32_t +#undef int_fast64_t +#undef uint_fast64_t +#define int_fast8_t long int +#define uint_fast8_t unsigned int_fast8_t +#define int_fast16_t long int +#define uint_fast16_t unsigned int_fast16_t +#define int_fast32_t long int +#define uint_fast32_t unsigned int_fast32_t +#ifdef GL_INT64_T +# define int_fast64_t int64_t +#endif +#ifdef GL_UINT64_T +# define uint_fast64_t uint64_t +#endif + +/* 7.18.1.4. Integer types capable of holding object pointers */ + +#undef intptr_t +#undef uintptr_t +#define intptr_t long int +#define uintptr_t unsigned long int + +/* 7.18.1.5. Greatest-width integer types */ + +/* Note: These types are compiler dependent. It may be unwise to use them in + public header files. */ + +#undef intmax_t +#if @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define intmax_t long long int +#elif defined GL_INT64_T +# define intmax_t int64_t +#else +# define intmax_t long int +#endif + +#undef uintmax_t +#if @HAVE_UNSIGNED_LONG_LONG_INT@ && ULONG_MAX >> 31 == 1 +# define uintmax_t unsigned long long int +#elif defined GL_UINT64_T +# define uintmax_t uint64_t +#else +# define uintmax_t unsigned long int +#endif + +/* 7.18.2. Limits of specified-width integer types */ + +#if ! defined __cplusplus || defined __STDC_LIMIT_MACROS + +/* 7.18.2.1. Limits of exact-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. */ + +#undef INT8_MIN +#undef INT8_MAX +#undef UINT8_MAX +#define INT8_MIN (~ INT8_MAX) +#define INT8_MAX 127 +#define UINT8_MAX 255 + +#undef INT16_MIN +#undef INT16_MAX +#undef UINT16_MAX +#define INT16_MIN (~ INT16_MAX) +#define INT16_MAX 32767 +#define UINT16_MAX 65535 + +#undef INT32_MIN +#undef INT32_MAX +#undef UINT32_MAX +#define INT32_MIN (~ INT32_MAX) +#define INT32_MAX 2147483647 +#define UINT32_MAX 4294967295U + +#undef INT64_MIN +#undef INT64_MAX +#ifdef GL_INT64_T +/* Prefer (- INTMAX_C (1) << 63) over (~ INT64_MAX) because SunPRO C 5.0 + evaluates the latter incorrectly in preprocessor expressions. */ +# define INT64_MIN (- INTMAX_C (1) << 63) +# define INT64_MAX INTMAX_C (9223372036854775807) +#endif + +#undef UINT64_MAX +#ifdef GL_UINT64_T +# define UINT64_MAX UINTMAX_C (18446744073709551615) +#endif + +/* 7.18.2.2. Limits of minimum-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the leastN_t types + are the same as the corresponding N_t types. */ + +#undef INT_LEAST8_MIN +#undef INT_LEAST8_MAX +#undef UINT_LEAST8_MAX +#define INT_LEAST8_MIN INT8_MIN +#define INT_LEAST8_MAX INT8_MAX +#define UINT_LEAST8_MAX UINT8_MAX + +#undef INT_LEAST16_MIN +#undef INT_LEAST16_MAX +#undef UINT_LEAST16_MAX +#define INT_LEAST16_MIN INT16_MIN +#define INT_LEAST16_MAX INT16_MAX +#define UINT_LEAST16_MAX UINT16_MAX + +#undef INT_LEAST32_MIN +#undef INT_LEAST32_MAX +#undef UINT_LEAST32_MAX +#define INT_LEAST32_MIN INT32_MIN +#define INT_LEAST32_MAX INT32_MAX +#define UINT_LEAST32_MAX UINT32_MAX + +#undef INT_LEAST64_MIN +#undef INT_LEAST64_MAX +#ifdef GL_INT64_T +# define INT_LEAST64_MIN INT64_MIN +# define INT_LEAST64_MAX INT64_MAX +#endif + +#undef UINT_LEAST64_MAX +#ifdef GL_UINT64_T +# define UINT_LEAST64_MAX UINT64_MAX +#endif + +/* 7.18.2.3. Limits of fastest minimum-width integer types */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits. Therefore the fastN_t types + are taken from the same list of types. */ + +#undef INT_FAST8_MIN +#undef INT_FAST8_MAX +#undef UINT_FAST8_MAX +#define INT_FAST8_MIN LONG_MIN +#define INT_FAST8_MAX LONG_MAX +#define UINT_FAST8_MAX ULONG_MAX + +#undef INT_FAST16_MIN +#undef INT_FAST16_MAX +#undef UINT_FAST16_MAX +#define INT_FAST16_MIN LONG_MIN +#define INT_FAST16_MAX LONG_MAX +#define UINT_FAST16_MAX ULONG_MAX + +#undef INT_FAST32_MIN +#undef INT_FAST32_MAX +#undef UINT_FAST32_MAX +#define INT_FAST32_MIN LONG_MIN +#define INT_FAST32_MAX LONG_MAX +#define UINT_FAST32_MAX ULONG_MAX + +#undef INT_FAST64_MIN +#undef INT_FAST64_MAX +#ifdef GL_INT64_T +# define INT_FAST64_MIN INT64_MIN +# define INT_FAST64_MAX INT64_MAX +#endif + +#undef UINT_FAST64_MAX +#ifdef GL_UINT64_T +# define UINT_FAST64_MAX UINT64_MAX +#endif + +/* 7.18.2.4. Limits of integer types capable of holding object pointers */ + +#undef INTPTR_MIN +#undef INTPTR_MAX +#undef UINTPTR_MAX +#define INTPTR_MIN LONG_MIN +#define INTPTR_MAX LONG_MAX +#define UINTPTR_MAX ULONG_MAX + +/* 7.18.2.5. Limits of greatest-width integer types */ + +#undef INTMAX_MIN +#undef INTMAX_MAX +#ifdef INT64_MAX +# define INTMAX_MIN INT64_MIN +# define INTMAX_MAX INT64_MAX +#else +# define INTMAX_MIN INT32_MIN +# define INTMAX_MAX INT32_MAX +#endif + +#undef UINTMAX_MAX +#ifdef UINT64_MAX +# define UINTMAX_MAX UINT64_MAX +#else +# define UINTMAX_MAX UINT32_MAX +#endif + +/* 7.18.3. Limits of other integer types */ + +/* ptrdiff_t limits */ +#undef PTRDIFF_MIN +#undef PTRDIFF_MAX +#define PTRDIFF_MIN \ + _STDINT_MIN (1, @BITSIZEOF_PTRDIFF_T@, 0@PTRDIFF_T_SUFFIX@) +#define PTRDIFF_MAX \ + _STDINT_MAX (1, @BITSIZEOF_PTRDIFF_T@, 0@PTRDIFF_T_SUFFIX@) + +/* sig_atomic_t limits */ +#undef SIG_ATOMIC_MIN +#undef SIG_ATOMIC_MAX +#define SIG_ATOMIC_MIN \ + _STDINT_MIN (@HAVE_SIGNED_SIG_ATOMIC_T@, @BITSIZEOF_SIG_ATOMIC_T@, \ + 0@SIG_ATOMIC_T_SUFFIX@) +#define SIG_ATOMIC_MAX \ + _STDINT_MAX (@HAVE_SIGNED_SIG_ATOMIC_T@, @BITSIZEOF_SIG_ATOMIC_T@, \ + 0@SIG_ATOMIC_T_SUFFIX@) + + +/* size_t limit */ +#undef SIZE_MAX +#define SIZE_MAX _STDINT_MAX (0, @BITSIZEOF_SIZE_T@, 0@SIZE_T_SUFFIX@) + +/* wchar_t limits */ +#undef WCHAR_MIN +#undef WCHAR_MAX +#define WCHAR_MIN \ + _STDINT_MIN (@HAVE_SIGNED_WCHAR_T@, @BITSIZEOF_WCHAR_T@, 0@WCHAR_T_SUFFIX@) +#define WCHAR_MAX \ + _STDINT_MAX (@HAVE_SIGNED_WCHAR_T@, @BITSIZEOF_WCHAR_T@, 0@WCHAR_T_SUFFIX@) + +/* wint_t limits */ +#undef WINT_MIN +#undef WINT_MAX +#define WINT_MIN \ + _STDINT_MIN (@HAVE_SIGNED_WINT_T@, @BITSIZEOF_WINT_T@, 0@WINT_T_SUFFIX@) +#define WINT_MAX \ + _STDINT_MAX (@HAVE_SIGNED_WINT_T@, @BITSIZEOF_WINT_T@, 0@WINT_T_SUFFIX@) + +#endif /* !defined __cplusplus || defined __STDC_LIMIT_MACROS */ + +/* 7.18.4. Macros for integer constants */ + +#if ! defined __cplusplus || defined __STDC_CONSTANT_MACROS + +/* 7.18.4.1. Macros for minimum-width integer constants */ +/* According to ISO C 99 Technical Corrigendum 1 */ + +/* Here we assume a standard architecture where the hardware integer + types have 8, 16, 32, optionally 64 bits, and int is 32 bits. */ + +#undef INT8_C +#undef UINT8_C +#define INT8_C(x) x +#define UINT8_C(x) x + +#undef INT16_C +#undef UINT16_C +#define INT16_C(x) x +#define UINT16_C(x) x + +#undef INT32_C +#undef UINT32_C +#define INT32_C(x) x +#define UINT32_C(x) x ## U + +#undef INT64_C +#undef UINT64_C +#if LONG_MAX >> 31 >> 31 == 1 +# define INT64_C(x) x##L +#elif defined _MSC_VER +# define INT64_C(x) x##i64 +#elif @HAVE_LONG_LONG_INT@ +# define INT64_C(x) x##LL +#endif +#if ULONG_MAX >> 31 >> 31 >> 1 == 1 +# define UINT64_C(x) x##UL +#elif defined _MSC_VER +# define UINT64_C(x) x##ui64 +#elif @HAVE_UNSIGNED_LONG_LONG_INT@ +# define UINT64_C(x) x##ULL +#endif + +/* 7.18.4.2. Macros for greatest-width integer constants */ + +#undef INTMAX_C +#if @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define INTMAX_C(x) x##LL +#elif defined GL_INT64_T +# define INTMAX_C(x) INT64_C(x) +#else +# define INTMAX_C(x) x##L +#endif + +#undef UINTMAX_C +#if @HAVE_UNSIGNED_LONG_LONG_INT@ && ULONG_MAX >> 31 == 1 +# define UINTMAX_C(x) x##ULL +#elif defined GL_UINT64_T +# define UINTMAX_C(x) UINT64_C(x) +#else +# define UINTMAX_C(x) x##UL +#endif + +#endif /* !defined __cplusplus || defined __STDC_CONSTANT_MACROS */ + +#endif /* _GL_STDINT_H */ +#endif /* _GL_STDINT_H */ diff --git a/lib/stdio_.h b/lib/stdio_.h new file mode 100644 index 0000000..7afaede --- /dev/null +++ b/lib/stdio_.h @@ -0,0 +1,320 @@ +/* A GNU-like <stdio.h>. + + Copyright (C) 2004, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if defined __need_FILE || defined __need___FILE +/* Special invocation convention inside glibc header files. */ + +#if @HAVE_INCLUDE_NEXT@ +# include_next <stdio.h> +#else +# include @ABSOLUTE_STDIO_H@ +#endif + +#else +/* Normal invocation convention. */ + +#ifndef _GL_STDIO_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <stdio.h> +#else +# include @ABSOLUTE_STDIO_H@ +#endif + +#ifndef _GL_STDIO_H +#define _GL_STDIO_H + +#include <stdarg.h> +#include <stddef.h> + +#if (@GNULIB_FSEEKO@ && @REPLACE_FSEEKO@) || (@GNULIB_FTELLO@ && @REPLACE_FTELLO@) +/* Get off_t. */ +# include <sys/types.h> +#endif + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + + +/* The definition of GL_LINK_WARNING is copied here. */ + + +#ifdef __cplusplus +extern "C" { +#endif + + +#if @GNULIB_FPRINTF_POSIX@ +# if @REPLACE_FPRINTF@ +# define fprintf rpl_fprintf +extern int fprintf (FILE *fp, const char *format, ...) + __attribute__ ((__format__ (__printf__, 2, 3))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef fprintf +# define fprintf \ + (GL_LINK_WARNING ("fprintf is not always POSIX compliant - " \ + "use gnulib module fprintf-posix for portable " \ + "POSIX compliance"), \ + fprintf) +#endif + +#if @GNULIB_VFPRINTF_POSIX@ +# if @REPLACE_VFPRINTF@ +# define vfprintf rpl_vfprintf +extern int vfprintf (FILE *fp, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 2, 0))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef vfprintf +# define vfprintf(s,f,a) \ + (GL_LINK_WARNING ("vfprintf is not always POSIX compliant - " \ + "use gnulib module vfprintf-posix for portable " \ + "POSIX compliance"), \ + vfprintf (s, f, a)) +#endif + +#if @GNULIB_PRINTF_POSIX@ +# if @REPLACE_PRINTF@ +/* Don't break __attribute__((format(printf,M,N))). */ +# define printf __printf__ +extern int printf (const char *format, ...) + __attribute__ ((__format__ (__printf__, 1, 2))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef printf +# define printf \ + (GL_LINK_WARNING ("printf is not always POSIX compliant - " \ + "use gnulib module printf-posix for portable " \ + "POSIX compliance"), \ + printf) +/* Don't break __attribute__((format(printf,M,N))). */ +# define format(kind,m,n) format (__##kind##__, m, n) +# define __format__(kind,m,n) __format__ (__##kind##__, m, n) +# define ____printf____ __printf__ +# define ____scanf____ __scanf__ +# define ____strftime____ __strftime__ +# define ____strfmon____ __strfmon__ +#endif + +#if @GNULIB_VPRINTF_POSIX@ +# if @REPLACE_VPRINTF@ +# define vprintf rpl_vprintf +extern int vprintf (const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 1, 0))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef vprintf +# define vprintf(f,a) \ + (GL_LINK_WARNING ("vprintf is not always POSIX compliant - " \ + "use gnulib module vprintf-posix for portable " \ + "POSIX compliance"), \ + vprintf (f, a)) +#endif + +#if @GNULIB_SNPRINTF@ +# if @REPLACE_SNPRINTF@ +# define snprintf rpl_snprintf +# endif +# if @REPLACE_SNPRINTF@ || !@HAVE_DECL_SNPRINTF@ +extern int snprintf (char *str, size_t size, const char *format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef snprintf +# define snprintf \ + (GL_LINK_WARNING ("snprintf is unportable - " \ + "use gnulib module snprintf for portability"), \ + snprintf) +#endif + +#if @GNULIB_VSNPRINTF@ +# if @REPLACE_VSNPRINTF@ +# define vsnprintf rpl_vsnprintf +# endif +# if @REPLACE_VSNPRINTF@ || !@HAVE_DECL_VSNPRINTF@ +extern int vsnprintf (char *str, size_t size, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 3, 0))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef vsnprintf +# define vsnprintf(b,s,f,a) \ + (GL_LINK_WARNING ("vsnprintf is unportable - " \ + "use gnulib module vsnprintf for portability"), \ + vsnprintf (b, s, f, a)) +#endif + +#if @GNULIB_SPRINTF_POSIX@ +# if @REPLACE_SPRINTF@ +# define sprintf rpl_sprintf +extern int sprintf (char *str, const char *format, ...) + __attribute__ ((__format__ (__printf__, 2, 3))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef sprintf +# define sprintf \ + (GL_LINK_WARNING ("sprintf is not always POSIX compliant - " \ + "use gnulib module sprintf-posix for portable " \ + "POSIX compliance"), \ + sprintf) +#endif + +#if @GNULIB_VSPRINTF_POSIX@ +# if @REPLACE_VSPRINTF@ +# define vsprintf rpl_vsprintf +extern int vsprintf (char *str, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 2, 0))); +# endif +#elif defined GNULIB_POSIXCHECK +# undef vsprintf +# define vsprintf(b,f,a) \ + (GL_LINK_WARNING ("vsprintf is not always POSIX compliant - " \ + "use gnulib module vsprintf-posix for portable " \ + "POSIX compliance"), \ + vsprintf (b, f, a)) +#endif + +#if @GNULIB_VASPRINTF@ +# if @REPLACE_VASPRINTF@ +# define asprintf rpl_asprintf +# define vasprintf rpl_vasprintf +# endif +# if @REPLACE_VASPRINTF@ || !@HAVE_VASPRINTF@ + /* Write formatted output to a string dynamically allocated with malloc(). + If the memory allocation succeeds, store the address of the string in + *RESULT and return the number of resulting bytes, excluding the trailing + NUL. Upon memory allocation error, or some other error, return -1. */ + extern int asprintf (char **result, const char *format, ...) + __attribute__ ((__format__ (__printf__, 2, 3))); + extern int vasprintf (char **result, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 2, 0))); +# endif +#endif + +#if @GNULIB_FSEEKO@ +# if @REPLACE_FSEEKO@ +/* Provide fseek, fseeko functions that are aware of a preceding + fflush(), and which detect pipes. */ +# define fseeko rpl_fseeko +extern int fseeko (FILE *fp, off_t offset, int whence); +# define fseek(fp, offset, whence) fseeko (fp, (off_t)(offset), whence) +# endif +#elif defined GNULIB_POSIXCHECK +# undef fseeko +# define fseeko(f,o,w) \ + (GL_LINK_WARNING ("fseeko is unportable - " \ + "use gnulib module fseeko for portability"), \ + fseeko (f, o, w)) +#endif + +#if @GNULIB_FSEEK@ && @REPLACE_FSEEK@ +extern int rpl_fseek (FILE *fp, long offset, int whence); +# undef fseek +# if defined GNULIB_POSIXCHECK +# define fseek(f,o,w) \ + (GL_LINK_WARNING ("fseek cannot handle files larger than 4 GB " \ + "on 32-bit platforms - " \ + "use fseeko function for handling of large files"), \ + rpl_fseek (f, o, w)) +# else +# define fseek rpl_fseek +# endif +#elif defined GNULIB_POSIXCHECK +# ifndef fseek +# define fseek(f,o,w) \ + (GL_LINK_WARNING ("fseek cannot handle files larger than 4 GB " \ + "on 32-bit platforms - " \ + "use fseeko function for handling of large files"), \ + fseek (f, o, w)) +# endif +#endif + +#if @GNULIB_FTELLO@ +# if @REPLACE_FTELLO@ +# define ftello rpl_ftello +extern off_t ftello (FILE *fp); +# define ftell(fp) ftello (fp) +# endif +#elif defined GNULIB_POSIXCHECK +# undef ftello +# define ftello(f) \ + (GL_LINK_WARNING ("ftello is unportable - " \ + "use gnulib module ftello for portability"), \ + ftello (f)) +#endif + +#if @GNULIB_FTELL@ && @REPLACE_FTELL@ +extern long rpl_ftell (FILE *fp); +# undef ftell +# if GNULIB_POSIXCHECK +# define ftell(f) \ + (GL_LINK_WARNING ("ftell cannot handle files larger than 4 GB " \ + "on 32-bit platforms - " \ + "use ftello function for handling of large files"), \ + rpl_ftell (f)) +# else +# define ftell rpl_ftell +# endif +#elif defined GNULIB_POSIXCHECK +# ifndef ftell +# define ftell(f) \ + (GL_LINK_WARNING ("ftell cannot handle files larger than 4 GB " \ + "on 32-bit platforms - " \ + "use ftello function for handling of large files"), \ + ftell (f)) +# endif +#endif + +#if @GNULIB_FFLUSH@ +# if @REPLACE_FFLUSH@ +# define fflush rpl_fflush + /* Flush all pending data on STREAM according to POSIX rules. Both + output and seekable input streams are supported. + Note! LOSS OF DATA can occur if fflush is applied on an input stream + that is _not_seekable_ or on an update stream that is _not_seekable_ + and in which the most recent operation was input. Seekability can + be tested with lseek(fileno(fp),0,SEEK_CUR). */ + extern int fflush (FILE *gl_stream); +# endif +#elif defined GNULIB_POSIXCHECK +# undef fflush +# define fflush(f) \ + (GL_LINK_WARNING ("fflush is not always POSIX compliant - " \ + "use gnulib module fflush for portable " \ + "POSIX compliance"), \ + fflush (f)) +#endif + +#ifdef __cplusplus +} +#endif + +#endif /* _GL_STDIO_H */ +#endif /* _GL_STDIO_H */ +#endif diff --git a/lib/stdlib_.h b/lib/stdlib_.h new file mode 100644 index 0000000..130f071 --- /dev/null +++ b/lib/stdlib_.h @@ -0,0 +1,140 @@ +/* A GNU-like <stdlib.h>. + + Copyright (C) 1995, 2001-2002, 2006-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if defined __need_malloc_and_calloc +/* Special invocation convention inside glibc header files. */ + +#if @HAVE_INCLUDE_NEXT@ +# include_next <stdlib.h> +#else +# include @ABSOLUTE_STDLIB_H@ +#endif + +#else +/* Normal invocation convention. */ + +#ifndef _GL_STDLIB_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <stdlib.h> +#else +# include @ABSOLUTE_STDLIB_H@ +#endif + +#ifndef _GL_STDLIB_H +#define _GL_STDLIB_H + + +/* The definition of GL_LINK_WARNING is copied here. */ + + +/* Some systems do not define EXIT_*, despite otherwise supporting C89. */ +#ifndef EXIT_SUCCESS +# define EXIT_SUCCESS 0 +#endif +/* Tandem/NSK and other platforms that define EXIT_FAILURE as -1 interfere + with proper operation of xargs. */ +#ifndef EXIT_FAILURE +# define EXIT_FAILURE 1 +#elif EXIT_FAILURE != 1 +# undef EXIT_FAILURE +# define EXIT_FAILURE 1 +#endif + + +#ifdef __cplusplus +extern "C" { +#endif + + +#if @GNULIB_GETSUBOPT@ +/* Assuming *OPTIONP is a comma separated list of elements of the form + "token" or "token=value", getsubopt parses the first of these elements. + If the first element refers to a "token" that is member of the given + NULL-terminated array of tokens: + - It replaces the comma with a NUL byte, updates *OPTIONP to point past + the first option and the comma, sets *VALUEP to the value of the + element (or NULL if it doesn't contain an "=" sign), + - It returns the index of the "token" in the given array of tokens. + Otherwise it returns -1, and *OPTIONP and *VALUEP are undefined. + For more details see the POSIX:2001 specification. + http://www.opengroup.org/susv3xsh/getsubopt.html */ +# if !@HAVE_GETSUBOPT@ +extern int getsubopt (char **optionp, char *const *tokens, char **valuep); +# endif +#elif defined GNULIB_POSIXCHECK +# undef getsubopt +# define getsubopt(o,t,v) \ + (GL_LINK_WARNING ("getsubopt is unportable - " \ + "use gnulib module getsubopt for portability"), \ + getsubopt (o, t, v)) +#endif + + +#if @GNULIB_MKDTEMP@ +# if !@HAVE_MKDTEMP@ +/* Create a unique temporary directory from TEMPLATE. + The last six characters of TEMPLATE must be "XXXXXX"; + they are replaced with a string that makes the directory name unique. + Returns TEMPLATE, or a null pointer if it cannot get a unique name. + The directory is created mode 700. */ +extern char * mkdtemp (char * /*template*/); +# endif +#elif defined GNULIB_POSIXCHECK +# undef mkdtemp +# define mkdtemp(t) \ + (GL_LINK_WARNING ("mkdtemp is unportable - " \ + "use gnulib module mkdtemp for portability"), \ + mkdtemp (t)) +#endif + + +#if @GNULIB_MKSTEMP@ +# if @REPLACE_MKSTEMP@ +/* Create a unique temporary file from TEMPLATE. + The last six characters of TEMPLATE must be "XXXXXX"; + they are replaced with a string that makes the file name unique. + The file is then created, ensuring it didn't exist before. + The file is created read-write (mask at least 0600 & ~umask), but it may be + world-readable and world-writable (mask 0666 & ~umask), depending on the + implementation. + Returns the open file descriptor if successful, otherwise -1 and errno + set. */ +# define mkstemp rpl_mkstemp +extern int mkstemp (char * /*template*/); +# else +/* On MacOS X 10.3, only <unistd.h> declares mkstemp. */ +# include <unistd.h> +# endif +#elif defined GNULIB_POSIXCHECK +# undef mkstemp +# define mkstemp(t) \ + (GL_LINK_WARNING ("mkstemp is unportable - " \ + "use gnulib module mkstemp for portability"), \ + mkstemp (t)) +#endif + + +#ifdef __cplusplus +} +#endif + +#endif /* _GL_STDLIB_H */ +#endif /* _GL_STDLIB_H */ +#endif diff --git a/lib/stdopen.c b/lib/stdopen.c new file mode 100644 index 0000000..92825b5 --- /dev/null +++ b/lib/stdopen.c @@ -0,0 +1,77 @@ +/* stdopen.c - ensure that the three standard file descriptors are in use + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and Jim Meyering. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#include "stdopen.h" + +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <unistd.h> +#include <errno.h> + +/* Try to ensure that all of the standard file numbers (0, 1, 2) + are in use. Without this, each application would have to guard + every call to open, dup, fopen, etc. with tests to ensure they + don't use one of the special file numbers when opening a file. + Return false if at least one of the file descriptors is initially + closed and an attempt to reopen it fails. Otherwise, return true. */ +bool +stdopen (void) +{ + int fd; + bool ok = true; + + for (fd = 0; fd <= 2; fd++) + { + if (fcntl (fd, F_GETFD) < 0) + { + if (errno != EBADF) + ok = false; + else + { + static const int contrary_mode[] + = { O_WRONLY, O_RDONLY, O_RDONLY }; + int mode = contrary_mode[fd]; + int new_fd; + /* Open /dev/null with the contrary mode so that the typical + read (stdin) or write (stdout, stderr) operation will fail. + With descriptor 0, we can do even better on systems that + have /dev/full, by opening that write-only instead of + /dev/null. The only drawback is that a write-provoked + failure comes with a misleading errno value, ENOSPC. */ + if (mode == O_RDONLY + || (new_fd = open ("/dev/full", mode) != fd)) + new_fd = open ("/dev/null", mode); + if (new_fd != fd) + { + if (0 <= new_fd) + close (new_fd); + ok = false; + } + } + } + } + + return ok; +} diff --git a/lib/stdopen.h b/lib/stdopen.h new file mode 100644 index 0000000..d54e5f1 --- /dev/null +++ b/lib/stdopen.h @@ -0,0 +1,16 @@ +#ifndef STDOPEN_H +# define STDOPEN_H 1 + +# include <stdbool.h> + +# ifdef __cplusplus +extern "C" { +# endif + +bool stdopen (void); + +# ifdef __cplusplus +} +# endif + +#endif diff --git a/lib/stpcpy.c b/lib/stpcpy.c new file mode 100644 index 0000000..4089d39 --- /dev/null +++ b/lib/stpcpy.c @@ -0,0 +1,49 @@ +/* stpcpy.c -- copy a string and return pointer to end of new string + Copyright (C) 1992, 1995, 1997-1998, 2006 Free Software Foundation, Inc. + + NOTE: The canonical source of this file is maintained with the GNU C Library. + Bugs can be reported to bug-glibc@prep.ai.mit.edu. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <string.h> + +#undef __stpcpy +#ifdef _LIBC +# undef stpcpy +#endif + +#ifndef weak_alias +# define __stpcpy stpcpy +#endif + +/* Copy SRC to DEST, returning the address of the terminating '\0' in DEST. */ +char * +__stpcpy (char *dest, const char *src) +{ + register char *d = dest; + register const char *s = src; + + do + *d++ = *s; + while (*s++ != '\0'); + + return d - 1; +} +#ifdef weak_alias +weak_alias (__stpcpy, stpcpy) +#endif diff --git a/lib/strcasecmp.c b/lib/strcasecmp.c new file mode 100644 index 0000000..c605fb0 --- /dev/null +++ b/lib/strcasecmp.c @@ -0,0 +1,63 @@ +/* Case-insensitive string comparison function. + Copyright (C) 1998-1999, 2005-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <string.h> + +#include <ctype.h> +#include <limits.h> + +#define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch)) + +/* Compare strings S1 and S2, ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less than, equal to or greater + than S2. + Note: This function does not work with multibyte strings! */ + +int +strcasecmp (const char *s1, const char *s2) +{ + const unsigned char *p1 = (const unsigned char *) s1; + const unsigned char *p2 = (const unsigned char *) s2; + unsigned char c1, c2; + + if (p1 == p2) + return 0; + + do + { + c1 = TOLOWER (*p1); + c2 = TOLOWER (*p2); + + if (c1 == '\0') + break; + + ++p1; + ++p2; + } + while (c1 == c2); + + if (UCHAR_MAX <= INT_MAX) + return c1 - c2; + else + /* On machines where 'char' and 'int' are types of the same size, the + difference of two 'unsigned char' values - including the sign bit - + doesn't fit in an 'int'. */ + return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0); +} diff --git a/lib/strchrnul.c b/lib/strchrnul.c new file mode 100644 index 0000000..07014be --- /dev/null +++ b/lib/strchrnul.c @@ -0,0 +1,32 @@ +/* Searching in a string. + Copyright (C) 2003, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <string.h> + +/* Find the first occurrence of C in S or the final NUL byte. */ +char * +strchrnul (const char *s, int c_in) +{ + char c = c_in; + while (*s && (*s != c)) + s++; + + return (char *) s; +} diff --git a/lib/strdup.c b/lib/strdup.c new file mode 100644 index 0000000..c614108 --- /dev/null +++ b/lib/strdup.c @@ -0,0 +1,55 @@ +/* Copyright (C) 1991, 1996, 1997, 1998, 2002, 2003, 2004, 2006, 2007 Free + Software Foundation, Inc. + + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _LIBC +# include <config.h> +#endif + +/* Get specification. */ +#include <string.h> + +#include <stdlib.h> + +#undef __strdup +#ifdef _LIBC +# undef strdup +#endif + +#ifndef weak_alias +# define __strdup strdup +#endif + +/* Duplicate S, returning an identical malloc'd string. */ +char * +__strdup (const char *s) +{ + size_t len = strlen (s) + 1; + void *new = malloc (len); + + if (new == NULL) + return NULL; + + return (char *) memcpy (new, s, len); +} +#ifdef libc_hidden_def +libc_hidden_def (__strdup) +#endif +#ifdef weak_alias +weak_alias (__strdup, strdup) +#endif diff --git a/lib/strerror.c b/lib/strerror.c new file mode 100644 index 0000000..54b851b --- /dev/null +++ b/lib/strerror.c @@ -0,0 +1,57 @@ +/* strerror.c --- ANSI C compatible system error routine + + Copyright (C) 1986, 1988, 1989, 1991, 2002, 2003, 2006, 2007 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#if !HAVE_STRERROR + +#include <limits.h> + +/* Don't include <stdio.h>, since it may or may not declare + sys_errlist and its declarations may collide with ours. Just + declare the stuff that we need directly. Standard hosted C89 + implementations define strerror and they don't need this strerror + function, so take some liberties with the standard to cater to + ancient or limited freestanding implementations. */ +int sprintf (char *, char const *, ...); +extern int sys_nerr; +extern char *sys_errlist[]; + +char * +strerror (int n) +{ + static char const fmt[] = "Unknown error (%d)"; + static char mesg[sizeof fmt + sizeof n * CHAR_BIT / 3]; + + if (n < 0 || n >= sys_nerr) + { + sprintf (mesg, fmt, n); + return mesg; + } + else + return sys_errlist[n]; +} + +#else + +/* This declaration is solely to ensure that after preprocessing + this file is never empty. */ +typedef int dummy; + +#endif diff --git a/lib/string_.h b/lib/string_.h new file mode 100644 index 0000000..6ec72c3 --- /dev/null +++ b/lib/string_.h @@ -0,0 +1,546 @@ +/* A GNU-like <string.h>. + + Copyright (C) 1995-1996, 2001-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_STRING_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <string.h> +#else +# include @ABSOLUTE_STRING_H@ +#endif + +#ifndef _GL_STRING_H +#define _GL_STRING_H + + +/* The definition of GL_LINK_WARNING is copied here. */ + + +#ifdef __cplusplus +extern "C" { +#endif + +/* Return the first occurrence of NEEDLE in HAYSTACK. */ +#if @GNULIB_MEMMEM@ +# if ! @HAVE_DECL_MEMMEM@ +extern void *memmem (void const *__haystack, size_t __haystack_len, + void const *__needle, size_t __needle_len); +# endif +#elif defined GNULIB_POSIXCHECK +# undef memmem +# define memmem(a,al,b,bl) \ + (GL_LINK_WARNING ("memmem is unportable - " \ + "use gnulib module memmem for portability"), \ + memmem (a, al, b, bl)) +#endif + +/* Copy N bytes of SRC to DEST, return pointer to bytes after the + last written byte. */ +#if @GNULIB_MEMPCPY@ +# if ! @HAVE_MEMPCPY@ +extern void *mempcpy (void *restrict __dest, void const *restrict __src, + size_t __n); +# endif +#elif defined GNULIB_POSIXCHECK +# undef mempcpy +# define mempcpy(a,b,n) \ + (GL_LINK_WARNING ("mempcpy is unportable - " \ + "use gnulib module mempcpy for portability"), \ + mempcpy (a, b, n)) +#endif + +/* Search backwards through a block for a byte (specified as an int). */ +#if @GNULIB_MEMRCHR@ +# if ! @HAVE_DECL_MEMRCHR@ +extern void *memrchr (void const *, int, size_t); +# endif +#elif defined GNULIB_POSIXCHECK +# undef memrchr +# define memrchr(a,b,c) \ + (GL_LINK_WARNING ("memrchr is unportable - " \ + "use gnulib module memrchr for portability"), \ + memrchr (a, b, c)) +#endif + +/* Copy SRC to DST, returning the address of the terminating '\0' in DST. */ +#if @GNULIB_STPCPY@ +# if ! @HAVE_STPCPY@ +extern char *stpcpy (char *restrict __dst, char const *restrict __src); +# endif +#elif defined GNULIB_POSIXCHECK +# undef stpcpy +# define stpcpy(a,b) \ + (GL_LINK_WARNING ("stpcpy is unportable - " \ + "use gnulib module stpcpy for portability"), \ + stpcpy (a, b)) +#endif + +/* Copy no more than N bytes of SRC to DST, returning a pointer past the + last non-NUL byte written into DST. */ +#if @GNULIB_STPNCPY@ +# if ! @HAVE_STPNCPY@ +# define stpncpy gnu_stpncpy +extern char *stpncpy (char *restrict __dst, char const *restrict __src, + size_t __n); +# endif +#elif defined GNULIB_POSIXCHECK +# undef stpncpy +# define stpncpy(a,b,n) \ + (GL_LINK_WARNING ("stpncpy is unportable - " \ + "use gnulib module stpncpy for portability"), \ + stpncpy (a, b, n)) +#endif + +/* Compare strings S1 and S2, ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less than, equal to or greater + than S2. + Note: This function does not work in multibyte locales. */ +#if ! @HAVE_STRCASECMP@ +extern int strcasecmp (char const *s1, char const *s2); +#endif +#if defined GNULIB_POSIXCHECK +/* strcasecmp() does not work with multibyte strings: + POSIX says that it operates on "strings", and "string" in POSIX is defined + as a sequence of bytes, not of characters. */ +# undef strcasecmp +# define strcasecmp(a,b) \ + (GL_LINK_WARNING ("strcasecmp cannot work correctly on character strings " \ + "in multibyte locales - " \ + "use mbscasecmp if you care about " \ + "internationalization, or use c_strcasecmp (from " \ + "gnulib module c-strcase) if you want a locale " \ + "independent function"), \ + strcasecmp (a, b)) +#endif + +/* Compare no more than N bytes of strings S1 and S2, ignoring case, + returning less than, equal to or greater than zero if S1 is + lexicographically less than, equal to or greater than S2. + Note: This function cannot work correctly in multibyte locales. */ +#if ! @HAVE_DECL_STRNCASECMP@ +extern int strncasecmp (char const *s1, char const *s2, size_t n); +#endif +#if defined GNULIB_POSIXCHECK +/* strncasecmp() does not work with multibyte strings: + POSIX says that it operates on "strings", and "string" in POSIX is defined + as a sequence of bytes, not of characters. */ +# undef strncasecmp +# define strncasecmp(a,b,n) \ + (GL_LINK_WARNING ("strncasecmp cannot work correctly on character " \ + "strings in multibyte locales - " \ + "use mbsncasecmp or mbspcasecmp if you care about " \ + "internationalization, or use c_strncasecmp (from " \ + "gnulib module c-strcase) if you want a locale " \ + "independent function"), \ + strncasecmp (a, b, n)) +#endif + +#if defined GNULIB_POSIXCHECK +/* strchr() does not work with multibyte strings if the locale encoding is + GB18030 and the character to be searched is a digit. */ +# undef strchr +# define strchr(s,c) \ + (GL_LINK_WARNING ("strchr cannot work correctly on character strings " \ + "in some multibyte locales - " \ + "use mbschr if you care about internationalization"), \ + strchr (s, c)) +#endif + +/* Find the first occurrence of C in S or the final NUL byte. */ +#if @GNULIB_STRCHRNUL@ +# if ! @HAVE_STRCHRNUL@ +extern char *strchrnul (char const *__s, int __c_in); +# endif +#elif defined GNULIB_POSIXCHECK +# undef strchrnul +# define strchrnul(a,b) \ + (GL_LINK_WARNING ("strchrnul is unportable - " \ + "use gnulib module strchrnul for portability"), \ + strchrnul (a, b)) +#endif + +/* Duplicate S, returning an identical malloc'd string. */ +#if @GNULIB_STRDUP@ +# if ! @HAVE_DECL_STRDUP@ && ! defined strdup +extern char *strdup (char const *__s); +# endif +#elif defined GNULIB_POSIXCHECK +# undef strdup +# define strdup(a) \ + (GL_LINK_WARNING ("strdup is unportable - " \ + "use gnulib module strdup for portability"), \ + strdup (a)) +#endif + +/* Return a newly allocated copy of at most N bytes of STRING. */ +#if @GNULIB_STRNDUP@ +# if ! @HAVE_STRNDUP@ +# undef strndup +# define strndup rpl_strndup +# endif +# if ! @HAVE_STRNDUP@ || ! @HAVE_DECL_STRNDUP@ +extern char *strndup (char const *__string, size_t __n); +# endif +#elif defined GNULIB_POSIXCHECK +# undef strndup +# define strndup(a,n) \ + (GL_LINK_WARNING ("strndup is unportable - " \ + "use gnulib module strndup for portability"), \ + strndup (a, n)) +#endif + +/* Find the length (number of bytes) of STRING, but scan at most + MAXLEN bytes. If no '\0' terminator is found in that many bytes, + return MAXLEN. */ +#if @GNULIB_STRNLEN@ +# if ! @HAVE_DECL_STRNLEN@ +extern size_t strnlen (char const *__string, size_t __maxlen); +# endif +#elif defined GNULIB_POSIXCHECK +# undef strnlen +# define strnlen(a,n) \ + (GL_LINK_WARNING ("strnlen is unportable - " \ + "use gnulib module strnlen for portability"), \ + strnlen (a, n)) +#endif + +#if defined GNULIB_POSIXCHECK +/* strcspn() assumes the second argument is a list of single-byte characters. + Even in this simple case, it does not work with multibyte strings if the + locale encoding is GB18030 and one of the characters to be searched is a + digit. */ +# undef strcspn +# define strcspn(s,a) \ + (GL_LINK_WARNING ("strcspn cannot work correctly on character strings " \ + "in multibyte locales - " \ + "use mbscspn if you care about internationalization"), \ + strcspn (s, a)) +#endif + +/* Find the first occurrence in S of any character in ACCEPT. */ +#if @GNULIB_STRPBRK@ +# if ! @HAVE_STRPBRK@ +extern char *strpbrk (char const *__s, char const *__accept); +# endif +# if defined GNULIB_POSIXCHECK +/* strpbrk() assumes the second argument is a list of single-byte characters. + Even in this simple case, it does not work with multibyte strings if the + locale encoding is GB18030 and one of the characters to be searched is a + digit. */ +# undef strpbrk +# define strpbrk(s,a) \ + (GL_LINK_WARNING ("strpbrk cannot work correctly on character strings " \ + "in multibyte locales - " \ + "use mbspbrk if you care about internationalization"), \ + strpbrk (s, a)) +# endif +#elif defined GNULIB_POSIXCHECK +# undef strpbrk +# define strpbrk(s,a) \ + (GL_LINK_WARNING ("strpbrk is unportable - " \ + "use gnulib module strpbrk for portability"), \ + strpbrk (s, a)) +#endif + +#if defined GNULIB_POSIXCHECK +/* strspn() assumes the second argument is a list of single-byte characters. + Even in this simple case, it cannot work with multibyte strings. */ +# undef strspn +# define strspn(s,a) \ + (GL_LINK_WARNING ("strspn cannot work correctly on character strings " \ + "in multibyte locales - " \ + "use mbsspn if you care about internationalization"), \ + strspn (s, a)) +#endif + +#if defined GNULIB_POSIXCHECK +/* strrchr() does not work with multibyte strings if the locale encoding is + GB18030 and the character to be searched is a digit. */ +# undef strrchr +# define strrchr(s,c) \ + (GL_LINK_WARNING ("strrchr cannot work correctly on character strings " \ + "in some multibyte locales - " \ + "use mbsrchr if you care about internationalization"), \ + strrchr (s, c)) +#endif + +/* Search the next delimiter (char listed in DELIM) starting at *STRINGP. + If one is found, overwrite it with a NUL, and advance *STRINGP + to point to the next char after it. Otherwise, set *STRINGP to NULL. + If *STRINGP was already NULL, nothing happens. + Return the old value of *STRINGP. + + This is a variant of strtok() that is multithread-safe and supports + empty fields. + + Caveat: It modifies the original string. + Caveat: These functions cannot be used on constant strings. + Caveat: The identity of the delimiting character is lost. + Caveat: It doesn't work with multibyte strings unless all of the delimiter + characters are ASCII characters < 0x30. + + See also strtok_r(). */ +#if @GNULIB_STRSEP@ +# if ! @HAVE_STRSEP@ +extern char *strsep (char **restrict __stringp, char const *restrict __delim); +# endif +# if defined GNULIB_POSIXCHECK +# undef strsep +# define strsep(s,d) \ + (GL_LINK_WARNING ("strsep cannot work correctly on character strings " \ + "in multibyte locales - " \ + "use mbssep if you care about internationalization"), \ + strsep (s, d)) +# endif +#elif defined GNULIB_POSIXCHECK +# undef strsep +# define strsep(s,d) \ + (GL_LINK_WARNING ("strsep is unportable - " \ + "use gnulib module strsep for portability"), \ + strsep (s, d)) +#endif + +#if defined GNULIB_POSIXCHECK +/* strstr() does not work with multibyte strings if the locale encoding is + different from UTF-8: + POSIX says that it operates on "strings", and "string" in POSIX is defined + as a sequence of bytes, not of characters. */ +# undef strstr +# define strstr(a,b) \ + (GL_LINK_WARNING ("strstr cannot work correctly on character strings " \ + "in most multibyte locales - " \ + "use mbsstr if you care about internationalization"), \ + strstr (a, b)) +#endif + +/* Find the first occurrence of NEEDLE in HAYSTACK, using case-insensitive + comparison. */ +#if ! @HAVE_STRCASESTR@ +extern char *strcasestr (const char *haystack, const char *needle); +#endif +#if defined GNULIB_POSIXCHECK +/* strcasestr() does not work with multibyte strings: + It is a glibc extension, and glibc implements it only for unibyte + locales. */ +# undef strcasestr +# define strcasestr(a,b) \ + (GL_LINK_WARNING ("strcasestr does work correctly on character strings " \ + "in multibyte locales - " \ + "use mbscasestr if you care about " \ + "internationalization, or use c-strcasestr if you want " \ + "a locale independent function"), \ + strcasestr (a, b)) +#endif + +/* Parse S into tokens separated by characters in DELIM. + If S is NULL, the saved pointer in SAVE_PTR is used as + the next starting point. For example: + char s[] = "-abc-=-def"; + char *sp; + x = strtok_r(s, "-", &sp); // x = "abc", sp = "=-def" + x = strtok_r(NULL, "-=", &sp); // x = "def", sp = NULL + x = strtok_r(NULL, "=", &sp); // x = NULL + // s = "abc\0-def\0" + + This is a variant of strtok() that is multithread-safe. + + For the POSIX documentation for this function, see: + http://www.opengroup.org/susv3xsh/strtok.html + + Caveat: It modifies the original string. + Caveat: These functions cannot be used on constant strings. + Caveat: The identity of the delimiting character is lost. + Caveat: It doesn't work with multibyte strings unless all of the delimiter + characters are ASCII characters < 0x30. + + See also strsep(). */ +#if @GNULIB_STRTOK_R@ +# if ! @HAVE_DECL_STRTOK_R@ +extern char *strtok_r (char *restrict s, char const *restrict delim, + char **restrict save_ptr); +# endif +# if defined GNULIB_POSIXCHECK +# undef strtok_r +# define strtok_r(s,d,p) \ + (GL_LINK_WARNING ("strtok_r cannot work correctly on character strings " \ + "in multibyte locales - " \ + "use mbstok_r if you care about internationalization"), \ + strtok_r (s, d, p)) +# endif +#elif defined GNULIB_POSIXCHECK +# undef strtok_r +# define strtok_r(s,d,p) \ + (GL_LINK_WARNING ("strtok_r is unportable - " \ + "use gnulib module strtok_r for portability"), \ + strtok_r (s, d, p)) +#endif + + +/* The following functions are not specified by POSIX. They are gnulib + extensions. */ + +#if @GNULIB_MBSLEN@ +/* Return the number of multibyte characters in the character string STRING. + This considers multibyte characters, unlike strlen, which counts bytes. */ +extern size_t mbslen (const char *string); +#endif + +#if @GNULIB_MBSCHR@ +/* Locate the first single-byte character C in the character string STRING, + and return a pointer to it. Return NULL if C is not found in STRING. + Unlike strchr(), this function works correctly in multibyte locales with + encodings such as GB18030. */ +# define mbschr rpl_mbschr /* avoid collision with HP-UX function */ +extern char * mbschr (const char *string, int c); +#endif + +#if @GNULIB_MBSRCHR@ +/* Locate the last single-byte character C in the character string STRING, + and return a pointer to it. Return NULL if C is not found in STRING. + Unlike strrchr(), this function works correctly in multibyte locales with + encodings such as GB18030. */ +# define mbsrchr rpl_mbsrchr /* avoid collision with HP-UX function */ +extern char * mbsrchr (const char *string, int c); +#endif + +#if @GNULIB_MBSSTR@ +/* Find the first occurrence of the character string NEEDLE in the character + string HAYSTACK. Return NULL if NEEDLE is not found in HAYSTACK. + Unlike strstr(), this function works correctly in multibyte locales with + encodings different from UTF-8. */ +extern char * mbsstr (const char *haystack, const char *needle); +#endif + +#if @GNULIB_MBSCASECMP@ +/* Compare the character strings S1 and S2, ignoring case, returning less than, + equal to or greater than zero if S1 is lexicographically less than, equal to + or greater than S2. + Note: This function may, in multibyte locales, return 0 for strings of + different lengths! + Unlike strcasecmp(), this function works correctly in multibyte locales. */ +extern int mbscasecmp (const char *s1, const char *s2); +#endif + +#if @GNULIB_MBSNCASECMP@ +/* Compare the initial segment of the character string S1 consisting of at most + N characters with the initial segment of the character string S2 consisting + of at most N characters, ignoring case, returning less than, equal to or + greater than zero if the initial segment of S1 is lexicographically less + than, equal to or greater than the initial segment of S2. + Note: This function may, in multibyte locales, return 0 for initial segments + of different lengths! + Unlike strncasecmp(), this function works correctly in multibyte locales. + But beware that N is not a byte count but a character count! */ +extern int mbsncasecmp (const char *s1, const char *s2, size_t n); +#endif + +#if @GNULIB_MBSPCASECMP@ +/* Compare the initial segment of the character string STRING consisting of + at most mbslen (PREFIX) characters with the character string PREFIX, + ignoring case, returning less than, equal to or greater than zero if this + initial segment is lexicographically less than, equal to or greater than + PREFIX. + Note: This function may, in multibyte locales, return 0 if STRING is of + smaller length than PREFIX! + Unlike strncasecmp(), this function works correctly in multibyte + locales. */ +extern char * mbspcasecmp (const char *string, const char *prefix); +#endif + +#if @GNULIB_MBSCASESTR@ +/* Find the first occurrence of the character string NEEDLE in the character + string HAYSTACK, using case-insensitive comparison. + Note: This function may, in multibyte locales, return success even if + strlen (haystack) < strlen (needle) ! + Unlike strcasestr(), this function works correctly in multibyte locales. */ +extern char * mbscasestr (const char *haystack, const char *needle); +#endif + +#if @GNULIB_MBSCSPN@ +/* Find the first occurrence in the character string STRING of any character + in the character string ACCEPT. Return the number of bytes from the + beginning of the string to this occurrence, or to the end of the string + if none exists. + Unlike strcspn(), this function works correctly in multibyte locales. */ +extern size_t mbscspn (const char *string, const char *accept); +#endif + +#if @GNULIB_MBSPBRK@ +/* Find the first occurrence in the character string STRING of any character + in the character string ACCEPT. Return the pointer to it, or NULL if none + exists. + Unlike strpbrk(), this function works correctly in multibyte locales. */ +# define mbspbrk rpl_mbspbrk /* avoid collision with HP-UX function */ +extern char * mbspbrk (const char *string, const char *accept); +#endif + +#if @GNULIB_MBSSPN@ +/* Find the first occurrence in the character string STRING of any character + not in the character string REJECT. Return the number of bytes from the + beginning of the string to this occurrence, or to the end of the string + if none exists. + Unlike strspn(), this function works correctly in multibyte locales. */ +extern size_t mbsspn (const char *string, const char *reject); +#endif + +#if @GNULIB_MBSSEP@ +/* Search the next delimiter (multibyte character listed in the character + string DELIM) starting at the character string *STRINGP. + If one is found, overwrite it with a NUL, and advance *STRINGP to point + to the next multibyte character after it. Otherwise, set *STRINGP to NULL. + If *STRINGP was already NULL, nothing happens. + Return the old value of *STRINGP. + + This is a variant of mbstok_r() that supports empty fields. + + Caveat: It modifies the original string. + Caveat: These functions cannot be used on constant strings. + Caveat: The identity of the delimiting character is lost. + + See also mbstok_r(). */ +extern char * mbssep (char **stringp, const char *delim); +#endif + +#if @GNULIB_MBSTOK_R@ +/* Parse the character string STRING into tokens separated by characters in + the character string DELIM. + If STRING is NULL, the saved pointer in SAVE_PTR is used as + the next starting point. For example: + char s[] = "-abc-=-def"; + char *sp; + x = mbstok_r(s, "-", &sp); // x = "abc", sp = "=-def" + x = mbstok_r(NULL, "-=", &sp); // x = "def", sp = NULL + x = mbstok_r(NULL, "=", &sp); // x = NULL + // s = "abc\0-def\0" + + Caveat: It modifies the original string. + Caveat: These functions cannot be used on constant strings. + Caveat: The identity of the delimiting character is lost. + + See also mbssep(). */ +extern char * mbstok_r (char *string, const char *delim, char **save_ptr); +#endif + + +#ifdef __cplusplus +} +#endif + +#endif /* _GL_STRING_H */ +#endif /* _GL_STRING_H */ diff --git a/lib/stripslash.c b/lib/stripslash.c new file mode 100644 index 0000000..342d497 --- /dev/null +++ b/lib/stripslash.c @@ -0,0 +1,45 @@ +/* stripslash.c -- remove redundant trailing slashes from a file name + + Copyright (C) 1990, 2001, 2003-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "dirname.h" + +/* Remove trailing slashes from FILE. Return true if a trailing slash + was removed. This is useful when using file name completion from a + shell that adds a "/" after directory names (such as tcsh and + bash), because on symlinks to directories, several system calls + have different semantics according to whether a trailing slash is + present. */ + +bool +strip_trailing_slashes (char *file) +{ + char *base = last_component (file); + char *base_lim; + bool had_slash; + + /* last_component returns "" for file system roots, but we need to turn + `///' into `/'. */ + if (! *base) + base = file; + base_lim = base + base_len (base); + had_slash = (*base_lim != '\0'); + *base_lim = '\0'; + return had_slash; +} diff --git a/lib/strncasecmp.c b/lib/strncasecmp.c new file mode 100644 index 0000000..473a610 --- /dev/null +++ b/lib/strncasecmp.c @@ -0,0 +1,63 @@ +/* strncasecmp.c -- case insensitive string comparator + Copyright (C) 1998-1999, 2005-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include <string.h> + +#include <ctype.h> +#include <limits.h> + +#define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch)) + +/* Compare no more than N bytes of strings S1 and S2, ignoring case, + returning less than, equal to or greater than zero if S1 is + lexicographically less than, equal to or greater than S2. + Note: This function cannot work correctly in multibyte locales. */ + +int +strncasecmp (const char *s1, const char *s2, size_t n) +{ + register const unsigned char *p1 = (const unsigned char *) s1; + register const unsigned char *p2 = (const unsigned char *) s2; + unsigned char c1, c2; + + if (p1 == p2 || n == 0) + return 0; + + do + { + c1 = TOLOWER (*p1); + c2 = TOLOWER (*p2); + + if (--n == 0 || c1 == '\0') + break; + + ++p1; + ++p2; + } + while (c1 == c2); + + if (UCHAR_MAX <= INT_MAX) + return c1 - c2; + else + /* On machines where 'char' and 'int' are types of the same size, the + difference of two 'unsigned char' values - including the sign bit - + doesn't fit in an 'int'. */ + return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0); +} diff --git a/lib/strndup.c b/lib/strndup.c new file mode 100644 index 0000000..3a1b0ea --- /dev/null +++ b/lib/strndup.c @@ -0,0 +1,37 @@ +/* A replacement function, for systems that lack strndup. + + Copyright (C) 1996, 1997, 1998, 2001, 2002, 2003, 2005, 2006, 2007 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <string.h> + +#include <stdlib.h> + +char * +strndup (char const *s, size_t n) +{ + size_t len = strnlen (s, n); + char *new = malloc (len + 1); + + if (new == NULL) + return NULL; + + new[len] = '\0'; + return memcpy (new, s, len); +} diff --git a/lib/strnlen.c b/lib/strnlen.c new file mode 100644 index 0000000..d346d32 --- /dev/null +++ b/lib/strnlen.c @@ -0,0 +1,31 @@ +/* Find the length of STRING, but scan at most MAXLEN characters. + Copyright (C) 2005, 2006, 2007 Free Software Foundation, Inc. + Written by Simon Josefsson. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <string.h> + +/* Find the length of STRING, but scan at most MAXLEN characters. + If no '\0' terminator is found in that many characters, return MAXLEN. */ + +size_t +strnlen (const char *string, size_t maxlen) +{ + const char *end = memchr (string, '\0', maxlen); + return end ? (size_t) (end - string) : maxlen; +} diff --git a/lib/strnlen1.c b/lib/strnlen1.c new file mode 100644 index 0000000..422ed9e --- /dev/null +++ b/lib/strnlen1.c @@ -0,0 +1,36 @@ +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + Copyright (C) 2005-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "strnlen1.h" + +#include <string.h> + +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + If no '\0' terminator is found in that many characters, return MAXLEN. */ +/* This is the same as strnlen (string, maxlen - 1) + 1. */ +size_t +strnlen1 (const char *string, size_t maxlen) +{ + const char *end = (const char *) memchr (string, '\0', maxlen); + if (end != NULL) + return end - string + 1; + else + return maxlen; +} diff --git a/lib/strnlen1.h b/lib/strnlen1.h new file mode 100644 index 0000000..7ce7d0c --- /dev/null +++ b/lib/strnlen1.h @@ -0,0 +1,40 @@ +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _STRNLEN1_H +#define _STRNLEN1_H + +#include <stddef.h> + + +#ifdef __cplusplus +extern "C" { +#endif + + +/* Find the length of STRING + 1, but scan at most MAXLEN bytes. + If no '\0' terminator is found in that many characters, return MAXLEN. */ +/* This is the same as strnlen (string, maxlen - 1) + 1. */ +extern size_t strnlen1 (const char *string, size_t maxlen); + + +#ifdef __cplusplus +} +#endif + + +#endif /* _STRNLEN1_H */ diff --git a/lib/strtoimax.c b/lib/strtoimax.c new file mode 100644 index 0000000..82dc4f0 --- /dev/null +++ b/lib/strtoimax.c @@ -0,0 +1,76 @@ +/* Convert string representation of a number into an intmax_t value. + + Copyright (C) 1999, 2001, 2002, 2003, 2004, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +/* Verify interface. */ +#include <inttypes.h> + +#include <stdlib.h> + +#include "verify.h" + +#ifdef UNSIGNED +# ifndef HAVE_DECL_STRTOULL +"this configure-time declaration test was not run" +# endif +# if !HAVE_DECL_STRTOULL && HAVE_UNSIGNED_LONG_LONG_INT +unsigned long long int strtoull (char const *, char **, int); +# endif + +#else + +# ifndef HAVE_DECL_STRTOLL +"this configure-time declaration test was not run" +# endif +# if !HAVE_DECL_STRTOLL && HAVE_LONG_LONG_INT +long long int strtoll (char const *, char **, int); +# endif +#endif + +#ifdef UNSIGNED +# define Have_long_long HAVE_UNSIGNED_LONG_LONG_INT +# define Int uintmax_t +# define Unsigned unsigned +# define strtoimax strtoumax +# define strtol strtoul +# define strtoll strtoull +#else +# define Have_long_long HAVE_LONG_LONG_INT +# define Int intmax_t +# define Unsigned +#endif + +Int +strtoimax (char const *ptr, char **endptr, int base) +{ +#if Have_long_long + verify (sizeof (Int) == sizeof (Unsigned long int) + || sizeof (Int) == sizeof (Unsigned long long int)); + + if (sizeof (Int) != sizeof (Unsigned long int)) + return strtoll (ptr, endptr, base); +#else + verify (sizeof (Int) == sizeof (Unsigned long int)); +#endif + + return strtol (ptr, endptr, base); +} diff --git a/lib/strtol.c b/lib/strtol.c new file mode 100644 index 0000000..e14d3cf --- /dev/null +++ b/lib/strtol.c @@ -0,0 +1,436 @@ +/* Convert string representation of a number into an integer value. + + Copyright (C) 1991, 1992, 1994, 1995, 1996, 1997, 1998, 1999, 2003, 2005, + 2006, 2007 + Free Software Foundation, Inc. + + NOTE: The canonical source of this file is maintained with the GNU C + Library. Bugs can be reported to bug-glibc@gnu.org. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef _LIBC +# define USE_NUMBER_GROUPING +#else +# include <config.h> +#endif + +#include <ctype.h> +#include <errno.h> +#ifndef __set_errno +# define __set_errno(Val) errno = (Val) +#endif + +#include <limits.h> +#include <stddef.h> +#include <stdlib.h> +#include <string.h> + +#ifdef USE_NUMBER_GROUPING +# include "../locale/localeinfo.h" +#endif + +/* Nonzero if we are defining `strtoul' or `strtoull', operating on + unsigned integers. */ +#ifndef UNSIGNED +# define UNSIGNED 0 +# define INT LONG int +#else +# define INT unsigned LONG int +#endif + +/* Determine the name. */ +#ifdef USE_IN_EXTENDED_LOCALE_MODEL +# if UNSIGNED +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol __wcstoull_l +# else +# define strtol __wcstoul_l +# endif +# else +# ifdef QUAD +# define strtol __strtoull_l +# else +# define strtol __strtoul_l +# endif +# endif +# else +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol __wcstoll_l +# else +# define strtol __wcstol_l +# endif +# else +# ifdef QUAD +# define strtol __strtoll_l +# else +# define strtol __strtol_l +# endif +# endif +# endif +#else +# if UNSIGNED +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol wcstoull +# else +# define strtol wcstoul +# endif +# else +# ifdef QUAD +# define strtol strtoull +# else +# define strtol strtoul +# endif +# endif +# else +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol wcstoll +# else +# define strtol wcstol +# endif +# else +# ifdef QUAD +# define strtol strtoll +# endif +# endif +# endif +#endif + +/* If QUAD is defined, we are defining `strtoll' or `strtoull', + operating on `long long int's. */ +#ifdef QUAD +# define LONG long long +# define STRTOL_LONG_MIN LONG_LONG_MIN +# define STRTOL_LONG_MAX LONG_LONG_MAX +# define STRTOL_ULONG_MAX ULONG_LONG_MAX + +/* The extra casts in the following macros work around compiler bugs, + e.g., in Cray C 5.0.3.0. */ + +/* True if negative values of the signed integer type T use two's + complement, ones' complement, or signed magnitude representation, + respectively. Much GNU code assumes two's complement, but some + people like to be portable to all possible C hosts. */ +# define TYPE_TWOS_COMPLEMENT(t) ((t) ~ (t) 0 == (t) -1) +# define TYPE_ONES_COMPLEMENT(t) ((t) ~ (t) 0 == 0) +# define TYPE_SIGNED_MAGNITUDE(t) ((t) ~ (t) 0 < (t) -1) + +/* True if the arithmetic type T is signed. */ +# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) + +/* The maximum and minimum values for the integer type T. These + macros have undefined behavior if T is signed and has padding bits. + If this is a problem for you, please let us know how to fix it for + your host. */ +# define TYPE_MINIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) 0 \ + : TYPE_SIGNED_MAGNITUDE (t) \ + ? ~ (t) 0 \ + : ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1))) +# define TYPE_MAXIMUM(t) \ + ((t) (! TYPE_SIGNED (t) \ + ? (t) -1 \ + : ~ (~ (t) 0 << (sizeof (t) * CHAR_BIT - 1)))) + +# ifndef ULONG_LONG_MAX +# define ULONG_LONG_MAX TYPE_MAXIMUM (unsigned long long) +# endif +# ifndef LONG_LONG_MAX +# define LONG_LONG_MAX TYPE_MAXIMUM (long long int) +# endif +# ifndef LONG_LONG_MIN +# define LONG_LONG_MIN TYPE_MINIMUM (long long int) +# endif + +# if __GNUC__ == 2 && __GNUC_MINOR__ < 7 + /* Work around gcc bug with using this constant. */ + static const unsigned long long int maxquad = ULONG_LONG_MAX; +# undef STRTOL_ULONG_MAX +# define STRTOL_ULONG_MAX maxquad +# endif +#else +# define LONG long +# define STRTOL_LONG_MIN LONG_MIN +# define STRTOL_LONG_MAX LONG_MAX +# define STRTOL_ULONG_MAX ULONG_MAX +#endif + + +/* We use this code also for the extended locale handling where the + function gets as an additional argument the locale which has to be + used. To access the values we have to redefine the _NL_CURRENT + macro. */ +#ifdef USE_IN_EXTENDED_LOCALE_MODEL +# undef _NL_CURRENT +# define _NL_CURRENT(category, item) \ + (current->values[_NL_ITEM_INDEX (item)].string) +# define LOCALE_PARAM , loc +# define LOCALE_PARAM_PROTO , __locale_t loc +#else +# define LOCALE_PARAM +# define LOCALE_PARAM_PROTO +#endif + +#include <wchar.h> + +#ifdef USE_WIDE_CHAR +# include <wctype.h> +# define L_(Ch) L##Ch +# define UCHAR_TYPE wint_t +# define STRING_TYPE wchar_t +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define ISSPACE(Ch) __iswspace_l ((Ch), loc) +# define ISALPHA(Ch) __iswalpha_l ((Ch), loc) +# define TOUPPER(Ch) __towupper_l ((Ch), loc) +# else +# define ISSPACE(Ch) iswspace (Ch) +# define ISALPHA(Ch) iswalpha (Ch) +# define TOUPPER(Ch) towupper (Ch) +# endif +#else +# define L_(Ch) Ch +# define UCHAR_TYPE unsigned char +# define STRING_TYPE char +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define ISSPACE(Ch) __isspace_l ((Ch), loc) +# define ISALPHA(Ch) __isalpha_l ((Ch), loc) +# define TOUPPER(Ch) __toupper_l ((Ch), loc) +# else +# define ISSPACE(Ch) isspace (Ch) +# define ISALPHA(Ch) isalpha (Ch) +# define TOUPPER(Ch) toupper (Ch) +# endif +#endif + +#define INTERNAL(X) INTERNAL1(X) +#define INTERNAL1(X) __##X##_internal +#define WEAKNAME(X) WEAKNAME1(X) + +#ifdef USE_NUMBER_GROUPING +/* This file defines a function to check for correct grouping. */ +# include "grouping.h" +#endif + + + +/* Convert NPTR to an `unsigned long int' or `long int' in base BASE. + If BASE is 0 the base is determined by the presence of a leading + zero, indicating octal or a leading "0x" or "0X", indicating hexadecimal. + If BASE is < 2 or > 36, it is reset to 10. + If ENDPTR is not NULL, a pointer to the character after the last + one converted is stored in *ENDPTR. */ + +INT +INTERNAL (strtol) (const STRING_TYPE *nptr, STRING_TYPE **endptr, + int base, int group LOCALE_PARAM_PROTO) +{ + int negative; + register unsigned LONG int cutoff; + register unsigned int cutlim; + register unsigned LONG int i; + register const STRING_TYPE *s; + register UCHAR_TYPE c; + const STRING_TYPE *save, *end; + int overflow; + +#ifdef USE_NUMBER_GROUPING +# ifdef USE_IN_EXTENDED_LOCALE_MODEL + struct locale_data *current = loc->__locales[LC_NUMERIC]; +# endif + /* The thousands character of the current locale. */ + wchar_t thousands = L'\0'; + /* The numeric grouping specification of the current locale, + in the format described in <locale.h>. */ + const char *grouping; + + if (group) + { + grouping = _NL_CURRENT (LC_NUMERIC, GROUPING); + if (*grouping <= 0 || *grouping == CHAR_MAX) + grouping = NULL; + else + { + /* Figure out the thousands separator character. */ +# if defined _LIBC || defined _HAVE_BTOWC + thousands = __btowc (*_NL_CURRENT (LC_NUMERIC, THOUSANDS_SEP)); + if (thousands == WEOF) + thousands = L'\0'; +# endif + if (thousands == L'\0') + grouping = NULL; + } + } + else + grouping = NULL; +#endif + + if (base < 0 || base == 1 || base > 36) + { + __set_errno (EINVAL); + return 0; + } + + save = s = nptr; + + /* Skip white space. */ + while (ISSPACE (*s)) + ++s; + if (*s == L_('\0')) + goto noconv; + + /* Check for a sign. */ + if (*s == L_('-')) + { + negative = 1; + ++s; + } + else if (*s == L_('+')) + { + negative = 0; + ++s; + } + else + negative = 0; + + /* Recognize number prefix and if BASE is zero, figure it out ourselves. */ + if (*s == L_('0')) + { + if ((base == 0 || base == 16) && TOUPPER (s[1]) == L_('X')) + { + s += 2; + base = 16; + } + else if (base == 0) + base = 8; + } + else if (base == 0) + base = 10; + + /* Save the pointer so we can check later if anything happened. */ + save = s; + +#ifdef USE_NUMBER_GROUPING + if (group) + { + /* Find the end of the digit string and check its grouping. */ + end = s; + for (c = *end; c != L_('\0'); c = *++end) + if ((wchar_t) c != thousands + && ((wchar_t) c < L_('0') || (wchar_t) c > L_('9')) + && (!ISALPHA (c) || (int) (TOUPPER (c) - L_('A') + 10) >= base)) + break; + if (*s == thousands) + end = s; + else + end = correctly_grouped_prefix (s, end, thousands, grouping); + } + else +#endif + end = NULL; + + cutoff = STRTOL_ULONG_MAX / (unsigned LONG int) base; + cutlim = STRTOL_ULONG_MAX % (unsigned LONG int) base; + + overflow = 0; + i = 0; + for (c = *s; c != L_('\0'); c = *++s) + { + if (s == end) + break; + if (c >= L_('0') && c <= L_('9')) + c -= L_('0'); + else if (ISALPHA (c)) + c = TOUPPER (c) - L_('A') + 10; + else + break; + if ((int) c >= base) + break; + /* Check for overflow. */ + if (i > cutoff || (i == cutoff && c > cutlim)) + overflow = 1; + else + { + i *= (unsigned LONG int) base; + i += c; + } + } + + /* Check if anything actually happened. */ + if (s == save) + goto noconv; + + /* Store in ENDPTR the address of one character + past the last character we converted. */ + if (endptr != NULL) + *endptr = (STRING_TYPE *) s; + +#if !UNSIGNED + /* Check for a value that is within the range of + `unsigned LONG int', but outside the range of `LONG int'. */ + if (overflow == 0 + && i > (negative + ? -((unsigned LONG int) (STRTOL_LONG_MIN + 1)) + 1 + : (unsigned LONG int) STRTOL_LONG_MAX)) + overflow = 1; +#endif + + if (overflow) + { + __set_errno (ERANGE); +#if UNSIGNED + return STRTOL_ULONG_MAX; +#else + return negative ? STRTOL_LONG_MIN : STRTOL_LONG_MAX; +#endif + } + + /* Return the result of the appropriate sign. */ + return negative ? -i : i; + +noconv: + /* We must handle a special case here: the base is 0 or 16 and the + first two characters are '0' and 'x', but the rest are no + hexadecimal digits. This is no error case. We return 0 and + ENDPTR points to the `x`. */ + if (endptr != NULL) + { + if (save - nptr >= 2 && TOUPPER (save[-1]) == L_('X') + && save[-2] == L_('0')) + *endptr = (STRING_TYPE *) &save[-1]; + else + /* There was no number to convert. */ + *endptr = (STRING_TYPE *) nptr; + } + + return 0L; +} + +/* External user entry point. */ + + +INT +#ifdef weak_function +weak_function +#endif +strtol (const STRING_TYPE *nptr, STRING_TYPE **endptr, + int base LOCALE_PARAM_PROTO) +{ + return INTERNAL (strtol) (nptr, endptr, base, 0 LOCALE_PARAM); +} diff --git a/lib/strtoll.c b/lib/strtoll.c new file mode 100644 index 0000000..f61f5ad --- /dev/null +++ b/lib/strtoll.c @@ -0,0 +1,33 @@ +/* Function to parse a `long long int' from text. + Copyright (C) 1995, 1996, 1997, 1999, 2001 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#define QUAD 1 + +#include <strtol.c> + +#ifdef _LIBC +# ifdef SHARED +# include <shlib-compat.h> + +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_2) +compat_symbol (libc, __strtoll_internal, __strtoq_internal, GLIBC_2_0); +# endif + +# endif +weak_alias (strtoll, strtoq) +#endif diff --git a/lib/strtoul.c b/lib/strtoul.c new file mode 100644 index 0000000..79ceed2 --- /dev/null +++ b/lib/strtoul.c @@ -0,0 +1,20 @@ +/* Copyright (C) 1991, 1997 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#define UNSIGNED 1 + +#include "strtol.c" diff --git a/lib/strtoull.c b/lib/strtoull.c new file mode 100644 index 0000000..6186bfb --- /dev/null +++ b/lib/strtoull.c @@ -0,0 +1,27 @@ +/* Function to parse an `unsigned long long int' from text. + Copyright (C) 1995, 1996, 1997, 1999 Free Software Foundation, Inc. + NOTE: The canonical source of this file is maintained with the GNU C + Library. Bugs can be reported to bug-glibc@gnu.org. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#define QUAD 1 + +#include "strtoul.c" + +#ifdef _LIBC +strong_alias (__strtoull_internal, __strtouq_internal) +weak_alias (strtoull, strtouq) +#endif diff --git a/lib/strtoumax.c b/lib/strtoumax.c new file mode 100644 index 0000000..dc395d6 --- /dev/null +++ b/lib/strtoumax.c @@ -0,0 +1,2 @@ +#define UNSIGNED 1 +#include "strtoimax.c" diff --git a/lib/sys_stat_.h b/lib/sys_stat_.h new file mode 100644 index 0000000..13333d3 --- /dev/null +++ b/lib/sys_stat_.h @@ -0,0 +1,284 @@ +/* Provide a more complete sys/stat header file. + Copyright (C) 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Eric Blake, Paul Eggert, and Jim Meyering. */ + +/* This file is supposed to be used on platforms where <sys/stat.h> is + incomplete. It is intended to provide definitions and prototypes + needed by an application. Start with what the system provides. */ + +#ifndef _GL_SYS_STAT_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <sys/stat.h> +#else +# include @ABSOLUTE_SYS_STAT_H@ +#endif + +#ifndef _GL_SYS_STAT_H +#define _GL_SYS_STAT_H + +#ifndef S_IFMT +# define S_IFMT 0170000 +#endif + +#if STAT_MACROS_BROKEN +# undef S_ISBLK +# undef S_ISCHR +# undef S_ISDIR +# undef S_ISFIFO +# undef S_ISLNK +# undef S_ISNAM +# undef S_ISMPB +# undef S_ISMPC +# undef S_ISNWK +# undef S_ISREG +# undef S_ISSOCK +#endif + +#ifndef S_ISBLK +# ifdef S_IFBLK +# define S_ISBLK(m) (((m) & S_IFMT) == S_IFBLK) +# else +# define S_ISBLK(m) 0 +# endif +#endif + +#ifndef S_ISCHR +# ifdef S_IFCHR +# define S_ISCHR(m) (((m) & S_IFMT) == S_IFCHR) +# else +# define S_ISCHR(m) 0 +# endif +#endif + +#ifndef S_ISDIR +# ifdef S_IFDIR +# define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +# else +# define S_ISDIR(m) 0 +# endif +#endif + +#ifndef S_ISDOOR /* Solaris 2.5 and up */ +# define S_ISDOOR(m) 0 +#endif + +#ifndef S_ISFIFO +# ifdef S_IFIFO +# define S_ISFIFO(m) (((m) & S_IFMT) == S_IFIFO) +# else +# define S_ISFIFO(m) 0 +# endif +#endif + +#ifndef S_ISLNK +# ifdef S_IFLNK +# define S_ISLNK(m) (((m) & S_IFMT) == S_IFLNK) +# else +# define S_ISLNK(m) 0 +# endif +#endif + +#ifndef S_ISMPB /* V7 */ +# ifdef S_IFMPB +# define S_ISMPB(m) (((m) & S_IFMT) == S_IFMPB) +# define S_ISMPC(m) (((m) & S_IFMT) == S_IFMPC) +# else +# define S_ISMPB(m) 0 +# define S_ISMPC(m) 0 +# endif +#endif + +#ifndef S_ISNAM /* Xenix */ +# ifdef S_IFNAM +# define S_ISNAM(m) (((m) & S_IFMT) == S_IFNAM) +# else +# define S_ISNAM(m) 0 +# endif +#endif + +#ifndef S_ISNWK /* HP/UX */ +# ifdef S_IFNWK +# define S_ISNWK(m) (((m) & S_IFMT) == S_IFNWK) +# else +# define S_ISNWK(m) 0 +# endif +#endif + +#ifndef S_ISPORT /* Solaris 10 and up */ +# define S_ISPORT(m) 0 +#endif + +#ifndef S_ISREG +# ifdef S_IFREG +# define S_ISREG(m) (((m) & S_IFMT) == S_IFREG) +# else +# define S_ISREG(m) 0 +# endif +#endif + +#ifndef S_ISSOCK +# ifdef S_IFSOCK +# define S_ISSOCK(m) (((m) & S_IFMT) == S_IFSOCK) +# else +# define S_ISSOCK(m) 0 +# endif +#endif + + +#ifndef S_TYPEISMQ +# define S_TYPEISMQ(p) 0 +#endif + +#ifndef S_TYPEISTMO +# define S_TYPEISTMO(p) 0 +#endif + + +#ifndef S_TYPEISSEM +# ifdef S_INSEM +# define S_TYPEISSEM(p) (S_ISNAM ((p)->st_mode) && (p)->st_rdev == S_INSEM) +# else +# define S_TYPEISSEM(p) 0 +# endif +#endif + +#ifndef S_TYPEISSHM +# ifdef S_INSHD +# define S_TYPEISSHM(p) (S_ISNAM ((p)->st_mode) && (p)->st_rdev == S_INSHD) +# else +# define S_TYPEISSHM(p) 0 +# endif +#endif + +/* high performance ("contiguous data") */ +#ifndef S_ISCTG +# define S_ISCTG(p) 0 +#endif + +/* Cray DMF (data migration facility): off line, with data */ +#ifndef S_ISOFD +# define S_ISOFD(p) 0 +#endif + +/* Cray DMF (data migration facility): off line, with no data */ +#ifndef S_ISOFL +# define S_ISOFL(p) 0 +#endif + +/* 4.4BSD whiteout */ +#ifndef S_ISWHT +# define S_ISWHT(m) 0 +#endif + +/* If any of the following are undefined, + define them to their de facto standard values. */ +#if !S_ISUID +# define S_ISUID 04000 +#endif +#if !S_ISGID +# define S_ISGID 02000 +#endif + +/* S_ISVTX is a common extension to POSIX. */ +#ifndef S_ISVTX +# define S_ISVTX 01000 +#endif + +#if !S_IRUSR && S_IREAD +# define S_IRUSR S_IREAD +#endif +#if !S_IRUSR +# define S_IRUSR 00400 +#endif +#if !S_IRGRP +# define S_IRGRP (S_IRUSR >> 3) +#endif +#if !S_IROTH +# define S_IROTH (S_IRUSR >> 6) +#endif + +#if !S_IWUSR && S_IWRITE +# define S_IWUSR S_IWRITE +#endif +#if !S_IWUSR +# define S_IWUSR 00200 +#endif +#if !S_IWGRP +# define S_IWGRP (S_IWUSR >> 3) +#endif +#if !S_IWOTH +# define S_IWOTH (S_IWUSR >> 6) +#endif + +#if !S_IXUSR && S_IEXEC +# define S_IXUSR S_IEXEC +#endif +#if !S_IXUSR +# define S_IXUSR 00100 +#endif +#if !S_IXGRP +# define S_IXGRP (S_IXUSR >> 3) +#endif +#if !S_IXOTH +# define S_IXOTH (S_IXUSR >> 6) +#endif + +#if !S_IRWXU +# define S_IRWXU (S_IRUSR | S_IWUSR | S_IXUSR) +#endif +#if !S_IRWXG +# define S_IRWXG (S_IRGRP | S_IWGRP | S_IXGRP) +#endif +#if !S_IRWXO +# define S_IRWXO (S_IROTH | S_IWOTH | S_IXOTH) +#endif + +/* S_IXUGO is a common extension to POSIX. */ +#if !S_IXUGO +# define S_IXUGO (S_IXUSR | S_IXGRP | S_IXOTH) +#endif + +#ifndef S_IRWXUGO +# define S_IRWXUGO (S_IRWXU | S_IRWXG | S_IRWXO) +#endif + +/* mingw does not support symlinks, therefore it does not have lstat. But + without links, stat does just fine. */ +#if ! @HAVE_LSTAT@ +# define lstat stat +#endif + +/* mingw's _mkdir() function has 1 argument, but we pass 2 arguments. + Additionally, it declares _mkdir (and depending on compile flags, an + alias mkdir), only in the nonstandard io.h. */ +#if ! @HAVE_DECL_MKDIR@ && @HAVE_IO_H@ +# include <io.h> + +static inline int +rpl_mkdir (char const *name, mode_t mode) +{ + return _mkdir (name); +} + +# define mkdir rpl_mkdir +#endif + +#endif /* _GL_SYS_STAT_H */ +#endif /* _GL_SYS_STAT_H */ diff --git a/lib/sys_time_.h b/lib/sys_time_.h new file mode 100644 index 0000000..253316f --- /dev/null +++ b/lib/sys_time_.h @@ -0,0 +1,52 @@ +/* Provide a more complete sys/time.h. + + Copyright (C) 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#ifndef _GL_SYS_TIME_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_SYS_TIME_H@ +# if @HAVE_INCLUDE_NEXT@ +# include_next <sys/time.h> +# else +# include @ABSOLUTE_SYS_TIME_H@ +# endif +#else +# include <time.h> +#endif + +#ifndef _GL_SYS_TIME_H +#define _GL_SYS_TIME_H + +#if ! @HAVE_STRUCT_TIMEVAL@ +struct timeval +{ + time_t tv_sec; + long int tv_usec; +}; +#endif + +#if @REPLACE_GETTIMEOFDAY@ +# undef gettimeofday +# define gettimeofday rpl_gettimeofday +int gettimeofday (struct timeval *restrict, void *restrict); +#endif + +#endif /* _GL_SYS_TIME_H */ +#endif /* _GL_SYS_TIME_H */ diff --git a/lib/sysexits_.h b/lib/sysexits_.h new file mode 100644 index 0000000..ba12aa2 --- /dev/null +++ b/lib/sysexits_.h @@ -0,0 +1,72 @@ +/* exit() exit codes for some BSD system programs. + Copyright (C) 2003, 2006-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Simon Josefsson based on sysexits(3) man page */ + +#ifndef _GL_SYSEXITS_H + +#if @HAVE_SYSEXITS_H@ + +/* IRIX 6.5 has an <unistd.h> that defines a macro EX_OK with a nonzero + value. Override it. See + <http://lists.gnu.org/archive/html/bug-gnulib/2007-03/msg00361.html> */ +# ifdef __sgi +# include <unistd.h> +# undef EX_OK +# endif + +/* The include_next requires a split double-inclusion guard. */ +# if @HAVE_INCLUDE_NEXT@ +# include_next <sysexits.h> +# else +# include @ABSOLUTE_SYSEXITS_H@ +# endif + +/* HP-UX 11 <sysexits.h> ends at EX_NOPERM. */ +# ifndef EX_CONFIG +# define EX_CONFIG 78 +# endif + +#endif + +#ifndef _GL_SYSEXITS_H +#define _GL_SYSEXITS_H + +#if !@HAVE_SYSEXITS_H@ + +# define EX_OK 0 /* same value as EXIT_SUCCESS */ + +# define EX_USAGE 64 +# define EX_DATAERR 65 +# define EX_NOINPUT 66 +# define EX_NOUSER 67 +# define EX_NOHOST 68 +# define EX_UNAVAILABLE 69 +# define EX_SOFTWARE 70 +# define EX_OSERR 71 +# define EX_OSFILE 72 +# define EX_CANTCREAT 73 +# define EX_IOERR 74 +# define EX_TEMPFAIL 75 +# define EX_PROTOCOL 76 +# define EX_NOPERM 77 +# define EX_CONFIG 78 + +#endif + +#endif /* _GL_SYSEXITS_H */ +#endif /* _GL_SYSEXITS_H */ diff --git a/lib/system-ioctl.h b/lib/system-ioctl.h new file mode 100644 index 0000000..a61c5fd --- /dev/null +++ b/lib/system-ioctl.h @@ -0,0 +1,55 @@ +/* System dependent definitions for GNU tar's use of ioctl macros. + + Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003, + 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* This is a real challenge to properly get MTIO* symbols :-(. ISC uses + <sys/gentape.h>. SCO and BSDi uses <sys/tape.h>; BSDi also requires + <sys/tprintf.h> and <sys/device.h> for defining tp_dev and tpr_t. It + seems that the rest use <sys/mtio.h>, which itself requires other files, + depending on systems. Pyramid defines _IOW in <sgtty.h>, for example. */ + +#if HAVE_SYS_GENTAPE_H +# include <sys/gentape.h> +#else +# if HAVE_SYS_TAPE_H +# if HAVE_SYS_DEVICE_H +# include <sys/device.h> +# endif +# if HAVE_SYS_PARAM_H +# include <sys/param.h> +# endif +# if HAVE_SYS_BUF_H +# include <sys/buf.h> +# endif +# if HAVE_SYS_TPRINTF_H +# include <sys/tprintf.h> +# endif +# include <sys/tape.h> +# else +# if HAVE_SYS_MTIO_H +# include <sys/ioctl.h> +# if HAVE_SGTTY_H +# include <sgtty.h> +# endif +# if HAVE_SYS_IO_TRIOCTL_H +# include <sys/io/trioctl.h> +# endif +# include <sys/mtio.h> +# endif +# endif +#endif diff --git a/lib/system.h b/lib/system.h new file mode 100644 index 0000000..7602263 --- /dev/null +++ b/lib/system.h @@ -0,0 +1,475 @@ +/* System dependent definitions for GNU tar. + + Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003, + 2004, 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#include <alloca.h> + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(spec) /* empty */ +# endif +#endif + +#include <sys/types.h> +#include <ctype.h> + +/* IN_CTYPE_DOMAIN (C) is nonzero if the unsigned char C can safely be given + as an argument to <ctype.h> macros like `isspace'. */ +#if STDC_HEADERS +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) ((unsigned) (c) <= 0177) +#endif + +#define ISDIGIT(c) ((unsigned) (c) - '0' <= 9) +#define ISODIGIT(c) ((unsigned) (c) - '0' <= 7) +#define ISPRINT(c) (IN_CTYPE_DOMAIN (c) && isprint (c)) +#define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c)) + +/* Declare string and memory handling routines. Take care that an ANSI + string.h and pre-ANSI memory.h might conflict, and that memory.h and + strings.h conflict on some systems. */ + +#if STDC_HEADERS || HAVE_STRING_H +# include <string.h> +# if !STDC_HEADERS && HAVE_MEMORY_H +# include <memory.h> +# endif +#else +# include <strings.h> +# ifndef strchr +# define strchr index +# endif +# ifndef strrchr +# define strrchr rindex +# endif +# ifndef memcpy +# define memcpy(d, s, n) bcopy ((char const *) (s), (char *) (d), n) +# endif +# ifndef memcmp +# define memcmp(a, b, n) bcmp ((char const *) (a), (char const *) (b), n) +# endif +#endif + +/* Declare errno. */ + +#include <errno.h> +#ifndef errno +extern int errno; +#endif + +/* Declare open parameters. */ + +#if HAVE_FCNTL_H +# include <fcntl.h> +#else +# include <sys/file.h> +#endif + /* Pick only one of the next three: */ +#ifndef O_RDONLY +# define O_RDONLY 0 /* only allow read */ +#endif +#ifndef O_WRONLY +# define O_WRONLY 1 /* only allow write */ +#endif +#ifndef O_RDWR +# define O_RDWR 2 /* both are allowed */ +#endif +#ifndef O_ACCMODE +# define O_ACCMODE (O_RDONLY | O_RDWR | O_WRONLY) +#endif + /* The rest can be OR-ed in to the above: */ +#ifndef O_CREAT +# define O_CREAT 8 /* create file if needed */ +#endif +#ifndef O_EXCL +# define O_EXCL 16 /* file cannot already exist */ +#endif +#ifndef O_TRUNC +# define O_TRUNC 32 /* truncate file on open */ +#endif + +#ifndef O_BINARY +# define O_BINARY 0 +#endif +#ifndef O_DIRECTORY +# define O_DIRECTORY 0 +#endif +#ifndef O_NOATIME +# define O_NOATIME 0 +#endif +#ifndef O_NONBLOCK +# define O_NONBLOCK 0 +#endif + +/* Declare file status routines and bits. */ + +#include <sys/stat.h> + +#if !HAVE_LSTAT && !defined lstat +# define lstat stat +#endif + +#if STX_HIDDEN && !_LARGE_FILES /* AIX */ +# ifdef stat +# undef stat +# endif +# define stat(file_name, buf) statx (file_name, buf, STATSIZE, STX_HIDDEN) +# ifdef lstat +# undef lstat +# endif +# define lstat(file_name, buf) statx (file_name, buf, STATSIZE, STX_HIDDEN | STX_LINK) +#endif + +#if STAT_MACROS_BROKEN +# undef S_ISBLK +# undef S_ISCHR +# undef S_ISCTG +# undef S_ISDIR +# undef S_ISFIFO +# undef S_ISLNK +# undef S_ISREG +# undef S_ISSOCK +#endif + +/* On MSDOS, there are missing things from <sys/stat.h>. */ +#if MSDOS +# define S_ISUID 0 +# define S_ISGID 0 +# define S_ISVTX 0 +#endif + +#ifndef S_ISDIR +# define S_ISDIR(mode) (((mode) & S_IFMT) == S_IFDIR) +#endif +#ifndef S_ISREG +# define S_ISREG(mode) (((mode) & S_IFMT) == S_IFREG) +#endif + +#ifndef S_ISBLK +# ifdef S_IFBLK +# define S_ISBLK(mode) (((mode) & S_IFMT) == S_IFBLK) +# else +# define S_ISBLK(mode) 0 +# endif +#endif +#ifndef S_ISCHR +# ifdef S_IFCHR +# define S_ISCHR(mode) (((mode) & S_IFMT) == S_IFCHR) +# else +# define S_ISCHR(mode) 0 +# endif +#endif +#ifndef S_ISCTG +# ifdef S_IFCTG +# define S_ISCTG(mode) (((mode) & S_IFMT) == S_IFCTG) +# else +# define S_ISCTG(mode) 0 +# endif +#endif +#ifndef S_ISDOOR +# define S_ISDOOR(mode) 0 +#endif +#ifndef S_ISFIFO +# ifdef S_IFIFO +# define S_ISFIFO(mode) (((mode) & S_IFMT) == S_IFIFO) +# else +# define S_ISFIFO(mode) 0 +# endif +#endif +#ifndef S_ISLNK +# ifdef S_IFLNK +# define S_ISLNK(mode) (((mode) & S_IFMT) == S_IFLNK) +# else +# define S_ISLNK(mode) 0 +# endif +#endif +#ifndef S_ISSOCK +# ifdef S_IFSOCK +# define S_ISSOCK(mode) (((mode) & S_IFMT) == S_IFSOCK) +# else +# define S_ISSOCK(mode) 0 +# endif +#endif + +#if !HAVE_MKFIFO && !defined mkfifo && defined S_IFIFO +# define mkfifo(file_name, mode) (mknod (file_name, (mode) | S_IFIFO, 0)) +#endif + +#ifndef S_ISUID +# define S_ISUID 0004000 +#endif +#ifndef S_ISGID +# define S_ISGID 0002000 +#endif +#ifndef S_ISVTX +# define S_ISVTX 0001000 +#endif +#ifndef S_IRUSR +# define S_IRUSR 0000400 +#endif +#ifndef S_IWUSR +# define S_IWUSR 0000200 +#endif +#ifndef S_IXUSR +# define S_IXUSR 0000100 +#endif +#ifndef S_IRGRP +# define S_IRGRP 0000040 +#endif +#ifndef S_IWGRP +# define S_IWGRP 0000020 +#endif +#ifndef S_IXGRP +# define S_IXGRP 0000010 +#endif +#ifndef S_IROTH +# define S_IROTH 0000004 +#endif +#ifndef S_IWOTH +# define S_IWOTH 0000002 +#endif +#ifndef S_IXOTH +# define S_IXOTH 0000001 +#endif + +#define MODE_WXUSR (S_IWUSR | S_IXUSR) +#define MODE_R (S_IRUSR | S_IRGRP | S_IROTH) +#define MODE_RW (S_IWUSR | S_IWGRP | S_IWOTH | MODE_R) +#define MODE_RWX (S_IXUSR | S_IXGRP | S_IXOTH | MODE_RW) +#define MODE_ALL (S_ISUID | S_ISGID | S_ISVTX | MODE_RWX) + +/* Include <unistd.h> before any preprocessor test of _POSIX_VERSION. */ +#include <unistd.h> + +#ifndef SEEK_SET +# define SEEK_SET 0 +#endif +#ifndef SEEK_CUR +# define SEEK_CUR 1 +#endif +#ifndef SEEK_END +# define SEEK_END 2 +#endif + +#ifndef STDIN_FILENO +# define STDIN_FILENO 0 +#endif +#ifndef STDOUT_FILENO +# define STDOUT_FILENO 1 +#endif +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +/* Declare make device, major and minor. Since major is a function on + SVR4, we have to resort to GOT_MAJOR instead of just testing if + major is #define'd. */ + +#if MAJOR_IN_MKDEV +# include <sys/mkdev.h> +# if !defined(makedev) && defined(mkdev) +# define makedev(a,b) mkdev((a),(b)) +# endif +# define GOT_MAJOR +#endif + +#if MAJOR_IN_SYSMACROS +# include <sys/sysmacros.h> +# define GOT_MAJOR +#endif + +/* Some <sys/types.h> defines the macros. */ +#ifdef major +# define GOT_MAJOR +#endif + +#ifndef GOT_MAJOR +# if MSDOS +# define major(device) (device) +# define minor(device) (device) +# define makedev(major, minor) (((major) << 8) | (minor)) +# define GOT_MAJOR +# endif +#endif + +/* For HP-UX before HP-UX 8, major/minor are not in <sys/sysmacros.h>. */ +#ifndef GOT_MAJOR +# if defined(hpux) || defined(__hpux__) || defined(__hpux) +# include <sys/mknod.h> +# define GOT_MAJOR +# endif +#endif + +#ifndef GOT_MAJOR +# define major(device) (((device) >> 8) & 0xff) +# define minor(device) ((device) & 0xff) +# define makedev(major, minor) (((major) << 8) | (minor)) +#endif + +#undef GOT_MAJOR + +/* Declare wait status. */ + +#if HAVE_SYS_WAIT_H +# include <sys/wait.h> +#endif +#ifndef WEXITSTATUS +# define WEXITSTATUS(s) (((s) >> 8) & 0xff) +#endif +#ifndef WIFSIGNALED +# define WIFSIGNALED(s) (((s) & 0xffff) - 1 < (unsigned) 0xff) +#endif +#ifndef WTERMSIG +# define WTERMSIG(s) ((s) & 0x7f) +#endif + +/* FIXME: It is wrong to use BLOCKSIZE for buffers when the logical block + size is greater than 512 bytes; so ST_BLKSIZE code below, in preparation + for some cleanup in this area, later. */ + +/* Extract or fake data from a `struct stat'. ST_BLKSIZE gives the + optimal I/O blocksize for the file, in bytes. Some systems, like + Sequents, return st_blksize of 0 on pipes. */ + +#define DEFAULT_ST_BLKSIZE 512 + +#if !HAVE_ST_BLKSIZE +# define ST_BLKSIZE(statbuf) DEFAULT_ST_BLKSIZE +#else +# define ST_BLKSIZE(statbuf) \ + ((statbuf).st_blksize > 0 ? (statbuf).st_blksize : DEFAULT_ST_BLKSIZE) +#endif + +/* Extract or fake data from a `struct stat'. ST_NBLOCKS gives the + number of ST_NBLOCKSIZE-byte blocks in the file (including indirect blocks). + HP-UX counts st_blocks in 1024-byte units, + this loses when mixing HP-UX and BSD filesystems with NFS. AIX PS/2 + counts st_blocks in 4K units. */ + +#if !HAVE_ST_BLOCKS +# if defined(_POSIX_SOURCE) || !defined(BSIZE) +# define ST_NBLOCKS(statbuf) ((statbuf).st_size / ST_NBLOCKSIZE + ((statbuf).st_size % ST_NBLOCKSIZE != 0)) +# else + off_t st_blocks (); +# define ST_NBLOCKS(statbuf) (st_blocks ((statbuf).st_size)) +# endif +#else +# define ST_NBLOCKS(statbuf) ((statbuf).st_blocks) +# if defined(hpux) || defined(__hpux__) || defined(__hpux) +# define ST_NBLOCKSIZE 1024 +# else +# if defined(_AIX) && defined(_I386) +# define ST_NBLOCKSIZE (4 * 1024) +# endif +# endif +#endif + +#ifndef ST_NBLOCKSIZE +# define ST_NBLOCKSIZE 512 +#endif + +#define ST_IS_SPARSE(st) \ + (ST_NBLOCKS (st) \ + < ((st).st_size / ST_NBLOCKSIZE + ((st).st_size % ST_NBLOCKSIZE != 0))) + +/* Declare standard functions. */ + +#if STDC_HEADERS +# include <stdlib.h> +#else +void *malloc (); +char *getenv (); +#endif + +#include <stdbool.h> +#include <stddef.h> + +#include <stdio.h> +#if !defined _POSIX_VERSION && MSDOS +# include <io.h> +#endif + +#if WITH_DMALLOC +# define DMALLOC_FUNC_CHECK +# include <dmalloc.h> +#endif + +#include <limits.h> + +#ifndef MB_LEN_MAX +# define MB_LEN_MAX 1 +#endif + +#include <inttypes.h> + +#include <intprops.h> + +#define UINTMAX_STRSIZE_BOUND INT_BUFSIZE_BOUND (uintmax_t) + +/* Prototypes for external functions. */ + +#if HAVE_LOCALE_H +# include <locale.h> +#endif +#if !HAVE_SETLOCALE +# define setlocale(category, locale) /* empty */ +#endif + +#include <time.h> +#ifdef TIME_WITH_SYS_TIME +# include <sys/time.h> +#endif + +/* Library modules. */ + +#include <dirname.h> +#include <error.h> +#include <savedir.h> +#include <unlocked-io.h> +#include <xalloc.h> + +#include <gettext.h> +#define _(msgid) gettext (msgid) +#define N_(msgid) msgid + +#if MSDOS +# include <process.h> +# define SET_BINARY_MODE(arc) setmode(arc, O_BINARY) +# define ERRNO_IS_EACCES errno == EACCES +# define mkdir(file, mode) (mkdir) (file) +# define TTY_NAME "con" +# define sys_reset_uid_gid() +#else +# include <pwd.h> +# include <grp.h> +# define SET_BINARY_MODE(arc) +# define ERRNO_IS_EACCES 0 +# define TTY_NAME "/dev/tty" +# define sys_reset_uid_gid() \ + do { setuid (getuid ()); setgid (getgid ()); } while (0) +#endif + +#if XENIX +# include <sys/inode.h> +#endif diff --git a/lib/tempname.c b/lib/tempname.c new file mode 100644 index 0000000..e213600 --- /dev/null +++ b/lib/tempname.c @@ -0,0 +1,315 @@ +/* tempname.c - generate the name of a temporary file. + + Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, + 2000, 2001, 2002, 2003, 2005, 2006, 2007 Free Software Foundation, + Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Extracted from glibc sysdeps/posix/tempname.c. See also tmpdir.c. */ + +#if !_LIBC +# include <config.h> +# include "tempname.h" +#endif + +#include <sys/types.h> +#include <assert.h> + +#include <errno.h> +#ifndef __set_errno +# define __set_errno(Val) errno = (Val) +#endif + +#include <stdio.h> +#ifndef P_tmpdir +# define P_tmpdir "/tmp" +#endif +#ifndef TMP_MAX +# define TMP_MAX 238328 +#endif +#ifndef __GT_FILE +# define __GT_FILE 0 +# define __GT_BIGFILE 1 +# define __GT_DIR 2 +# define __GT_NOCREATE 3 +#endif + +#include <stddef.h> +#include <stdlib.h> +#include <string.h> + +#include <fcntl.h> +#include <sys/time.h> +#include <stdint.h> +#include <unistd.h> + +#include <sys/stat.h> + +#if _LIBC +# define struct_stat64 struct stat64 +# define small_open __open +# define large_open __open64 +#else +# define struct_stat64 struct stat +# define small_open open +# define large_open open +# define __gen_tempname gen_tempname +# define __getpid getpid +# define __gettimeofday gettimeofday +# define __mkdir mkdir +# define __lxstat64(version, file, buf) lstat (file, buf) +# define __xstat64(version, file, buf) stat (file, buf) +#endif + +#if ! (HAVE___SECURE_GETENV || _LIBC) +# define __secure_getenv getenv +#endif + +#ifdef _LIBC +# include <hp-timing.h> +# if HP_TIMING_AVAIL +# define RANDOM_BITS(Var) \ + if (__builtin_expect (value == UINT64_C (0), 0)) \ + { \ + /* If this is the first time this function is used initialize \ + the variable we accumulate the value in to some somewhat \ + random value. If we'd not do this programs at startup time \ + might have a reduced set of possible names, at least on slow \ + machines. */ \ + struct timeval tv; \ + __gettimeofday (&tv, NULL); \ + value = ((uint64_t) tv.tv_usec << 16) ^ tv.tv_sec; \ + } \ + HP_TIMING_NOW (Var) +# endif +#endif + +/* Use the widest available unsigned type if uint64_t is not + available. The algorithm below extracts a number less than 62**6 + (approximately 2**35.725) from uint64_t, so ancient hosts where + uintmax_t is only 32 bits lose about 3.725 bits of randomness, + which is better than not having mkstemp at all. */ +#if !defined UINT64_MAX && !defined uint64_t +# define uint64_t uintmax_t +#endif + +#if _LIBC +/* Return nonzero if DIR is an existent directory. */ +static int +direxists (const char *dir) +{ + struct_stat64 buf; + return __xstat64 (_STAT_VER, dir, &buf) == 0 && S_ISDIR (buf.st_mode); +} + +/* Path search algorithm, for tmpnam, tmpfile, etc. If DIR is + non-null and exists, uses it; otherwise uses the first of $TMPDIR, + P_tmpdir, /tmp that exists. Copies into TMPL a template suitable + for use with mk[s]temp. Will fail (-1) if DIR is non-null and + doesn't exist, none of the searched dirs exists, or there's not + enough space in TMPL. */ +int +__path_search (char *tmpl, size_t tmpl_len, const char *dir, const char *pfx, + int try_tmpdir) +{ + const char *d; + size_t dlen, plen; + + if (!pfx || !pfx[0]) + { + pfx = "file"; + plen = 4; + } + else + { + plen = strlen (pfx); + if (plen > 5) + plen = 5; + } + + if (try_tmpdir) + { + d = __secure_getenv ("TMPDIR"); + if (d != NULL && direxists (d)) + dir = d; + else if (dir != NULL && direxists (dir)) + /* nothing */ ; + else + dir = NULL; + } + if (dir == NULL) + { + if (direxists (P_tmpdir)) + dir = P_tmpdir; + else if (strcmp (P_tmpdir, "/tmp") != 0 && direxists ("/tmp")) + dir = "/tmp"; + else + { + __set_errno (ENOENT); + return -1; + } + } + + dlen = strlen (dir); + while (dlen > 1 && dir[dlen - 1] == '/') + dlen--; /* remove trailing slashes */ + + /* check we have room for "${dir}/${pfx}XXXXXX\0" */ + if (tmpl_len < dlen + 1 + plen + 6 + 1) + { + __set_errno (EINVAL); + return -1; + } + + sprintf (tmpl, "%.*s/%.*sXXXXXX", (int) dlen, dir, (int) plen, pfx); + return 0; +} +#endif /* _LIBC */ + +/* These are the characters used in temporary file names. */ +static const char letters[] = +"abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789"; + +/* Generate a temporary file name based on TMPL. TMPL must match the + rules for mk[s]temp (i.e. end in "XXXXXX"). The name constructed + does not exist at the time of the call to __gen_tempname. TMPL is + overwritten with the result. + + KIND may be one of: + __GT_NOCREATE: simply verify that the name does not exist + at the time of the call. + __GT_FILE: create the file using open(O_CREAT|O_EXCL) + and return a read-write fd. The file is mode 0600. + __GT_BIGFILE: same as __GT_FILE but use open64(). + __GT_DIR: create a directory, which will be mode 0700. + + We use a clever algorithm to get hard-to-predict names. */ +int +__gen_tempname (char *tmpl, int kind) +{ + int len; + char *XXXXXX; + static uint64_t value; + uint64_t random_time_bits; + unsigned int count; + int fd = -1; + int save_errno = errno; + struct_stat64 st; + + /* A lower bound on the number of temporary files to attempt to + generate. The maximum total number of temporary file names that + can exist for a given template is 62**6. It should never be + necessary to try all these combinations. Instead if a reasonable + number of names is tried (we define reasonable as 62**3) fail to + give the system administrator the chance to remove the problems. */ +#define ATTEMPTS_MIN (62 * 62 * 62) + + /* The number of times to attempt to generate a temporary file. To + conform to POSIX, this must be no smaller than TMP_MAX. */ +#if ATTEMPTS_MIN < TMP_MAX + unsigned int attempts = TMP_MAX; +#else + unsigned int attempts = ATTEMPTS_MIN; +#endif + + len = strlen (tmpl); + if (len < 6 || strcmp (&tmpl[len - 6], "XXXXXX")) + { + __set_errno (EINVAL); + return -1; + } + + /* This is where the Xs start. */ + XXXXXX = &tmpl[len - 6]; + + /* Get some more or less random data. */ +#ifdef RANDOM_BITS + RANDOM_BITS (random_time_bits); +#else + { + struct timeval tv; + __gettimeofday (&tv, NULL); + random_time_bits = ((uint64_t) tv.tv_usec << 16) ^ tv.tv_sec; + } +#endif + value += random_time_bits ^ __getpid (); + + for (count = 0; count < attempts; value += 7777, ++count) + { + uint64_t v = value; + + /* Fill in the random bits. */ + XXXXXX[0] = letters[v % 62]; + v /= 62; + XXXXXX[1] = letters[v % 62]; + v /= 62; + XXXXXX[2] = letters[v % 62]; + v /= 62; + XXXXXX[3] = letters[v % 62]; + v /= 62; + XXXXXX[4] = letters[v % 62]; + v /= 62; + XXXXXX[5] = letters[v % 62]; + + switch (kind) + { + case __GT_FILE: + fd = small_open (tmpl, O_RDWR | O_CREAT | O_EXCL, S_IRUSR | S_IWUSR); + break; + + case __GT_BIGFILE: + fd = large_open (tmpl, O_RDWR | O_CREAT | O_EXCL, S_IRUSR | S_IWUSR); + break; + + case __GT_DIR: + fd = __mkdir (tmpl, S_IRUSR | S_IWUSR | S_IXUSR); + break; + + case __GT_NOCREATE: + /* This case is backward from the other three. __gen_tempname + succeeds if __xstat fails because the name does not exist. + Note the continue to bypass the common logic at the bottom + of the loop. */ + if (__lxstat64 (_STAT_VER, tmpl, &st) < 0) + { + if (errno == ENOENT) + { + __set_errno (save_errno); + return 0; + } + else + /* Give up now. */ + return -1; + } + continue; + + default: + assert (! "invalid KIND in __gen_tempname"); + } + + if (fd >= 0) + { + __set_errno (save_errno); + return fd; + } + else if (errno != EEXIST) + return -1; + } + + /* We got out of the loop because we ran out of combinations to try. */ + __set_errno (EEXIST); + return -1; +} diff --git a/lib/tempname.h b/lib/tempname.h new file mode 100644 index 0000000..c51fa69 --- /dev/null +++ b/lib/tempname.h @@ -0,0 +1,40 @@ +/* Create a temporary file or directory. + + Copyright (C) 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* header written by Eric Blake */ + +/* In gnulib, always prefer large files. GT_FILE maps to + __GT_BIGFILE, not __GT_FILE, for a reason. */ +#define GT_FILE 1 +#define GT_DIR 2 +#define GT_NOCREATE 3 + +/* Generate a temporary file name based on TMPL. TMPL must match the + rules for mk[s]temp (i.e. end in "XXXXXX"). The name constructed + does not exist at the time of the call to gen_tempname. TMPL is + overwritten with the result. + + KIND may be one of: + GT_NOCREATE: simply verify that the name does not exist + at the time of the call. + GT_FILE: create a large file using open(O_CREAT|O_EXCL) + and return a read-write fd. The file is mode 0600. + GT_DIR: create a directory, which will be mode 0700. + + We use a clever algorithm to get hard-to-predict names. */ +extern int gen_tempname (char *tmpl, int kind); diff --git a/lib/time_.h b/lib/time_.h new file mode 100644 index 0000000..9a0507e --- /dev/null +++ b/lib/time_.h @@ -0,0 +1,125 @@ +/* A more-standard <time.h>. + + Copyright (C) 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Don't get in the way of glibc when it includes time.h merely to + declare a few standard symbols, rather than to declare all the + symbols. */ +#if defined __need_time_t || defined __need_clock_t || defined __need_timespec + +# if @HAVE_INCLUDE_NEXT@ +# include_next <time.h> +# else +# include @ABSOLUTE_TIME_H@ +# endif + +#else +/* Normal invocation convention. */ + +# if ! defined _GL_TIME_H + +/* The include_next requires a split double-inclusion guard. */ +# if @HAVE_INCLUDE_NEXT@ +# include_next <time.h> +# else +# include @ABSOLUTE_TIME_H@ +# endif + +# if ! defined _GL_TIME_H +# define _GL_TIME_H + +# ifdef __cplusplus +extern "C" { +# endif + +/* Some systems don't define struct timespec (e.g., AIX 4.1, Ultrix 4.3). + Or they define it with the wrong member names or define it in <sys/time.h> + (e.g., FreeBSD circa 1997). */ +# if ! @TIME_H_DEFINES_STRUCT_TIMESPEC@ +# if @SYS_TIME_H_DEFINES_STRUCT_TIMESPEC@ +# include <sys/time.h> +# else +# undef timespec +# define timespec rpl_timespec +struct timespec +{ + time_t tv_sec; + long int tv_nsec; +}; +# endif +# endif + +/* Sleep for at least RQTP seconds unless interrupted, If interrupted, + return -1 and store the remaining time into RMTP. See + <http://www.opengroup.org/susv3xsh/nanosleep.html>. */ +# if @REPLACE_NANOSLEEP@ +# define nanosleep rpl_nanosleep +int nanosleep (struct timespec const *__rqtp, struct timespec *__rmtp); +# endif + +/* Convert TIMER to RESULT, assuming local time and UTC respectively. See + <http://www.opengroup.org/susv3xsh/localtime_r.html> and + <http://www.opengroup.org/susv3xsh/gmtime_r.html>. */ +# if @REPLACE_LOCALTIME_R@ +# undef localtime_r +# define localtime_r rpl_localtime_r +# undef gmtime_r +# define gmtime_r rpl_gmtime_r +struct tm *localtime_r (time_t const *restrict __timer, + struct tm *restrict __result); +struct tm *gmtime_r (time_t const *restrict __timer, + struct tm *restrict __result); +# endif + +/* Parse BUF as a time stamp, assuming FORMAT specifies its layout, and store + the resulting broken-down time into TM. See + <http://www.opengroup.org/susv3xsh/strptime.html>. */ +# if @REPLACE_STRPTIME@ +# undef strptime +# define strptime rpl_strptime +char *strptime (char const *restrict __buf, char const *restrict __format, + struct tm *restrict __tm); +# endif + +/* Convert TM to a time_t value, assuming UTC. */ +# if @REPLACE_TIMEGM@ +# undef timegm +# define timegm rpl_timegm +time_t timegm (struct tm *__tm); +# endif + +/* Encourage applications to avoid unsafe functions that can overrun + buffers when given outlandish struct tm values. Portable + applications should use strftime (or even sprintf) instead. */ +# if GNULIB_PORTCHECK +# undef asctime +# define asctime eschew_asctime +# undef asctime_r +# define asctime_r eschew_asctime_r +# undef ctime +# define ctime eschew_ctime +# undef ctime_r +# define ctime_r eschew_ctime_r +# endif + +# ifdef __cplusplus +} +# endif + +# endif /* _GL_TIME_H */ +# endif /* _GL_TIME_H */ +#endif diff --git a/lib/time_r.c b/lib/time_r.c new file mode 100644 index 0000000..2b72692 --- /dev/null +++ b/lib/time_r.c @@ -0,0 +1,47 @@ +/* Reentrant time functions like localtime_r. + + Copyright (C) 2003, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <config.h> + +#include <time.h> + +#include <string.h> + +static struct tm * +copy_tm_result (struct tm *dest, struct tm const *src) +{ + if (! src) + return 0; + *dest = *src; + return dest; +} + + +struct tm * +gmtime_r (time_t const * restrict t, struct tm * restrict tp) +{ + return copy_tm_result (tp, gmtime (t)); +} + +struct tm * +localtime_r (time_t const * restrict t, struct tm * restrict tp) +{ + return copy_tm_result (tp, localtime (t)); +} diff --git a/lib/timespec.h b/lib/timespec.h new file mode 100644 index 0000000..cce2d66 --- /dev/null +++ b/lib/timespec.h @@ -0,0 +1,37 @@ +/* timespec -- System time interface + + Copyright (C) 2000, 2002, 2004, 2005, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if ! defined TIMESPEC_H +# define TIMESPEC_H + +# include <time.h> + +/* Return negative, zero, positive if A < B, A == B, A > B, respectively. + Assume the nanosecond components are in range, or close to it. */ +static inline int +timespec_cmp (struct timespec a, struct timespec b) +{ + return (a.tv_sec < b.tv_sec ? -1 + : a.tv_sec > b.tv_sec ? 1 + : a.tv_nsec - b.tv_nsec); +} + +void gettime (struct timespec *); +int settime (struct timespec const *); + +#endif diff --git a/lib/uinttostr.c b/lib/uinttostr.c new file mode 100644 index 0000000..52d288e --- /dev/null +++ b/lib/uinttostr.c @@ -0,0 +1,3 @@ +#define inttostr uinttostr +#define inttype unsigned int +#include "inttostr.c" diff --git a/lib/umaxtostr.c b/lib/umaxtostr.c new file mode 100644 index 0000000..4f49a7f --- /dev/null +++ b/lib/umaxtostr.c @@ -0,0 +1,3 @@ +#define inttostr umaxtostr +#define inttype uintmax_t +#include "inttostr.c" diff --git a/lib/unistd--.h b/lib/unistd--.h new file mode 100644 index 0000000..1fe6ce8 --- /dev/null +++ b/lib/unistd--.h @@ -0,0 +1,28 @@ +/* Like unistd.h, but redefine some names to avoid glitches. + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +#include <unistd.h> +#include "unistd-safer.h" + +#undef dup +#define dup dup_safer + +#undef pipe +#define pipe pipe_safer diff --git a/lib/unistd-safer.h b/lib/unistd-safer.h new file mode 100644 index 0000000..f95999d --- /dev/null +++ b/lib/unistd-safer.h @@ -0,0 +1,23 @@ +/* Invoke unistd-like functions, but avoid some glitches. + + Copyright (C) 2001, 2003, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +int dup_safer (int); +int fd_safer (int); +int pipe_safer (int[2]); diff --git a/lib/unistd_.h b/lib/unistd_.h new file mode 100644 index 0000000..f34bc66 --- /dev/null +++ b/lib/unistd_.h @@ -0,0 +1,242 @@ +/* Substitute for and wrapper around <unistd.h>. + Copyright (C) 2004-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _GL_UNISTD_H + +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_UNISTD_H@ +# if @HAVE_INCLUDE_NEXT@ +# include_next <unistd.h> +# else +# include @ABSOLUTE_UNISTD_H@ +# endif +#endif + +#ifndef _GL_UNISTD_H +#define _GL_UNISTD_H + +/* mingw doesn't define the SEEK_* macros in <unistd.h>. */ +#if !(defined SEEK_CUR && defined SEEK_END && defined SEEK_SET) +# include <stdio.h> +#endif + +/* mingw fails to declare _exit in <unistd.h>. */ +#include <stdlib.h> + +/* The definition of GL_LINK_WARNING is copied here. */ + + +/* Declare overridden functions. */ + +#ifdef __cplusplus +extern "C" { +#endif + + +#if @GNULIB_CHOWN@ +# if @REPLACE_CHOWN@ +/* Change the owner of FILE to UID (if UID is not -1) and the group of FILE + to GID (if GID is not -1). + Return 0 if successful, otherwise -1 and errno set. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/chown.html>. */ +# define chown rpl_chown +extern int chown (const char *file, uid_t uid, gid_t gid); +# endif +#elif defined GNULIB_POSIXCHECK +# undef chown +# define chown(f,u,g) \ + (GL_LINK_WARNING ("chown fails to follow symlinks on some systems and " \ + "doesn't treat a uid or gid of -1 on some systems - " \ + "use gnulib module chown for portability"), \ + chown (f, u, g)) +#endif + + +#if @GNULIB_DUP2@ +# if !@HAVE_DUP2@ +/* Copy the file descriptor OLDFD into file descriptor NEWFD. Do nothing if + NEWFD = OLDFD, otherwise close NEWFD first if it is open. + Return 0 if successful, otherwise -1 and errno set. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/dup2.html>. */ +extern int dup2 (int oldfd, int newfd); +# endif +#elif defined GNULIB_POSIXCHECK +# undef dup2 +# define dup2(o,n) \ + (GL_LINK_WARNING ("dup2 is unportable - " \ + "use gnulib module dup2 for portability"), \ + dup2 (o, n)) +#endif + + +#if @GNULIB_FCHDIR@ +# if @REPLACE_FCHDIR@ + +/* Change the process' current working directory to the directory on which + the given file descriptor is open. + Return 0 if successful, otherwise -1 and errno set. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/fchdir.html>. */ +extern int fchdir (int /*fd*/); + +# define close rpl_close +extern int close (int); +# define dup rpl_dup +extern int dup (int); +# define dup2 rpl_dup2 +extern int dup2 (int, int); + +# endif +#elif defined GNULIB_POSIXCHECK +# undef fchdir +# define fchdir(f) \ + (GL_LINK_WARNING ("fchdir is unportable - " \ + "use gnulib module fchdir for portability"), \ + fchdir (f)) +#endif + + +#if @GNULIB_FTRUNCATE@ +# if !@HAVE_FTRUNCATE@ +/* Change the size of the file to which FD is opened to become equal to LENGTH. + Return 0 if successful, otherwise -1 and errno set. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/ftruncate.html>. */ +extern int ftruncate (int fd, off_t length); +# endif +#elif defined GNULIB_POSIXCHECK +# undef ftruncate +# define ftruncate(f,l) \ + (GL_LINK_WARNING ("ftruncate is unportable - " \ + "use gnulib module ftruncate for portability"), \ + ftruncate (f, l)) +#endif + + +#if @GNULIB_GETCWD@ +/* Include the headers that might declare getcwd so that they will not + cause confusion if included after this file. */ +# include <stdlib.h> +# if @REPLACE_GETCWD@ +/* Get the name of the current working directory, and put it in SIZE bytes + of BUF. + Return BUF if successful, or NULL if the directory couldn't be determined + or SIZE was too small. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/getcwd.html>. + Additionally, the gnulib module 'getcwd' guarantees the following GNU + extension: If BUF is NULL, an array is allocated with 'malloc'; the array + is SIZE bytes long, unless SIZE == 0, in which case it is as big as + necessary. */ +# define getcwd rpl_getcwd +extern char * getcwd (char *buf, size_t size); +# endif +#elif defined GNULIB_POSIXCHECK +# undef getcwd +# define getcwd(b,s) \ + (GL_LINK_WARNING ("getcwd is unportable - " \ + "use gnulib module getcwd for portability"), \ + getcwd (b, s)) +#endif + + +#if @GNULIB_GETLOGIN_R@ +/* Copies the user's login name to NAME. + The array pointed to by NAME has room for SIZE bytes. + + Returns 0 if successful. Upon error, an error number is returned, or -1 in + the case that the login name cannot be found but no specific error is + provided (this case is hopefully rare but is left open by the POSIX spec). + + See <http://www.opengroup.org/susv3xsh/getlogin.html>. + */ +# if !@HAVE_DECL_GETLOGIN_R@ +# include <stddef.h> +extern int getlogin_r (char *name, size_t size); +# endif +#elif defined GNULIB_POSIXCHECK +# undef getlogin_r +# define getlogin_r(n,s) \ + (GL_LINK_WARNING ("getlogin_r is unportable - " \ + "use gnulib module getlogin_r for portability"), \ + getlogin_r (n, s)) +#endif + + +#if @GNULIB_LSEEK@ +# if @REPLACE_LSEEK@ +/* Set the offset of FD relative to SEEK_SET, SEEK_CUR, or SEEK_END. + Return the new offset if successful, otherwise -1 and errno set. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/lseek.html>. */ +# define lseek rpl_lseek + extern off_t lseek (int fd, off_t offset, int whence); +# endif +#elif defined GNULIB_POSIXCHECK +# undef lseek +# define lseek(f,o,w) \ + (GL_LINK_WARNING ("lseek does not fail with ESPIPE on pipes on some " \ + "systems - use gnulib module lseek for portability"), \ + lseek (f, o, w)) +#endif + + +#if @GNULIB_READLINK@ +/* Read the contents of the symbolic link FILE and place the first BUFSIZE + bytes of it into BUF. Return the number of bytes placed into BUF if + successful, otherwise -1 and errno set. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/readlink.html>. */ +# if !@HAVE_READLINK@ +# include <stddef.h> +extern int readlink (const char *file, char *buf, size_t bufsize); +# endif +#elif defined GNULIB_POSIXCHECK +# undef readlink +# define readlink(f,b,s) \ + (GL_LINK_WARNING ("readlink is unportable - " \ + "use gnulib module readlink for portability"), \ + readlink (f, b, s)) +#endif + + +#if @GNULIB_SLEEP@ +/* Pause the execution of the current thread for N seconds. + Returns the number of seconds left to sleep. + See the POSIX:2001 specification + <http://www.opengroup.org/susv3xsh/sleep.html>. */ +# if !@HAVE_SLEEP@ +extern unsigned int sleep (unsigned int n); +# endif +#elif defined GNULIB_POSIXCHECK +# undef sleep +# define sleep(n) \ + (GL_LINK_WARNING ("sleep is unportable - " \ + "use gnulib module sleep for portability"), \ + sleep (n)) +#endif + + +#ifdef __cplusplus +} +#endif + + +#endif /* _GL_UNISTD_H */ +#endif /* _GL_UNISTD_H */ diff --git a/lib/unlinkdir.c b/lib/unlinkdir.c new file mode 100644 index 0000000..07620db --- /dev/null +++ b/lib/unlinkdir.c @@ -0,0 +1,68 @@ +/* unlinkdir.c - determine (and maybe change) whether we can unlink directories + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and Jim Meyering. */ + +#include <config.h> + +#include "unlinkdir.h" + +#if HAVE_PRIV_H +# include <priv.h> +#endif +#include <unistd.h> + +#if ! UNLINK_CANNOT_UNLINK_DIR + +/* Return true if we cannot unlink directories, false if we might be + able to unlink directories. If possible, tell the kernel we don't + want to be able to unlink directories, so that we can return true. */ + +bool +cannot_unlink_dir (void) +{ + static bool initialized; + static bool cannot; + + if (! initialized) + { +# if defined PRIV_EFFECTIVE && defined PRIV_SYS_LINKDIR + /* We might be able to unlink directories if we cannot + determine our privileges, or if we have the + PRIV_SYS_LINKDIR privilege and cannot delete it. */ + priv_set_t *pset = priv_allocset (); + if (pset) + { + cannot = + (getppriv (PRIV_EFFECTIVE, pset) == 0 + && (! priv_ismember (pset, PRIV_SYS_LINKDIR) + || (priv_delset (pset, PRIV_SYS_LINKDIR) == 0 + && setppriv (PRIV_SET, PRIV_EFFECTIVE, pset) == 0))); + priv_freeset (pset); + } +# else + /* In traditional Unix, only root can unlink directories. */ + cannot = (geteuid () != 0); +# endif + initialized = true; + } + + return cannot; +} + +#endif diff --git a/lib/unlinkdir.h b/lib/unlinkdir.h new file mode 100644 index 0000000..6582418 --- /dev/null +++ b/lib/unlinkdir.h @@ -0,0 +1,27 @@ +/* unlinkdir.h - determine (and maybe change) whether we can unlink directories + + Copyright (C) 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert and Jim Meyering. */ + +#include <stdbool.h> + +#if UNLINK_CANNOT_UNLINK_DIR +# define cannot_unlink_dir() true +#else +bool cannot_unlink_dir (void); +#endif diff --git a/lib/unlocked-io.h b/lib/unlocked-io.h new file mode 100644 index 0000000..d009303 --- /dev/null +++ b/lib/unlocked-io.h @@ -0,0 +1,137 @@ +/* Prefer faster, non-thread-safe stdio functions if available. + + Copyright (C) 2001, 2002, 2003, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#ifndef UNLOCKED_IO_H +# define UNLOCKED_IO_H 1 + +/* These are wrappers for functions/macros from the GNU C library, and + from other C libraries supporting POSIX's optional thread-safe functions. + + The standard I/O functions are thread-safe. These *_unlocked ones are + more efficient but not thread-safe. That they're not thread-safe is + fine since all of the applications in this package are single threaded. + + Also, some code that is shared with the GNU C library may invoke + the *_unlocked functions directly. On hosts that lack those + functions, invoke the non-thread-safe versions instead. */ + +# include <stdio.h> + +# if HAVE_DECL_CLEARERR_UNLOCKED +# undef clearerr +# define clearerr(x) clearerr_unlocked (x) +# else +# define clearerr_unlocked(x) clearerr (x) +# endif + +# if HAVE_DECL_FEOF_UNLOCKED +# undef feof +# define feof(x) feof_unlocked (x) +# else +# define feof_unlocked(x) feof (x) +# endif + +# if HAVE_DECL_FERROR_UNLOCKED +# undef ferror +# define ferror(x) ferror_unlocked (x) +# else +# define ferror_unlocked(x) ferror (x) +# endif + +# if HAVE_DECL_FFLUSH_UNLOCKED +# undef fflush +# define fflush(x) fflush_unlocked (x) +# else +# define fflush_unlocked(x) fflush (x) +# endif + +# if HAVE_DECL_FGETS_UNLOCKED +# undef fgets +# define fgets(x,y,z) fgets_unlocked (x,y,z) +# else +# define fgets_unlocked(x,y,z) fgets (x,y,z) +# endif + +# if HAVE_DECL_FPUTC_UNLOCKED +# undef fputc +# define fputc(x,y) fputc_unlocked (x,y) +# else +# define fputc_unlocked(x,y) fputc (x,y) +# endif + +# if HAVE_DECL_FPUTS_UNLOCKED +# undef fputs +# define fputs(x,y) fputs_unlocked (x,y) +# else +# define fputs_unlocked(x,y) fputs (x,y) +# endif + +# if HAVE_DECL_FREAD_UNLOCKED +# undef fread +# define fread(w,x,y,z) fread_unlocked (w,x,y,z) +# else +# define fread_unlocked(w,x,y,z) fread (w,x,y,z) +# endif + +# if HAVE_DECL_FWRITE_UNLOCKED +# undef fwrite +# define fwrite(w,x,y,z) fwrite_unlocked (w,x,y,z) +# else +# define fwrite_unlocked(w,x,y,z) fwrite (w,x,y,z) +# endif + +# if HAVE_DECL_GETC_UNLOCKED +# undef getc +# define getc(x) getc_unlocked (x) +# else +# define getc_unlocked(x) getc (x) +# endif + +# if HAVE_DECL_GETCHAR_UNLOCKED +# undef getchar +# define getchar() getchar_unlocked () +# else +# define getchar_unlocked() getchar () +# endif + +# if HAVE_DECL_PUTC_UNLOCKED +# undef putc +# define putc(x,y) putc_unlocked (x,y) +# else +# define putc_unlocked(x,y) putc (x,y) +# endif + +# if HAVE_DECL_PUTCHAR_UNLOCKED +# undef putchar +# define putchar(x) putchar_unlocked (x) +# else +# define putchar_unlocked(x) putchar (x) +# endif + +# undef flockfile +# define flockfile(x) ((void) 0) + +# undef ftrylockfile +# define ftrylockfile(x) 0 + +# undef funlockfile +# define funlockfile(x) ((void) 0) + +#endif /* UNLOCKED_IO_H */ diff --git a/lib/unsetenv.c b/lib/unsetenv.c new file mode 100644 index 0000000..0f83744 --- /dev/null +++ b/lib/unsetenv.c @@ -0,0 +1,92 @@ +/* Copyright (C) 1992,1995-1999,2000-2002,2005-2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include <errno.h> +#if !_LIBC +# define __set_errno(ev) ((errno) = (ev)) +#endif + +#include <stdlib.h> +#include <string.h> +#include <unistd.h> + +#if !_LIBC +# define __environ environ +# ifndef HAVE_ENVIRON_DECL +extern char **environ; +# endif +#endif + +#if _LIBC +/* This lock protects against simultaneous modifications of `environ'. */ +# include <bits/libc-lock.h> +__libc_lock_define_initialized (static, envlock) +# define LOCK __libc_lock_lock (envlock) +# define UNLOCK __libc_lock_unlock (envlock) +#else +# define LOCK +# define UNLOCK +#endif + +/* In the GNU C library we must keep the namespace clean. */ +#ifdef _LIBC +# define unsetenv __unsetenv +#endif + + +int +unsetenv (const char *name) +{ + size_t len; + char **ep; + + if (name == NULL || *name == '\0' || strchr (name, '=') != NULL) + { + __set_errno (EINVAL); + return -1; + } + + len = strlen (name); + + LOCK; + + ep = __environ; + while (*ep != NULL) + if (!strncmp (*ep, name, len) && (*ep)[len] == '=') + { + /* Found it. Remove this pointer by moving later ones back. */ + char **dp = ep; + + do + dp[0] = dp[1]; + while (*dp++); + /* Continue the loop in case NAME appears again. */ + } + else + ++ep; + + UNLOCK; + + return 0; +} + +#ifdef _LIBC +# undef unsetenv +weak_alias (__unsetenv, unsetenv) +#endif diff --git a/lib/utime.c b/lib/utime.c new file mode 100644 index 0000000..273a5fb --- /dev/null +++ b/lib/utime.c @@ -0,0 +1,109 @@ +/* Copyright (C) 1998, 2001, 2002, 2003, 2004, 2006 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* derived from a function in touch.c */ + +#include <config.h> +#undef utime + +#include <sys/types.h> + +#ifdef HAVE_UTIME_H +# include <utime.h> +#endif + +#if !HAVE_UTIMES_NULL +# include <sys/stat.h> +# include <fcntl.h> +#endif + +#include <unistd.h> +#include <errno.h> + +#include "full-write.h" +#include "safe-read.h" + +/* Some systems (even some that do have <utime.h>) don't declare this + structure anywhere. */ +#ifndef HAVE_STRUCT_UTIMBUF +struct utimbuf +{ + long actime; + long modtime; +}; +#endif + +/* The results of open() in this file are not used with fchdir, + therefore save some unnecessary work in fchdir.c. */ +#undef open +#undef close + +/* Emulate utime (file, NULL) for systems (like 4.3BSD) that do not + interpret it to set the access and modification times of FILE to + the current time. Return 0 if successful, -1 if not. */ + +static int +utime_null (const char *file) +{ +#if HAVE_UTIMES_NULL + return utimes (file, 0); +#else + int fd; + char c; + int status = 0; + struct stat st; + int saved_errno = 0; + + fd = open (file, O_RDWR); + if (fd < 0 + || fstat (fd, &st) < 0 + || safe_read (fd, &c, sizeof c) == SAFE_READ_ERROR + || lseek (fd, (off_t) 0, SEEK_SET) < 0 + || full_write (fd, &c, sizeof c) != sizeof c + /* Maybe do this -- it's necessary on SunOS 4.1.3 with some combination + of patches, but that system doesn't use this code: it has utimes. + || fsync (fd) < 0 + */ + || (st.st_size == 0 && ftruncate (fd, st.st_size) < 0)) + { + saved_errno = errno; + status = -1; + } + + if (0 <= fd) + { + if (close (fd) < 0) + status = -1; + + /* If there was a prior failure, use the saved errno value. + But if the only failure was in the close, don't change errno. */ + if (saved_errno) + errno = saved_errno; + } + + return status; +#endif +} + +int +rpl_utime (const char *file, const struct utimbuf *times) +{ + if (times) + return utime (file, times); + + return utime_null (file); +} diff --git a/lib/utimens.c b/lib/utimens.c new file mode 100644 index 0000000..7e3175a --- /dev/null +++ b/lib/utimens.c @@ -0,0 +1,189 @@ +/* Set file access and modification times. + + Copyright (C) 2003, 2004, 2005, 2006, 2007 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert. */ + +/* derived from a function in touch.c */ + +#include <config.h> + +#include "utimens.h" + +#include <errno.h> +#include <fcntl.h> +#include <sys/time.h> +#include <unistd.h> + +#if HAVE_UTIME_H +# include <utime.h> +#endif + +/* Some systems (even some that do have <utime.h>) don't declare this + structure anywhere. */ +#ifndef HAVE_STRUCT_UTIMBUF +struct utimbuf +{ + long actime; + long modtime; +}; +#endif + +/* Some systems don't have ENOSYS. */ +#ifndef ENOSYS +# ifdef ENOTSUP +# define ENOSYS ENOTSUP +# else +/* Some systems don't have ENOTSUP either. */ +# define ENOSYS EINVAL +# endif +#endif + +#ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +#endif + +#ifndef ATTRIBUTE_UNUSED +# define ATTRIBUTE_UNUSED __attribute__ ((__unused__)) +#endif + +/* Set the access and modification time stamps of FD (a.k.a. FILE) to be + TIMESPEC[0] and TIMESPEC[1], respectively. + FD must be either negative -- in which case it is ignored -- + or a file descriptor that is open on FILE. + If FD is nonnegative, then FILE can be NULL, which means + use just futimes (or equivalent) instead of utimes (or equivalent), + and fail if on an old system without futimes (or equivalent). + If TIMESPEC is null, set the time stamps to the current time. + Return 0 on success, -1 (setting errno) on failure. */ + +int +gl_futimens (int fd ATTRIBUTE_UNUSED, + char const *file, struct timespec const timespec[2]) +{ + /* Some Linux-based NFS clients are buggy, and mishandle time stamps + of files in NFS file systems in some cases. We have no + configure-time test for this, but please see + <http://bugs.gentoo.org/show_bug.cgi?id=132673> for references to + some of the problems with Linux 2.6.16. If this affects you, + compile with -DHAVE_BUGGY_NFS_TIME_STAMPS; this is reported to + help in some cases, albeit at a cost in performance. But you + really should upgrade your kernel to a fixed version, since the + problem affects many applications. */ + +#if HAVE_BUGGY_NFS_TIME_STAMPS + if (fd < 0) + sync (); + else + fsync (fd); +#endif + + /* There's currently no interface to set file timestamps with + nanosecond resolution, so do the best we can, discarding any + fractional part of the timestamp. */ +#if HAVE_FUTIMESAT || HAVE_WORKING_UTIMES + struct timeval timeval[2]; + struct timeval const *t; + if (timespec) + { + timeval[0].tv_sec = timespec[0].tv_sec; + timeval[0].tv_usec = timespec[0].tv_nsec / 1000; + timeval[1].tv_sec = timespec[1].tv_sec; + timeval[1].tv_usec = timespec[1].tv_nsec / 1000; + t = timeval; + } + else + t = NULL; + + + if (fd < 0) + { +# if HAVE_FUTIMESAT + return futimesat (AT_FDCWD, file, t); +# endif + } + else + { + /* If futimesat or futimes fails here, don't try to speed things + up by returning right away. glibc can incorrectly fail with + errno == ENOENT if /proc isn't mounted. Also, Mandrake 10.0 + in high security mode doesn't allow ordinary users to read + /proc/self, so glibc incorrectly fails with errno == EACCES. + If errno == EIO, EPERM, or EROFS, it's probably safe to fail + right away, but these cases are rare enough that they're not + worth optimizing, and who knows what other messed-up systems + are out there? So play it safe and fall back on the code + below. */ +# if HAVE_FUTIMESAT + if (futimesat (fd, NULL, t) == 0) + return 0; +# elif HAVE_FUTIMES + if (futimes (fd, t) == 0) + return 0; +# endif + } +#endif + + if (!file) + { +#if ! (HAVE_FUTIMESAT || (HAVE_WORKING_UTIMES && HAVE_FUTIMES)) + errno = ENOSYS; +#endif + + /* Prefer EBADF to ENOSYS if both error numbers apply. */ + if (errno == ENOSYS) + { + int fd2 = dup (fd); + int dup_errno = errno; + if (0 <= fd2) + close (fd2); + errno = (fd2 < 0 && dup_errno == EBADF ? EBADF : ENOSYS); + } + + return -1; + } + +#if HAVE_WORKING_UTIMES + return utimes (file, t); +#else + { + struct utimbuf utimbuf; + struct utimbuf const *ut; + if (timespec) + { + utimbuf.actime = timespec[0].tv_sec; + utimbuf.modtime = timespec[1].tv_sec; + ut = &utimbuf; + } + else + ut = NULL; + + return utime (file, ut); + } +#endif +} + +/* Set the access and modification time stamps of FILE to be + TIMESPEC[0] and TIMESPEC[1], respectively. */ +int +utimens (char const *file, struct timespec const timespec[2]) +{ + return gl_futimens (-1, file, timespec); +} diff --git a/lib/utimens.h b/lib/utimens.h new file mode 100644 index 0000000..169521d --- /dev/null +++ b/lib/utimens.h @@ -0,0 +1,3 @@ +#include <time.h> +int gl_futimens (int, char const *, struct timespec const [2]); +int utimens (char const *, struct timespec const [2]); diff --git a/lib/vasnprintf.c b/lib/vasnprintf.c new file mode 100644 index 0000000..78ead8e --- /dev/null +++ b/lib/vasnprintf.c @@ -0,0 +1,918 @@ +/* vsprintf with automatic memory allocation. + Copyright (C) 1999, 2002-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Tell glibc's <stdio.h> to provide a prototype for snprintf(). + This must come before <config.h> because <config.h> may include + <features.h>, and once <features.h> has been included, it's too late. */ +#ifndef _GNU_SOURCE +# define _GNU_SOURCE 1 +#endif + +#include <config.h> + +#ifndef IN_LIBINTL +# include <alloca.h> +#endif + +/* Specification. */ +#if WIDE_CHAR_VERSION +# include "vasnwprintf.h" +#else +# include "vasnprintf.h" +#endif + +#include <stdio.h> /* snprintf(), sprintf() */ +#include <stdlib.h> /* abort(), malloc(), realloc(), free() */ +#include <string.h> /* memcpy(), strlen() */ +#include <errno.h> /* errno */ +#include <limits.h> /* CHAR_BIT, INT_MAX */ +#include <float.h> /* DBL_MAX_EXP, LDBL_MAX_EXP */ +#if WIDE_CHAR_VERSION +# include "wprintf-parse.h" +#else +# include "printf-parse.h" +#endif + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* Some systems, like OSF/1 4.0 and Woe32, don't have EOVERFLOW. */ +#ifndef EOVERFLOW +# define EOVERFLOW E2BIG +#endif + +#ifdef HAVE_WCHAR_T +# ifdef HAVE_WCSLEN +# define local_wcslen wcslen +# else + /* Solaris 2.5.1 has wcslen() in a separate library libw.so. To avoid + a dependency towards this library, here is a local substitute. + Define this substitute only once, even if this file is included + twice in the same compilation unit. */ +# ifndef local_wcslen_defined +# define local_wcslen_defined 1 +static size_t +local_wcslen (const wchar_t *s) +{ + const wchar_t *ptr; + + for (ptr = s; *ptr != (wchar_t) 0; ptr++) + ; + return ptr - s; +} +# endif +# endif +#endif + +#if WIDE_CHAR_VERSION +# define VASNPRINTF vasnwprintf +# define CHAR_T wchar_t +# define DIRECTIVE wchar_t_directive +# define DIRECTIVES wchar_t_directives +# define PRINTF_PARSE wprintf_parse +# define USE_SNPRINTF 1 +# if HAVE_DECL__SNWPRINTF + /* On Windows, the function swprintf() has a different signature than + on Unix; we use the _snwprintf() function instead. */ +# define SNPRINTF _snwprintf +# else + /* Unix. */ +# define SNPRINTF swprintf +# endif +#else +# define VASNPRINTF vasnprintf +# define CHAR_T char +# define DIRECTIVE char_directive +# define DIRECTIVES char_directives +# define PRINTF_PARSE printf_parse +# define USE_SNPRINTF (HAVE_DECL__SNPRINTF || HAVE_SNPRINTF) +# if HAVE_DECL__SNPRINTF + /* Windows. */ +# define SNPRINTF _snprintf +# else + /* Unix. */ +# define SNPRINTF snprintf +# endif +#endif + +CHAR_T * +VASNPRINTF (CHAR_T *resultbuf, size_t *lengthp, const CHAR_T *format, va_list args) +{ + DIRECTIVES d; + arguments a; + + if (PRINTF_PARSE (format, &d, &a) < 0) + { + errno = EINVAL; + return NULL; + } + +#define CLEANUP() \ + free (d.dir); \ + if (a.arg) \ + free (a.arg); + + if (printf_fetchargs (args, &a) < 0) + { + CLEANUP (); + errno = EINVAL; + return NULL; + } + + { + size_t buf_neededlength; + CHAR_T *buf; + CHAR_T *buf_malloced; + const CHAR_T *cp; + size_t i; + DIRECTIVE *dp; + /* Output string accumulator. */ + CHAR_T *result; + size_t allocated; + size_t length; + + /* Allocate a small buffer that will hold a directive passed to + sprintf or snprintf. */ + buf_neededlength = 7 + d.max_width_length + d.max_precision_length + 6; +#if HAVE_ALLOCA + if (buf_neededlength < 4000 / sizeof (CHAR_T)) + { + buf = (CHAR_T *) alloca (buf_neededlength * sizeof (CHAR_T)); + buf_malloced = NULL; + } + else +#endif + { + if (SIZE_MAX / sizeof (CHAR_T) < buf_neededlength) + goto out_of_memory_1; + buf = (CHAR_T *) malloc (buf_neededlength * sizeof (CHAR_T)); + if (buf == NULL) + goto out_of_memory_1; + buf_malloced = buf; + } + + if (resultbuf != NULL) + { + result = resultbuf; + allocated = *lengthp; + } + else + { + result = NULL; + allocated = 0; + } + length = 0; + /* Invariants: + result is either == resultbuf or == NULL or malloc-allocated. + If length > 0, then result != NULL. */ + + /* Ensures that allocated >= length + extra. Aborts through a jump to + out_of_memory if size is too big. */ +#define ENSURE_ALLOCATION(extra) \ + { \ + size_t needed = length + (extra); \ + if (needed < length) \ + goto out_of_memory; \ + if (needed > allocated) \ + { \ + size_t memory_size; \ + CHAR_T *memory; \ + \ + allocated = (allocated > 0 ? 2 * allocated : 12); \ + if (needed > allocated) \ + allocated = needed; \ + if (SIZE_MAX / sizeof (CHAR_T) < allocated) \ + goto out_of_memory; \ + memory_size = allocated * sizeof (CHAR_T); \ + if (result == resultbuf || result == NULL) \ + memory = (CHAR_T *) malloc (memory_size); \ + else \ + memory = (CHAR_T *) realloc (result, memory_size); \ + if (memory == NULL) \ + goto out_of_memory; \ + if (result == resultbuf && length > 0) \ + memcpy (memory, result, length * sizeof (CHAR_T)); \ + result = memory; \ + } \ + } + + for (cp = format, i = 0, dp = &d.dir[0]; ; cp = dp->dir_end, i++, dp++) + { + if (cp != dp->dir_start) + { + size_t n = dp->dir_start - cp; + + ENSURE_ALLOCATION (n); + memcpy (result + length, cp, n * sizeof (CHAR_T)); + length += n; + } + if (i == d.count) + break; + + /* Execute a single directive. */ + if (dp->conversion == '%') + { + if (!(dp->arg_index == ARG_NONE)) + abort (); + ENSURE_ALLOCATION (1); + result[length] = '%'; + length += 1; + } + else + { + if (!(dp->arg_index != ARG_NONE)) + abort (); + + if (dp->conversion == 'n') + { + switch (a.arg[dp->arg_index].type) + { + case TYPE_COUNT_SCHAR_POINTER: + *a.arg[dp->arg_index].a.a_count_schar_pointer = length; + break; + case TYPE_COUNT_SHORT_POINTER: + *a.arg[dp->arg_index].a.a_count_short_pointer = length; + break; + case TYPE_COUNT_INT_POINTER: + *a.arg[dp->arg_index].a.a_count_int_pointer = length; + break; + case TYPE_COUNT_LONGINT_POINTER: + *a.arg[dp->arg_index].a.a_count_longint_pointer = length; + break; +#ifdef HAVE_LONG_LONG + case TYPE_COUNT_LONGLONGINT_POINTER: + *a.arg[dp->arg_index].a.a_count_longlongint_pointer = length; + break; +#endif + default: + abort (); + } + } + else + { + arg_type type = a.arg[dp->arg_index].type; + CHAR_T *p; + unsigned int prefix_count; + int prefixes[2]; +#if !USE_SNPRINTF + size_t tmp_length; + CHAR_T tmpbuf[700]; + CHAR_T *tmp; + + /* Allocate a temporary buffer of sufficient size for calling + sprintf. */ + { + size_t width; + size_t precision; + + width = 0; + if (dp->width_start != dp->width_end) + { + if (dp->width_arg_index != ARG_NONE) + { + int arg; + + if (!(a.arg[dp->width_arg_index].type == TYPE_INT)) + abort (); + arg = a.arg[dp->width_arg_index].a.a_int; + width = (arg < 0 ? (unsigned int) (-arg) : arg); + } + else + { + const CHAR_T *digitp = dp->width_start; + + do + { + size_t w_tmp = width * 10 + (*digitp++ - '0'); + if (SIZE_MAX / 10 < width || w_tmp < width) + goto out_of_memory; + width = w_tmp; + } + while (digitp != dp->width_end); + } + } + + precision = 6; + if (dp->precision_start != dp->precision_end) + { + if (dp->precision_arg_index != ARG_NONE) + { + int arg; + + if (!(a.arg[dp->precision_arg_index].type == TYPE_INT)) + abort (); + arg = a.arg[dp->precision_arg_index].a.a_int; + precision = (arg < 0 ? 0 : arg); + } + else + { + const CHAR_T *digitp = dp->precision_start + 1; + + precision = 0; + while (digitp != dp->precision_end) + { + size_t p1 = 10 * precision + (*digitp++ - '0'); + precision = ((SIZE_MAX / 10 < precision + || p1 < precision) + ? SIZE_MAX : p1); + } + } + } + + switch (dp->conversion) + { + + case 'd': case 'i': case 'u': +# ifdef HAVE_LONG_LONG + if (type == TYPE_LONGLONGINT || type == TYPE_ULONGLONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long long) * CHAR_BIT + * 0.30103 /* binary -> decimal */ + ) + + 1; /* turn floor into ceil */ + else +# endif + if (type == TYPE_LONGINT || type == TYPE_ULONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long) * CHAR_BIT + * 0.30103 /* binary -> decimal */ + ) + + 1; /* turn floor into ceil */ + else + tmp_length = + (unsigned int) (sizeof (unsigned int) * CHAR_BIT + * 0.30103 /* binary -> decimal */ + ) + + 1; /* turn floor into ceil */ + if (tmp_length < precision) + tmp_length = precision; + /* Multiply by 2, as an estimate for FLAG_GROUP. */ + /* Add 1, to account for a leading sign. */ + tmp_length = (tmp_length < SIZE_MAX / 2 + ? 2 * tmp_length + 1 + : SIZE_MAX); + break; + + case 'o': +# ifdef HAVE_LONG_LONG + if (type == TYPE_LONGLONGINT || type == TYPE_ULONGLONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long long) * CHAR_BIT + * 0.333334 /* binary -> octal */ + ) + + 1; /* turn floor into ceil */ + else +# endif + if (type == TYPE_LONGINT || type == TYPE_ULONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long) * CHAR_BIT + * 0.333334 /* binary -> octal */ + ) + + 1; /* turn floor into ceil */ + else + tmp_length = + (unsigned int) (sizeof (unsigned int) * CHAR_BIT + * 0.333334 /* binary -> octal */ + ) + + 1; /* turn floor into ceil */ + if (tmp_length < precision) + tmp_length = precision; + /* Add 1, to account for a leading sign. */ + tmp_length += (tmp_length < SIZE_MAX); + break; + + case 'x': case 'X': +# ifdef HAVE_LONG_LONG + if (type == TYPE_LONGLONGINT || type == TYPE_ULONGLONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long long) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1; /* turn floor into ceil */ + else +# endif + if (type == TYPE_LONGINT || type == TYPE_ULONGINT) + tmp_length = + (unsigned int) (sizeof (unsigned long) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1; /* turn floor into ceil */ + else + tmp_length = + (unsigned int) (sizeof (unsigned int) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1; /* turn floor into ceil */ + if (tmp_length < precision) + tmp_length = precision; + /* Add 2, to account for a leading sign or alternate form. */ + if (tmp_length <= SIZE_MAX / 2) + tmp_length *= 2; + break; + + case 'f': case 'F': +# ifdef HAVE_LONG_DOUBLE + if (type == TYPE_LONGDOUBLE) + tmp_length = + (unsigned int) (LDBL_MAX_EXP + * 0.30103 /* binary -> decimal */ + * 2 /* estimate for FLAG_GROUP */ + ) + + 1 /* turn floor into ceil */ + + 10; /* sign, decimal point etc. */ + else +# endif + tmp_length = + (unsigned int) (DBL_MAX_EXP + * 0.30103 /* binary -> decimal */ + * 2 /* estimate for FLAG_GROUP */ + ) + + 1 /* turn floor into ceil */ + + 10; /* sign, decimal point etc. */ + tmp_length += precision; + if (tmp_length < precision) + goto out_of_memory; + break; + + case 'e': case 'E': case 'g': case 'G': + case 'a': case 'A': + tmp_length = + 12; /* sign, decimal point, exponent etc. */ + tmp_length += precision; + if (tmp_length < precision) + goto out_of_memory; + break; + + case 'c': +# if defined HAVE_WINT_T && !WIDE_CHAR_VERSION + if (type == TYPE_WIDE_CHAR) + tmp_length = MB_CUR_MAX; + else +# endif + tmp_length = 1; + break; + + case 's': +# ifdef HAVE_WCHAR_T + if (type == TYPE_WIDE_STRING) + { + tmp_length = + local_wcslen (a.arg[dp->arg_index].a.a_wide_string); + +# if !WIDE_CHAR_VERSION + if (SIZE_MAX / MB_CUR_MAX < tmp_length) + goto out_of_memory; + tmp_length *= MB_CUR_MAX; +# endif + } + else +# endif + tmp_length = strlen (a.arg[dp->arg_index].a.a_string); + break; + + case 'p': + tmp_length = + (unsigned int) (sizeof (void *) * CHAR_BIT + * 0.25 /* binary -> hexadecimal */ + ) + + 1 /* turn floor into ceil */ + + 2; /* account for leading 0x */ + break; + + default: + abort (); + } + + if (tmp_length < width) + tmp_length = width; + + tmp_length++; /* account for trailing NUL */ + if (!tmp_length) + goto out_of_memory; + } + + if (tmp_length <= sizeof (tmpbuf) / sizeof (CHAR_T)) + tmp = tmpbuf; + else + { + if (SIZE_MAX / sizeof (CHAR_T) < tmp_length) + /* Overflow, would lead to out of memory. */ + goto out_of_memory; + tmp = (CHAR_T *) malloc (tmp_length * sizeof (CHAR_T)); + if (tmp == NULL) + /* Out of memory. */ + goto out_of_memory; + } +#endif + + /* Construct the format string for calling snprintf or + sprintf. */ + p = buf; + *p++ = '%'; + if (dp->flags & FLAG_GROUP) + *p++ = '\''; + if (dp->flags & FLAG_LEFT) + *p++ = '-'; + if (dp->flags & FLAG_SHOWSIGN) + *p++ = '+'; + if (dp->flags & FLAG_SPACE) + *p++ = ' '; + if (dp->flags & FLAG_ALT) + *p++ = '#'; + if (dp->flags & FLAG_ZERO) + *p++ = '0'; + if (dp->width_start != dp->width_end) + { + size_t n = dp->width_end - dp->width_start; + memcpy (p, dp->width_start, n * sizeof (CHAR_T)); + p += n; + } + if (dp->precision_start != dp->precision_end) + { + size_t n = dp->precision_end - dp->precision_start; + memcpy (p, dp->precision_start, n * sizeof (CHAR_T)); + p += n; + } + + switch (type) + { +#ifdef HAVE_LONG_LONG + case TYPE_LONGLONGINT: + case TYPE_ULONGLONGINT: + *p++ = 'l'; + /*FALLTHROUGH*/ +#endif + case TYPE_LONGINT: + case TYPE_ULONGINT: +#ifdef HAVE_WINT_T + case TYPE_WIDE_CHAR: +#endif +#ifdef HAVE_WCHAR_T + case TYPE_WIDE_STRING: +#endif + *p++ = 'l'; + break; +#ifdef HAVE_LONG_DOUBLE + case TYPE_LONGDOUBLE: + *p++ = 'L'; + break; +#endif + default: + break; + } + *p = dp->conversion; +#if USE_SNPRINTF + p[1] = '%'; + p[2] = 'n'; + p[3] = '\0'; +#else + p[1] = '\0'; +#endif + + /* Construct the arguments for calling snprintf or sprintf. */ + prefix_count = 0; + if (dp->width_arg_index != ARG_NONE) + { + if (!(a.arg[dp->width_arg_index].type == TYPE_INT)) + abort (); + prefixes[prefix_count++] = a.arg[dp->width_arg_index].a.a_int; + } + if (dp->precision_arg_index != ARG_NONE) + { + if (!(a.arg[dp->precision_arg_index].type == TYPE_INT)) + abort (); + prefixes[prefix_count++] = a.arg[dp->precision_arg_index].a.a_int; + } + +#if USE_SNPRINTF + /* Prepare checking whether snprintf returns the count + via %n. */ + ENSURE_ALLOCATION (1); + result[length] = '\0'; +#endif + + for (;;) + { + size_t maxlen; + int count; + int retcount; + + maxlen = allocated - length; + count = -1; + retcount = 0; + +#if USE_SNPRINTF +# define SNPRINTF_BUF(arg) \ + switch (prefix_count) \ + { \ + case 0: \ + retcount = SNPRINTF (result + length, maxlen, buf, \ + arg, &count); \ + break; \ + case 1: \ + retcount = SNPRINTF (result + length, maxlen, buf, \ + prefixes[0], arg, &count); \ + break; \ + case 2: \ + retcount = SNPRINTF (result + length, maxlen, buf, \ + prefixes[0], prefixes[1], arg, \ + &count); \ + break; \ + default: \ + abort (); \ + } +#else +# define SNPRINTF_BUF(arg) \ + switch (prefix_count) \ + { \ + case 0: \ + count = sprintf (tmp, buf, arg); \ + break; \ + case 1: \ + count = sprintf (tmp, buf, prefixes[0], arg); \ + break; \ + case 2: \ + count = sprintf (tmp, buf, prefixes[0], prefixes[1],\ + arg); \ + break; \ + default: \ + abort (); \ + } +#endif + + switch (type) + { + case TYPE_SCHAR: + { + int arg = a.arg[dp->arg_index].a.a_schar; + SNPRINTF_BUF (arg); + } + break; + case TYPE_UCHAR: + { + unsigned int arg = a.arg[dp->arg_index].a.a_uchar; + SNPRINTF_BUF (arg); + } + break; + case TYPE_SHORT: + { + int arg = a.arg[dp->arg_index].a.a_short; + SNPRINTF_BUF (arg); + } + break; + case TYPE_USHORT: + { + unsigned int arg = a.arg[dp->arg_index].a.a_ushort; + SNPRINTF_BUF (arg); + } + break; + case TYPE_INT: + { + int arg = a.arg[dp->arg_index].a.a_int; + SNPRINTF_BUF (arg); + } + break; + case TYPE_UINT: + { + unsigned int arg = a.arg[dp->arg_index].a.a_uint; + SNPRINTF_BUF (arg); + } + break; + case TYPE_LONGINT: + { + long int arg = a.arg[dp->arg_index].a.a_longint; + SNPRINTF_BUF (arg); + } + break; + case TYPE_ULONGINT: + { + unsigned long int arg = a.arg[dp->arg_index].a.a_ulongint; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_LONG_LONG + case TYPE_LONGLONGINT: + { + long long int arg = a.arg[dp->arg_index].a.a_longlongint; + SNPRINTF_BUF (arg); + } + break; + case TYPE_ULONGLONGINT: + { + unsigned long long int arg = a.arg[dp->arg_index].a.a_ulonglongint; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_DOUBLE: + { + double arg = a.arg[dp->arg_index].a.a_double; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_LONG_DOUBLE + case TYPE_LONGDOUBLE: + { + long double arg = a.arg[dp->arg_index].a.a_longdouble; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_CHAR: + { + int arg = a.arg[dp->arg_index].a.a_char; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_WINT_T + case TYPE_WIDE_CHAR: + { + wint_t arg = a.arg[dp->arg_index].a.a_wide_char; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_STRING: + { + const char *arg = a.arg[dp->arg_index].a.a_string; + SNPRINTF_BUF (arg); + } + break; +#ifdef HAVE_WCHAR_T + case TYPE_WIDE_STRING: + { + const wchar_t *arg = a.arg[dp->arg_index].a.a_wide_string; + SNPRINTF_BUF (arg); + } + break; +#endif + case TYPE_POINTER: + { + void *arg = a.arg[dp->arg_index].a.a_pointer; + SNPRINTF_BUF (arg); + } + break; + default: + abort (); + } + +#if USE_SNPRINTF + /* Portability: Not all implementations of snprintf() + are ISO C 99 compliant. Determine the number of + bytes that snprintf() has produced or would have + produced. */ + if (count >= 0) + { + /* Verify that snprintf() has NUL-terminated its + result. */ + if (count < maxlen && result[length + count] != '\0') + abort (); + /* Portability hack. */ + if (retcount > count) + count = retcount; + } + else + { + /* snprintf() doesn't understand the '%n' + directive. */ + if (p[1] != '\0') + { + /* Don't use the '%n' directive; instead, look + at the snprintf() return value. */ + p[1] = '\0'; + continue; + } + else + { + /* Look at the snprintf() return value. */ + if (retcount < 0) + { + /* HP-UX 10.20 snprintf() is doubly deficient: + It doesn't understand the '%n' directive, + *and* it returns -1 (rather than the length + that would have been required) when the + buffer is too small. */ + size_t bigger_need = + (allocated > 12 ? allocated : 12); + ENSURE_ALLOCATION (bigger_need); + continue; + } + else + count = retcount; + } + } +#endif + + /* Attempt to handle failure. */ + if (count < 0) + { + if (!(result == resultbuf || result == NULL)) + free (result); + if (buf_malloced != NULL) + free (buf_malloced); + CLEANUP (); + errno = EINVAL; + return NULL; + } + +#if !USE_SNPRINTF + if (count >= tmp_length) + /* tmp_length was incorrectly calculated - fix the + code above! */ + abort (); +#endif + + /* Make room for the result. */ + if (count >= maxlen) + { + /* Need at least count bytes. But allocate + proportionally, to avoid looping eternally if + snprintf() reports a too small count. */ + ENSURE_ALLOCATION (count < allocated + ? allocated : count); +#if USE_SNPRINTF + continue; +#endif + } + +#if USE_SNPRINTF + /* The snprintf() result did fit. */ +#else + /* Append the sprintf() result. */ + memcpy (result + length, tmp, count * sizeof (CHAR_T)); + if (tmp != tmpbuf) + free (tmp); +#endif + + length += count; + break; + } + } + } + } + + /* Add the final NUL. */ + ENSURE_ALLOCATION (1); + result[length] = '\0'; + + if (result != resultbuf && length + 1 < allocated) + { + /* Shrink the allocated memory if possible. */ + CHAR_T *memory; + + memory = (CHAR_T *) realloc (result, (length + 1) * sizeof (CHAR_T)); + if (memory != NULL) + result = memory; + } + + if (buf_malloced != NULL) + free (buf_malloced); + CLEANUP (); + *lengthp = length; + if (length > INT_MAX) + goto length_overflow; + return result; + + length_overflow: + /* We could produce such a big string, but its length doesn't fit into + an 'int'. POSIX says that snprintf() fails with errno = EOVERFLOW in + this case. */ + if (result != resultbuf) + free (result); + errno = EOVERFLOW; + return NULL; + + out_of_memory: + if (!(result == resultbuf || result == NULL)) + free (result); + if (buf_malloced != NULL) + free (buf_malloced); + out_of_memory_1: + CLEANUP (); + errno = ENOMEM; + return NULL; + } +} + +#undef SNPRINTF +#undef USE_SNPRINTF +#undef PRINTF_PARSE +#undef DIRECTIVES +#undef DIRECTIVE +#undef CHAR_T +#undef VASNPRINTF diff --git a/lib/vasnprintf.h b/lib/vasnprintf.h new file mode 100644 index 0000000..7a0c01f --- /dev/null +++ b/lib/vasnprintf.h @@ -0,0 +1,81 @@ +/* vsprintf with automatic memory allocation. + Copyright (C) 2002-2004, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _VASNPRINTF_H +#define _VASNPRINTF_H + +/* Get va_list. */ +#include <stdarg.h> + +/* Get size_t. */ +#include <stddef.h> + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) || __STRICT_ANSI__ +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* Write formatted output to a string dynamically allocated with malloc(). + You can pass a preallocated buffer for the result in RESULTBUF and its + size in *LENGTHP; otherwise you pass RESULTBUF = NULL. + If successful, return the address of the string (this may be = RESULTBUF + if no dynamic memory allocation was necessary) and set *LENGTHP to the + number of resulting bytes, excluding the trailing NUL. Upon error, set + errno and return NULL. + + When dynamic memory allocation occurs, the preallocated buffer is left + alone (with possibly modified contents). This makes it possible to use + a statically allocated or stack-allocated buffer, like this: + + char buf[100]; + size_t len = sizeof (buf); + char *output = vasnprintf (buf, &len, format, args); + if (output == NULL) + ... error handling ...; + else + { + ... use the output string ...; + if (output != buf) + free (output); + } + */ +#if REPLACE_VASNPRINTF +# define asnprintf rpl_asnprintf +# define vasnprintf rpl_vasnprintf +#endif +extern char * asnprintf (char *resultbuf, size_t *lengthp, const char *format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); +extern char * vasnprintf (char *resultbuf, size_t *lengthp, const char *format, va_list args) + __attribute__ ((__format__ (__printf__, 3, 0))); + +#ifdef __cplusplus +} +#endif + +#endif /* _VASNPRINTF_H */ diff --git a/lib/verify.h b/lib/verify.h new file mode 100644 index 0000000..d603b17 --- /dev/null +++ b/lib/verify.h @@ -0,0 +1,141 @@ +/* Compile-time assert-like macros. + + Copyright (C) 2005, 2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Paul Eggert, Bruno Haible, and Jim Meyering. */ + +#ifndef VERIFY_H +# define VERIFY_H 1 + +/* Each of these macros verifies that its argument R is nonzero. To + be portable, R should be an integer constant expression. Unlike + assert (R), there is no run-time overhead. + + There are two macros, since no single macro can be used in all + contexts in C. verify_true (R) is for scalar contexts, including + integer constant expression contexts. verify (R) is for declaration + contexts, e.g., the top level. + + Symbols ending in "__" are private to this header. + + The code below uses several ideas. + + * The first step is ((R) ? 1 : -1). Given an expression R, of + integral or boolean or floating-point type, this yields an + expression of integral type, whose value is later verified to be + constant and nonnegative. + + * Next this expression W is wrapped in a type + struct verify_type__ { unsigned int verify_error_if_negative_size__: W; }. + If W is negative, this yields a compile-time error. No compiler can + deal with a bit-field of negative size. + + One might think that an array size check would have the same + effect, that is, that the type struct { unsigned int dummy[W]; } + would work as well. However, inside a function, some compilers + (such as C++ compilers and GNU C) allow local parameters and + variables inside array size expressions. With these compilers, + an array size check would not properly diagnose this misuse of + the verify macro: + + void function (int n) { verify (n < 0); } + + * For the verify macro, the struct verify_type__ will need to + somehow be embedded into a declaration. To be portable, this + declaration must declare an object, a constant, a function, or a + typedef name. If the declared entity uses the type directly, + such as in + + struct dummy {...}; + typedef struct {...} dummy; + extern struct {...} *dummy; + extern void dummy (struct {...} *); + extern struct {...} *dummy (void); + + two uses of the verify macro would yield colliding declarations + if the entity names are not disambiguated. A workaround is to + attach the current line number to the entity name: + + #define GL_CONCAT0(x, y) x##y + #define GL_CONCAT(x, y) GL_CONCAT0 (x, y) + extern struct {...} * GL_CONCAT(dummy,__LINE__); + + But this has the problem that two invocations of verify from + within the same macro would collide, since the __LINE__ value + would be the same for both invocations. + + A solution is to use the sizeof operator. It yields a number, + getting rid of the identity of the type. Declarations like + + extern int dummy [sizeof (struct {...})]; + extern void dummy (int [sizeof (struct {...})]); + extern int (*dummy (void)) [sizeof (struct {...})]; + + can be repeated. + + * Should the implementation use a named struct or an unnamed struct? + Which of the following alternatives can be used? + + extern int dummy [sizeof (struct {...})]; + extern int dummy [sizeof (struct verify_type__ {...})]; + extern void dummy (int [sizeof (struct {...})]); + extern void dummy (int [sizeof (struct verify_type__ {...})]); + extern int (*dummy (void)) [sizeof (struct {...})]; + extern int (*dummy (void)) [sizeof (struct verify_type__ {...})]; + + In the second and sixth case, the struct type is exported to the + outer scope; two such declarations therefore collide. GCC warns + about the first, third, and fourth cases. So the only remaining + possibility is the fifth case: + + extern int (*dummy (void)) [sizeof (struct {...})]; + + * This implementation exploits the fact that GCC does not warn about + the last declaration mentioned above. If a future version of GCC + introduces a warning for this, the problem could be worked around + by using code specialized to GCC, e.g.,: + + #if 4 <= __GNUC__ + # define verify(R) \ + extern int (* verify_function__ (void)) \ + [__builtin_constant_p (R) && (R) ? 1 : -1] + #endif + + * In C++, any struct definition inside sizeof is invalid. + Use a template type to work around the problem. */ + + +/* Verify requirement R at compile-time, as an integer constant expression. + Return 1. */ + +# ifdef __cplusplus +template <int w> + struct verify_type__ { unsigned int verify_error_if_negative_size__: w; }; +# define verify_true(R) \ + (!!sizeof (verify_type__<(R) ? 1 : -1>)) +# else +# define verify_true(R) \ + (!!sizeof \ + (struct { unsigned int verify_error_if_negative_size__: (R) ? 1 : -1; })) +# endif + +/* Verify requirement R at compile-time, as a declaration without a + trailing ';'. */ + +# define verify(R) extern int (* verify_function__ (void)) [verify_true (R)] + +#endif diff --git a/lib/version-etc-fsf.c b/lib/version-etc-fsf.c new file mode 100644 index 0000000..f25eb65 --- /dev/null +++ b/lib/version-etc-fsf.c @@ -0,0 +1,31 @@ +/* Variable with FSF copyright information, for version-etc. + Copyright (C) 1999-2006 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +/* Specification. */ +#include "version-etc.h" + +/* Default copyright goes to the FSF. */ + +const char version_etc_copyright[] = + /* Do *not* mark this string for translation. %s is a copyright + symbol suitable for this locale, and %d is the copyright + year. */ + "Copyright %s %d Free Software Foundation, Inc."; diff --git a/lib/version-etc.c b/lib/version-etc.c new file mode 100644 index 0000000..1ae82c4 --- /dev/null +++ b/lib/version-etc.c @@ -0,0 +1,173 @@ +/* Utility to help print --version output in a consistent format. + Copyright (C) 1999-2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +/* Specification. */ +#include "version-etc.h" + +#include <stdarg.h> +#include <stdio.h> +#include <stdlib.h> + +#if USE_UNLOCKED_IO +# include "unlocked-io.h" +#endif + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +enum { COPYRIGHT_YEAR = 2007 }; + +/* Like version_etc, below, but with the NULL-terminated author list + provided via a variable of type va_list. */ +void +version_etc_va (FILE *stream, + const char *command_name, const char *package, + const char *version, va_list authors) +{ + size_t n_authors; + + /* Count the number of authors. */ + { + va_list tmp_authors; + + va_copy (tmp_authors, authors); + + n_authors = 0; + while (va_arg (tmp_authors, const char *) != NULL) + ++n_authors; + } + + if (command_name) + fprintf (stream, "%s (%s) %s\n", command_name, package, version); + else + fprintf (stream, "%s %s\n", package, version); + + /* TRANSLATORS: Translate "(C)" to the copyright symbol + (C-in-a-circle), if this symbol is available in the user's + locale. Otherwise, do not translate "(C)"; leave it as-is. */ + fprintf (stream, version_etc_copyright, _("(C)"), COPYRIGHT_YEAR); + + fputs (_("\ +\n\ +License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>\n\ +This is free software: you are free to change and redistribute it.\n\ +There is NO WARRANTY, to the extent permitted by law.\n\ +\n\ +"), + stream); + + switch (n_authors) + { + case 0: + /* The caller must provide at least one author name. */ + abort (); + case 1: + /* TRANSLATORS: %s denotes an author name. */ + vfprintf (stream, _("Written by %s.\n"), authors); + break; + case 2: + /* TRANSLATORS: Each %s denotes an author name. */ + vfprintf (stream, _("Written by %s and %s.\n"), authors); + break; + case 3: + /* TRANSLATORS: Each %s denotes an author name. */ + vfprintf (stream, _("Written by %s, %s, and %s.\n"), authors); + break; + case 4: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\nand %s.\n"), authors); + break; + case 5: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\n%s, and %s.\n"), authors); + break; + case 6: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\n%s, %s, and %s.\n"), + authors); + break; + case 7: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\n%s, %s, %s, and %s.\n"), + authors); + break; + case 8: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("\ +Written by %s, %s, %s,\n%s, %s, %s, %s,\nand %s.\n"), + authors); + break; + case 9: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("\ +Written by %s, %s, %s,\n%s, %s, %s, %s,\n%s, and %s.\n"), + authors); + break; + default: + /* 10 or more authors. Use an abbreviation, since the human reader + will probably not want to read the entire list anyway. */ + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("\ +Written by %s, %s, %s,\n%s, %s, %s, %s,\n%s, %s, and others.\n"), + authors); + break; + } + va_end (authors); +} + + +/* Display the --version information the standard way. + + If COMMAND_NAME is NULL, the PACKAGE is asumed to be the name of + the program. The formats are therefore: + + PACKAGE VERSION + + or + + COMMAND_NAME (PACKAGE) VERSION. + + The author names are passed as separate arguments, with an additional + NULL argument at the end. */ +void +version_etc (FILE *stream, + const char *command_name, const char *package, + const char *version, /* const char *author1, ...*/ ...) +{ + va_list authors; + + va_start (authors, version); + version_etc_va (stream, command_name, package, version, authors); +} diff --git a/lib/version-etc.h b/lib/version-etc.h new file mode 100644 index 0000000..84da535 --- /dev/null +++ b/lib/version-etc.h @@ -0,0 +1,37 @@ +/* Utility to help print --version output in a consistent format. + Copyright (C) 1999, 2003, 2005 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#ifndef VERSION_ETC_H +# define VERSION_ETC_H 1 + +# include <stdarg.h> +# include <stdio.h> + +extern const char version_etc_copyright[]; + +extern void version_etc_va (FILE *stream, + const char *command_name, const char *package, + const char *version, va_list authors); + +extern void version_etc (FILE *stream, + const char *command_name, const char *package, + const char *version, + /* const char *author1, ...*/ ...); + +#endif /* VERSION_ETC_H */ diff --git a/lib/vsnprintf.c b/lib/vsnprintf.c new file mode 100644 index 0000000..4f841fc --- /dev/null +++ b/lib/vsnprintf.c @@ -0,0 +1,76 @@ +/* Formatted output to strings. + Copyright (C) 2004, 2006-2007 Free Software Foundation, Inc. + Written by Simon Josefsson and Yoann Vandoorselaere <yoann@prelude-ids.org>. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +/* Specification. */ +#include <stdio.h> + +#include <errno.h> +#include <limits.h> +#include <stdarg.h> +#include <stdlib.h> +#include <string.h> + +#include "vasnprintf.h" + +/* Some systems, like OSF/1 4.0 and Woe32, don't have EOVERFLOW. */ +#ifndef EOVERFLOW +# define EOVERFLOW E2BIG +#endif + +/* Print formatted output to string STR. Similar to vsprintf, but + additional length SIZE limit how much is written into STR. Returns + string length of formatted string (which may be larger than SIZE). + STR may be NULL, in which case nothing will be written. On error, + return a negative value. */ +int +vsnprintf (char *str, size_t size, const char *format, va_list args) +{ + char *output; + size_t len; + size_t lenbuf = size; + + output = vasnprintf (str, &lenbuf, format, args); + len = lenbuf; + + if (!output) + return -1; + + if (output != str) + { + if (size) + { + size_t pruned_len = (len < size ? len : size - 1); + memcpy (str, output, pruned_len); + str[pruned_len] = '\0'; + } + + free (output); + } + + if (len > INT_MAX) + { + errno = EOVERFLOW; + return -1; + } + + return len; +} diff --git a/lib/waitpid.c b/lib/waitpid.c new file mode 100644 index 0000000..b6f585e --- /dev/null +++ b/lib/waitpid.c @@ -0,0 +1,75 @@ +/* Emulate waitpid on systems that just have wait. + Copyright 1994, 1995, 1998, 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif + +#include <sys/types.h> +#include <sys/stat.h> + +#include <errno.h> +#ifndef errno +extern int errno; +#endif + +#define WAITPID_CHILDREN 8 +static pid_t waited_pid[WAITPID_CHILDREN]; +static int waited_status[WAITPID_CHILDREN]; + +pid_t +waitpid (pid_t pid, int *stat_loc, int options) +{ + int i; + pid_t p; + + if (!options && (pid == -1 || 0 < pid)) + { + /* If we have already waited for this child, return it immediately. */ + for (i = 0; i < WAITPID_CHILDREN; i++) + { + p = waited_pid[i]; + if (p && (p == pid || pid == -1)) + { + waited_pid[i] = 0; + goto success; + } + } + + /* The child has not returned yet; wait for it, accumulating status. */ + for (i = 0; i < WAITPID_CHILDREN; i++) + if (! waited_pid[i]) + { + p = wait (&waited_status[i]); + if (p < 0) + return p; + if (p == pid || pid == -1) + goto success; + waited_pid[i] = p; + } + } + + /* We cannot emulate this wait call, e.g. because of too many children. */ + errno = EINVAL; + return -1; + +success: + if (stat_loc) + *stat_loc = waited_status[i]; + return p; +} diff --git a/lib/wchar_.h b/lib/wchar_.h new file mode 100644 index 0000000..2a71b8b --- /dev/null +++ b/lib/wchar_.h @@ -0,0 +1,50 @@ +/* A substitute for ISO C99 <wchar.h>, for platforms that have issues. + + Copyright (C) 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Eric Blake. */ + +/* + * ISO C 99 <wchar.h> for platforms that have issues. + * <http://www.opengroup.org/susv3xbd/wchar.h.html> + * + * For now, this just ensures proper prerequisite inclusion order. + */ + +#ifndef _GL_WCHAR_H + +/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be + included before <wchar.h>. */ +#include <stddef.h> +#include <stdio.h> +#include <time.h> + +/* Include the original <wchar.h>. */ +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_INCLUDE_NEXT@ +# include_next <wchar.h> +#else +# include @ABSOLUTE_WCHAR_H@ +#endif + +#ifndef _GL_WCHAR_H +#define _GL_WCHAR_H + +#endif /* _GL_WCHAR_H */ +#endif /* _GL_WCHAR_H */ diff --git a/lib/wctype_.h b/lib/wctype_.h new file mode 100644 index 0000000..ebbfca4 --- /dev/null +++ b/lib/wctype_.h @@ -0,0 +1,165 @@ +/* A substitute for ISO C99 <wctype.h>, for platforms that lack it. + + Copyright (C) 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Bruno Haible and Paul Eggert. */ + +/* + * ISO C 99 <wctype.h> for platforms that lack it. + * <http://www.opengroup.org/susv3xbd/wctype.h.html> + * + * iswctype, towctrans, towlower, towupper, wctrans, wctype, + * wctrans_t, and wctype_t are not yet implemented. + */ + +#ifndef _GL_WCTYPE_H + +#if @HAVE_WINT_T@ +/* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>. + Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before + <wchar.h>. + BSD/OS 4.0.1 has a bug: <stddef.h>, <stdio.h> and <time.h> must be + included before <wchar.h>. */ +# include <stddef.h> +# include <stdio.h> +# include <time.h> +# include <wchar.h> +#endif + +/* Include the original <wctype.h> if it exists. + BeOS 5 has the functions but no <wctype.h>. */ +/* The include_next requires a split double-inclusion guard. */ +#if @HAVE_WCTYPE_H@ +# if @HAVE_INCLUDE_NEXT@ +# include_next <wctype.h> +# else +# include @ABSOLUTE_WCTYPE_H@ +# endif +#endif + +#ifndef _GL_WCTYPE_H +#define _GL_WCTYPE_H + +#if @HAVE_WINT_T@ +typedef wint_t __wctype_wint_t; +#else +typedef int __wctype_wint_t; +#endif + +/* FreeBSD 4.4 to 4.11 has <wctype.h> but lacks the functions. + Assume all 12 functions are implemented the same way, or not at all. */ +#if ! @HAVE_ISWCNTRL@ + +/* IRIX 5.3 has macros but no functions, its isw* macros refer to an + undefined variable _ctmp_ and to <ctype.h> macros like _P, and they + refer to system functions like _iswctype that are not in the + standard C library. Rather than try to get ancient buggy + implementations like this to work, just disable them. */ +# undef iswalnum +# undef iswalpha +# undef iswblank +# undef iswcntrl +# undef iswdigit +# undef iswgraph +# undef iswlower +# undef iswprint +# undef iswpunct +# undef iswspace +# undef iswupper +# undef iswxdigit + +static inline int +iswalnum (__wctype_wint_t wc) +{ + return ((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z')); +} + +static inline int +iswalpha (__wctype_wint_t wc) +{ + return (wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z'; +} + +static inline int +iswblank (__wctype_wint_t wc) +{ + return wc == ' ' || wc == '\t'; +} + +static inline int +iswcntrl (__wctype_wint_t wc) +{ + return (wc & ~0x1f) == 0 || wc == 0x7f; +} + +static inline int +iswdigit (__wctype_wint_t wc) +{ + return wc >= '0' && wc <= '9'; +} + +static inline int +iswgraph (__wctype_wint_t wc) +{ + return wc >= '!' && wc <= '~'; +} + +static inline int +iswlower (__wctype_wint_t wc) +{ + return wc >= 'a' && wc <= 'z'; +} + +static inline int +iswprint (__wctype_wint_t wc) +{ + return wc >= ' ' && wc <= '~'; +} + +static inline int +iswpunct (__wctype_wint_t wc) +{ + return (wc >= '!' && wc <= '~' + && !((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z'))); +} + +static inline int +iswspace (__wctype_wint_t wc) +{ + return (wc == ' ' || wc == '\t' + || wc == '\n' || wc == '\v' || wc == '\f' || wc == '\r'); +} + +static inline int +iswupper (__wctype_wint_t wc) +{ + return wc >= 'A' && wc <= 'Z'; +} + +static inline int +iswxdigit (__wctype_wint_t wc) +{ + return ((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'F')); +} + +# endif /* ! HAVE_ISWCNTRL */ + +#endif /* _GL_WCTYPE_H */ +#endif /* _GL_WCTYPE_H */ diff --git a/lib/wcwidth.h b/lib/wcwidth.h new file mode 100644 index 0000000..8ed5ff8 --- /dev/null +++ b/lib/wcwidth.h @@ -0,0 +1,57 @@ +/* Determine the number of screen columns needed for a character. + Copyright (C) 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef _gl_WCWIDTH_H +#define _gl_WCWIDTH_H + +#if HAVE_WCHAR_T + +/* Get wcwidth if available, along with wchar_t. */ +# include <wchar.h> + +/* Get iswprint. */ +# include <wctype.h> + +# ifndef HAVE_DECL_WCWIDTH +"this configure-time declaration test was not run" +# endif +# ifndef wcwidth +# if !HAVE_WCWIDTH + +/* wcwidth doesn't exist, so assume all printable characters have + width 1. */ +static inline int +wcwidth (wchar_t wc) +{ + return wc == 0 ? 0 : iswprint (wc) ? 1 : -1; +} + +# elif !HAVE_DECL_WCWIDTH + +/* wcwidth exists but is not declared. */ +extern +# ifdef __cplusplus +"C" +# endif +int wcwidth (int /* actually wchar_t */); + +# endif +# endif + +#endif /* HAVE_WCHAR_T */ + +#endif /* _gl_WCWIDTH_H */ diff --git a/lib/xalloc-die.c b/lib/xalloc-die.c new file mode 100644 index 0000000..090f060 --- /dev/null +++ b/lib/xalloc-die.c @@ -0,0 +1,42 @@ +/* Report a memory allocation failure and exit. + + Copyright (C) 1997, 1998, 1999, 2000, 2002, 2003, 2004, 2006 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#include "xalloc.h" + +#include <stdlib.h> + +#include "error.h" +#include "exitfail.h" + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +void +xalloc_die (void) +{ + error (exit_failure, 0, "%s", _("memory exhausted")); + + /* The `noreturn' cannot be given to error, since it may return if + its first argument is 0. To help compilers understand the + xalloc_die does not return, call abort. Also, the abort is a + safety feature if exit_failure is 0 (which shouldn't happen). */ + abort (); +} diff --git a/lib/xalloc.h b/lib/xalloc.h new file mode 100644 index 0000000..0c6d8dc --- /dev/null +++ b/lib/xalloc.h @@ -0,0 +1,271 @@ +/* xalloc.h -- malloc with out-of-memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 2003, 2004, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef XALLOC_H_ +# define XALLOC_H_ + +# include <stddef.h> + + +# ifdef __cplusplus +extern "C" { +# endif + + +# ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +# endif + +# ifndef ATTRIBUTE_NORETURN +# define ATTRIBUTE_NORETURN __attribute__ ((__noreturn__)) +# endif + +/* This function is always triggered when memory is exhausted. + It must be defined by the application, either explicitly + or by using gnulib's xalloc-die module. This is the + function to call when one wants the program to die because of a + memory allocation failure. */ +extern void xalloc_die (void) ATTRIBUTE_NORETURN; + +void *xmalloc (size_t s); +void *xzalloc (size_t s); +void *xcalloc (size_t n, size_t s); +void *xrealloc (void *p, size_t s); +void *x2realloc (void *p, size_t *pn); +void *xmemdup (void const *p, size_t s); +char *xstrdup (char const *str); + +/* Return 1 if an array of N objects, each of size S, cannot exist due + to size arithmetic overflow. S must be positive and N must be + nonnegative. This is a macro, not an inline function, so that it + works correctly even when SIZE_MAX < N. + + By gnulib convention, SIZE_MAX represents overflow in size + calculations, so the conservative dividend to use here is + SIZE_MAX - 1, since SIZE_MAX might represent an overflowed value. + However, malloc (SIZE_MAX) fails on all known hosts where + sizeof (ptrdiff_t) <= sizeof (size_t), so do not bother to test for + exactly-SIZE_MAX allocations on such hosts; this avoids a test and + branch when S is known to be 1. */ +# define xalloc_oversized(n, s) \ + ((size_t) (sizeof (ptrdiff_t) <= sizeof (size_t) ? -1 : -2) / (s) < (n)) + + +/* In the following macros, T must be an elementary or structure/union or + typedef'ed type, or a pointer to such a type. To apply one of the + following macros to a function pointer or array type, you need to typedef + it first and use the typedef name. */ + +/* Allocate an object of type T dynamically, with error checking. */ +/* extern t *XMALLOC (typename t); */ +# define XMALLOC(t) ((t *) xmalloc (sizeof (t))) + +/* Allocate memory for N elements of type T, with error checking. */ +/* extern t *XNMALLOC (size_t n, typename t); */ +# define XNMALLOC(n, t) \ + ((t *) (sizeof (t) == 1 ? xmalloc (n) : xnmalloc (n, sizeof (t)))) + +/* Allocate an object of type T dynamically, with error checking, + and zero it. */ +/* extern t *XZALLOC (typename t); */ +# define XZALLOC(t) ((t *) xzalloc (sizeof (t))) + +/* Allocate memory for N elements of type T, with error checking, + and zero it. */ +/* extern t *XCALLOC (size_t n, typename t); */ +# define XCALLOC(n, t) \ + ((t *) (sizeof (t) == 1 ? xzalloc (n) : xcalloc (n, sizeof (t)))) + + +# if HAVE_INLINE +# define static_inline static inline +# else + void *xnmalloc (size_t n, size_t s); + void *xnrealloc (void *p, size_t n, size_t s); + void *x2nrealloc (void *p, size_t *pn, size_t s); + char *xcharalloc (size_t n); +# endif + +# ifdef static_inline + +/* Allocate an array of N objects, each with S bytes of memory, + dynamically, with error checking. S must be nonzero. */ + +static_inline void * +xnmalloc (size_t n, size_t s) +{ + if (xalloc_oversized (n, s)) + xalloc_die (); + return xmalloc (n * s); +} + +/* Change the size of an allocated block of memory P to an array of N + objects each of S bytes, with error checking. S must be nonzero. */ + +static_inline void * +xnrealloc (void *p, size_t n, size_t s) +{ + if (xalloc_oversized (n, s)) + xalloc_die (); + return xrealloc (p, n * s); +} + +/* If P is null, allocate a block of at least *PN such objects; + otherwise, reallocate P so that it contains more than *PN objects + each of S bytes. *PN must be nonzero unless P is null, and S must + be nonzero. Set *PN to the new number of objects, and return the + pointer to the new block. *PN is never set to zero, and the + returned pointer is never null. + + Repeated reallocations are guaranteed to make progress, either by + allocating an initial block with a nonzero size, or by allocating a + larger block. + + In the following implementation, nonzero sizes are increased by a + factor of approximately 1.5 so that repeated reallocations have + O(N) overall cost rather than O(N**2) cost, but the + specification for this function does not guarantee that rate. + + Here is an example of use: + + int *p = NULL; + size_t used = 0; + size_t allocated = 0; + + void + append_int (int value) + { + if (used == allocated) + p = x2nrealloc (p, &allocated, sizeof *p); + p[used++] = value; + } + + This causes x2nrealloc to allocate a block of some nonzero size the + first time it is called. + + To have finer-grained control over the initial size, set *PN to a + nonzero value before calling this function with P == NULL. For + example: + + int *p = NULL; + size_t used = 0; + size_t allocated = 0; + size_t allocated1 = 1000; + + void + append_int (int value) + { + if (used == allocated) + { + p = x2nrealloc (p, &allocated1, sizeof *p); + allocated = allocated1; + } + p[used++] = value; + } + + */ + +static_inline void * +x2nrealloc (void *p, size_t *pn, size_t s) +{ + size_t n = *pn; + + if (! p) + { + if (! n) + { + /* The approximate size to use for initial small allocation + requests, when the invoking code specifies an old size of + zero. 64 bytes is the largest "small" request for the + GNU C library malloc. */ + enum { DEFAULT_MXFAST = 64 }; + + n = DEFAULT_MXFAST / s; + n += !n; + } + } + else + { + /* Set N = ceil (1.5 * N) so that progress is made if N == 1. + Check for overflow, so that N * S stays in size_t range. + The check is slightly conservative, but an exact check isn't + worth the trouble. */ + if ((size_t) -1 / 3 * 2 / s <= n) + xalloc_die (); + n += (n + 1) / 2; + } + + *pn = n; + return xrealloc (p, n * s); +} + +/* Return a pointer to a new buffer of N bytes. This is like xmalloc, + except it returns char *. */ + +static_inline char * +xcharalloc (size_t n) +{ + return XNMALLOC (n, char); +} + +# endif + +# ifdef __cplusplus +} + +/* C++ does not allow conversions from void * to other pointer types + without a cast. Use templates to work around the problem when + possible. */ + +template <typename T> inline T * +xrealloc (T *p, size_t s) +{ + return (T *) xrealloc ((void *) p, s); +} + +template <typename T> inline T * +xnrealloc (T *p, size_t n, size_t s) +{ + return (T *) xnrealloc ((void *) p, n, s); +} + +template <typename T> inline T * +x2realloc (T *p, size_t *pn) +{ + return (T *) x2realloc ((void *) p, pn); +} + +template <typename T> inline T * +x2nrealloc (T *p, size_t *pn, size_t s) +{ + return (T *) x2nrealloc ((void *) p, pn, s); +} + +template <typename T> inline T * +xmemdup (T const *p, size_t s) +{ + return (T *) xmemdup ((void const *) p, s); +} + +# endif + + +#endif /* !XALLOC_H_ */ diff --git a/lib/xgetcwd.c b/lib/xgetcwd.c new file mode 100644 index 0000000..26ea5da --- /dev/null +++ b/lib/xgetcwd.c @@ -0,0 +1,41 @@ +/* xgetcwd.c -- return current directory with unlimited length + + Copyright (C) 2001, 2003, 2004, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#include <config.h> + +#include "xgetcwd.h" + +#include <errno.h> +#include <unistd.h> + +#include "xalloc.h" + +/* Return the current directory, newly allocated. + Upon an out-of-memory error, call xalloc_die. + Upon any other type of error, return NULL. */ + +char * +xgetcwd (void) +{ + char *cwd = getcwd (NULL, 0); + if (! cwd && errno == ENOMEM) + xalloc_die (); + return cwd; +} diff --git a/lib/xgetcwd.h b/lib/xgetcwd.h new file mode 100644 index 0000000..70afe35 --- /dev/null +++ b/lib/xgetcwd.h @@ -0,0 +1,18 @@ +/* prototype for xgetcwd + Copyright (C) 1995, 2001, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +extern char *xgetcwd (void); diff --git a/lib/xmalloc.c b/lib/xmalloc.c new file mode 100644 index 0000000..318e0dd --- /dev/null +++ b/lib/xmalloc.c @@ -0,0 +1,123 @@ +/* xmalloc.c -- malloc with out of memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 2002, 2003, 2004, 2005, 2006 Free Software Foundation, + Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +#if ! HAVE_INLINE +# define static_inline +#endif +#include "xalloc.h" +#undef static_inline + +#include <stdlib.h> +#include <string.h> + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* 1 if calloc is known to be compatible with GNU calloc. This + matters if we are not also using the calloc module, which defines + HAVE_CALLOC and supports the GNU API even on non-GNU platforms. */ +#if defined HAVE_CALLOC || defined __GLIBC__ +enum { HAVE_GNU_CALLOC = 1 }; +#else +enum { HAVE_GNU_CALLOC = 0 }; +#endif + +/* Allocate N bytes of memory dynamically, with error checking. */ + +void * +xmalloc (size_t n) +{ + void *p = malloc (n); + if (!p && n != 0) + xalloc_die (); + return p; +} + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. */ + +void * +xrealloc (void *p, size_t n) +{ + p = realloc (p, n); + if (!p && n != 0) + xalloc_die (); + return p; +} + +/* If P is null, allocate a block of at least *PN bytes; otherwise, + reallocate P so that it contains more than *PN bytes. *PN must be + nonzero unless P is null. Set *PN to the new block's size, and + return the pointer to the new block. *PN is never set to zero, and + the returned pointer is never null. */ + +void * +x2realloc (void *p, size_t *pn) +{ + return x2nrealloc (p, pn, 1); +} + +/* Allocate S bytes of zeroed memory dynamically, with error checking. + There's no need for xnzalloc (N, S), since it would be equivalent + to xcalloc (N, S). */ + +void * +xzalloc (size_t s) +{ + return memset (xmalloc (s), 0, s); +} + +/* Allocate zeroed memory for N elements of S bytes, with error + checking. S must be nonzero. */ + +void * +xcalloc (size_t n, size_t s) +{ + void *p; + /* Test for overflow, since some calloc implementations don't have + proper overflow checks. But omit overflow and size-zero tests if + HAVE_GNU_CALLOC, since GNU calloc catches overflow and never + returns NULL if successful. */ + if ((! HAVE_GNU_CALLOC && xalloc_oversized (n, s)) + || (! (p = calloc (n, s)) && (HAVE_GNU_CALLOC || n != 0))) + xalloc_die (); + return p; +} + +/* Clone an object P of size S, with error checking. There's no need + for xnmemdup (P, N, S), since xmemdup (P, N * S) works without any + need for an arithmetic overflow check. */ + +void * +xmemdup (void const *p, size_t s) +{ + return memcpy (xmalloc (s), p, s); +} + +/* Clone STRING. */ + +char * +xstrdup (char const *string) +{ + return xmemdup (string, strlen (string) + 1); +} diff --git a/lib/xstrndup.c b/lib/xstrndup.c new file mode 100644 index 0000000..7ccefd7 --- /dev/null +++ b/lib/xstrndup.c @@ -0,0 +1,37 @@ +/* Duplicate a bounded initial segment of a string, with out-of-memory + checking. + Copyright (C) 2003, 2006, 2007 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <config.h> + +/* Specification. */ +#include "xstrndup.h" + +#include <string.h> +#include "xalloc.h" + +/* Return a newly allocated copy of at most N bytes of STRING. + In other words, return a copy of the initial segment of length N of + STRING. */ +char * +xstrndup (const char *string, size_t n) +{ + char *s = strndup (string, n); + if (! s) + xalloc_die (); + return s; +} diff --git a/lib/xstrndup.h b/lib/xstrndup.h new file mode 100644 index 0000000..88354cf --- /dev/null +++ b/lib/xstrndup.h @@ -0,0 +1,24 @@ +/* Duplicate a bounded initial segment of a string, with out-of-memory + checking. + Copyright (C) 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#include <stddef.h> + +/* Return a newly allocated copy of at most N bytes of STRING. + In other words, return a copy of the initial segment of length N of + STRING. */ +extern char *xstrndup (const char *string, size_t n); diff --git a/lib/xstrtol.c b/lib/xstrtol.c new file mode 100644 index 0000000..c4557a0 --- /dev/null +++ b/lib/xstrtol.c @@ -0,0 +1,263 @@ +/* A more useful interface to strtol. + + Copyright (C) 1995, 1996, 1998, 1999, 2000, 2001, 2003, 2004, 2005, 2006 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +/* Written by Jim Meyering. */ + +#ifndef __strtol +# define __strtol strtol +# define __strtol_t long int +# define __xstrtol xstrtol +# define STRTOL_T_MINIMUM LONG_MIN +# define STRTOL_T_MAXIMUM LONG_MAX +#endif + +#include <config.h> + +#include "xstrtol.h" + +/* Some pre-ANSI implementations (e.g. SunOS 4) + need stderr defined if assertion checking is enabled. */ +#include <stdio.h> + +#include <assert.h> +#include <ctype.h> +#include <errno.h> +#include <limits.h> +#include <stdlib.h> +#include <string.h> + +#include "intprops.h" + +static strtol_error +bkm_scale (__strtol_t *x, int scale_factor) +{ + if (TYPE_SIGNED (__strtol_t) && *x < STRTOL_T_MINIMUM / scale_factor) + { + *x = STRTOL_T_MINIMUM; + return LONGINT_OVERFLOW; + } + if (STRTOL_T_MAXIMUM / scale_factor < *x) + { + *x = STRTOL_T_MAXIMUM; + return LONGINT_OVERFLOW; + } + *x *= scale_factor; + return LONGINT_OK; +} + +static strtol_error +bkm_scale_by_power (__strtol_t *x, int base, int power) +{ + strtol_error err = LONGINT_OK; + while (power--) + err |= bkm_scale (x, base); + return err; +} + +/* FIXME: comment. */ + +strtol_error +__xstrtol (const char *s, char **ptr, int strtol_base, + __strtol_t *val, const char *valid_suffixes) +{ + char *t_ptr; + char **p; + __strtol_t tmp; + strtol_error err = LONGINT_OK; + + assert (0 <= strtol_base && strtol_base <= 36); + + p = (ptr ? ptr : &t_ptr); + + if (! TYPE_SIGNED (__strtol_t)) + { + const char *q = s; + unsigned char ch = *q; + while (isspace (ch)) + ch = *++q; + if (ch == '-') + return LONGINT_INVALID; + } + + errno = 0; + tmp = __strtol (s, p, strtol_base); + + if (*p == s) + { + /* If there is no number but there is a valid suffix, assume the + number is 1. The string is invalid otherwise. */ + if (valid_suffixes && **p && strchr (valid_suffixes, **p)) + tmp = 1; + else + return LONGINT_INVALID; + } + else if (errno != 0) + { + if (errno != ERANGE) + return LONGINT_INVALID; + err = LONGINT_OVERFLOW; + } + + /* Let valid_suffixes == NULL mean `allow any suffix'. */ + /* FIXME: update all callers except the ones that allow suffixes + after the number, changing last parameter NULL to `""'. */ + if (!valid_suffixes) + { + *val = tmp; + return err; + } + + if (**p != '\0') + { + int base = 1024; + int suffixes = 1; + strtol_error overflow; + + if (!strchr (valid_suffixes, **p)) + { + *val = tmp; + return err | LONGINT_INVALID_SUFFIX_CHAR; + } + + if (strchr (valid_suffixes, '0')) + { + /* The ``valid suffix'' '0' is a special flag meaning that + an optional second suffix is allowed, which can change + the base. A suffix "B" (e.g. "100MB") stands for a power + of 1000, whereas a suffix "iB" (e.g. "100MiB") stands for + a power of 1024. If no suffix (e.g. "100M"), assume + power-of-1024. */ + + switch (p[0][1]) + { + case 'i': + if (p[0][2] == 'B') + suffixes += 2; + break; + + case 'B': + case 'D': /* 'D' is obsolescent */ + base = 1000; + suffixes++; + break; + } + } + + switch (**p) + { + case 'b': + overflow = bkm_scale (&tmp, 512); + break; + + case 'B': + overflow = bkm_scale (&tmp, 1024); + break; + + case 'c': + overflow = 0; + break; + + case 'E': /* exa or exbi */ + overflow = bkm_scale_by_power (&tmp, base, 6); + break; + + case 'G': /* giga or gibi */ + case 'g': /* 'g' is undocumented; for compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 3); + break; + + case 'k': /* kilo */ + case 'K': /* kibi */ + overflow = bkm_scale_by_power (&tmp, base, 1); + break; + + case 'M': /* mega or mebi */ + case 'm': /* 'm' is undocumented; for compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 2); + break; + + case 'P': /* peta or pebi */ + overflow = bkm_scale_by_power (&tmp, base, 5); + break; + + case 'T': /* tera or tebi */ + case 't': /* 't' is undocumented; for compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 4); + break; + + case 'w': + overflow = bkm_scale (&tmp, 2); + break; + + case 'Y': /* yotta or 2**80 */ + overflow = bkm_scale_by_power (&tmp, base, 8); + break; + + case 'Z': /* zetta or 2**70 */ + overflow = bkm_scale_by_power (&tmp, base, 7); + break; + + default: + *val = tmp; + return err | LONGINT_INVALID_SUFFIX_CHAR; + } + + err |= overflow; + *p += suffixes; + if (**p) + err |= LONGINT_INVALID_SUFFIX_CHAR; + } + + *val = tmp; + return err; +} + +#ifdef TESTING_XSTRTO + +# include <stdio.h> +# include "error.h" + +char *program_name; + +int +main (int argc, char **argv) +{ + strtol_error s_err; + int i; + + program_name = argv[0]; + for (i=1; i<argc; i++) + { + char *p; + __strtol_t val; + + s_err = __xstrtol (argv[i], &p, 0, &val, "bckmw"); + if (s_err == LONGINT_OK) + { + printf ("%s->%lu (%s)\n", argv[i], val, p); + } + else + { + STRTOL_FATAL_ERROR (argv[i], "arg", s_err); + } + } + exit (0); +} + +#endif /* TESTING_XSTRTO */ diff --git a/lib/xstrtol.h b/lib/xstrtol.h new file mode 100644 index 0000000..475728a --- /dev/null +++ b/lib/xstrtol.h @@ -0,0 +1,87 @@ +/* A more useful interface to strtol. + + Copyright (C) 1995, 1996, 1998, 1999, 2001, 2002, 2003, 2004, 2006 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */ + +#ifndef XSTRTOL_H_ +# define XSTRTOL_H_ 1 + +# include "exitfail.h" + +# include <inttypes.h> + +# include "gettext.h" + +# ifndef _STRTOL_ERROR +enum strtol_error + { + LONGINT_OK = 0, + + /* These two values can be ORed together, to indicate that both + errors occurred. */ + LONGINT_OVERFLOW = 1, + LONGINT_INVALID_SUFFIX_CHAR = 2, + + LONGINT_INVALID_SUFFIX_CHAR_WITH_OVERFLOW = (LONGINT_INVALID_SUFFIX_CHAR + | LONGINT_OVERFLOW), + LONGINT_INVALID = 4 + }; +typedef enum strtol_error strtol_error; +# endif + +# define _DECLARE_XSTRTOL(name, type) \ + strtol_error name (const char *, char **, int, type *, const char *); +_DECLARE_XSTRTOL (xstrtol, long int) +_DECLARE_XSTRTOL (xstrtoul, unsigned long int) +_DECLARE_XSTRTOL (xstrtoimax, intmax_t) +_DECLARE_XSTRTOL (xstrtoumax, uintmax_t) + +# define _STRTOL_ERROR(Exit_code, Str, Argument_type_string, Err) \ + do \ + { \ + switch ((Err)) \ + { \ + default: \ + abort (); \ + \ + case LONGINT_INVALID: \ + error ((Exit_code), 0, gettext ("invalid %s `%s'"), \ + (Argument_type_string), (Str)); \ + break; \ + \ + case LONGINT_INVALID_SUFFIX_CHAR: \ + case LONGINT_INVALID_SUFFIX_CHAR | LONGINT_OVERFLOW: \ + error ((Exit_code), 0, \ + gettext ("invalid character following %s in `%s'"), \ + (Argument_type_string), (Str)); \ + break; \ + \ + case LONGINT_OVERFLOW: \ + error ((Exit_code), 0, gettext ("%s `%s' too large"), \ + (Argument_type_string), (Str)); \ + break; \ + } \ + } \ + while (0) + +# define STRTOL_FATAL_ERROR(Str, Argument_type_string, Err) \ + _STRTOL_ERROR (exit_failure, Str, Argument_type_string, Err) + +# define STRTOL_FAIL_WARN(Str, Argument_type_string, Err) \ + _STRTOL_ERROR (0, Str, Argument_type_string, Err) + +#endif /* not XSTRTOL_H_ */ diff --git a/lib/xstrtoul.c b/lib/xstrtoul.c new file mode 100644 index 0000000..285f7b9 --- /dev/null +++ b/lib/xstrtoul.c @@ -0,0 +1,6 @@ +#define __strtol strtoul +#define __strtol_t unsigned long int +#define __xstrtol xstrtoul +#define STRTOL_T_MINIMUM 0 +#define STRTOL_T_MAXIMUM ULONG_MAX +#include "xstrtol.c" diff --git a/lib/xstrtoumax.c b/lib/xstrtoumax.c new file mode 100644 index 0000000..9a2349f --- /dev/null +++ b/lib/xstrtoumax.c @@ -0,0 +1,6 @@ +#define __strtol strtoumax +#define __strtol_t uintmax_t +#define __xstrtol xstrtoumax +#define STRTOL_T_MINIMUM 0 +#define STRTOL_T_MAXIMUM UINTMAX_MAX +#include "xstrtol.c" |