summaryrefslogtreecommitdiff
path: root/pristine-gz
blob: 8caf59d26b58f95272b299ea1c6d44f9cef253fc (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
#!/usr/bin/perl

=head1 NAME

pristine-gz - regenerate pristine gz files

=head1 SYNOPSIS

B<pristine-gz> [-vdk] gendelta I<file.gz> I<delta>

B<pristine-gz> [-vdk] gengz I<delta> I<file>

=head1 DESCRIPTION

This is a complement to the pristine-tar(1) command. Normally you don't
need to run it by hand, since pristine-tar calls it as necessary to handle
.tar.gz files.

pristine-gz gendelta takes the specified I<gz> file, and generates a
small binary I<delta> file that can later be used by pristine-gz gengz
to recreate the original file.

pristine-gz gengz takes the specified I<delta> file, and compresses
the specified input I<file> (which must be identical to the contents
of the original gz file). The resulting file will be identical to
the original gz file used to create the delta.

The approach used to regenerate the original gz file is to figure out how
it was produced -- what compression level was used, whether it was built
with GNU gzip(1) or with a library or BSD version, whether the --rsyncable
option was used, etc, and to reproduce this build environment when
regenerating the gz.

This approach will work for about 99.5% of cases. One example of a case it
cannot currently support is a gz file that has been produced by appending
together multiple gz files.

For the few where it doesn't work, a binary diff will be included in the
delta between the closest regneratable gz file and the original. In
the worst case, the diff will include the entire content of the original
gz file, resulting in a larger than usual delta. If the delta is much
larger than usual, pristine-gz will print a warning.

If the delta filename is "-", pristine-gz reads or writes it to stdio.

=head1 OPTIONS

=over 4

=item -v

=item --verbose

Verbose mode, show each command that is run.

=item -d

=item --debug

Debug mode.

=item -k

=item --keep

Don't clean up the temporary directory on exit.

=back

=head1 ENVIRONMENT

=over 4

=item B<TMPDIR>

Specifies a location to place temporary files, other than the default.

=back

=head1 AUTHOR

Joey Hess <joeyh@debian.org>,
Faidon Liambotis <paravoid@debian.org>
Josh Triplett <josh@joshtriplett.org>

Licensed under the GPL, version 2.

=cut

use warnings;
use strict;
use Pristine::Tar;
use Pristine::Tar::Delta;
use Pristine::Tar::Formats;
use File::Basename qw/basename/;

dispatch(
	commands => {
		usage => [\&usage],
		gendelta => [\&gendelta, 2],
		gengz => [\&gengz, 2],
	},
);

sub usage {
	print STDERR "Usage: pristine-gz [-vdk] gendelta file.gz delta\n";
	print STDERR "       pristine-gz [-vdk] gengz delta file\n";
}

sub readgzip {
	my $filename = shift;
	
	if (! is_gz($filename)) {
		error "This is not a valid GZip archive.";
	}

	open(GZIP, "< $filename")
		or die("Could not open '$filename' for reading: $!\n");

	my $chars;
	if (read(GZIP, $chars, 10) != 10) {
		die("Unable to read 10 bytes from input\n");
	}

	my ($id1, $id2, $method, $flags, $timestamp, $level, $os, $name)
		= (unpack("CCCb8VCC", $chars), '');

	my @flags = split(//, $flags);
	
	if ($flags[$fconstants{GZIP_FLAG_FNAME}]) {
		# read a null-terminated string
		$name .= $chars
			while (read(GZIP, $chars, 1) == 1 && ord($chars) != 0);
	}
	close(GZIP);

	return (\@flags, $timestamp, $level, $os, $name);
}

sub predictgzipargs {
	my ($flags, $timestamp, $level) = @_;
	my @flags = @$flags;

	my @args;
	unless ($flags[$fconstants{GZIP_FLAG_FNAME}]) {
		push @args, '-n';
		push @args, '-M' if $timestamp;
	}

	if ($level == $fconstants{GZIP_COMPRESSION_BEST}) {
		push @args, '-9'
	}
	elsif ($level == $fconstants{GZIP_COMPRESSION_FAST}) {
		push @args, '-1'
	}

	return @args;
}

sub comparefiles {
	my ($old, $new) = (shift, shift);
	system('cmp', '-s', $old, $new);

	if ($? == -1 || $? & 127) {
		die("Failed to execute cmp: $!\n");
	}

	return $? >> 8;
}

sub reproducegz {
	my ($orig, $tempdir, $tempin) = @_;
	my $tempout="$tempdir/test.gz";
	doit_redir($orig, $tempin, "gzip", "-dc");

	# read fields from gzip headers
	my ($flags, $timestamp, $level, $os, $name) = readgzip($orig);
	debug("flags: [".join(", ", @$flags).
		"] timestamp: $timestamp level: $level os: $os name: $name");

	# try to guess the gzip arguments that are needed by the header
	# information
	my @args = predictgzipargs($flags, $timestamp, $level);
	my @extraargs = ("-F", $name, "-T", $timestamp);

	my @try;

	if ($os == $fconstants{GZIP_OS_UNIX}) {
		# for 98% of the cases the simple heuristic above works
		# and it was produced by gnu gzip.
		push @try, ['--gnu', @args];
		push @try, ['--gnu', @args, '--rsyncable'];
	}

	if ($name =~ /\//) {
		push @args, "--original-name", $name;
		@extraargs = ("-T", $timestamp);
		$name = basename($name);
	}

	# set the Operating System flag to the one found in the original
	# archive
	push @args, ("--osflag", $os) if $os != $fconstants{GZIP_OS_UNIX};

	# many of the .gz out there are created using the BSD version of
	# gzip which is using the zlib library; try with our version of
	# bsd-gzip with added support for the undocumented GNU gzip options
	# -m and -M
	push @try, [@args];

	# apparently, there is an old version of bsd-gzip (or a similar tool
	# based on zlib) that creates gz using maximum compression (-9) but
	# does not indicate so in the headers. surprisingly, there are many
	# .gz out there.
	push @try, [@args, '--quirk', 'buggy-bsd'];

	# Windows' NTFS gzip implementation; quirk is really really evil
	# it should be the last test: it can result in a corrupted archive!
	if ($os == $fconstants{GZIP_OS_NTFS}) {
		pop @args; pop @args; # ntfs quirk implies NTFS osflag
		push @try, [@args, '--quirk', 'ntfs'];
	}

	my $origsize=(stat($orig))[7];
	my ($bestvariant, $bestsize);

	foreach my $variant (@try) {
		doit_redir($tempin, $tempout, 'zgz', @$variant, @extraargs, '-c');
		if (!comparefiles($orig, $tempout)) {
			# success
			return $name, $timestamp, undef, @$variant;
		}
		else {
			# generate a binary delta and see if this is the
			# best variant so far
			my $ret=system("xdelta delta -0 --pristine $tempout $orig $tempdir/tmpdelta 2>/dev/null") >> 8;
			# xdelta exits 1 on success
			if ($ret == 1) {
				my $size=(stat("$tempdir/tmpdelta"))[7];
				if (! defined $bestsize || $size < $bestsize) {
					$bestvariant = $variant;
					$bestsize=$size;
					rename("$tempdir/tmpdelta", "$tempdir/bestdelta") || die "rename: $!";
				}
			}
		}
	}

	# Nothing worked perfectly, so use the delta that was generated for
	# the best variant
	my $percentover=100 - int (($origsize-$bestsize)/$origsize*100);
	debug("Using delta to best variant, bloating $percentover%: @$bestvariant");
	if ($percentover > 10) {
		print STDERR "warning: pristine-gz cannot reproduce build of $orig; ";
		if ($percentover >= 100) {
			print STDERR "storing entire file in delta!\n";
		}
		else {
			print STDERR "storing $percentover% size diff in delta\n";
		}
		print STDERR "(Please consider filing a bug report so the delta size can be improved.)\n";
	}
	return $name, $timestamp, "$tempdir/bestdelta", @$bestvariant;
}

sub gengz {
	my $deltafile=shift;
	my $file=shift;

	my $delta=Pristine::Tar::Delta::read(Tarball => $deltafile);
	Pristine::Tar::Delta::assert($delta, type => "gz", maxversion => 3,
		fields => [qw{params filename timestamp}]);

	my @params=split(' ', $delta->{params});
	while (@params) {
		$_=shift @params;
		next if /^(--gnu|--rsyncable|-[nmM1-9])$/;
		if (/^(--original-name|--quirk|--osflag)$/) {
			shift @params;
			next;
		}
		die "paranoia check failed on params from delta (@params)";
	}
	@params=split(' ', $delta->{params});

	my $filename=$delta->{filename};
	$filename=~s/^.*\///; # basename isn't strong enough

	my @zgz=("zgz", @params, "-T", $delta->{timestamp});
	if (! grep { $_ eq "--original-name" } @params) {
		push @zgz, "-F", $filename;
	}
	push @zgz, "-c";

	if (exists $delta->{delta}) {
		my $tempdir=tempdir();
		my $tfile="$tempdir/".basename($file).".gz";
		doit_redir($file, $tfile, @zgz);
		doit("xdelta", "patch", "--pristine", $delta->{delta}, $tfile, "$file.gz");
	}
	else {
		doit_redir("$file", "$file.gz", @zgz);
	}
	doit("rm", "-f", $file);
}

sub gendelta {
	my $gzfile=shift;
	my $deltafile=shift;

	my $tempdir=tempdir();
	my ($filename, $timestamp, $xdelta, @params)=
		reproducegz($gzfile, $tempdir, "$tempdir/test");
	
	Pristine::Tar::Delta::write(Tarball => $deltafile, {
		version => (defined $xdelta ? "3.0" : "2.0"),
		type => 'gz',
		params => "@params",
		filename => basename($filename),
		timestamp => $timestamp,

		(defined $xdelta ? (delta => $xdelta) : ()),
	});
}