diff options
Diffstat (limited to 'lib')
67 files changed, 23321 insertions, 0 deletions
diff --git a/lib/Makefile.am b/lib/Makefile.am new file mode 100644 index 0000000..6629320 --- /dev/null +++ b/lib/Makefile.am @@ -0,0 +1,16 @@ +## Process this file with automake to produce Makefile.in + +noinst_LIBRARIES = +EXTRA_DIST = +BUILT_SOURCES = +CLEANFILES = +MOSTLYCLEANFILES = +MOSTLYCLEANDIRS = + +# Pacify automake +SUFFIXES = .a .o .c .y .l .h .sh .elc .el + +AM_CPPFLAGS = -I$(top_srcdir)/lib -I$(top_srcdir)/intl -I$(top_srcdir) \ + -I$(top_builddir)/lib -I$(top_builddir)/intl + +include gnulib.mk diff --git a/lib/Makefile.in b/lib/Makefile.in new file mode 100644 index 0000000..ba87d2a --- /dev/null +++ b/lib/Makefile.in @@ -0,0 +1,780 @@ +# Makefile.in generated by automake 1.9.6 from Makefile.am. +# @configure_input@ + +# Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, +# 2003, 2004, 2005 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@ + +# Copyright (C) 2004-2006 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=. --lib=libsed --source-base=lib --m4-base=build-aux --doc-base=doc --aux-dir=build-aux --makefile_name=gnulib.mk --no-libtool --macro-prefix=gl acl alloca getline getopt gettext localcharset memchr memcmp memmove mkstemp obstack pathmax regex stat-macros stdbool strerror strverscmp unlocked-io verify + +srcdir = @srcdir@ +top_srcdir = @top_srcdir@ +VPATH = @srcdir@ +pkgdatadir = $(datadir)/@PACKAGE@ +pkglibdir = $(libdir)/@PACKAGE@ +pkgincludedir = $(includedir)/@PACKAGE@ +top_builddir = .. +am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd +INSTALL = @INSTALL@ +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 = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ + $(srcdir)/gnulib.mk acl.c alloca.c error.c error.h exitfail.c \ + getdelim.c getline.c getopt.c getopt1.c malloc.c mbchar.c \ + memchr.c memcmp.c memmove.c mkstemp.c obstack.c obstack.h \ + quote.c quotearg.c regex.c strcasecmp.c strerror.c \ + strncasecmp.c strverscmp.c tempname.c xmalloc.c +subdir = lib +ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 +am__aclocal_m4_deps = $(top_srcdir)/build-aux/absolute-header.m4 \ + $(top_srcdir)/build-aux/acl.m4 \ + $(top_srcdir)/build-aux/alloca.m4 \ + $(top_srcdir)/build-aux/codeset.m4 \ + $(top_srcdir)/build-aux/error.m4 \ + $(top_srcdir)/build-aux/exitfail.m4 \ + $(top_srcdir)/build-aux/extensions.m4 \ + $(top_srcdir)/build-aux/getdelim.m4 \ + $(top_srcdir)/build-aux/getline.m4 \ + $(top_srcdir)/build-aux/getopt.m4 \ + $(top_srcdir)/build-aux/gettext.m4 \ + $(top_srcdir)/build-aux/glibc2.m4 \ + $(top_srcdir)/build-aux/glibc21.m4 \ + $(top_srcdir)/build-aux/gnulib-comp.m4 \ + $(top_srcdir)/build-aux/iconv.m4 \ + $(top_srcdir)/build-aux/intdiv0.m4 \ + $(top_srcdir)/build-aux/intmax.m4 \ + $(top_srcdir)/build-aux/inttypes-pri.m4 \ + $(top_srcdir)/build-aux/inttypes_h.m4 \ + $(top_srcdir)/build-aux/lcmessage.m4 \ + $(top_srcdir)/build-aux/lib-ld.m4 \ + $(top_srcdir)/build-aux/lib-link.m4 \ + $(top_srcdir)/build-aux/lib-prefix.m4 \ + $(top_srcdir)/build-aux/localcharset.m4 \ + $(top_srcdir)/build-aux/lock.m4 \ + $(top_srcdir)/build-aux/longdouble.m4 \ + $(top_srcdir)/build-aux/longlong.m4 \ + $(top_srcdir)/build-aux/mbchar.m4 \ + $(top_srcdir)/build-aux/mbiter.m4 \ + $(top_srcdir)/build-aux/mbrtowc.m4 \ + $(top_srcdir)/build-aux/mbstate_t.m4 \ + $(top_srcdir)/build-aux/memchr.m4 \ + $(top_srcdir)/build-aux/memcmp.m4 \ + $(top_srcdir)/build-aux/memmove.m4 \ + $(top_srcdir)/build-aux/mkstemp.m4 \ + $(top_srcdir)/build-aux/nls.m4 \ + $(top_srcdir)/build-aux/pathmax.m4 \ + $(top_srcdir)/build-aux/po.m4 \ + $(top_srcdir)/build-aux/printf-posix.m4 \ + $(top_srcdir)/build-aux/progtest.m4 \ + $(top_srcdir)/build-aux/quote.m4 \ + $(top_srcdir)/build-aux/quotearg.m4 \ + $(top_srcdir)/build-aux/regex.m4 \ + $(top_srcdir)/build-aux/signed.m4 \ + $(top_srcdir)/build-aux/size_max.m4 \ + $(top_srcdir)/build-aux/ssize_t.m4 \ + $(top_srcdir)/build-aux/stat-macros.m4 \ + $(top_srcdir)/build-aux/stdbool.m4 \ + $(top_srcdir)/build-aux/stdint.m4 \ + $(top_srcdir)/build-aux/stdint_h.m4 \ + $(top_srcdir)/build-aux/strcase.m4 \ + $(top_srcdir)/build-aux/strerror.m4 \ + $(top_srcdir)/build-aux/strverscmp.m4 \ + $(top_srcdir)/build-aux/sys_stat_h.m4 \ + $(top_srcdir)/build-aux/uintmax_t.m4 \ + $(top_srcdir)/build-aux/ulonglong.m4 \ + $(top_srcdir)/build-aux/unistd_h.m4 \ + $(top_srcdir)/build-aux/unlocked-io.m4 \ + $(top_srcdir)/build-aux/visibility.m4 \ + $(top_srcdir)/build-aux/wchar_t.m4 \ + $(top_srcdir)/build-aux/wcwidth.m4 \ + $(top_srcdir)/build-aux/wint_t.m4 \ + $(top_srcdir)/build-aux/xalloc.m4 \ + $(top_srcdir)/build-aux/xsize.m4 $(top_srcdir)/configure.ac +am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ + $(ACLOCAL_M4) +mkinstalldirs = $(install_sh) -d +CONFIG_HEADER = $(top_builddir)/config.h +CONFIG_CLEAN_FILES = +LIBRARIES = $(noinst_LIBRARIES) +AR = ar +ARFLAGS = cru +libsed_a_AR = $(AR) $(ARFLAGS) +am__DEPENDENCIES_1 = @LIBOBJS@ +libsed_a_DEPENDENCIES = $(am__DEPENDENCIES_1) @ALLOCA@ +am_libsed_a_OBJECTS = localcharset.$(OBJEXT) strnlen1.$(OBJEXT) \ + xalloc-die.$(OBJEXT) +libsed_a_OBJECTS = $(am_libsed_a_OBJECTS) +DEFAULT_INCLUDES = -I. -I$(srcdir) -I$(top_builddir) +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 $@ +SOURCES = $(libsed_a_SOURCES) +DIST_SOURCES = $(libsed_a_SOURCES) +ETAGS = etags +CTAGS = ctags +DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) +ABSOLUTE_STDINT_H = @ABSOLUTE_STDINT_H@ +ABSOLUTE_SYS_STAT_H = @ABSOLUTE_SYS_STAT_H@ +ACLOCAL = @ACLOCAL@ +ALLOCA = @ALLOCA@ +ALLOCA_H = @ALLOCA_H@ +AMDEP_FALSE = @AMDEP_FALSE@ +AMDEP_TRUE = @AMDEP_TRUE@ +AMTAR = @AMTAR@ +AUTOCONF = @AUTOCONF@ +AUTOHEADER = @AUTOHEADER@ +AUTOMAKE = @AUTOMAKE@ +AWK = @AWK@ +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@ +BUILD_HTML_FALSE = @BUILD_HTML_FALSE@ +BUILD_HTML_TRUE = @BUILD_HTML_TRUE@ +BUILD_INCLUDED_LIBINTL = @BUILD_INCLUDED_LIBINTL@ +CATOBJEXT = @CATOBJEXT@ +CC = @CC@ +CCDEPMODE = @CCDEPMODE@ +CFLAGS = @CFLAGS@ +CFLAG_VISIBILITY = @CFLAG_VISIBILITY@ +CPP = @CPP@ +CPPFLAGS = @CPPFLAGS@ +CYGPATH_W = @CYGPATH_W@ +DATADIRNAME = @DATADIRNAME@ +DEFS = @DEFS@ +DEPDIR = @DEPDIR@ +ECHO_C = @ECHO_C@ +ECHO_N = @ECHO_N@ +ECHO_T = @ECHO_T@ +EGREP = @EGREP@ +EXEEXT = @EXEEXT@ +GENCAT = @GENCAT@ +GETOPT_H = @GETOPT_H@ +GLIBC2 = @GLIBC2@ +GLIBC21 = @GLIBC21@ +GL_COND_LIBTOOL_FALSE = @GL_COND_LIBTOOL_FALSE@ +GL_COND_LIBTOOL_TRUE = @GL_COND_LIBTOOL_TRUE@ +GMSGFMT = @GMSGFMT@ +GMSGFMT_015 = @GMSGFMT_015@ +GREP = @GREP@ +HAVE_ASPRINTF = @HAVE_ASPRINTF@ +HAVE_INTTYPES_H = @HAVE_INTTYPES_H@ +HAVE_LONG_LONG_INT = @HAVE_LONG_LONG_INT@ +HAVE_POSIX_PRINTF = @HAVE_POSIX_PRINTF@ +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_SNPRINTF = @HAVE_SNPRINTF@ +HAVE_STDINT_H = @HAVE_STDINT_H@ +HAVE_SYS_BITYPES_H = @HAVE_SYS_BITYPES_H@ +HAVE_SYS_INTTYPES_H = @HAVE_SYS_INTTYPES_H@ +HAVE_SYS_TYPES_H = @HAVE_SYS_TYPES_H@ +HAVE_VISIBILITY = @HAVE_VISIBILITY@ +HAVE_WCHAR_H = @HAVE_WCHAR_H@ +HAVE_WPRINTF = @HAVE_WPRINTF@ +HAVE__BOOL = @HAVE__BOOL@ +INSTALL_DATA = @INSTALL_DATA@ +INSTALL_PROGRAM = @INSTALL_PROGRAM@ +INSTALL_SCRIPT = @INSTALL_SCRIPT@ +INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ +INSTOBJEXT = @INSTOBJEXT@ +INTLBISON = @INTLBISON@ +INTLLIBS = @INTLLIBS@ +INTLOBJS = @INTLOBJS@ +INTL_LIBTOOL_SUFFIX_PREFIX = @INTL_LIBTOOL_SUFFIX_PREFIX@ +INTL_MACOSX_LIBS = @INTL_MACOSX_LIBS@ +LDFLAGS = @LDFLAGS@ +LIBICONV = @LIBICONV@ +LIBINTL = @LIBINTL@ +LIBMULTITHREAD = @LIBMULTITHREAD@ +LIBOBJS = @LIBOBJS@ +LIBPTH = @LIBPTH@ +LIBS = @LIBS@ +LIBSED_LIBDEPS = @LIBSED_LIBDEPS@ +LIBSED_LTLIBDEPS = @LIBSED_LTLIBDEPS@ +LIBTHREAD = @LIBTHREAD@ +LIB_ACL = @LIB_ACL@ +LTLIBICONV = @LTLIBICONV@ +LTLIBINTL = @LTLIBINTL@ +LTLIBMULTITHREAD = @LTLIBMULTITHREAD@ +LTLIBOBJS = @LTLIBOBJS@ +LTLIBPTH = @LTLIBPTH@ +LTLIBTHREAD = @LTLIBTHREAD@ +MAKEINFO = @MAKEINFO@ +MAKEINFO_HTML_FALSE = @MAKEINFO_HTML_FALSE@ +MAKEINFO_HTML_TRUE = @MAKEINFO_HTML_TRUE@ +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@ +PRI_MACROS_BROKEN = @PRI_MACROS_BROKEN@ +PTRDIFF_T_SUFFIX = @PTRDIFF_T_SUFFIX@ +RANLIB = @RANLIB@ +SED_FEATURE_VERSION = @SED_FEATURE_VERSION@ +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@ +SYS_STAT_H = @SYS_STAT_H@ +TEST_REGEX_FALSE = @TEST_REGEX_FALSE@ +TEST_REGEX_TRUE = @TEST_REGEX_TRUE@ +TEXI2HTML = @TEXI2HTML@ +TEXI2HTML_HTML_FALSE = @TEXI2HTML_HTML_FALSE@ +TEXI2HTML_HTML_TRUE = @TEXI2HTML_HTML_TRUE@ +UNISTD_H = @UNISTD_H@ +USE_INCLUDED_LIBINTL = @USE_INCLUDED_LIBINTL@ +USE_NLS = @USE_NLS@ +VERSION = @VERSION@ +WCHAR_T_SUFFIX = @WCHAR_T_SUFFIX@ +WINT_T_SUFFIX = @WINT_T_SUFFIX@ +WOE32DLL = @WOE32DLL@ +XGETTEXT = @XGETTEXT@ +XGETTEXT_015 = @XGETTEXT_015@ +ac_ct_CC = @ac_ct_CC@ +am__fastdepCC_FALSE = @am__fastdepCC_FALSE@ +am__fastdepCC_TRUE = @am__fastdepCC_TRUE@ +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@ +datadir = @datadir@ +datarootdir = @datarootdir@ +docdir = @docdir@ +dvidir = @dvidir@ +exec_prefix = @exec_prefix@ +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@ +sysconfdir = @sysconfdir@ +target_alias = @target_alias@ +noinst_LIBRARIES = libsed.a +EXTRA_DIST = acl.h alloca_.h exitfail.h getdelim.h getline.h getopt_.h \ + getopt_int.h config.charset ref-add.sin ref-del.sin mkstemp.h \ + pathmax.h quote.h quotearg.h regcomp.c regex.h \ + regex_internal.c regex_internal.h regexec.c stat-macros.h \ + stdbool_.h stdint_.h strverscmp.h stat_.h unlocked-io.h \ + xalloc.h +BUILT_SOURCES = $(ALLOCA_H) configmake.h $(GETOPT_H) $(STDBOOL_H) \ + $(STDINT_H) $(SYS_STAT_H) $(UNISTD_H) +CLEANFILES = configmake.h configmake.h-t charset.alias ref-add.sed \ + ref-del.sed +MOSTLYCLEANFILES = core *.stackdump alloca.h alloca.h-t getopt.h \ + getopt.h-t stdbool.h stdbool.h-t stdint.h stdint.h-t \ + sys/stat.h sys/stat.h-t unistd.h +MOSTLYCLEANDIRS = sys + +# Pacify automake +SUFFIXES = .a .o .c .y .l .h .sh .elc .el .sed .sin + +# 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_srcdir)/lib -I$(top_srcdir)/intl -I$(top_srcdir) \ + -I$(top_builddir)/lib -I$(top_builddir)/intl \ + -I$(top_builddir)/intl +libsed_a_SOURCES = exit.h gettext.h localcharset.h localcharset.c \ + mbchar.h mbuiter.h strcase.h strnlen1.h strnlen1.c verify.h \ + wcwidth.h xalloc-die.c +libsed_a_LIBADD = $(LIBOBJS) @ALLOCA@ +charset_alias = $(DESTDIR)$(libdir)/charset.alias +charset_tmp = $(DESTDIR)$(libdir)/charset.tmp +all: $(BUILT_SOURCES) + $(MAKE) $(AM_MAKEFLAGS) all-am + +.SUFFIXES: +.SUFFIXES: .a .o .c .y .l .h .sh .elc .el .sed .sin .obj +$(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) +libsed.a: $(libsed_a_OBJECTS) $(libsed_a_DEPENDENCIES) + -rm -f libsed.a + $(libsed_a_AR) libsed.a $(libsed_a_OBJECTS) $(libsed_a_LIBADD) + $(RANLIB) libsed.a + +mostlyclean-compile: + -rm -f *.$(OBJEXT) + +distclean-compile: + -rm -f *.tab.c + +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/acl.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/alloca.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/error.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/exitfail.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@$(DEPDIR)/malloc.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/mbchar.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/memchr.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/memcmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/memmove.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/mkstemp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/obstack.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/quote.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/quotearg.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/regex.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/strcasecmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/strerror.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/strncasecmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/strverscmp.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/tempname.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@$(DEPDIR)/xmalloc.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/localcharset.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/strnlen1.Po@am__quote@ +@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/xalloc-die.Po@am__quote@ + +.c.o: +@am__fastdepCC_TRUE@ if $(COMPILE) -MT $@ -MD -MP -MF "$(DEPDIR)/$*.Tpo" -c -o $@ $<; \ +@am__fastdepCC_TRUE@ then mv -f "$(DEPDIR)/$*.Tpo" "$(DEPDIR)/$*.Po"; else rm -f "$(DEPDIR)/$*.Tpo"; exit 1; fi +@AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ +@AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ +@am__fastdepCC_FALSE@ $(COMPILE) -c $< + +.c.obj: +@am__fastdepCC_TRUE@ if $(COMPILE) -MT $@ -MD -MP -MF "$(DEPDIR)/$*.Tpo" -c -o $@ `$(CYGPATH_W) '$<'`; \ +@am__fastdepCC_TRUE@ then mv -f "$(DEPDIR)/$*.Tpo" "$(DEPDIR)/$*.Po"; else rm -f "$(DEPDIR)/$*.Tpo"; exit 1; fi +@AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ +@AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ +@am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` +uninstall-info-am: + +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) $$here + +distclean-tags: + -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags + +distdir: $(DISTFILES) + @srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; \ + topsrcdirstrip=`echo "$(top_srcdir)" | sed 's|.|.|g'`; \ + list='$(DISTFILES)'; for file in $$list; do \ + case $$file in \ + $(srcdir)/*) file=`echo "$$file" | sed "s|^$$srcdirstrip/||"`;; \ + $(top_srcdir)/*) file=`echo "$$file" | sed "s|^$$topsrcdirstrip/|$(top_builddir)/|"`;; \ + esac; \ + if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ + dir=`echo "$$file" | sed -e 's,/[^/]*$$,,'`; \ + if test "$$dir" != "$$file" && test "$$dir" != "."; then \ + dir="/$$dir"; \ + $(mkdir_p) "$(distdir)$$dir"; \ + else \ + dir=''; \ + fi; \ + if test -d $$d/$$file; then \ + 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) 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) + +maintainer-clean-generic: + @echo "This command is intended for maintainers to use" + @echo "it deletes files that may require special tools to rebuild." + -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) +clean: clean-am + +clean-am: clean-generic clean-noinstLIBRARIES 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-exec-am: install-exec-local + +install-info: install-info-am + +install-man: + +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-info-am uninstall-local + +.PHONY: CTAGS GTAGS all all-am all-local check check-am clean \ + clean-generic clean-noinstLIBRARIES 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-exec install-exec-am \ + install-exec-local install-info install-info-am install-man \ + 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-info-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 + cp -f $(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 documentation. +configmake.h: Makefile + rm -f $@-t $@ + ( \ + 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 PKGLIBDIR "$(pkglibdir)"'; \ + echo '#define PKGINCLUDEDIR "$(pkgincludedir)"'; \ + echo '#define PKGDATADIR "$(pkgdatadir)"'; \ + :) | sed '/""/d' >$@-t + mv $@-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 + cp -f $(srcdir)/getopt_.h $@-t + mv -f $@-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 $@ + sed -e 's/@''HAVE__BOOL''@/$(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 $@ + sed -e 's/@''HAVE_WCHAR_H''@/$(HAVE_WCHAR_H)/g' \ + -e 's/@''HAVE_STDINT_H''@/$(HAVE_STDINT_H)/g' \ + -e 's|@''ABSOLUTE_STDINT_H''@|$(ABSOLUTE_STDINT_H)|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''@/$(HAVE_LONG_LONG_INT)/g' \ + -e 's/@''BITSIZEOF_PTRDIFF_T''@/$(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 <sys/stat.h> when the system +# has one that is incomplete. +sys/stat.h: stat_.h + test -d sys || mkdir sys + rm -f $@-t $@ + sed -e 's|@''ABSOLUTE_SYS_STAT_H''@|$(ABSOLUTE_SYS_STAT_H)|g' \ + < $(srcdir)/stat_.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: + echo '/* Empty placeholder for $@. */' >$@ + +mostlyclean-local: mostlyclean-generic + @test -z "$(MOSTLYCLEANDIRS)" || \ + for dir in $(MOSTLYCLEANDIRS); do \ + if test -d $$dir; then \ + echo "rmdir $$dir"; rmdir $$dir; \ + fi; \ + done +# 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/acl.c b/lib/acl.c new file mode 100644 index 0000000..803be0d --- /dev/null +++ b/lib/acl.c @@ -0,0 +1,407 @@ +/* acl.c - access control lists + + Copyright (C) 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; if not, write to the Free Software Foundation, + Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + + Written by Paul Eggert and Andreas Gruenbacher. */ + +#include <config.h> + +#include <stdbool.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#ifndef S_ISLNK +# define S_ISLNK(Mode) 0 +#endif + +#ifdef HAVE_ACL_LIBACL_H +# include <acl/libacl.h> +#endif + +#include "acl.h" +#include "error.h" +#include "quote.h" + +#include <errno.h> +#ifndef ENOSYS +# define ENOSYS (-1) +#endif +#ifndef ENOTSUP +# define ENOTSUP (-1) +#endif + +#if ENABLE_NLS +# include <libintl.h> +# define _(Text) gettext (Text) +#else +# define _(Text) Text +#endif + +#ifndef HAVE_FCHMOD +# define HAVE_FCHMOD false +# define fchmod(fd, mode) (-1) +#endif + +/* POSIX 1003.1e (draft 17) */ +#ifndef HAVE_ACL_GET_FD +# define HAVE_ACL_GET_FD false +# define acl_get_fd(fd) (NULL) +#endif + +/* POSIX 1003.1e (draft 17) */ +#ifndef HAVE_ACL_SET_FD +# define HAVE_ACL_SET_FD false +# define acl_set_fd(fd, acl) (-1) +#endif + +/* Linux-specific */ +#ifndef HAVE_ACL_EXTENDED_FILE +# define HAVE_ACL_EXTENDED_FILE false +# define acl_extended_file(name) (-1) +#endif + +/* Linux-specific */ +#ifndef HAVE_ACL_FROM_MODE +# define HAVE_ACL_FROM_MODE false +# define acl_from_mode(mode) (NULL) +#endif + +/* We detect the presence of POSIX 1003.1e (draft 17 -- abandoned) support + by checking for HAVE_ACL_GET_FILE, HAVE_ACL_SET_FILE, and HAVE_ACL_FREE. + Systems that have acl_get_file, acl_set_file, and acl_free must also + have acl_to_text, acl_from_text, and acl_delete_def_file (all defined + in the draft); systems that don't would hit #error statements here. */ + +#if USE_ACL && HAVE_ACL_GET_FILE && !HAVE_ACL_ENTRIES +# ifndef HAVE_ACL_TO_TEXT +# error Must have acl_to_text (see POSIX 1003.1e draft 17). +# endif + +/* Return the number of entries in ACL. Linux implements acl_entries + as a more efficient extension than using this workaround. */ + +static int +acl_entries (acl_t acl) +{ + char *text = acl_to_text (acl, NULL), *t; + int entries; + if (text == NULL) + return -1; + for (entries = 0, t = text; ; t++, entries++) { + t = strchr (t, '\n'); + if (t == NULL) + break; + } + acl_free (text); + return entries; +} +#endif + +/* If DESC is a valid file descriptor use fchmod to change the + file's mode to MODE on systems that have fchown. On systems + that don't have fchown and if DESC is invalid, use chown on + NAME instead. */ + +int +chmod_or_fchmod (const char *name, int desc, mode_t mode) +{ + if (HAVE_FCHMOD && desc != -1) + return fchmod (desc, mode); + else + return chmod (name, mode); +} + +/* Return 1 if NAME has a nontrivial access control list, 0 if + NAME only has no or a base access control list, and -1 on + error. SB must be set to the stat buffer of FILE. */ + +int +file_has_acl (char const *name, struct stat const *sb) +{ +#if USE_ACL && HAVE_ACL && defined GETACLCNT + /* This implementation should work on recent-enough versions of HP-UX, + Solaris, and Unixware. */ + +# ifndef MIN_ACL_ENTRIES +# define MIN_ACL_ENTRIES 4 +# endif + + if (! S_ISLNK (sb->st_mode)) + { + int n = acl (name, GETACLCNT, 0, NULL); + return n < 0 ? (errno == ENOSYS ? 0 : -1) : (MIN_ACL_ENTRIES < n); + } +#elif USE_ACL && HAVE_ACL_GET_FILE && HAVE_ACL_FREE + /* POSIX 1003.1e (draft 17 -- abandoned) specific version. */ + + if (! S_ISLNK (sb->st_mode)) + { + int ret; + + if (HAVE_ACL_EXTENDED_FILE) + ret = acl_extended_file (name); + else + { + acl_t acl = acl_get_file (name, ACL_TYPE_ACCESS); + if (acl) + { + ret = (3 < acl_entries (acl)); + acl_free (acl); + if (ret == 0 && S_ISDIR (sb->st_mode)) + { + acl = acl_get_file (name, ACL_TYPE_DEFAULT); + if (acl) + { + ret = (0 < acl_entries (acl)); + acl_free (acl); + } + else + ret = -1; + } + } + else + ret = -1; + } + if (ret < 0) + return (errno == ENOSYS || errno == ENOTSUP) ? 0 : -1; + return ret; + } +#endif + + /* FIXME: Add support for AIX, Irix, and Tru64. Please see Samba's + source/lib/sysacls.c file for fix-related ideas. */ + + return 0; +} + +/* Copy access control lists from one file to another. If SOURCE_DESC is + a valid file descriptor, use file descriptor operations, else use + filename based operations on SRC_NAME. Likewise for DEST_DESC and + DEST_NAME. + If access control lists are not available, fchmod the target file to + MODE. Also sets the non-permission bits of the destination file + (S_ISUID, S_ISGID, S_ISVTX) to those from MODE if any are set. + System call return value semantics. */ + +int +copy_acl (const char *src_name, int source_desc, const char *dst_name, + int dest_desc, mode_t mode) +{ + int ret; + +#if USE_ACL && HAVE_ACL_GET_FILE && HAVE_ACL_SET_FILE && HAVE_ACL_FREE + /* POSIX 1003.1e (draft 17 -- abandoned) specific version. */ + + acl_t acl; + if (HAVE_ACL_GET_FD && source_desc != -1) + acl = acl_get_fd (source_desc); + else + acl = acl_get_file (src_name, ACL_TYPE_ACCESS); + if (acl == NULL) + { + if (errno == ENOSYS || errno == ENOTSUP) + return set_acl (dst_name, dest_desc, mode); + else + { + error (0, errno, "%s", quote (src_name)); + return -1; + } + } + + if (HAVE_ACL_SET_FD && dest_desc != -1) + ret = acl_set_fd (dest_desc, acl); + else + ret = acl_set_file (dst_name, ACL_TYPE_ACCESS, acl); + if (ret != 0) + { + int saved_errno = errno; + + if (errno == ENOSYS || errno == ENOTSUP) + { + int n = acl_entries (acl); + + acl_free (acl); + if (n == 3) + { + if (chmod_or_fchmod (dst_name, dest_desc, mode) != 0) + saved_errno = errno; + else + return 0; + } + else + chmod_or_fchmod (dst_name, dest_desc, mode); + } + else + { + acl_free (acl); + chmod_or_fchmod (dst_name, dest_desc, mode); + } + error (0, saved_errno, _("preserving permissions for %s"), + quote (dst_name)); + return -1; + } + else + acl_free (acl); + + if (mode & (S_ISUID | S_ISGID | S_ISVTX)) + { + /* We did not call chmod so far, so the special bits have not yet + been set. */ + + if (chmod_or_fchmod (dst_name, dest_desc, mode) != 0) + { + error (0, errno, _("preserving permissions for %s"), + quote (dst_name)); + return -1; + } + } + + if (S_ISDIR (mode)) + { + acl = acl_get_file (src_name, ACL_TYPE_DEFAULT); + if (acl == NULL) + { + error (0, errno, "%s", quote (src_name)); + return -1; + } + + if (acl_set_file (dst_name, ACL_TYPE_DEFAULT, acl)) + { + error (0, errno, _("preserving permissions for %s"), + quote (dst_name)); + acl_free (acl); + return -1; + } + else + acl_free (acl); + } + return 0; +#else + ret = chmod_or_fchmod (dst_name, dest_desc, mode); + if (ret != 0) + error (0, errno, _("preserving permissions for %s"), quote (dst_name)); + return ret; +#endif +} + +/* Set the access control lists of a file. If DESC is a valid file + descriptor, use file descriptor operations where available, else use + filename based operations on NAME. If access control lists are not + available, fchmod the target file to MODE. Also sets the + non-permission bits of the destination file (S_ISUID, S_ISGID, S_ISVTX) + to those from MODE if any are set. System call return value + semantics. */ + +int +set_acl (char const *name, int desc, mode_t mode) +{ +#if USE_ACL && HAVE_ACL_SET_FILE && HAVE_ACL_FREE + /* POSIX 1003.1e draft 17 (abandoned) specific version. */ + + /* We must also have have_acl_from_text and acl_delete_def_file. + (acl_delete_def_file could be emulated with acl_init followed + by acl_set_file, but acl_set_file with an empty acl is + unspecified.) */ + +# ifndef HAVE_ACL_FROM_TEXT +# error Must have acl_from_text (see POSIX 1003.1e draft 17). +# endif +# ifndef HAVE_ACL_DELETE_DEF_FILE +# error Must have acl_delete_def_file (see POSIX 1003.1e draft 17). +# endif + + acl_t acl; + int ret; + + if (HAVE_ACL_FROM_MODE) + { + acl = acl_from_mode (mode); + if (!acl) + { + error (0, errno, "%s", quote (name)); + return -1; + } + } + else + { + char acl_text[] = "u::---,g::---,o::---"; + + if (mode & S_IRUSR) acl_text[ 3] = 'r'; + if (mode & S_IWUSR) acl_text[ 4] = 'w'; + if (mode & S_IXUSR) acl_text[ 5] = 'x'; + if (mode & S_IRGRP) acl_text[10] = 'r'; + if (mode & S_IWGRP) acl_text[11] = 'w'; + if (mode & S_IXGRP) acl_text[12] = 'x'; + if (mode & S_IROTH) acl_text[17] = 'r'; + if (mode & S_IWOTH) acl_text[18] = 'w'; + if (mode & S_IXOTH) acl_text[19] = 'x'; + + acl = acl_from_text (acl_text); + if (!acl) + { + error (0, errno, "%s", quote (name)); + return -1; + } + } + if (HAVE_ACL_SET_FD && desc != -1) + ret = acl_set_fd (desc, acl); + else + ret = acl_set_file (name, ACL_TYPE_ACCESS, acl); + if (ret != 0) + { + int saved_errno = errno; + acl_free (acl); + + if (errno == ENOTSUP || errno == ENOSYS) + { + if (chmod_or_fchmod (name, desc, mode) != 0) + saved_errno = errno; + else + return 0; + } + error (0, saved_errno, _("setting permissions for %s"), quote (name)); + return -1; + } + else + acl_free (acl); + + if (S_ISDIR (mode) && acl_delete_def_file (name)) + { + error (0, errno, _("setting permissions for %s"), quote (name)); + return -1; + } + + if (mode & (S_ISUID | S_ISGID | S_ISVTX)) + { + /* We did not call chmod so far, so the special bits have not yet + been set. */ + + if (chmod_or_fchmod (name, desc, mode)) + { + error (0, errno, _("preserving permissions for %s"), quote (name)); + return -1; + } + } + return 0; +#else + int ret = chmod_or_fchmod (name, desc, mode); + if (ret) + error (0, errno, _("setting permissions for %s"), quote (name)); + return ret; +#endif +} diff --git a/lib/acl.h b/lib/acl.h new file mode 100644 index 0000000..a7b4f9e --- /dev/null +++ b/lib/acl.h @@ -0,0 +1,31 @@ +/* acl.c - access control lists + + 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. + + Written by Paul Eggert. */ + +#if HAVE_SYS_ACL_H +# include <sys/acl.h> +#endif +#if defined HAVE_ACL && ! defined GETACLCNT && defined ACL_CNT +# define GETACLCNT ACL_CNT +#endif + +int file_has_acl (char const *, struct stat const *); +int copy_acl (char const *, int, char const *, int, mode_t); +int set_acl (char const *, int, mode_t); +int chmod_or_fchmod (char const *, int, mode_t); 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..dd0b3e9 --- /dev/null +++ b/lib/alloca_.h @@ -0,0 +1,54 @@ +/* Memory allocation on the stack. + + Copyright (C) 1995, 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. */ + +/* Avoid using the symbol _ALLOCA_H here, as Bison assumes _ALLOCA_H + means there is a real alloca function. */ +#ifndef _GNULIB_ALLOCA_H +# define _GNULIB_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 /* _GNULIB_ALLOCA_H */ 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/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/exit.h b/lib/exit.h new file mode 100644 index 0000000..e8f7738 --- /dev/null +++ b/lib/exit.h @@ -0,0 +1,32 @@ +/* exit() function. + Copyright (C) 1995, 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. */ + +#ifndef _EXIT_H +#define _EXIT_H + +/* Get exit() declaration. */ +#include <stdlib.h> + +/* Some systems do not define EXIT_*, despite otherwise supporting C89. */ +#ifndef EXIT_SUCCESS +# define EXIT_SUCCESS 0 +#endif +#ifndef EXIT_FAILURE +# define EXIT_FAILURE 1 +#endif + +#endif /* _EXIT_H */ diff --git a/lib/exitfail.c b/lib/exitfail.c new file mode 100644 index 0000000..97abc67 --- /dev/null +++ b/lib/exitfail.c @@ -0,0 +1,25 @@ +/* Failure exit status + + Copyright (C) 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. */ + +#include <config.h> + +#include "exitfail.h" +#include "exit.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/getdelim.c b/lib/getdelim.c new file mode 100644 index 0000000..a131b0c --- /dev/null +++ b/lib/getdelim.c @@ -0,0 +1,128 @@ +/* getdelim.c --- Implementation of replacement getdelim function. + Copyright (C) 1994, 1996, 1997, 1998, 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. */ + +/* Ported from glibc by Simon Josefsson. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#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..ab054fb --- /dev/null +++ b/lib/getline.c @@ -0,0 +1,32 @@ +/* getline.c --- Implementation of 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. */ + +#ifdef HAVE_CONFIG_H +# include <config.h> +#endif + +#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..3c406e5 --- /dev/null +++ b/lib/getopt_.h @@ -0,0 +1,225 @@ +/* Declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004,2005 + 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 '\0'. This behavior is specific to the GNU + `getopt'. */ + +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/gettext.h b/lib/gettext.h new file mode 100644 index 0000000..deeaf79 --- /dev/null +++ b/lib/gettext.h @@ -0,0 +1,242 @@ +/* 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> + +#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) ((const char *) (Msgid)) +# define dcgettext(Domainname, Msgid, Category) ((const char *) (Msgid)) +# define ngettext(Msgid1, Msgid2, N) \ + ((N) == 1 ? (const char *) (Msgid1) : (const char *) (Msgid2)) +# define dngettext(Domainname, Msgid1, Msgid2, N) \ + ((N) == 1 ? (const char *) (Msgid1) : (const char *) (Msgid2)) +# define dcngettext(Domainname, Msgid1, Msgid2, N, Category) \ + ((N) == 1 ? (const char *) (Msgid1) : (const char *) (Msgid2)) +# define textdomain(Domainname) ((const char *) (Domainname)) +# define bindtextdomain(Domainname, Dirname) ((const char *) (Dirname)) +# define bind_textdomain_codeset(Domainname, Codeset) ((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'. */ +#define pgettext(Msgctxt, Msgid) \ + pgettext_aux (NULL, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, LC_MESSAGES) +#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) +#define npgettext(Msgctxt, Msgid, MsgidPlural, N) \ + npgettext_aux (NULL, Msgctxt GETTEXT_CONTEXT_GLUE Msgid, Msgid, MsgidPlural, N, LC_MESSAGES) +#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 || defined __cplusplus) + +#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/gnulib.mk b/lib/gnulib.mk new file mode 100644 index 0000000..c10e764 --- /dev/null +++ b/lib/gnulib.mk @@ -0,0 +1,400 @@ +## Process this file with automake to produce Makefile.in. +# Copyright (C) 2004-2006 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=. --lib=libsed --source-base=lib --m4-base=build-aux --doc-base=doc --aux-dir=build-aux --makefile_name=gnulib.mk --no-libtool --macro-prefix=gl acl alloca getline getopt gettext localcharset memchr memcmp memmove mkstemp obstack pathmax regex stat-macros stdbool strerror strverscmp unlocked-io verify + + +noinst_LIBRARIES += libsed.a + +libsed_a_SOURCES = +libsed_a_LIBADD = $(LIBOBJS) +MOSTLYCLEANFILES += core *.stackdump + +## begin gnulib module acl + +EXTRA_DIST += acl.h + +## end gnulib module acl + +## begin gnulib module alloca + + +libsed_a_LIBADD += @ALLOCA@ +## end gnulib module alloca + +## begin gnulib module alloca-opt + +BUILT_SOURCES += $(ALLOCA_H) +EXTRA_DIST += 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 + cp -f $(srcdir)/alloca_.h $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += alloca.h alloca.h-t + +## end gnulib module alloca-opt + +## begin gnulib module configmake + +# Retrieve values of the variables through 'configure' followed by +# 'make', not directly through 'configure', so that a user who +# sets some of these variables consistently on the 'make' command +# line gets correct results. +# +# One advantage of this approach, compared to the classical +# approach of adding -DLIBDIR=\"$(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 documentation. +configmake.h: Makefile + rm -f $@-t $@ + ( \ + echo '#define PREFIX "$(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 PKGLIBDIR "$(pkglibdir)"'; \ + echo '#define PKGINCLUDEDIR "$(pkgincludedir)"'; \ + echo '#define PKGDATADIR "$(pkgdatadir)"'; \ + :) | sed '/""/d' >$@-t + mv $@-t $@ +BUILT_SOURCES += configmake.h +CLEANFILES += configmake.h configmake.h-t + +## end gnulib module configmake + +## begin gnulib module exit + +libsed_a_SOURCES += exit.h + +## end gnulib module exit + +## begin gnulib module exitfail + +EXTRA_DIST += exitfail.h + +## end gnulib module exitfail + +## begin gnulib module getdelim + +EXTRA_DIST += getdelim.h + +## end gnulib module getdelim + +## begin gnulib module getline + +EXTRA_DIST += getline.h + +## end gnulib module getline + +## begin gnulib module getopt + +BUILT_SOURCES += $(GETOPT_H) +EXTRA_DIST += getopt_.h getopt_int.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 + cp -f $(srcdir)/getopt_.h $@-t + mv -f $@-t $@ +MOSTLYCLEANFILES += getopt.h getopt.h-t + +## end gnulib module getopt + +## begin gnulib module gettext + +# This is for those projects which use "gettextize --intl" to put a source-code +# copy of libintl into their package. In such projects, every Makefile.am needs +# -I$(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 + +libsed_a_SOURCES += gettext.h + +## end gnulib module gettext-h + +## begin gnulib module localcharset + +libsed_a_SOURCES += localcharset.h localcharset.c +EXTRA_DIST += config.charset ref-add.sin ref-del.sin + +# 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 + +## end gnulib module localcharset + +## begin gnulib module mbchar + +libsed_a_SOURCES += mbchar.h + +## end gnulib module mbchar + +## begin gnulib module mbuiter + +libsed_a_SOURCES += mbuiter.h + +## end gnulib module mbuiter + +## begin gnulib module mkstemp + +EXTRA_DIST += mkstemp.h + +## end gnulib module mkstemp + +## begin gnulib module pathmax + +EXTRA_DIST += pathmax.h + +## end gnulib module pathmax + +## begin gnulib module quote + +EXTRA_DIST += quote.h + +## end gnulib module quote + +## begin gnulib module quotearg + +EXTRA_DIST += quotearg.h + +## end gnulib module quotearg + +## begin gnulib module regex + +EXTRA_DIST += regcomp.c regex.h regex_internal.c regex_internal.h regexec.c + +## end gnulib module regex + +## begin gnulib module stat-macros + +EXTRA_DIST += stat-macros.h + +## end gnulib module stat-macros + +## begin gnulib module stdbool + +BUILT_SOURCES += $(STDBOOL_H) +EXTRA_DIST += 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 $@ + sed -e 's/@''HAVE__BOOL''@/$(HAVE__BOOL)/g' < $(srcdir)/stdbool_.h > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdbool.h stdbool.h-t + +## end gnulib module stdbool + +## begin gnulib module stdint + +BUILT_SOURCES += $(STDINT_H) +EXTRA_DIST += 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 $@ + sed -e 's/@''HAVE_WCHAR_H''@/$(HAVE_WCHAR_H)/g' \ + -e 's/@''HAVE_STDINT_H''@/$(HAVE_STDINT_H)/g' \ + -e 's|@''ABSOLUTE_STDINT_H''@|$(ABSOLUTE_STDINT_H)|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/@''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 + +## end gnulib module stdint + +## begin gnulib module strcase + +libsed_a_SOURCES += strcase.h + +## end gnulib module strcase + +## begin gnulib module strnlen1 + +libsed_a_SOURCES += strnlen1.h strnlen1.c + +## end gnulib module strnlen1 + +## begin gnulib module strverscmp + +EXTRA_DIST += strverscmp.h + +## end gnulib module strverscmp + +## begin gnulib module sys_stat + +BUILT_SOURCES += $(SYS_STAT_H) +EXTRA_DIST += stat_.h + +# We need the following in order to create <sys/stat.h> when the system +# has one that is incomplete. +sys/stat.h: stat_.h + test -d sys || mkdir sys + rm -f $@-t $@ + sed -e 's|@''ABSOLUTE_SYS_STAT_H''@|$(ABSOLUTE_SYS_STAT_H)|g' \ + < $(srcdir)/stat_.h > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += sys/stat.h sys/stat.h-t +MOSTLYCLEANDIRS += sys + +## end gnulib module sys_stat + +## 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: + echo '/* Empty placeholder for $@. */' >$@ +MOSTLYCLEANFILES += unistd.h + +## end gnulib module unistd + +## begin gnulib module unlocked-io + +EXTRA_DIST += unlocked-io.h + +## end gnulib module unlocked-io + +## begin gnulib module verify + +libsed_a_SOURCES += verify.h + +## end gnulib module verify + +## begin gnulib module wcwidth + +libsed_a_SOURCES += wcwidth.h + +## end gnulib module wcwidth + +## begin gnulib module xalloc + +EXTRA_DIST += xalloc.h + +## end gnulib module xalloc + +## begin gnulib module xalloc-die + +libsed_a_SOURCES += xalloc-die.c + +## end gnulib module xalloc-die + + +mostlyclean-local: mostlyclean-generic + @test -z "$(MOSTLYCLEANDIRS)" || \ + for dir in $(MOSTLYCLEANDIRS); do \ + if test -d $$dir; then \ + echo "rmdir $$dir"; rmdir $$dir; \ + fi; \ + done diff --git a/lib/localcharset.c b/lib/localcharset.c new file mode 100644 index 0000000..a2fc5dc --- /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 HAVE_SETLOCALE +# 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 HAVE_SETLOCALE && 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/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..590109b --- /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". */ +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..bf31027 --- /dev/null +++ b/lib/mbchar.h @@ -0,0 +1,478 @@ +/* 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> + +/* BeOS 5 has the functions but no <wctype.h>. */ +#if HAVE_WCTYPE_H +# include <wctype.h> +#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 !defined iswalnum && !HAVE_ISWCNTRL +static inline int +iswalnum (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? (wc >= '0' && wc <= '9') || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z') + : 0); +} +# define iswalnum iswalnum +#endif +#if !defined iswalpha && !HAVE_ISWCNTRL +static inline int +iswalpha (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? (wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z' + : 0); +} +# define iswalpha iswalpha +#endif +#if !defined iswblank && !HAVE_ISWCNTRL +static inline int +iswblank (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc == ' ' || wc == '\t' + : 0); +} +# define iswblank iswblank +#endif +#if !defined iswcntrl && !HAVE_ISWCNTRL +static inline int +iswcntrl (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? (wc & ~0x1f) == 0 || wc == 0x7f + : 0); +} +# define iswcntrl iswcntrl +#endif +#if !defined iswdigit && !HAVE_ISWCNTRL +static inline int +iswdigit (wint_t wc) +{ + return (wc >= '0' && wc <= '9'); +} +# define iswdigit iswdigit +#endif +#if !defined iswgraph && !HAVE_ISWCNTRL +static inline int +iswgraph (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc >= '!' && wc <= '~' + : 1); +} +# define iswgraph iswgraph +#endif +#if !defined iswlower && !HAVE_ISWCNTRL +static inline int +iswlower (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc >= 'a' && wc <= 'z' + : 0); +} +# define iswlower iswlower +#endif +#if !defined iswprint && !HAVE_ISWCNTRL +static inline int +iswprint (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc >= ' ' && wc <= '~' + : 1); +} +# define iswprint iswprint +#endif +#if !defined iswpunct && !HAVE_ISWCNTRL +static inline int +iswpunct (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc >= '!' && wc <= '~' + && !((wc >= '0' && wc <= '9') + || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'Z')) + : 1); +} +# define iswpunct iswpunct +#endif +#if !defined iswspace && !HAVE_ISWCNTRL +static inline int +iswspace (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc == ' ' || wc == '\t' + || wc == '\n' || wc == '\v' || wc == '\f' || wc == '\r' + : 0); +} +# define iswspace iswspace +#endif +#if !defined iswupper && !HAVE_ISWCNTRL +static inline int +iswupper (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc >= 'A' && wc <= 'Z' + : 0); +} +# define iswupper iswupper +#endif +#if !defined iswxdigit && !HAVE_ISWCNTRL +static inline int +iswxdigit (wint_t wc) +{ + return (wc >= '0' && wc <= '9') || ((wc & ~0x20) >= 'A' && (wc & ~0x20) <= 'F'); +} +# define iswxdigit iswxdigit +#endif + +#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, const mbchar_t *old) +{ + if (old->ptr == &old->buf[0]) + { + memcpy (&new->buf[0], &old->buf[0], old->bytes); + new->ptr = &new->buf[0]; + } + else + new->ptr = old->ptr; + new->bytes = old->bytes; + if ((new->wc_valid = old->wc_valid)) + new->wc = old->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 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/mbuiter.h b/lib/mbuiter.h new file mode 100644 index 0000000..9da3a6c --- /dev/null +++ b/lib/mbuiter.h @@ -0,0 +1,203 @@ +/* Iterating through multibyte strings: macros for multi-byte encodings. + Copyright (C) 2001, 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., 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. + + 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); + */ + +#ifndef _MBUITER_H +#define _MBUITER_H 1 + +#include <assert.h> +#include <stdbool.h> +#include <stdlib.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; +} + +/* 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) + +#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/memcmp.c b/lib/memcmp.c new file mode 100644 index 0000000..cf98bfa --- /dev/null +++ b/lib/memcmp.c @@ -0,0 +1,363 @@ +/* Copyright (C) 1991, 1993, 1995, 1997, 1998, 2003, 2006 Free Software + Foundation, Inc. + + Contributed by Torbjorn Granlund (tege@sics.se). + + 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> + +#undef memcmp + +#ifdef _LIBC + +# include <memcopy.h> +# include <endian.h> + +# if __BYTE_ORDER == __BIG_ENDIAN +# define WORDS_BIGENDIAN +# endif + +#else /* Not in the GNU C library. */ + +# include <sys/types.h> + +/* Type to use for aligned memory operations. + This should normally be the biggest type supported by a single load + and store. Must be an unsigned type. */ +# define op_t unsigned long int +# define OPSIZ (sizeof(op_t)) + +/* Threshold value for when to enter the unrolled loops. */ +# define OP_T_THRES 16 + +/* Type to use for unaligned operations. */ +typedef unsigned char byte; + +# ifndef WORDS_BIGENDIAN +# define MERGE(w0, sh_1, w1, sh_2) (((w0) >> (sh_1)) | ((w1) << (sh_2))) +# else +# define MERGE(w0, sh_1, w1, sh_2) (((w0) << (sh_1)) | ((w1) >> (sh_2))) +# endif + +#endif /* In the GNU C library. */ + +#ifdef WORDS_BIGENDIAN +# define CMP_LT_OR_GT(a, b) ((a) > (b) ? 1 : -1) +#else +# define CMP_LT_OR_GT(a, b) memcmp_bytes ((a), (b)) +#endif + +/* BE VERY CAREFUL IF YOU CHANGE THIS CODE! */ + +/* The strategy of this memcmp is: + + 1. Compare bytes until one of the block pointers is aligned. + + 2. Compare using memcmp_common_alignment or + memcmp_not_common_alignment, regarding the alignment of the other + block after the initial byte operations. The maximum number of + full words (of type op_t) are compared in this way. + + 3. Compare the few remaining bytes. */ + +#ifndef WORDS_BIGENDIAN +/* memcmp_bytes -- Compare A and B bytewise in the byte order of the machine. + A and B are known to be different. + This is needed only on little-endian machines. */ + +# ifdef __GNUC__ +__inline +# endif +static int +memcmp_bytes (long unsigned int a, long unsigned int b) +{ + long int srcp1 = (long int) &a; + long int srcp2 = (long int) &b; + op_t a0, b0; + + do + { + a0 = ((byte *) srcp1)[0]; + b0 = ((byte *) srcp2)[0]; + srcp1 += 1; + srcp2 += 1; + } + while (a0 == b0); + return a0 - b0; +} +#endif + +/* memcmp_common_alignment -- Compare blocks at SRCP1 and SRCP2 with LEN `op_t' + objects (not LEN bytes!). Both SRCP1 and SRCP2 should be aligned for + memory operations on `op_t's. */ +#ifdef __GNUC__ +__inline +#endif +static int +memcmp_common_alignment (long int srcp1, long int srcp2, size_t len) +{ + op_t a0, a1; + op_t b0, b1; + + switch (len % 4) + { + default: /* Avoid warning about uninitialized local variables. */ + case 2: + a0 = ((op_t *) srcp1)[0]; + b0 = ((op_t *) srcp2)[0]; + srcp1 -= 2 * OPSIZ; + srcp2 -= 2 * OPSIZ; + len += 2; + goto do1; + case 3: + a1 = ((op_t *) srcp1)[0]; + b1 = ((op_t *) srcp2)[0]; + srcp1 -= OPSIZ; + srcp2 -= OPSIZ; + len += 1; + goto do2; + case 0: + if (OP_T_THRES <= 3 * OPSIZ && len == 0) + return 0; + a0 = ((op_t *) srcp1)[0]; + b0 = ((op_t *) srcp2)[0]; + goto do3; + case 1: + a1 = ((op_t *) srcp1)[0]; + b1 = ((op_t *) srcp2)[0]; + srcp1 += OPSIZ; + srcp2 += OPSIZ; + len -= 1; + if (OP_T_THRES <= 3 * OPSIZ && len == 0) + goto do0; + /* Fall through. */ + } + + do + { + a0 = ((op_t *) srcp1)[0]; + b0 = ((op_t *) srcp2)[0]; + if (a1 != b1) + return CMP_LT_OR_GT (a1, b1); + + do3: + a1 = ((op_t *) srcp1)[1]; + b1 = ((op_t *) srcp2)[1]; + if (a0 != b0) + return CMP_LT_OR_GT (a0, b0); + + do2: + a0 = ((op_t *) srcp1)[2]; + b0 = ((op_t *) srcp2)[2]; + if (a1 != b1) + return CMP_LT_OR_GT (a1, b1); + + do1: + a1 = ((op_t *) srcp1)[3]; + b1 = ((op_t *) srcp2)[3]; + if (a0 != b0) + return CMP_LT_OR_GT (a0, b0); + + srcp1 += 4 * OPSIZ; + srcp2 += 4 * OPSIZ; + len -= 4; + } + while (len != 0); + + /* This is the right position for do0. Please don't move + it into the loop. */ + do0: + if (a1 != b1) + return CMP_LT_OR_GT (a1, b1); + return 0; +} + +/* memcmp_not_common_alignment -- Compare blocks at SRCP1 and SRCP2 with LEN + `op_t' objects (not LEN bytes!). SRCP2 should be aligned for memory + operations on `op_t', but SRCP1 *should be unaligned*. */ +#ifdef __GNUC__ +__inline +#endif +static int +memcmp_not_common_alignment (long int srcp1, long int srcp2, size_t len) +{ + op_t a0, a1, a2, a3; + op_t b0, b1, b2, b3; + op_t x; + int shl, shr; + + /* Calculate how to shift a word read at the memory operation + aligned srcp1 to make it aligned for comparison. */ + + shl = 8 * (srcp1 % OPSIZ); + shr = 8 * OPSIZ - shl; + + /* Make SRCP1 aligned by rounding it down to the beginning of the `op_t' + it points in the middle of. */ + srcp1 &= -OPSIZ; + + switch (len % 4) + { + default: /* Avoid warning about uninitialized local variables. */ + case 2: + a1 = ((op_t *) srcp1)[0]; + a2 = ((op_t *) srcp1)[1]; + b2 = ((op_t *) srcp2)[0]; + srcp1 -= 1 * OPSIZ; + srcp2 -= 2 * OPSIZ; + len += 2; + goto do1; + case 3: + a0 = ((op_t *) srcp1)[0]; + a1 = ((op_t *) srcp1)[1]; + b1 = ((op_t *) srcp2)[0]; + srcp2 -= 1 * OPSIZ; + len += 1; + goto do2; + case 0: + if (OP_T_THRES <= 3 * OPSIZ && len == 0) + return 0; + a3 = ((op_t *) srcp1)[0]; + a0 = ((op_t *) srcp1)[1]; + b0 = ((op_t *) srcp2)[0]; + srcp1 += 1 * OPSIZ; + goto do3; + case 1: + a2 = ((op_t *) srcp1)[0]; + a3 = ((op_t *) srcp1)[1]; + b3 = ((op_t *) srcp2)[0]; + srcp1 += 2 * OPSIZ; + srcp2 += 1 * OPSIZ; + len -= 1; + if (OP_T_THRES <= 3 * OPSIZ && len == 0) + goto do0; + /* Fall through. */ + } + + do + { + a0 = ((op_t *) srcp1)[0]; + b0 = ((op_t *) srcp2)[0]; + x = MERGE(a2, shl, a3, shr); + if (x != b3) + return CMP_LT_OR_GT (x, b3); + + do3: + a1 = ((op_t *) srcp1)[1]; + b1 = ((op_t *) srcp2)[1]; + x = MERGE(a3, shl, a0, shr); + if (x != b0) + return CMP_LT_OR_GT (x, b0); + + do2: + a2 = ((op_t *) srcp1)[2]; + b2 = ((op_t *) srcp2)[2]; + x = MERGE(a0, shl, a1, shr); + if (x != b1) + return CMP_LT_OR_GT (x, b1); + + do1: + a3 = ((op_t *) srcp1)[3]; + b3 = ((op_t *) srcp2)[3]; + x = MERGE(a1, shl, a2, shr); + if (x != b2) + return CMP_LT_OR_GT (x, b2); + + srcp1 += 4 * OPSIZ; + srcp2 += 4 * OPSIZ; + len -= 4; + } + while (len != 0); + + /* This is the right position for do0. Please don't move + it into the loop. */ + do0: + x = MERGE(a2, shl, a3, shr); + if (x != b3) + return CMP_LT_OR_GT (x, b3); + return 0; +} + +int +rpl_memcmp (const void *s1, const void *s2, size_t len) +{ + op_t a0; + op_t b0; + long int srcp1 = (long int) s1; + long int srcp2 = (long int) s2; + op_t res; + + if (len >= OP_T_THRES) + { + /* There are at least some bytes to compare. No need to test + for LEN == 0 in this alignment loop. */ + while (srcp2 % OPSIZ != 0) + { + a0 = ((byte *) srcp1)[0]; + b0 = ((byte *) srcp2)[0]; + srcp1 += 1; + srcp2 += 1; + res = a0 - b0; + if (res != 0) + return res; + len -= 1; + } + + /* SRCP2 is now aligned for memory operations on `op_t'. + SRCP1 alignment determines if we can do a simple, + aligned compare or need to shuffle bits. */ + + if (srcp1 % OPSIZ == 0) + res = memcmp_common_alignment (srcp1, srcp2, len / OPSIZ); + else + res = memcmp_not_common_alignment (srcp1, srcp2, len / OPSIZ); + if (res != 0) + return res; + + /* Number of bytes remaining in the interval [0..OPSIZ-1]. */ + srcp1 += len & -OPSIZ; + srcp2 += len & -OPSIZ; + len %= OPSIZ; + } + + /* There are just a few bytes to compare. Use byte memory operations. */ + while (len != 0) + { + a0 = ((byte *) srcp1)[0]; + b0 = ((byte *) srcp2)[0]; + srcp1 += 1; + srcp2 += 1; + res = a0 - b0; + if (res != 0) + return res; + len -= 1; + } + + return 0; +} + +#ifdef weak_alias +# undef bcmp +weak_alias (memcmp, bcmp) +#endif diff --git a/lib/memmove.c b/lib/memmove.c new file mode 100644 index 0000000..c5ff8b5 --- /dev/null +++ b/lib/memmove.c @@ -0,0 +1,26 @@ +/* memmove.c -- copy memory. + Copy LENGTH bytes from SOURCE to DEST. Does not null-terminate. + In the public domain. + By David MacKenzie <djm@gnu.ai.mit.edu>. */ + +#include <config.h> + +#include <stddef.h> + +void * +memmove (void *dest0, void const *source0, size_t length) +{ + char *dest = dest0; + char const *source = source0; + if (source < dest) + /* Moving from low mem to hi mem; start at end. */ + for (source += length, dest += length; length; --length) + *--dest = *--source; + else if (source != dest) + { + /* Moving from hi mem to low mem; start at beginning. */ + for (; length; --length) + *dest++ = *source++; + } + return dest0; +} diff --git a/lib/mkstemp.c b/lib/mkstemp.c new file mode 100644 index 0000000..eabee0c --- /dev/null +++ b/lib/mkstemp.c @@ -0,0 +1,40 @@ +/* Copyright (C) 1998, 1999, 2001, 2005, 2006 Free Software Foundation, Inc. + This file is derived from the one in 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 "mkstemp.h" +int __gen_tempname (); +#endif + +#include <stdio.h> +#include <stdlib.h> + +#ifndef __GT_FILE +# define __GT_FILE 0 +#endif + +/* Generate a unique temporary file name from TEMPLATE. + The last six characters of TEMPLATE must be "XXXXXX"; + they are replaced with a string that makes the file name unique. + Then open the file and return a fd. */ +int +mkstemp (template) + char *template; +{ + return __gen_tempname (template, __GT_FILE); +} diff --git a/lib/mkstemp.h b/lib/mkstemp.h new file mode 100644 index 0000000..69e6fec --- /dev/null +++ b/lib/mkstemp.h @@ -0,0 +1,30 @@ +/* Create a unique temporary 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. */ + +/* written by Jim Meyering */ + +#include <stdlib.h> + +#ifdef __MKSTEMP_PREFIX +# define _GL_CONCAT(x, y) x ## y +# define _GL_XCONCAT(x, y) _GL_CONCAT (x, y) +# define __MKSTEMP_ID(y) _GL_XCONCAT (__MKSTEMP_PREFIX, y) +# undef mkstemp +# define mkstemp __MKSTEMP_ID (mkstemp) +int mkstemp (char *); +#endif diff --git a/lib/obstack.c b/lib/obstack.c new file mode 100644 index 0000000..1bbf62b --- /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..95dd438 --- /dev/null +++ b/lib/obstack.h @@ -0,0 +1,508 @@ +/* obstack.h - object stack macros + Copyright (C) 1988-1994,1996-1999,2003,2004,2005 + 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 *); + +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/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/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..381f00a --- /dev/null +++ b/lib/quotearg.c @@ -0,0 +1,688 @@ +/* quotearg.c - quote arguments for output + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 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 <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 "gettext.h" +#define _(msgid) gettext (msgid) +#define N_(msgid) msgid + +#if HAVE_WCHAR_H + +/* BSD/OS 4.1 wchar.h requires FILE and struct tm to be declared. */ +# include <stdio.h> +# include <time.h> + +# include <wchar.h> +#endif + +#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 +# 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 iswprint +# if HAVE_WCTYPE_H +# include <wctype.h> +# endif +# if !defined iswprint && !HAVE_ISWPRINT +# define iswprint(wc) 1 +# endif +#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 = xmalloc (sizeof *p); + *p = *(o ? o : &default_quoting_options); + 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 = xmalloc (bufsize); + quotearg_buffer (buf, bufsize, arg, argsize, o); + errno = e; + return buf; +} + +/* 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; + + /* 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; + unsigned int n0 = n; + struct slotvec + { + size_t size; + char *val; + }; + static struct slotvec slotvec0 = {sizeof slot0, slot0}; + static struct slotvec *slotvec = &slotvec0; + + 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; + + if (xalloc_oversized (n1, sizeof *slotvec)) + xalloc_die (); + + if (slotvec == &slotvec0) + { + slotvec = xmalloc (sizeof *slotvec); + *slotvec = slotvec0; + } + slotvec = xrealloc (slotvec, n1 * sizeof *slotvec); + memset (slotvec + nslots, 0, (n1 - nslots) * sizeof *slotvec); + nslots = n1; + } + + { + size_t size = slotvec[n].size; + char *val = slotvec[n].val; + size_t qsize = quotearg_buffer (val, size, arg, argsize, options); + + if (size <= qsize) + { + slotvec[n].size = size = qsize + 1; + if (val != slot0) + free (val); + slotvec[n].val = val = xmalloc (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..24f26f7 --- /dev/null +++ b/lib/quotearg.h @@ -0,0 +1,137 @@ +/* quotearg.h - quote arguments for output + + Copyright (C) 1998, 1999, 2000, 2001, 2002, 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 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); + +#endif /* !QUOTEARG_H_ */ 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..6e317f5 --- /dev/null +++ b/lib/regcomp.c @@ -0,0 +1,3858 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002,2003,2004,2005,2006 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. */ + +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)) + { + if (BE (msg_size > errbuf_size, 0)) + { +#if defined HAVE_MEMPCPY || defined _LIBC + *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0'; +#else + memcpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; +#endif + } + else + memcpy (errbuf, msg, msg_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; +#ifndef _LIBC + char *codeset_name; +#endif +#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 +# ifdef HAVE_LANGINFO_CODESET + codeset_name = nl_langinfo (CODESET); +# else + codeset_name = getenv ("LC_ALL"); + if (codeset_name == NULL || codeset_name[0] == '\0') + codeset_name = getenv ("LC_CTYPE"); + if (codeset_name == NULL || codeset_name[0] == '\0') + codeset_name = getenv ("LANG"); + if (codeset_name == NULL) + codeset_name = ""; + else if (strchr (codeset_name, '.') != NULL) + codeset_name = strchr (codeset_name, '.') + 1; +# endif + + if (strcasecmp (codeset_name, "UTF-8") == 0 + || strcasecmp (codeset_name, "UTF8") == 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, *exp; + 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)) + { + exp = parse_expression (regexp, preg, token, syntax, nest, err); + if (BE (*err != REG_NOERROR && exp == NULL, 0)) + { + return NULL; + } + if (tree != NULL && exp != NULL) + { + tree = create_tree (dfa, tree, exp, CONCAT); + if (tree == NULL) + { + *err = REG_ESPACE; + return NULL; + } + } + else if (tree == NULL) + tree = exp; + /* Otherwise exp == 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, '\0'); + 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 + /* + if (syntax & RE_HAT_LISTS_NOT_NEWLINE) + bitset_set(cset->sbcset, '\0'); + */ + 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..e8748e1 --- /dev/null +++ b/lib/regex.h @@ -0,0 +1,668 @@ +/* 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". */ +#ifndef __restrict +# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) +# if defined restrict || 199901L <= __STDC_VERSION__ +# define __restrict restrict +# else +# define __restrict +# endif +# endif +#endif +/* gcc 3.1 and up support the [restrict] syntax. */ +#ifndef __restrict_arr +# if (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)) \ + && !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..78e16f3 --- /dev/null +++ b/lib/regex_internal.c @@ -0,0 +1,1742 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 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. */ + +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, *q, *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; + + q = p; + if (BE (pstr->trans != NULL, 0)) + { + int i = mlen < 6 ? mlen : 6; + while (--i >= 0) + buf[i] = pstr->trans[p[i]]; + q = buf; + } + /* 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..cc9b5b7 --- /dev/null +++ b/lib/regex_internal.h @@ -0,0 +1,865 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 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. */ + +#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> + +#ifndef _LIBC +# include "strcase.h" +#endif + +#if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC +# include <langinfo.h> +#endif +#if defined HAVE_LOCALE_H || defined _LIBC +# include <locale.h> +#endif +#if defined HAVE_WCHAR_H || defined _LIBC +# include <wchar.h> +#endif /* HAVE_WCHAR_H || _LIBC */ +#if defined HAVE_WCTYPE_H || defined _LIBC +# include <wctype.h> +#endif /* HAVE_WCTYPE_H || _LIBC */ +#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_WCHAR_H && 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 +# ifndef __mempcpy +# define __mempcpy mempcpy +# endif +# 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 + 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..84c3aed --- /dev/null +++ b/lib/regexec.c @@ -0,0 +1,4398 @@ +/* Extended regular expression matching and search library. + Copyright (C) 2002, 2003, 2004, 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. */ + +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[]; + 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 */ +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 + 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/stat-macros.h b/lib/stat-macros.h new file mode 100644 index 0000000..829ddd3 --- /dev/null +++ b/lib/stat-macros.h @@ -0,0 +1,260 @@ +/* stat-related macros + + Copyright (C) 1993, 1994, 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 and Jim Meyering. */ + +#ifndef STAT_MACROS_H +# define STAT_MACROS_H 1 + +# if ! defined S_ISREG && ! defined S_IFREG +# error "you must include <sys/stat.h> before including this file" +# endif + +# 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 + +/* 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) + +#endif /* STAT_MACROS_H */ diff --git a/lib/stat_.h b/lib/stat_.h new file mode 100644 index 0000000..8c2e552 --- /dev/null +++ b/lib/stat_.h @@ -0,0 +1,48 @@ +/* Provide a complete sys/stat header file. + Copyright (C) 2006 Free Software Foundation, Inc. + Written by Eric Blake. + + 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_SYS_STAT_H +#define _gl_SYS_STAT_H + +/* 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. */ +#include @ABSOLUTE_SYS_STAT_H@ + +/* 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 */ diff --git a/lib/stdbool_.h b/lib/stdbool_.h new file mode 100644 index 0000000..efa80ba --- /dev/null +++ b/lib/stdbool_.h @@ -0,0 +1,115 @@ +/* Copyright (C) 2001, 2002, 2003, 2006 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 _STDBOOL_H +#define _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. + + - 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 /* _STDBOOL_H */ diff --git a/lib/stdint_.h b/lib/stdint_.h new file mode 100644 index 0000000..1ada949 --- /dev/null +++ b/lib/stdint_.h @@ -0,0 +1,456 @@ +/* Copyright (C) 2001-2002, 2004-2006 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. */ + +#ifndef _GL_STDINT_H +#define _GL_STDINT_H + +/* + * ISO C 99 <stdint.h> for platforms that lack it. + * <http://www.opengroup.org/susv3xbd/stdint.h.html> + */ + +/* 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. */ +# include @ABSOLUTE_STDINT_H@ +#endif + +/* <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>). + 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@ +# 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 @HAVE_WCHAR_H@ && ! (defined WCHAR_MIN && defined WCHAR_MAX) + /* 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> +# 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 + +#undef int64_t +#undef uint64_t +#if LONG_MAX >> 31 >> 31 == 1 +# define int64_t long int +# define uint64_t unsigned long int +#elif defined _MSC_VER +# define int64_t __int64 +# define uint64_t unsigned __int64 +#elif @HAVE_LONG_LONG_INT@ +# define int64_t long long int +# define uint64_t unsigned long long int +#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 int64_t +# define int_least64_t int64_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 int64_t +# define int_fast64_t int64_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 +#undef uintmax_t +#if @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define intmax_t long long int +# define uintmax_t unsigned long long int +#elif defined int64_t +# define intmax_t int64_t +# define uintmax_t uint64_t +#else +# define intmax_t long int +# 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 +#undef UINT64_MAX +#ifdef int64_t +# define INT64_MIN (~ INT64_MAX) +# define INT64_MAX INTMAX_C (9223372036854775807) +# 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 +#undef UINT_LEAST64_MAX +#ifdef int64_t +# define INT_LEAST64_MIN INT64_MIN +# define INT_LEAST64_MAX INT64_MAX +# 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 +#undef UINT_FAST64_MAX +#ifdef int64_t +# define INT_FAST64_MIN INT64_MIN +# define INT_FAST64_MAX INT64_MAX +# 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 +#undef UINTMAX_MAX +#define INTMAX_MIN (~ INTMAX_MAX) +#ifdef INT64_MAX +# define INTMAX_MAX INT64_MAX +# define UINTMAX_MAX UINT64_MAX +#else +# define INTMAX_MAX INT32_MAX +# 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 +# define UINT64_C(x) x##UL +#elif defined _MSC_VER +# define INT64_C(x) x##i64 +# define UINT64_C(x) x##ui64 +#elif @HAVE_LONG_LONG_INT@ +# define INT64_C(x) x##LL +# define UINT64_C(x) x##ULL +#endif + +/* 7.18.4.2. Macros for greatest-width integer constants */ + +#undef INTMAX_C +#undef UINTMAX_C +#if @HAVE_LONG_LONG_INT@ && LONG_MAX >> 30 == 1 +# define INTMAX_C(x) x##LL +# define UINTMAX_C(x) x##ULL +#elif defined int64_t +# define INTMAX_C(x) INT64_C(x) +# define UINTMAX_C(x) UINT64_C(x) +#else +# define INTMAX_C(x) x##L +# define UINTMAX_C(x) x##UL +#endif + +#endif /* !defined __cplusplus || defined __STDC_CONSTANT_MACROS */ + +#endif /* _GL_STDINT_H */ diff --git a/lib/strcase.h b/lib/strcase.h new file mode 100644 index 0000000..e420798 --- /dev/null +++ b/lib/strcase.h @@ -0,0 +1,48 @@ +/* Case-insensitive string comparison functions. + Copyright (C) 1995-1996, 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 _STRCASE_H +#define _STRCASE_H + +#include <stddef.h> + + +#ifdef __cplusplus +extern "C" { +#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 may, in multibyte locales, return 0 for strings of + different lengths! */ +extern int strcasecmp (const char *s1, const char *s2); + +/* Compare no more than N characters 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 can not work correctly in multibyte locales. */ +extern int strncasecmp (const char *s1, const char *s2, size_t n); + + +#ifdef __cplusplus +} +#endif + + +#endif /* _STRCASE_H */ diff --git a/lib/strcasecmp.c b/lib/strcasecmp.c new file mode 100644 index 0000000..99d5dd2 --- /dev/null +++ b/lib/strcasecmp.c @@ -0,0 +1,103 @@ +/* Case-insensitive string comparison function. + Copyright (C) 1998-1999, 2005-2006 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 "strcase.h" + +#include <ctype.h> +#include <limits.h> + +#if HAVE_MBRTOWC +# include "mbuiter.h" +#endif + +#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 may, in multibyte locales, return 0 for strings of + different lengths! */ +int +strcasecmp (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/strerror.c b/lib/strerror.c new file mode 100644 index 0000000..24e2e1b --- /dev/null +++ b/lib/strerror.c @@ -0,0 +1,47 @@ +/* strerror.c --- ANSI C compatible system error routine + + Copyright (C) 1986, 1988, 1989, 1991, 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 <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]; +} diff --git a/lib/strncasecmp.c b/lib/strncasecmp.c new file mode 100644 index 0000000..f59b953 --- /dev/null +++ b/lib/strncasecmp.c @@ -0,0 +1,63 @@ +/* strncasecmp.c -- case insensitive string comparator + Copyright (C) 1998-1999, 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 "strcase.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. */ + +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/strnlen1.c b/lib/strnlen1.c new file mode 100644 index 0000000..f74e427 --- /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 = 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/strverscmp.c b/lib/strverscmp.c new file mode 100644 index 0000000..f077651 --- /dev/null +++ b/lib/strverscmp.c @@ -0,0 +1,131 @@ +/* Compare strings while treating digits characters numerically. + Copyright (C) 1997, 2000, 2002, 2004, 2006 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Jean-François Bignolles <bignolle@ecoledoc.ibp.fr>, 1997. + + 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 <string.h> +#include <ctype.h> + +/* states: S_N: normal, S_I: comparing integral part, S_F: comparing + fractional parts, S_Z: idem but with leading Zeroes only */ +#define S_N 0x0 +#define S_I 0x4 +#define S_F 0x8 +#define S_Z 0xC + +/* result_type: CMP: return diff; LEN: compare using len_diff/diff */ +#define CMP 2 +#define LEN 3 + + +/* 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) + +#undef __strverscmp +#undef strverscmp + +#ifndef weak_alias +# define __strverscmp strverscmp +#endif + +/* Compare S1 and S2 as strings holding indices/version numbers, + returning less than, equal to or greater than zero if S1 is less than, + equal to or greater than S2 (for more info, see the texinfo doc). +*/ + +int +__strverscmp (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; + int state; + int diff; + + /* Symbol(s) 0 [1-9] others (padding) + Transition (10) 0 (01) d (00) x (11) - */ + static const unsigned int next_state[] = + { + /* state x d 0 - */ + /* S_N */ S_N, S_I, S_Z, S_N, + /* S_I */ S_N, S_I, S_I, S_I, + /* S_F */ S_N, S_F, S_F, S_F, + /* S_Z */ S_N, S_F, S_Z, S_Z + }; + + static const int result_type[] = + { + /* state x/x x/d x/0 x/- d/x d/d d/0 d/- + 0/x 0/d 0/0 0/- -/x -/d -/0 -/- */ + + /* S_N */ CMP, CMP, CMP, CMP, CMP, LEN, CMP, CMP, + CMP, CMP, CMP, CMP, CMP, CMP, CMP, CMP, + /* S_I */ CMP, -1, -1, CMP, 1, LEN, LEN, CMP, + 1, LEN, LEN, CMP, CMP, CMP, CMP, CMP, + /* S_F */ CMP, CMP, CMP, CMP, CMP, LEN, CMP, CMP, + CMP, CMP, CMP, CMP, CMP, CMP, CMP, CMP, + /* S_Z */ CMP, 1, 1, CMP, -1, CMP, CMP, CMP, + -1, CMP, CMP, CMP + }; + + if (p1 == p2) + return 0; + + c1 = *p1++; + c2 = *p2++; + /* Hint: '0' is a digit too. */ + state = S_N | ((c1 == '0') + (ISDIGIT (c1) != 0)); + + while ((diff = c1 - c2) == 0 && c1 != '\0') + { + state = next_state[state]; + c1 = *p1++; + c2 = *p2++; + state |= (c1 == '0') + (ISDIGIT (c1) != 0); + } + + state = result_type[state << 2 | ((c2 == '0') + (ISDIGIT (c2) != 0))]; + + switch (state) + { + case CMP: + return diff; + + case LEN: + while (ISDIGIT (*p1++)) + if (!ISDIGIT (*p2++)) + return 1; + + return ISDIGIT (*p2) ? -1 : diff; + + default: + return state; + } +} +#ifdef weak_alias +weak_alias (__strverscmp, strverscmp) +#endif diff --git a/lib/strverscmp.h b/lib/strverscmp.h new file mode 100644 index 0000000..7edeac5 --- /dev/null +++ b/lib/strverscmp.h @@ -0,0 +1,24 @@ +/* Compare strings while treating digits characters numerically. + + Copyright (C) 1997, 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 STRVERSCMP_H_ +# define STRVERSCMP_H_ + +int strverscmp (const char *, const char *); + +#endif /* not STRVERSCMP_H_ */ diff --git a/lib/tempname.c b/lib/tempname.c new file mode 100644 index 0000000..6fef5e9 --- /dev/null +++ b/lib/tempname.c @@ -0,0 +1,317 @@ +/* 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 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 !_LIBC +# include <config.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> + +#if HAVE_SYS_TIME_H || _LIBC +# include <sys/time.h> +#endif + +#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 +# include "stat-macros.h" +# define struct_stat64 struct stat +# define small_open open +# define large_open open +# 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 + +/* 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; +} + +/* 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 +# if HAVE_GETTIMEOFDAY || _LIBC + { + struct timeval tv; + __gettimeofday (&tv, NULL); + random_time_bits = ((uint64_t) tv.tv_usec << 16) ^ tv.tv_sec; + } +# else + random_time_bits = time (NULL); +# endif +#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/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/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/wcwidth.h b/lib/wcwidth.h new file mode 100644 index 0000000..37b83cb --- /dev/null +++ b/lib/wcwidth.h @@ -0,0 +1,77 @@ +/* Determine the number of screen columns needed for a character. + 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 _gl_WCWIDTH_H +#define _gl_WCWIDTH_H + +#if HAVE_WCHAR_T + +/* Get wcwidth if available, along with wchar_t. */ +# if HAVE_WCHAR_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> +# endif + +/* Get iswprint. */ +# if HAVE_WCTYPE_H +# include <wctype.h> +# endif +# if !defined iswprint && !HAVE_ISWPRINT +static inline int +iswprint (wint_t wc) +{ + return (wc >= 0 && wc < 128 + ? wc >= ' ' && wc <= '~' + : 1); +} +# define iswprint iswprint +# endif + +# 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_H */ + +#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..f80977e --- /dev/null +++ b/lib/xalloc.h @@ -0,0 +1,79 @@ +/* xalloc.h -- malloc with out-of-memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 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. */ + +#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 *xnmalloc (size_t n, size_t s); +void *xzalloc (size_t s); +void *xcalloc (size_t n, size_t s); +void *xrealloc (void *p, size_t s); +void *xnrealloc (void *p, size_t n, size_t s); +void *x2realloc (void *p, size_t *pn); +void *x2nrealloc (void *p, size_t *pn, size_t s); +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)) + +# ifdef __cplusplus +} +# endif + + +#endif /* !XALLOC_H_ */ diff --git a/lib/xmalloc.c b/lib/xmalloc.c new file mode 100644 index 0000000..fe84a0b --- /dev/null +++ b/lib/xmalloc.c @@ -0,0 +1,240 @@ +/* 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> + +#include "xalloc.h" + +#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 an array of N objects, each with S bytes of memory, + dynamically, with error checking. S must be nonzero. */ + +static inline void * +xnmalloc_inline (size_t n, size_t s) +{ + void *p; + if (xalloc_oversized (n, s) || (! (p = malloc (n * s)) && n != 0)) + xalloc_die (); + return p; +} + +void * +xnmalloc (size_t n, size_t s) +{ + return xnmalloc_inline (n, s); +} + +/* Allocate N bytes of memory dynamically, with error checking. */ + +void * +xmalloc (size_t n) +{ + return xnmalloc_inline (n, 1); +} + +/* 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_inline (void *p, size_t n, size_t s) +{ + if (xalloc_oversized (n, s) || (! (p = realloc (p, n * s)) && n != 0)) + xalloc_die (); + return p; +} + +void * +xnrealloc (void *p, size_t n, size_t s) +{ + return xnrealloc_inline (p, n, s); +} + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. */ + +void * +xrealloc (void *p, size_t n) +{ + return xnrealloc_inline (p, n, 1); +} + + +/* 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 doubled so that + repeated reallocations have O(N log N) overall cost rather than + O(N**2) cost, but the specification for this function does not + guarantee that sizes are doubled. + + 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_inline (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 + { + if (SIZE_MAX / 2 / s < n) + xalloc_die (); + n *= 2; + } + + *pn = n; + return xrealloc (p, n * s); +} + +void * +x2nrealloc (void *p, size_t *pn, size_t s) +{ + return x2nrealloc_inline (p, pn, s); +} + +/* 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_inline (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); +} |