Commit | Line | Data |
25f0751f |
1 | |
2 | package Compress::Zlib; |
3 | |
4 | require 5.004 ; |
5 | require Exporter; |
6 | use AutoLoader; |
7 | use Carp ; |
8 | use IO::Handle ; |
9 | use Scalar::Util qw(dualvar); |
10 | |
93d092e2 |
11 | use IO::Compress::Base::Common 2.005 ; |
12 | use Compress::Raw::Zlib 2.005 ; |
13 | use IO::Compress::Gzip 2.005 ; |
14 | use IO::Uncompress::Gunzip 2.005 ; |
25f0751f |
15 | |
16 | use strict ; |
17 | use warnings ; |
18 | use bytes ; |
19 | our ($VERSION, $XS_VERSION, @ISA, @EXPORT, $AUTOLOAD); |
20 | |
93d092e2 |
21 | $VERSION = '2.005'; |
25f0751f |
22 | $XS_VERSION = $VERSION; |
23 | $VERSION = eval $VERSION; |
24 | |
25 | @ISA = qw(Exporter); |
26 | # Items to export into callers namespace by default. Note: do not export |
27 | # names by default without a very good reason. Use EXPORT_OK instead. |
28 | # Do not simply export all your public functions/methods/constants. |
29 | @EXPORT = qw( |
30 | deflateInit inflateInit |
31 | |
32 | compress uncompress |
33 | |
34 | gzopen $gzerrno |
35 | ); |
36 | |
37 | push @EXPORT, @Compress::Raw::Zlib::EXPORT ; |
38 | |
39 | BEGIN |
40 | { |
41 | *zlib_version = \&Compress::Raw::Zlib::zlib_version; |
42 | } |
43 | |
44 | sub AUTOLOAD { |
45 | my($constname); |
46 | ($constname = $AUTOLOAD) =~ s/.*:://; |
47 | my ($error, $val) = Compress::Raw::Zlib::constant($constname); |
48 | Carp::croak $error if $error; |
49 | no strict 'refs'; |
50 | *{$AUTOLOAD} = sub { $val }; |
51 | goto &{$AUTOLOAD}; |
52 | } |
53 | |
54 | use constant FLAG_APPEND => 1 ; |
55 | use constant FLAG_CRC => 2 ; |
56 | use constant FLAG_ADLER => 4 ; |
57 | use constant FLAG_CONSUME_INPUT => 8 ; |
58 | |
59 | our (@my_z_errmsg); |
60 | |
61 | @my_z_errmsg = ( |
62 | "need dictionary", # Z_NEED_DICT 2 |
63 | "stream end", # Z_STREAM_END 1 |
64 | "", # Z_OK 0 |
65 | "file error", # Z_ERRNO (-1) |
66 | "stream error", # Z_STREAM_ERROR (-2) |
67 | "data error", # Z_DATA_ERROR (-3) |
68 | "insufficient memory", # Z_MEM_ERROR (-4) |
69 | "buffer error", # Z_BUF_ERROR (-5) |
70 | "incompatible version",# Z_VERSION_ERROR(-6) |
71 | ); |
72 | |
73 | |
74 | sub _set_gzerr |
75 | { |
76 | my $value = shift ; |
77 | |
78 | if ($value == 0) { |
79 | $Compress::Zlib::gzerrno = 0 ; |
80 | } |
81 | elsif ($value == Z_ERRNO() || $value > 2) { |
82 | $Compress::Zlib::gzerrno = $! ; |
83 | } |
84 | else { |
85 | $Compress::Zlib::gzerrno = dualvar($value+0, $my_z_errmsg[2 - $value]); |
86 | } |
87 | |
88 | return $value ; |
89 | } |
90 | |
91 | sub _save_gzerr |
92 | { |
93 | my $gz = shift ; |
94 | my $test_eof = shift ; |
95 | |
96 | my $value = $gz->errorNo() || 0 ; |
97 | |
98 | if ($test_eof) { |
99 | #my $gz = $self->[0] ; |
100 | # gzread uses Z_STREAM_END to denote a successful end |
101 | $value = Z_STREAM_END() if $gz->eof() && $value == 0 ; |
102 | } |
103 | |
104 | _set_gzerr($value) ; |
105 | } |
106 | |
107 | sub gzopen($$) |
108 | { |
109 | my ($file, $mode) = @_ ; |
110 | |
111 | my $gz ; |
112 | my %defOpts = (Level => Z_DEFAULT_COMPRESSION(), |
113 | Strategy => Z_DEFAULT_STRATEGY(), |
114 | ); |
115 | |
116 | my $writing ; |
117 | $writing = ! ($mode =~ /r/i) ; |
118 | $writing = ($mode =~ /[wa]/i) ; |
119 | |
120 | $defOpts{Level} = $1 if $mode =~ /(\d)/; |
121 | $defOpts{Strategy} = Z_FILTERED() if $mode =~ /f/i; |
122 | $defOpts{Strategy} = Z_HUFFMAN_ONLY() if $mode =~ /h/i; |
f6fd7794 |
123 | $defOpts{Append} = 1 if $mode =~ /a/i; |
25f0751f |
124 | |
125 | my $infDef = $writing ? 'deflate' : 'inflate'; |
126 | my @params = () ; |
127 | |
128 | croak "gzopen: file parameter is not a filehandle or filename" |
f6fd7794 |
129 | unless isaFilehandle $file || isaFilename $file || |
130 | (ref $file && ref $file eq 'SCALAR'); |
25f0751f |
131 | |
132 | return undef unless $mode =~ /[rwa]/i ; |
133 | |
134 | _set_gzerr(0) ; |
135 | |
136 | if ($writing) { |
137 | $gz = new IO::Compress::Gzip($file, Minimal => 1, AutoClose => 1, |
138 | %defOpts) |
139 | or $Compress::Zlib::gzerrno = $IO::Compress::Gzip::GzipError; |
140 | } |
141 | else { |
142 | $gz = new IO::Uncompress::Gunzip($file, |
143 | Transparent => 1, |
144 | Append => 0, |
145 | AutoClose => 1, |
f6fd7794 |
146 | MultiStream => 1, |
25f0751f |
147 | Strict => 0) |
148 | or $Compress::Zlib::gzerrno = $IO::Uncompress::Gunzip::GunzipError; |
149 | } |
150 | |
151 | return undef |
152 | if ! defined $gz ; |
153 | |
154 | bless [$gz, $infDef], 'Compress::Zlib::gzFile'; |
155 | } |
156 | |
157 | sub Compress::Zlib::gzFile::gzread |
158 | { |
159 | my $self = shift ; |
160 | |
161 | return _set_gzerr(Z_STREAM_ERROR()) |
162 | if $self->[1] ne 'inflate'; |
163 | |
98719077 |
164 | my $len = defined $_[1] ? $_[1] : 4096 ; |
165 | |
166 | if ($self->gzeof() || $len == 0) { |
cb7abd7f |
167 | # Zap the output buffer to match ver 1 behaviour. |
168 | $_[0] = "" ; |
169 | return 0 ; |
170 | } |
25f0751f |
171 | |
172 | my $gz = $self->[0] ; |
98719077 |
173 | my $status = $gz->read($_[0], $len) ; |
25f0751f |
174 | _save_gzerr($gz, 1); |
175 | return $status ; |
176 | } |
177 | |
178 | sub Compress::Zlib::gzFile::gzreadline |
179 | { |
180 | my $self = shift ; |
181 | |
182 | my $gz = $self->[0] ; |
183 | $_[0] = $gz->getline() ; |
184 | _save_gzerr($gz, 1); |
185 | return defined $_[0] ? length $_[0] : 0 ; |
186 | } |
187 | |
188 | sub Compress::Zlib::gzFile::gzwrite |
189 | { |
190 | my $self = shift ; |
191 | my $gz = $self->[0] ; |
192 | |
193 | return _set_gzerr(Z_STREAM_ERROR()) |
194 | if $self->[1] ne 'deflate'; |
195 | |
93d092e2 |
196 | $] >= 5.008 and (utf8::downgrade($_[0], 1) |
197 | or croak "Wide character in gzwrite"); |
198 | |
25f0751f |
199 | my $status = $gz->write($_[0]) ; |
200 | _save_gzerr($gz); |
201 | return $status ; |
202 | } |
203 | |
204 | sub Compress::Zlib::gzFile::gztell |
205 | { |
206 | my $self = shift ; |
207 | my $gz = $self->[0] ; |
208 | my $status = $gz->tell() ; |
209 | _save_gzerr($gz); |
210 | return $status ; |
211 | } |
212 | |
213 | sub Compress::Zlib::gzFile::gzseek |
214 | { |
215 | my $self = shift ; |
216 | my $offset = shift ; |
217 | my $whence = shift ; |
218 | |
219 | my $gz = $self->[0] ; |
220 | my $status ; |
221 | eval { $status = $gz->seek($offset, $whence) ; }; |
222 | if ($@) |
223 | { |
224 | my $error = $@; |
225 | $error =~ s/^.*: /gzseek: /; |
226 | $error =~ s/ at .* line \d+\s*$//; |
227 | croak $error; |
228 | } |
229 | _save_gzerr($gz); |
230 | return $status ; |
231 | } |
232 | |
233 | sub Compress::Zlib::gzFile::gzflush |
234 | { |
235 | my $self = shift ; |
236 | my $f = shift ; |
237 | |
238 | my $gz = $self->[0] ; |
239 | my $status = $gz->flush($f) ; |
240 | _save_gzerr($gz); |
241 | return $status ; |
242 | } |
243 | |
244 | sub Compress::Zlib::gzFile::gzclose |
245 | { |
246 | my $self = shift ; |
247 | my $gz = $self->[0] ; |
248 | |
249 | my $status = $gz->close() ; |
250 | _save_gzerr($gz); |
251 | return ! $status ; |
252 | } |
253 | |
254 | sub Compress::Zlib::gzFile::gzeof |
255 | { |
256 | my $self = shift ; |
257 | my $gz = $self->[0] ; |
258 | |
259 | return 0 |
260 | if $self->[1] ne 'inflate'; |
261 | |
262 | my $status = $gz->eof() ; |
263 | _save_gzerr($gz); |
264 | return $status ; |
265 | } |
266 | |
267 | sub Compress::Zlib::gzFile::gzsetparams |
268 | { |
269 | my $self = shift ; |
270 | croak "Usage: Compress::Zlib::gzFile::gzsetparams(file, level, strategy)" |
271 | unless @_ eq 2 ; |
272 | |
273 | my $gz = $self->[0] ; |
274 | my $level = shift ; |
275 | my $strategy = shift; |
276 | |
277 | return _set_gzerr(Z_STREAM_ERROR()) |
278 | if $self->[1] ne 'deflate'; |
279 | |
280 | my $status = *$gz->{Compress}->deflateParams(-Level => $level, |
281 | -Strategy => $strategy); |
282 | _save_gzerr($gz); |
283 | return $status ; |
284 | } |
285 | |
286 | sub Compress::Zlib::gzFile::gzerror |
287 | { |
288 | my $self = shift ; |
289 | my $gz = $self->[0] ; |
290 | |
291 | return $Compress::Zlib::gzerrno ; |
292 | } |
293 | |
294 | |
295 | sub compress($;$) |
296 | { |
297 | my ($x, $output, $err, $in) =('', '', '', '') ; |
298 | |
299 | if (ref $_[0] ) { |
300 | $in = $_[0] ; |
301 | croak "not a scalar reference" unless ref $in eq 'SCALAR' ; |
302 | } |
303 | else { |
304 | $in = \$_[0] ; |
305 | } |
306 | |
93d092e2 |
307 | $] >= 5.008 and (utf8::downgrade($$in, 1) |
308 | or croak "Wide character in compress"); |
309 | |
25f0751f |
310 | my $level = (@_ == 2 ? $_[1] : Z_DEFAULT_COMPRESSION() ); |
311 | |
312 | $x = new Compress::Raw::Zlib::Deflate -AppendOutput => 1, -Level => $level |
313 | or return undef ; |
314 | |
315 | $err = $x->deflate($in, $output) ; |
316 | return undef unless $err == Z_OK() ; |
317 | |
318 | $err = $x->flush($output) ; |
319 | return undef unless $err == Z_OK() ; |
320 | |
321 | return $output ; |
322 | |
323 | } |
324 | |
325 | sub uncompress($) |
326 | { |
327 | my ($x, $output, $err, $in) =('', '', '', '') ; |
328 | |
329 | if (ref $_[0] ) { |
330 | $in = $_[0] ; |
331 | croak "not a scalar reference" unless ref $in eq 'SCALAR' ; |
332 | } |
333 | else { |
334 | $in = \$_[0] ; |
335 | } |
336 | |
93d092e2 |
337 | $] >= 5.008 and (utf8::downgrade($$in, 1) |
338 | or croak "Wide character in uncompress"); |
339 | |
25f0751f |
340 | $x = new Compress::Raw::Zlib::Inflate -ConsumeInput => 0 or return undef ; |
341 | |
342 | $err = $x->inflate($in, $output) ; |
343 | return undef unless $err == Z_STREAM_END() ; |
344 | |
345 | return $output ; |
346 | } |
347 | |
348 | |
349 | |
350 | sub deflateInit(@) |
351 | { |
352 | my ($got) = ParseParameters(0, |
353 | { |
354 | 'Bufsize' => [1, 1, Parse_unsigned, 4096], |
355 | 'Level' => [1, 1, Parse_signed, Z_DEFAULT_COMPRESSION()], |
356 | 'Method' => [1, 1, Parse_unsigned, Z_DEFLATED()], |
357 | 'WindowBits' => [1, 1, Parse_signed, MAX_WBITS()], |
358 | 'MemLevel' => [1, 1, Parse_unsigned, MAX_MEM_LEVEL()], |
359 | 'Strategy' => [1, 1, Parse_unsigned, Z_DEFAULT_STRATEGY()], |
360 | 'Dictionary' => [1, 1, Parse_any, ""], |
361 | }, @_ ) ; |
362 | |
363 | croak "Compress::Zlib::deflateInit: Bufsize must be >= 1, you specified " . |
364 | $got->value('Bufsize') |
365 | unless $got->value('Bufsize') >= 1; |
366 | |
367 | my $obj ; |
368 | |
369 | my $status = 0 ; |
370 | ($obj, $status) = |
371 | Compress::Raw::Zlib::_deflateInit(0, |
372 | $got->value('Level'), |
373 | $got->value('Method'), |
374 | $got->value('WindowBits'), |
375 | $got->value('MemLevel'), |
376 | $got->value('Strategy'), |
377 | $got->value('Bufsize'), |
378 | $got->value('Dictionary')) ; |
379 | |
380 | my $x = ($status == Z_OK() ? bless $obj, "Zlib::OldDeflate" : undef) ; |
381 | return wantarray ? ($x, $status) : $x ; |
382 | } |
383 | |
384 | sub inflateInit(@) |
385 | { |
386 | my ($got) = ParseParameters(0, |
387 | { |
388 | 'Bufsize' => [1, 1, Parse_unsigned, 4096], |
389 | 'WindowBits' => [1, 1, Parse_signed, MAX_WBITS()], |
390 | 'Dictionary' => [1, 1, Parse_any, ""], |
391 | }, @_) ; |
392 | |
393 | |
394 | croak "Compress::Zlib::inflateInit: Bufsize must be >= 1, you specified " . |
395 | $got->value('Bufsize') |
396 | unless $got->value('Bufsize') >= 1; |
397 | |
398 | my $status = 0 ; |
399 | my $obj ; |
400 | ($obj, $status) = Compress::Raw::Zlib::_inflateInit(FLAG_CONSUME_INPUT, |
401 | $got->value('WindowBits'), |
402 | $got->value('Bufsize'), |
403 | $got->value('Dictionary')) ; |
404 | |
405 | my $x = ($status == Z_OK() ? bless $obj, "Zlib::OldInflate" : undef) ; |
406 | |
407 | wantarray ? ($x, $status) : $x ; |
408 | } |
409 | |
410 | package Zlib::OldDeflate ; |
411 | |
412 | our (@ISA); |
413 | @ISA = qw(Compress::Raw::Zlib::deflateStream); |
414 | |
415 | |
416 | sub deflate |
417 | { |
418 | my $self = shift ; |
419 | my $output ; |
420 | |
421 | my $status = $self->SUPER::deflate($_[0], $output) ; |
422 | wantarray ? ($output, $status) : $output ; |
423 | } |
424 | |
425 | sub flush |
426 | { |
427 | my $self = shift ; |
428 | my $output ; |
429 | my $flag = shift || Compress::Zlib::Z_FINISH(); |
430 | my $status = $self->SUPER::flush($output, $flag) ; |
431 | |
432 | wantarray ? ($output, $status) : $output ; |
433 | } |
434 | |
435 | package Zlib::OldInflate ; |
436 | |
437 | our (@ISA); |
438 | @ISA = qw(Compress::Raw::Zlib::inflateStream); |
439 | |
440 | sub inflate |
441 | { |
442 | my $self = shift ; |
443 | my $output ; |
444 | my $status = $self->SUPER::inflate($_[0], $output) ; |
445 | wantarray ? ($output, $status) : $output ; |
446 | } |
447 | |
448 | package Compress::Zlib ; |
449 | |
93d092e2 |
450 | use IO::Compress::Gzip::Constants 2.005 ; |
25f0751f |
451 | |
452 | sub memGzip($) |
453 | { |
4e7676c7 |
454 | my $out; |
455 | |
25f0751f |
456 | # if the deflation buffer isn't a reference, make it one |
457 | my $string = (ref $_[0] ? $_[0] : \$_[0]) ; |
458 | |
93d092e2 |
459 | $] >= 5.008 and (utf8::downgrade($$string, 1) |
460 | or croak "Wide character in memGzip"); |
461 | |
4e7676c7 |
462 | IO::Compress::Gzip::gzip($string, \$out, Minimal => 1) |
25f0751f |
463 | or return undef ; |
4e7676c7 |
464 | |
465 | return $out; |
25f0751f |
466 | } |
467 | |
468 | |
469 | sub _removeGzipHeader($) |
470 | { |
471 | my $string = shift ; |
472 | |
473 | return Z_DATA_ERROR() |
474 | if length($$string) < GZIP_MIN_HEADER_SIZE ; |
475 | |
476 | my ($magic1, $magic2, $method, $flags, $time, $xflags, $oscode) = |
477 | unpack ('CCCCVCC', $$string); |
478 | |
479 | return Z_DATA_ERROR() |
480 | unless $magic1 == GZIP_ID1 and $magic2 == GZIP_ID2 and |
481 | $method == Z_DEFLATED() and !($flags & GZIP_FLG_RESERVED) ; |
482 | substr($$string, 0, GZIP_MIN_HEADER_SIZE) = '' ; |
483 | |
484 | # skip extra field |
485 | if ($flags & GZIP_FLG_FEXTRA) |
486 | { |
487 | return Z_DATA_ERROR() |
488 | if length($$string) < GZIP_FEXTRA_HEADER_SIZE ; |
489 | |
490 | my ($extra_len) = unpack ('v', $$string); |
491 | $extra_len += GZIP_FEXTRA_HEADER_SIZE; |
492 | return Z_DATA_ERROR() |
493 | if length($$string) < $extra_len ; |
494 | |
495 | substr($$string, 0, $extra_len) = ''; |
496 | } |
497 | |
498 | # skip orig name |
499 | if ($flags & GZIP_FLG_FNAME) |
500 | { |
501 | my $name_end = index ($$string, GZIP_NULL_BYTE); |
502 | return Z_DATA_ERROR() |
503 | if $name_end == -1 ; |
504 | substr($$string, 0, $name_end + 1) = ''; |
505 | } |
506 | |
507 | # skip comment |
508 | if ($flags & GZIP_FLG_FCOMMENT) |
509 | { |
510 | my $comment_end = index ($$string, GZIP_NULL_BYTE); |
511 | return Z_DATA_ERROR() |
512 | if $comment_end == -1 ; |
513 | substr($$string, 0, $comment_end + 1) = ''; |
514 | } |
515 | |
516 | # skip header crc |
517 | if ($flags & GZIP_FLG_FHCRC) |
518 | { |
519 | return Z_DATA_ERROR() |
520 | if length ($$string) < GZIP_FHCRC_SIZE ; |
521 | substr($$string, 0, GZIP_FHCRC_SIZE) = ''; |
522 | } |
523 | |
524 | return Z_OK(); |
525 | } |
526 | |
527 | |
528 | sub memGunzip($) |
529 | { |
530 | # if the buffer isn't a reference, make it one |
531 | my $string = (ref $_[0] ? $_[0] : \$_[0]); |
532 | |
93d092e2 |
533 | $] >= 5.008 and (utf8::downgrade($$string, 1) |
534 | or croak "Wide character in memGunzip"); |
535 | |
25f0751f |
536 | _removeGzipHeader($string) == Z_OK() |
537 | or return undef; |
538 | |
539 | my $bufsize = length $$string > 4096 ? length $$string : 4096 ; |
540 | my $x = new Compress::Raw::Zlib::Inflate({-WindowBits => - MAX_WBITS(), |
541 | -Bufsize => $bufsize}) |
542 | |
543 | or return undef; |
544 | |
545 | my $output = "" ; |
546 | my $status = $x->inflate($string, $output); |
547 | return undef |
548 | unless $status == Z_STREAM_END(); |
549 | |
550 | if (length $$string >= 8) |
551 | { |
552 | my ($crc, $len) = unpack ("VV", substr($$string, 0, 8)); |
553 | substr($$string, 0, 8) = ''; |
554 | return undef |
555 | unless $len == length($output) and |
556 | $crc == crc32($output); |
557 | } |
558 | else |
559 | { |
560 | $$string = ''; |
561 | } |
562 | return $output; |
563 | } |
564 | |
565 | # Autoload methods go after __END__, and are processed by the autosplit program. |
566 | |
567 | 1; |
568 | __END__ |
569 | |
570 | |
571 | =head1 NAME |
572 | |
573 | Compress::Zlib - Interface to zlib compression library |
574 | |
575 | =head1 SYNOPSIS |
576 | |
577 | use Compress::Zlib ; |
578 | |
579 | ($d, $status) = deflateInit( [OPT] ) ; |
580 | $status = $d->deflate($input, $output) ; |
581 | $status = $d->flush($output [, $flush_type]) ; |
582 | $d->deflateParams(OPTS) ; |
583 | $d->deflateTune(OPTS) ; |
584 | $d->dict_adler() ; |
585 | $d->crc32() ; |
586 | $d->adler32() ; |
587 | $d->total_in() ; |
588 | $d->total_out() ; |
589 | $d->msg() ; |
590 | $d->get_Strategy(); |
591 | $d->get_Level(); |
592 | $d->get_BufSize(); |
593 | |
594 | ($i, $status) = inflateInit( [OPT] ) ; |
595 | $status = $i->inflate($input, $output [, $eof]) ; |
596 | $status = $i->inflateSync($input) ; |
597 | $i->dict_adler() ; |
598 | $d->crc32() ; |
599 | $d->adler32() ; |
600 | $i->total_in() ; |
601 | $i->total_out() ; |
602 | $i->msg() ; |
603 | $d->get_BufSize(); |
604 | |
605 | $dest = compress($source) ; |
606 | $dest = uncompress($source) ; |
607 | |
608 | $gz = gzopen($filename or filehandle, $mode) ; |
609 | $bytesread = $gz->gzread($buffer [,$size]) ; |
610 | $bytesread = $gz->gzreadline($line) ; |
611 | $byteswritten = $gz->gzwrite($buffer) ; |
612 | $status = $gz->gzflush($flush) ; |
613 | $offset = $gz->gztell() ; |
614 | $status = $gz->gzseek($offset, $whence) ; |
615 | $status = $gz->gzclose() ; |
616 | $status = $gz->gzeof() ; |
617 | $status = $gz->gzsetparams($level, $strategy) ; |
618 | $errstring = $gz->gzerror() ; |
619 | $gzerrno |
620 | |
621 | $dest = Compress::Zlib::memGzip($buffer) ; |
622 | $dest = Compress::Zlib::memGunzip($buffer) ; |
623 | |
624 | $crc = adler32($buffer [,$crc]) ; |
625 | $crc = crc32($buffer [,$crc]) ; |
626 | |
627 | $crc = adler32_combine($crc1, $crc2, $len2)l |
628 | $crc = crc32_combine($adler1, $adler2, $len2) |
629 | |
630 | ZLIB_VERSION |
631 | ZLIB_VERNUM |
632 | |
633 | |
634 | |
635 | =head1 DESCRIPTION |
636 | |
637 | The I<Compress::Zlib> module provides a Perl interface to the I<zlib> |
638 | compression library (see L</AUTHOR> for details about where to get |
639 | I<zlib>). |
640 | |
641 | The C<Compress::Zlib> module can be split into two general areas of |
642 | functionality, namely a simple read/write interface to I<gzip> files |
643 | and a low-level in-memory compression/decompression interface. |
644 | |
f6fd7794 |
645 | Each of these areas will be discussed in the following sections. |
25f0751f |
646 | |
647 | =head2 Notes for users of Compress::Zlib version 1 |
648 | |
f6fd7794 |
649 | The main change in C<Compress::Zlib> version 2.x is that it does not now |
650 | interface directly to the zlib library. Instead it uses the |
651 | C<IO::Compress::Gzip> and C<IO::Uncompress::Gunzip> modules for |
652 | reading/writing gzip files, and the C<Compress::Raw::Zlib> module for some |
653 | low-level zlib access. |
25f0751f |
654 | |
f6fd7794 |
655 | The interface provided by version 2 should be 100% backward compatible with |
656 | version 1. If you find a difference in the expected behaviour please |
657 | contact the author (See L</AUTHOR>). See L<GZIP INTERFACE> |
25f0751f |
658 | |
659 | If you are writing new code, your first port of call should be to use one |
660 | these new modules. |
661 | |
662 | =head1 GZIP INTERFACE |
663 | |
664 | A number of functions are supplied in I<zlib> for reading and writing |
665 | I<gzip> files that conform to RFC 1952. This module provides an interface |
666 | to most of them. |
667 | |
f6fd7794 |
668 | If you have previously used C<Compress::Zlib> 1.x, the following |
25f0751f |
669 | enhancements/changes have been made to the C<gzopen> interface: |
670 | |
671 | =over 5 |
672 | |
673 | =item 1 |
674 | |
675 | If you want to to open either STDIN or STDOUT with C<gzopen>, you can now |
676 | optionally use the special filename "C<->" as a synonym for C<\*STDIN> and |
677 | C<\*STDOUT>. |
678 | |
679 | =item 2 |
680 | |
681 | In C<Compress::Zlib> version 1.x, C<gzopen> used the zlib library to open |
682 | the underlying file. This made things especially tricky when a Perl |
683 | filehandle was passed to C<gzopen>. Behind the scenes the numeric C file |
684 | descriptor had to be extracted from the Perl filehandle and this passed to |
685 | the zlib library. |
686 | |
687 | Apart from being non-portable to some operating systems, this made it |
688 | difficult to use C<gzopen> in situations where you wanted to extract/create |
689 | a gzip data stream that is embedded in a larger file, without having to |
690 | resort to opening and closing the file multiple times. |
691 | |
692 | It also made it impossible to pass a perl filehandle that wasn't associated |
693 | with a real filesystem file, like, say, an C<IO::String>. |
694 | |
695 | In C<Compress::Zlib> version 2.x, the C<gzopen> interface has been |
696 | completely rewritten to use the L<IO::Compress::Gzip|IO::Compress::Gzip> |
697 | for writing gzip files and L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip> |
698 | for reading gzip files. None of the limitations mentioned above apply. |
699 | |
700 | =item 3 |
701 | |
702 | Addition of C<gzseek> to provide a restricted C<seek> interface. |
703 | |
704 | =item 4. |
705 | |
706 | Added C<gztell>. |
707 | |
708 | =back |
709 | |
710 | A more complete and flexible interface for reading/writing gzip |
e7d45986 |
711 | files/buffers is included with the module C<IO-Compress-Zlib>. See |
25f0751f |
712 | L<IO::Compress::Gzip|IO::Compress::Gzip> and |
713 | L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip> for more details. |
714 | |
715 | =over 5 |
716 | |
717 | =item B<$gz = gzopen($filename, $mode)> |
718 | |
719 | =item B<$gz = gzopen($filehandle, $mode)> |
720 | |
721 | This function opens either the I<gzip> file C<$filename> for reading or |
722 | writing or attaches to the opened filehandle, C<$filehandle>. |
723 | It returns an object on success and C<undef> on failure. |
724 | |
f6fd7794 |
725 | When writing a gzip file this interface will I<always> create the smallest |
25f0751f |
726 | possible gzip header (exactly 10 bytes). If you want greater control over |
f6fd7794 |
727 | what gets stored in the gzip header (like the original filename or a |
728 | comment) use L<IO::Compress::Gzip|IO::Compress::Gzip> instead. Similarly if |
729 | you want to read the contents of the gzip header use |
730 | L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip>. |
25f0751f |
731 | |
732 | The second parameter, C<$mode>, is used to specify whether the file is |
733 | opened for reading or writing and to optionally specify a compression |
734 | level and compression strategy when writing. The format of the C<$mode> |
735 | parameter is similar to the mode parameter to the 'C' function C<fopen>, |
f6fd7794 |
736 | so "rb" is used to open for reading, "wb" for writing and "ab" for |
737 | appending (writing at the end of the file). |
25f0751f |
738 | |
739 | To specify a compression level when writing, append a digit between 0 |
740 | and 9 to the mode string -- 0 means no compression and 9 means maximum |
741 | compression. |
742 | If no compression level is specified Z_DEFAULT_COMPRESSION is used. |
743 | |
744 | To specify the compression strategy when writing, append 'f' for filtered |
745 | data, 'h' for Huffman only compression, or 'R' for run-length encoding. |
746 | If no strategy is specified Z_DEFAULT_STRATEGY is used. |
747 | |
748 | So, for example, "wb9" means open for writing with the maximum compression |
749 | using the default strategy and "wb4R" means open for writing with compression |
750 | level 4 and run-length encoding. |
751 | |
752 | Refer to the I<zlib> documentation for the exact format of the C<$mode> |
753 | parameter. |
754 | |
25f0751f |
755 | =item B<$bytesread = $gz-E<gt>gzread($buffer [, $size]) ;> |
756 | |
757 | Reads C<$size> bytes from the compressed file into C<$buffer>. If |
758 | C<$size> is not specified, it will default to 4096. If the scalar |
759 | C<$buffer> is not large enough, it will be extended automatically. |
760 | |
761 | Returns the number of bytes actually read. On EOF it returns 0 and in |
762 | the case of an error, -1. |
763 | |
764 | =item B<$bytesread = $gz-E<gt>gzreadline($line) ;> |
765 | |
766 | Reads the next line from the compressed file into C<$line>. |
767 | |
768 | Returns the number of bytes actually read. On EOF it returns 0 and in |
769 | the case of an error, -1. |
770 | |
771 | It is legal to intermix calls to C<gzread> and C<gzreadline>. |
772 | |
773 | In addition, C<gzreadline> fully supports the use of of the variable C<$/> |
774 | (C<$INPUT_RECORD_SEPARATOR> or C<$RS> when C<English> is in use) to |
775 | determine what constitutes an end of line. Both paragraph mode and file |
776 | slurp mode are supported. |
777 | |
778 | |
779 | =item B<$byteswritten = $gz-E<gt>gzwrite($buffer) ;> |
780 | |
781 | Writes the contents of C<$buffer> to the compressed file. Returns the |
782 | number of bytes actually written, or 0 on error. |
783 | |
784 | =item B<$status = $gz-E<gt>gzflush($flush_type) ;> |
785 | |
786 | Flushes all pending output into the compressed file. |
787 | |
788 | This method takes an optional parameter, C<$flush_type>, that controls |
789 | how the flushing will be carried out. By default the C<$flush_type> |
790 | used is C<Z_FINISH>. Other valid values for C<$flush_type> are |
791 | C<Z_NO_FLUSH>, C<Z_SYNC_FLUSH>, C<Z_FULL_FLUSH> and C<Z_BLOCK>. It is |
792 | strongly recommended that you only set the C<flush_type> parameter if |
793 | you fully understand the implications of what it does - overuse of C<flush> |
794 | can seriously degrade the level of compression achieved. See the C<zlib> |
795 | documentation for details. |
796 | |
797 | Returns 1 on success, 0 on failure. |
798 | |
799 | |
800 | =item B<$offset = $gz-E<gt>gztell() ;> |
801 | |
802 | Returns the uncompressed file offset. |
803 | |
804 | =item B<$status = $gz-E<gt>gzseek($offset, $whence) ;> |
805 | |
806 | Provides a sub-set of the C<seek> functionality, with the restriction |
807 | that it is only legal to seek forward in the compressed file. |
808 | It is a fatal error to attempt to seek backward. |
809 | |
810 | When opened for writing, empty parts of the file will have NULL (0x00) |
811 | bytes written to them. |
812 | |
813 | The C<$whence> parameter should be one of SEEK_SET, SEEK_CUR or SEEK_END. |
814 | |
815 | Returns 1 on success, 0 on failure. |
816 | |
817 | =item B<$gz-E<gt>gzclose> |
818 | |
819 | Closes the compressed file. Any pending data is flushed to the file |
820 | before it is closed. |
821 | |
822 | Returns 1 on success, 0 on failure. |
823 | |
824 | =item B<$gz-E<gt>gzsetparams($level, $strategy> |
825 | |
826 | Change settings for the deflate stream C<$gz>. |
827 | |
828 | The list of the valid options is shown below. Options not specified |
829 | will remain unchanged. |
830 | |
831 | Note: This method is only available if you are running zlib 1.0.6 or better. |
832 | |
833 | =over 5 |
834 | |
835 | =item B<$level> |
836 | |
837 | Defines the compression level. Valid values are 0 through 9, |
838 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
839 | C<Z_DEFAULT_COMPRESSION>. |
840 | |
841 | =item B<$strategy> |
842 | |
843 | Defines the strategy used to tune the compression. The valid values are |
844 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
845 | |
846 | =back |
847 | |
848 | =item B<$gz-E<gt>gzerror> |
849 | |
850 | Returns the I<zlib> error message or number for the last operation |
851 | associated with C<$gz>. The return value will be the I<zlib> error |
852 | number when used in a numeric context and the I<zlib> error message |
853 | when used in a string context. The I<zlib> error number constants, |
854 | shown below, are available for use. |
855 | |
856 | Z_OK |
857 | Z_STREAM_END |
858 | Z_ERRNO |
859 | Z_STREAM_ERROR |
860 | Z_DATA_ERROR |
861 | Z_MEM_ERROR |
862 | Z_BUF_ERROR |
863 | |
864 | =item B<$gzerrno> |
865 | |
866 | The C<$gzerrno> scalar holds the error code associated with the most |
867 | recent I<gzip> routine. Note that unlike C<gzerror()>, the error is |
868 | I<not> associated with a particular file. |
869 | |
870 | As with C<gzerror()> it returns an error number in numeric context and |
871 | an error message in string context. Unlike C<gzerror()> though, the |
872 | error message will correspond to the I<zlib> message when the error is |
873 | associated with I<zlib> itself, or the UNIX error message when it is |
874 | not (i.e. I<zlib> returned C<Z_ERRORNO>). |
875 | |
876 | As there is an overlap between the error numbers used by I<zlib> and |
877 | UNIX, C<$gzerrno> should only be used to check for the presence of |
878 | I<an> error in numeric context. Use C<gzerror()> to check for specific |
879 | I<zlib> errors. The I<gzcat> example below shows how the variable can |
880 | be used safely. |
881 | |
882 | =back |
883 | |
884 | |
885 | =head2 Examples |
886 | |
887 | Here is an example script which uses the interface. It implements a |
888 | I<gzcat> function. |
889 | |
890 | use strict ; |
891 | use warnings ; |
892 | |
893 | use Compress::Zlib ; |
894 | |
895 | # use stdin if no files supplied |
896 | @ARGV = '-' unless @ARGV ; |
897 | |
898 | foreach my $file (@ARGV) { |
899 | my $buffer ; |
900 | |
901 | my $gz = gzopen($file, "rb") |
902 | or die "Cannot open $file: $gzerrno\n" ; |
903 | |
904 | print $buffer while $gz->gzread($buffer) > 0 ; |
905 | |
906 | die "Error reading from $file: $gzerrno" . ($gzerrno+0) . "\n" |
907 | if $gzerrno != Z_STREAM_END ; |
908 | |
909 | $gz->gzclose() ; |
910 | } |
911 | |
912 | Below is a script which makes use of C<gzreadline>. It implements a |
913 | very simple I<grep> like script. |
914 | |
915 | use strict ; |
916 | use warnings ; |
917 | |
918 | use Compress::Zlib ; |
919 | |
920 | die "Usage: gzgrep pattern [file...]\n" |
921 | unless @ARGV >= 1; |
922 | |
923 | my $pattern = shift ; |
924 | |
925 | # use stdin if no files supplied |
926 | @ARGV = '-' unless @ARGV ; |
927 | |
928 | foreach my $file (@ARGV) { |
929 | my $gz = gzopen($file, "rb") |
930 | or die "Cannot open $file: $gzerrno\n" ; |
931 | |
932 | while ($gz->gzreadline($_) > 0) { |
933 | print if /$pattern/ ; |
934 | } |
935 | |
936 | die "Error reading from $file: $gzerrno\n" |
937 | if $gzerrno != Z_STREAM_END ; |
938 | |
939 | $gz->gzclose() ; |
940 | } |
941 | |
942 | This script, I<gzstream>, does the opposite of the I<gzcat> script |
943 | above. It reads from standard input and writes a gzip data stream to |
944 | standard output. |
945 | |
946 | use strict ; |
947 | use warnings ; |
948 | |
949 | use Compress::Zlib ; |
950 | |
951 | binmode STDOUT; # gzopen only sets it on the fd |
952 | |
953 | my $gz = gzopen(\*STDOUT, "wb") |
954 | or die "Cannot open stdout: $gzerrno\n" ; |
955 | |
956 | while (<>) { |
957 | $gz->gzwrite($_) |
958 | or die "error writing: $gzerrno\n" ; |
959 | } |
960 | |
961 | $gz->gzclose ; |
962 | |
963 | =head2 Compress::Zlib::memGzip |
964 | |
965 | This function is used to create an in-memory gzip file with the minimum |
966 | possible gzip header (exactly 10 bytes). |
967 | |
968 | $dest = Compress::Zlib::memGzip($buffer) ; |
969 | |
970 | If successful, it returns the in-memory gzip file, otherwise it returns |
971 | undef. |
972 | |
973 | The C<$buffer> parameter can either be a scalar or a scalar reference. |
974 | |
975 | See L<IO::Compress::Gzip|IO::Compress::Gzip> for an alternative way to |
976 | carry out in-memory gzip compression. |
977 | |
978 | =head2 Compress::Zlib::memGunzip |
979 | |
980 | This function is used to uncompress an in-memory gzip file. |
981 | |
982 | $dest = Compress::Zlib::memGunzip($buffer) ; |
983 | |
984 | If successful, it returns the uncompressed gzip file, otherwise it |
985 | returns undef. |
986 | |
987 | The C<$buffer> parameter can either be a scalar or a scalar reference. The |
988 | contents of the C<$buffer> parameter are destroyed after calling this function. |
989 | |
990 | See L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip> for an alternative way |
991 | to carry out in-memory gzip uncompression. |
992 | |
993 | =head1 COMPRESS/UNCOMPRESS |
994 | |
995 | Two functions are provided to perform in-memory compression/uncompression of |
996 | RFC 1950 data streams. They are called C<compress> and C<uncompress>. |
997 | |
998 | =over 5 |
999 | |
1000 | =item B<$dest = compress($source [, $level] ) ;> |
1001 | |
1002 | Compresses C<$source>. If successful it returns the compressed |
1003 | data. Otherwise it returns I<undef>. |
1004 | |
1005 | The source buffer, C<$source>, can either be a scalar or a scalar |
1006 | reference. |
1007 | |
1008 | The C<$level> parameter defines the compression level. Valid values are |
1009 | 0 through 9, C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, |
1010 | C<Z_BEST_COMPRESSION>, and C<Z_DEFAULT_COMPRESSION>. |
1011 | If C<$level> is not specified C<Z_DEFAULT_COMPRESSION> will be used. |
1012 | |
1013 | |
1014 | =item B<$dest = uncompress($source) ;> |
1015 | |
1016 | Uncompresses C<$source>. If successful it returns the uncompressed |
1017 | data. Otherwise it returns I<undef>. |
1018 | |
1019 | The source buffer can either be a scalar or a scalar reference. |
1020 | |
1021 | =back |
1022 | |
1023 | Please note: the two functions defined above are I<not> compatible with |
1024 | the Unix commands of the same name. |
1025 | |
1026 | See L<IO::Deflate|IO::Deflate> and L<IO::Inflate|IO::Inflate> included with |
1027 | this distribution for an alternative interface for reading/writing RFC 1950 |
1028 | files/buffers. |
1029 | |
1030 | |
1031 | =head1 Deflate Interface |
1032 | |
1033 | This section defines an interface that allows in-memory compression using |
1034 | the I<deflate> interface provided by zlib. |
1035 | |
1036 | Here is a definition of the interface available: |
1037 | |
1038 | |
1039 | =head2 B<($d, $status) = deflateInit( [OPT] )> |
1040 | |
1041 | Initialises a deflation stream. |
1042 | |
1043 | It combines the features of the I<zlib> functions C<deflateInit>, |
1044 | C<deflateInit2> and C<deflateSetDictionary>. |
1045 | |
1046 | If successful, it will return the initialised deflation stream, C<$d> |
1047 | and C<$status> of C<Z_OK> in a list context. In scalar context it |
1048 | returns the deflation stream, C<$d>, only. |
1049 | |
1050 | If not successful, the returned deflation stream (C<$d>) will be |
1051 | I<undef> and C<$status> will hold the exact I<zlib> error code. |
1052 | |
1053 | The function optionally takes a number of named options specified as |
e7d45986 |
1054 | C<< -Name=>value >> pairs. This allows individual options to be |
25f0751f |
1055 | tailored without having to specify them all in the parameter list. |
1056 | |
1057 | For backward compatibility, it is also possible to pass the parameters |
1058 | as a reference to a hash containing the name=>value pairs. |
1059 | |
1060 | The function takes one optional parameter, a reference to a hash. The |
1061 | contents of the hash allow the deflation interface to be tailored. |
1062 | |
1063 | Here is a list of the valid options: |
1064 | |
1065 | =over 5 |
1066 | |
1067 | =item B<-Level> |
1068 | |
1069 | Defines the compression level. Valid values are 0 through 9, |
1070 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
1071 | C<Z_DEFAULT_COMPRESSION>. |
1072 | |
e7d45986 |
1073 | The default is Z_DEFAULT_COMPRESSION. |
25f0751f |
1074 | |
1075 | =item B<-Method> |
1076 | |
1077 | Defines the compression method. The only valid value at present (and |
e7d45986 |
1078 | the default) is Z_DEFLATED. |
25f0751f |
1079 | |
1080 | =item B<-WindowBits> |
1081 | |
1082 | To create an RFC 1950 data stream, set C<WindowBits> to a positive number. |
1083 | |
1084 | To create an RFC 1951 data stream, set C<WindowBits> to C<-MAX_WBITS>. |
1085 | |
1086 | For a full definition of the meaning and valid values for C<WindowBits> refer |
1087 | to the I<zlib> documentation for I<deflateInit2>. |
1088 | |
e7d45986 |
1089 | Defaults to MAX_WBITS. |
25f0751f |
1090 | |
1091 | =item B<-MemLevel> |
1092 | |
1093 | For a definition of the meaning and valid values for C<MemLevel> |
1094 | refer to the I<zlib> documentation for I<deflateInit2>. |
1095 | |
e7d45986 |
1096 | Defaults to MAX_MEM_LEVEL. |
25f0751f |
1097 | |
1098 | =item B<-Strategy> |
1099 | |
1100 | Defines the strategy used to tune the compression. The valid values are |
1101 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
1102 | |
e7d45986 |
1103 | The default is Z_DEFAULT_STRATEGY. |
25f0751f |
1104 | |
1105 | =item B<-Dictionary> |
1106 | |
1107 | When a dictionary is specified I<Compress::Zlib> will automatically |
1108 | call C<deflateSetDictionary> directly after calling C<deflateInit>. The |
1109 | Adler32 value for the dictionary can be obtained by calling the method |
1110 | C<$d->dict_adler()>. |
1111 | |
1112 | The default is no dictionary. |
1113 | |
1114 | =item B<-Bufsize> |
1115 | |
1116 | Sets the initial size for the deflation buffer. If the buffer has to be |
1117 | reallocated to increase the size, it will grow in increments of |
1118 | C<Bufsize>. |
1119 | |
1120 | The default is 4096. |
1121 | |
1122 | =back |
1123 | |
1124 | Here is an example of using the C<deflateInit> optional parameter list |
1125 | to override the default buffer size and compression level. All other |
1126 | options will take their default values. |
1127 | |
1128 | deflateInit( -Bufsize => 300, |
1129 | -Level => Z_BEST_SPEED ) ; |
1130 | |
1131 | |
1132 | =head2 B<($out, $status) = $d-E<gt>deflate($buffer)> |
1133 | |
1134 | |
1135 | Deflates the contents of C<$buffer>. The buffer can either be a scalar |
1136 | or a scalar reference. When finished, C<$buffer> will be |
1137 | completely processed (assuming there were no errors). If the deflation |
1138 | was successful it returns the deflated output, C<$out>, and a status |
1139 | value, C<$status>, of C<Z_OK>. |
1140 | |
1141 | On error, C<$out> will be I<undef> and C<$status> will contain the |
1142 | I<zlib> error code. |
1143 | |
1144 | In a scalar context C<deflate> will return C<$out> only. |
1145 | |
1146 | As with the I<deflate> function in I<zlib>, it is not necessarily the |
1147 | case that any output will be produced by this method. So don't rely on |
1148 | the fact that C<$out> is empty for an error test. |
1149 | |
1150 | |
1151 | =head2 B<($out, $status) = $d-E<gt>flush([flush_type])> |
1152 | |
1153 | Typically used to finish the deflation. Any pending output will be |
1154 | returned via C<$out>. |
1155 | C<$status> will have a value C<Z_OK> if successful. |
1156 | |
1157 | In a scalar context C<flush> will return C<$out> only. |
1158 | |
1159 | Note that flushing can seriously degrade the compression ratio, so it |
1160 | should only be used to terminate a decompression (using C<Z_FINISH>) or |
1161 | when you want to create a I<full flush point> (using C<Z_FULL_FLUSH>). |
1162 | |
1163 | By default the C<flush_type> used is C<Z_FINISH>. Other valid values |
1164 | for C<flush_type> are C<Z_NO_FLUSH>, C<Z_PARTIAL_FLUSH>, C<Z_SYNC_FLUSH> |
1165 | and C<Z_FULL_FLUSH>. It is strongly recommended that you only set the |
1166 | C<flush_type> parameter if you fully understand the implications of |
1167 | what it does. See the C<zlib> documentation for details. |
1168 | |
1169 | =head2 B<$status = $d-E<gt>deflateParams([OPT])> |
1170 | |
1171 | Change settings for the deflate stream C<$d>. |
1172 | |
1173 | The list of the valid options is shown below. Options not specified |
1174 | will remain unchanged. |
1175 | |
1176 | =over 5 |
1177 | |
1178 | =item B<-Level> |
1179 | |
1180 | Defines the compression level. Valid values are 0 through 9, |
1181 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
1182 | C<Z_DEFAULT_COMPRESSION>. |
1183 | |
1184 | =item B<-Strategy> |
1185 | |
1186 | Defines the strategy used to tune the compression. The valid values are |
1187 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
1188 | |
1189 | =back |
1190 | |
1191 | =head2 B<$d-E<gt>dict_adler()> |
1192 | |
1193 | Returns the adler32 value for the dictionary. |
1194 | |
1195 | =head2 B<$d-E<gt>msg()> |
1196 | |
1197 | Returns the last error message generated by zlib. |
1198 | |
1199 | =head2 B<$d-E<gt>total_in()> |
1200 | |
1201 | Returns the total number of bytes uncompressed bytes input to deflate. |
1202 | |
1203 | =head2 B<$d-E<gt>total_out()> |
1204 | |
1205 | Returns the total number of compressed bytes output from deflate. |
1206 | |
1207 | =head2 Example |
1208 | |
1209 | |
1210 | Here is a trivial example of using C<deflate>. It simply reads standard |
1211 | input, deflates it and writes it to standard output. |
1212 | |
1213 | use strict ; |
1214 | use warnings ; |
1215 | |
1216 | use Compress::Zlib ; |
1217 | |
1218 | binmode STDIN; |
1219 | binmode STDOUT; |
1220 | my $x = deflateInit() |
1221 | or die "Cannot create a deflation stream\n" ; |
1222 | |
1223 | my ($output, $status) ; |
1224 | while (<>) |
1225 | { |
1226 | ($output, $status) = $x->deflate($_) ; |
1227 | |
1228 | $status == Z_OK |
1229 | or die "deflation failed\n" ; |
1230 | |
1231 | print $output ; |
1232 | } |
1233 | |
1234 | ($output, $status) = $x->flush() ; |
1235 | |
1236 | $status == Z_OK |
1237 | or die "deflation failed\n" ; |
1238 | |
1239 | print $output ; |
1240 | |
1241 | =head1 Inflate Interface |
1242 | |
1243 | This section defines the interface available that allows in-memory |
1244 | uncompression using the I<deflate> interface provided by zlib. |
1245 | |
1246 | Here is a definition of the interface: |
1247 | |
1248 | |
1249 | =head2 B<($i, $status) = inflateInit()> |
1250 | |
1251 | Initialises an inflation stream. |
1252 | |
1253 | In a list context it returns the inflation stream, C<$i>, and the |
1254 | I<zlib> status code in C<$status>. In a scalar context it returns the |
1255 | inflation stream only. |
1256 | |
1257 | If successful, C<$i> will hold the inflation stream and C<$status> will |
1258 | be C<Z_OK>. |
1259 | |
1260 | If not successful, C<$i> will be I<undef> and C<$status> will hold the |
1261 | I<zlib> error code. |
1262 | |
1263 | The function optionally takes a number of named options specified as |
e7d45986 |
1264 | C<< -Name=>value >> pairs. This allows individual options to be |
25f0751f |
1265 | tailored without having to specify them all in the parameter list. |
1266 | |
1267 | For backward compatibility, it is also possible to pass the parameters |
1268 | as a reference to a hash containing the name=>value pairs. |
1269 | |
1270 | The function takes one optional parameter, a reference to a hash. The |
1271 | contents of the hash allow the deflation interface to be tailored. |
1272 | |
1273 | Here is a list of the valid options: |
1274 | |
1275 | =over 5 |
1276 | |
1277 | =item B<-WindowBits> |
1278 | |
1279 | To uncompress an RFC 1950 data stream, set C<WindowBits> to a positive number. |
1280 | |
1281 | To uncompress an RFC 1951 data stream, set C<WindowBits> to C<-MAX_WBITS>. |
1282 | |
1283 | For a full definition of the meaning and valid values for C<WindowBits> refer |
1284 | to the I<zlib> documentation for I<inflateInit2>. |
1285 | |
e7d45986 |
1286 | Defaults to MAX_WBITS. |
25f0751f |
1287 | |
1288 | =item B<-Bufsize> |
1289 | |
1290 | Sets the initial size for the inflation buffer. If the buffer has to be |
1291 | reallocated to increase the size, it will grow in increments of |
1292 | C<Bufsize>. |
1293 | |
1294 | Default is 4096. |
1295 | |
1296 | =item B<-Dictionary> |
1297 | |
1298 | The default is no dictionary. |
1299 | |
1300 | =back |
1301 | |
1302 | Here is an example of using the C<inflateInit> optional parameter to |
1303 | override the default buffer size. |
1304 | |
1305 | inflateInit( -Bufsize => 300 ) ; |
1306 | |
1307 | =head2 B<($out, $status) = $i-E<gt>inflate($buffer)> |
1308 | |
1309 | Inflates the complete contents of C<$buffer>. The buffer can either be |
1310 | a scalar or a scalar reference. |
1311 | |
1312 | Returns C<Z_OK> if successful and C<Z_STREAM_END> if the end of the |
1313 | compressed data has been successfully reached. |
1314 | If not successful, C<$out> will be I<undef> and C<$status> will hold |
1315 | the I<zlib> error code. |
1316 | |
1317 | The C<$buffer> parameter is modified by C<inflate>. On completion it |
1318 | will contain what remains of the input buffer after inflation. This |
1319 | means that C<$buffer> will be an empty string when the return status is |
1320 | C<Z_OK>. When the return status is C<Z_STREAM_END> the C<$buffer> |
1321 | parameter will contains what (if anything) was stored in the input |
1322 | buffer after the deflated data stream. |
1323 | |
1324 | This feature is useful when processing a file format that encapsulates |
1325 | a compressed data stream (e.g. gzip, zip). |
1326 | |
1327 | =head2 B<$status = $i-E<gt>inflateSync($buffer)> |
1328 | |
1329 | Scans C<$buffer> until it reaches either a I<full flush point> or the |
1330 | end of the buffer. |
1331 | |
1332 | If a I<full flush point> is found, C<Z_OK> is returned and C<$buffer> |
1333 | will be have all data up to the flush point removed. This can then be |
1334 | passed to the C<deflate> method. |
1335 | |
1336 | Any other return code means that a flush point was not found. If more |
1337 | data is available, C<inflateSync> can be called repeatedly with more |
1338 | compressed data until the flush point is found. |
1339 | |
1340 | |
1341 | =head2 B<$i-E<gt>dict_adler()> |
1342 | |
1343 | Returns the adler32 value for the dictionary. |
1344 | |
1345 | =head2 B<$i-E<gt>msg()> |
1346 | |
1347 | Returns the last error message generated by zlib. |
1348 | |
1349 | =head2 B<$i-E<gt>total_in()> |
1350 | |
1351 | Returns the total number of bytes compressed bytes input to inflate. |
1352 | |
1353 | =head2 B<$i-E<gt>total_out()> |
1354 | |
1355 | Returns the total number of uncompressed bytes output from inflate. |
1356 | |
1357 | =head2 Example |
1358 | |
1359 | Here is an example of using C<inflate>. |
1360 | |
1361 | use strict ; |
1362 | use warnings ; |
1363 | |
1364 | use Compress::Zlib ; |
1365 | |
1366 | my $x = inflateInit() |
1367 | or die "Cannot create a inflation stream\n" ; |
1368 | |
1369 | my $input = '' ; |
1370 | binmode STDIN; |
1371 | binmode STDOUT; |
1372 | |
1373 | my ($output, $status) ; |
1374 | while (read(STDIN, $input, 4096)) |
1375 | { |
1376 | ($output, $status) = $x->inflate(\$input) ; |
1377 | |
1378 | print $output |
1379 | if $status == Z_OK or $status == Z_STREAM_END ; |
1380 | |
1381 | last if $status != Z_OK ; |
1382 | } |
1383 | |
1384 | die "inflation failed\n" |
1385 | unless $status == Z_STREAM_END ; |
1386 | |
1387 | =head1 CHECKSUM FUNCTIONS |
1388 | |
1389 | Two functions are provided by I<zlib> to calculate checksums. For the |
1390 | Perl interface, the order of the two parameters in both functions has |
1391 | been reversed. This allows both running checksums and one off |
1392 | calculations to be done. |
1393 | |
1394 | $crc = adler32($buffer [,$crc]) ; |
1395 | $crc = crc32($buffer [,$crc]) ; |
1396 | |
1397 | The buffer parameters can either be a scalar or a scalar reference. |
1398 | |
1399 | If the $crc parameters is C<undef>, the crc value will be reset. |
1400 | |
1401 | If you have built this module with zlib 1.2.3 or better, two more |
1402 | CRC-related functions are available. |
1403 | |
1404 | $crc = adler32_combine($crc1, $crc2, $len2)l |
1405 | $crc = crc32_combine($adler1, $adler2, $len2) |
1406 | |
1407 | These functions allow checksums to be merged. |
1408 | |
25f0751f |
1409 | =head1 CONSTANTS |
1410 | |
1411 | All the I<zlib> constants are automatically imported when you make use |
1412 | of I<Compress::Zlib>. |
1413 | |
1414 | |
1415 | =head1 SEE ALSO |
1416 | |
258133d1 |
1417 | L<IO::Compress::Gzip>, L<IO::Uncompress::Gunzip>, L<IO::Compress::Deflate>, L<IO::Uncompress::Inflate>, L<IO::Compress::RawDeflate>, L<IO::Uncompress::RawInflate>, L<IO::Compress::Bzip2>, L<IO::Uncompress::Bunzip2>, L<IO::Compress::Lzop>, L<IO::Uncompress::UnLzop>, L<IO::Compress::Lzf>, L<IO::Uncompress::UnLzf>, L<IO::Uncompress::AnyInflate>, L<IO::Uncompress::AnyUncompress> |
25f0751f |
1418 | |
1419 | L<Compress::Zlib::FAQ|Compress::Zlib::FAQ> |
1420 | |
1421 | L<File::GlobMapper|File::GlobMapper>, L<Archive::Zip|Archive::Zip>, |
1422 | L<Archive::Tar|Archive::Tar>, |
1423 | L<IO::Zlib|IO::Zlib> |
1424 | |
1425 | |
1426 | For RFC 1950, 1951 and 1952 see |
1427 | F<http://www.faqs.org/rfcs/rfc1950.html>, |
1428 | F<http://www.faqs.org/rfcs/rfc1951.html> and |
1429 | F<http://www.faqs.org/rfcs/rfc1952.html> |
1430 | |
1431 | The I<zlib> compression library was written by Jean-loup Gailly |
1432 | F<gzip@prep.ai.mit.edu> and Mark Adler F<madler@alumni.caltech.edu>. |
1433 | |
1434 | The primary site for the I<zlib> compression library is |
1435 | F<http://www.zlib.org>. |
1436 | |
1437 | The primary site for gzip is F<http://www.gzip.org>. |
1438 | |
1439 | |
1440 | |
1441 | |
25f0751f |
1442 | =head1 AUTHOR |
1443 | |
cb7abd7f |
1444 | This module was written by Paul Marquess, F<pmqs@cpan.org>. |
25f0751f |
1445 | |
1446 | |
1447 | |
1448 | =head1 MODIFICATION HISTORY |
1449 | |
1450 | See the Changes file. |
1451 | |
1452 | =head1 COPYRIGHT AND LICENSE |
25f0751f |
1453 | |
b0cda13f |
1454 | Copyright (c) 1995-2007 Paul Marquess. All rights reserved. |
25f0751f |
1455 | |
1456 | This program is free software; you can redistribute it and/or |
1457 | modify it under the same terms as Perl itself. |
1458 | |
1459 | |
1460 | |
93d092e2 |
1461 | |