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 | |
11 | use IO::Compress::Base::Common ; |
12 | use Compress::Raw::Zlib; |
13 | use IO::Compress::Gzip; |
14 | use IO::Uncompress::Gunzip; |
15 | |
16 | use strict ; |
17 | use warnings ; |
18 | use bytes ; |
19 | our ($VERSION, $XS_VERSION, @ISA, @EXPORT, $AUTOLOAD); |
20 | |
c70c1701 |
21 | $VERSION = '2.000_12'; |
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; |
123 | |
124 | my $infDef = $writing ? 'deflate' : 'inflate'; |
125 | my @params = () ; |
126 | |
127 | croak "gzopen: file parameter is not a filehandle or filename" |
128 | unless isaFilehandle $file || isaFilename $file ; |
129 | |
130 | return undef unless $mode =~ /[rwa]/i ; |
131 | |
132 | _set_gzerr(0) ; |
133 | |
134 | if ($writing) { |
135 | $gz = new IO::Compress::Gzip($file, Minimal => 1, AutoClose => 1, |
136 | %defOpts) |
137 | or $Compress::Zlib::gzerrno = $IO::Compress::Gzip::GzipError; |
138 | } |
139 | else { |
140 | $gz = new IO::Uncompress::Gunzip($file, |
141 | Transparent => 1, |
142 | Append => 0, |
143 | AutoClose => 1, |
144 | Strict => 0) |
145 | or $Compress::Zlib::gzerrno = $IO::Uncompress::Gunzip::GunzipError; |
146 | } |
147 | |
148 | return undef |
149 | if ! defined $gz ; |
150 | |
151 | bless [$gz, $infDef], 'Compress::Zlib::gzFile'; |
152 | } |
153 | |
154 | sub Compress::Zlib::gzFile::gzread |
155 | { |
156 | my $self = shift ; |
157 | |
158 | return _set_gzerr(Z_STREAM_ERROR()) |
159 | if $self->[1] ne 'inflate'; |
160 | |
98719077 |
161 | my $len = defined $_[1] ? $_[1] : 4096 ; |
162 | |
163 | if ($self->gzeof() || $len == 0) { |
cb7abd7f |
164 | # Zap the output buffer to match ver 1 behaviour. |
165 | $_[0] = "" ; |
166 | return 0 ; |
167 | } |
25f0751f |
168 | |
169 | my $gz = $self->[0] ; |
98719077 |
170 | my $status = $gz->read($_[0], $len) ; |
25f0751f |
171 | _save_gzerr($gz, 1); |
172 | return $status ; |
173 | } |
174 | |
175 | sub Compress::Zlib::gzFile::gzreadline |
176 | { |
177 | my $self = shift ; |
178 | |
179 | my $gz = $self->[0] ; |
180 | $_[0] = $gz->getline() ; |
181 | _save_gzerr($gz, 1); |
182 | return defined $_[0] ? length $_[0] : 0 ; |
183 | } |
184 | |
185 | sub Compress::Zlib::gzFile::gzwrite |
186 | { |
187 | my $self = shift ; |
188 | my $gz = $self->[0] ; |
189 | |
190 | return _set_gzerr(Z_STREAM_ERROR()) |
191 | if $self->[1] ne 'deflate'; |
192 | |
193 | my $status = $gz->write($_[0]) ; |
194 | _save_gzerr($gz); |
195 | return $status ; |
196 | } |
197 | |
198 | sub Compress::Zlib::gzFile::gztell |
199 | { |
200 | my $self = shift ; |
201 | my $gz = $self->[0] ; |
202 | my $status = $gz->tell() ; |
203 | _save_gzerr($gz); |
204 | return $status ; |
205 | } |
206 | |
207 | sub Compress::Zlib::gzFile::gzseek |
208 | { |
209 | my $self = shift ; |
210 | my $offset = shift ; |
211 | my $whence = shift ; |
212 | |
213 | my $gz = $self->[0] ; |
214 | my $status ; |
215 | eval { $status = $gz->seek($offset, $whence) ; }; |
216 | if ($@) |
217 | { |
218 | my $error = $@; |
219 | $error =~ s/^.*: /gzseek: /; |
220 | $error =~ s/ at .* line \d+\s*$//; |
221 | croak $error; |
222 | } |
223 | _save_gzerr($gz); |
224 | return $status ; |
225 | } |
226 | |
227 | sub Compress::Zlib::gzFile::gzflush |
228 | { |
229 | my $self = shift ; |
230 | my $f = shift ; |
231 | |
232 | my $gz = $self->[0] ; |
233 | my $status = $gz->flush($f) ; |
234 | _save_gzerr($gz); |
235 | return $status ; |
236 | } |
237 | |
238 | sub Compress::Zlib::gzFile::gzclose |
239 | { |
240 | my $self = shift ; |
241 | my $gz = $self->[0] ; |
242 | |
243 | my $status = $gz->close() ; |
244 | _save_gzerr($gz); |
245 | return ! $status ; |
246 | } |
247 | |
248 | sub Compress::Zlib::gzFile::gzeof |
249 | { |
250 | my $self = shift ; |
251 | my $gz = $self->[0] ; |
252 | |
253 | return 0 |
254 | if $self->[1] ne 'inflate'; |
255 | |
256 | my $status = $gz->eof() ; |
257 | _save_gzerr($gz); |
258 | return $status ; |
259 | } |
260 | |
261 | sub Compress::Zlib::gzFile::gzsetparams |
262 | { |
263 | my $self = shift ; |
264 | croak "Usage: Compress::Zlib::gzFile::gzsetparams(file, level, strategy)" |
265 | unless @_ eq 2 ; |
266 | |
267 | my $gz = $self->[0] ; |
268 | my $level = shift ; |
269 | my $strategy = shift; |
270 | |
271 | return _set_gzerr(Z_STREAM_ERROR()) |
272 | if $self->[1] ne 'deflate'; |
273 | |
274 | my $status = *$gz->{Compress}->deflateParams(-Level => $level, |
275 | -Strategy => $strategy); |
276 | _save_gzerr($gz); |
277 | return $status ; |
278 | } |
279 | |
280 | sub Compress::Zlib::gzFile::gzerror |
281 | { |
282 | my $self = shift ; |
283 | my $gz = $self->[0] ; |
284 | |
285 | return $Compress::Zlib::gzerrno ; |
286 | } |
287 | |
288 | |
289 | sub compress($;$) |
290 | { |
291 | my ($x, $output, $err, $in) =('', '', '', '') ; |
292 | |
293 | if (ref $_[0] ) { |
294 | $in = $_[0] ; |
295 | croak "not a scalar reference" unless ref $in eq 'SCALAR' ; |
296 | } |
297 | else { |
298 | $in = \$_[0] ; |
299 | } |
300 | |
301 | my $level = (@_ == 2 ? $_[1] : Z_DEFAULT_COMPRESSION() ); |
302 | |
303 | $x = new Compress::Raw::Zlib::Deflate -AppendOutput => 1, -Level => $level |
304 | or return undef ; |
305 | |
306 | $err = $x->deflate($in, $output) ; |
307 | return undef unless $err == Z_OK() ; |
308 | |
309 | $err = $x->flush($output) ; |
310 | return undef unless $err == Z_OK() ; |
311 | |
312 | return $output ; |
313 | |
314 | } |
315 | |
316 | sub uncompress($) |
317 | { |
318 | my ($x, $output, $err, $in) =('', '', '', '') ; |
319 | |
320 | if (ref $_[0] ) { |
321 | $in = $_[0] ; |
322 | croak "not a scalar reference" unless ref $in eq 'SCALAR' ; |
323 | } |
324 | else { |
325 | $in = \$_[0] ; |
326 | } |
327 | |
328 | $x = new Compress::Raw::Zlib::Inflate -ConsumeInput => 0 or return undef ; |
329 | |
330 | $err = $x->inflate($in, $output) ; |
331 | return undef unless $err == Z_STREAM_END() ; |
332 | |
333 | return $output ; |
334 | } |
335 | |
336 | |
337 | |
338 | sub deflateInit(@) |
339 | { |
340 | my ($got) = ParseParameters(0, |
341 | { |
342 | 'Bufsize' => [1, 1, Parse_unsigned, 4096], |
343 | 'Level' => [1, 1, Parse_signed, Z_DEFAULT_COMPRESSION()], |
344 | 'Method' => [1, 1, Parse_unsigned, Z_DEFLATED()], |
345 | 'WindowBits' => [1, 1, Parse_signed, MAX_WBITS()], |
346 | 'MemLevel' => [1, 1, Parse_unsigned, MAX_MEM_LEVEL()], |
347 | 'Strategy' => [1, 1, Parse_unsigned, Z_DEFAULT_STRATEGY()], |
348 | 'Dictionary' => [1, 1, Parse_any, ""], |
349 | }, @_ ) ; |
350 | |
351 | croak "Compress::Zlib::deflateInit: Bufsize must be >= 1, you specified " . |
352 | $got->value('Bufsize') |
353 | unless $got->value('Bufsize') >= 1; |
354 | |
355 | my $obj ; |
356 | |
357 | my $status = 0 ; |
358 | ($obj, $status) = |
359 | Compress::Raw::Zlib::_deflateInit(0, |
360 | $got->value('Level'), |
361 | $got->value('Method'), |
362 | $got->value('WindowBits'), |
363 | $got->value('MemLevel'), |
364 | $got->value('Strategy'), |
365 | $got->value('Bufsize'), |
366 | $got->value('Dictionary')) ; |
367 | |
368 | my $x = ($status == Z_OK() ? bless $obj, "Zlib::OldDeflate" : undef) ; |
369 | return wantarray ? ($x, $status) : $x ; |
370 | } |
371 | |
372 | sub inflateInit(@) |
373 | { |
374 | my ($got) = ParseParameters(0, |
375 | { |
376 | 'Bufsize' => [1, 1, Parse_unsigned, 4096], |
377 | 'WindowBits' => [1, 1, Parse_signed, MAX_WBITS()], |
378 | 'Dictionary' => [1, 1, Parse_any, ""], |
379 | }, @_) ; |
380 | |
381 | |
382 | croak "Compress::Zlib::inflateInit: Bufsize must be >= 1, you specified " . |
383 | $got->value('Bufsize') |
384 | unless $got->value('Bufsize') >= 1; |
385 | |
386 | my $status = 0 ; |
387 | my $obj ; |
388 | ($obj, $status) = Compress::Raw::Zlib::_inflateInit(FLAG_CONSUME_INPUT, |
389 | $got->value('WindowBits'), |
390 | $got->value('Bufsize'), |
391 | $got->value('Dictionary')) ; |
392 | |
393 | my $x = ($status == Z_OK() ? bless $obj, "Zlib::OldInflate" : undef) ; |
394 | |
395 | wantarray ? ($x, $status) : $x ; |
396 | } |
397 | |
398 | package Zlib::OldDeflate ; |
399 | |
400 | our (@ISA); |
401 | @ISA = qw(Compress::Raw::Zlib::deflateStream); |
402 | |
403 | |
404 | sub deflate |
405 | { |
406 | my $self = shift ; |
407 | my $output ; |
408 | |
409 | my $status = $self->SUPER::deflate($_[0], $output) ; |
410 | wantarray ? ($output, $status) : $output ; |
411 | } |
412 | |
413 | sub flush |
414 | { |
415 | my $self = shift ; |
416 | my $output ; |
417 | my $flag = shift || Compress::Zlib::Z_FINISH(); |
418 | my $status = $self->SUPER::flush($output, $flag) ; |
419 | |
420 | wantarray ? ($output, $status) : $output ; |
421 | } |
422 | |
423 | package Zlib::OldInflate ; |
424 | |
425 | our (@ISA); |
426 | @ISA = qw(Compress::Raw::Zlib::inflateStream); |
427 | |
428 | sub inflate |
429 | { |
430 | my $self = shift ; |
431 | my $output ; |
432 | my $status = $self->SUPER::inflate($_[0], $output) ; |
433 | wantarray ? ($output, $status) : $output ; |
434 | } |
435 | |
436 | package Compress::Zlib ; |
437 | |
438 | use IO::Compress::Gzip::Constants; |
439 | |
440 | sub memGzip($) |
441 | { |
442 | my $x = new Compress::Raw::Zlib::Deflate( |
443 | -AppendOutput => 1, |
444 | -CRC32 => 1, |
445 | -ADLER32 => 0, |
446 | -Level => Z_BEST_COMPRESSION(), |
447 | -WindowBits => - MAX_WBITS(), |
448 | ) |
449 | or return undef ; |
450 | |
451 | # write a minimal gzip header |
452 | my $output = GZIP_MINIMUM_HEADER ; |
453 | |
454 | # if the deflation buffer isn't a reference, make it one |
455 | my $string = (ref $_[0] ? $_[0] : \$_[0]) ; |
456 | |
457 | my $status = $x->deflate($string, \$output) ; |
458 | $status == Z_OK() |
459 | or return undef ; |
460 | |
461 | $status = $x->flush(\$output) ; |
462 | $status == Z_OK() |
463 | or return undef ; |
464 | |
465 | return $output . pack("V V", $x->crc32(), $x->total_in()) ; |
466 | |
467 | } |
468 | |
469 | |
470 | sub _removeGzipHeader($) |
471 | { |
472 | my $string = shift ; |
473 | |
474 | return Z_DATA_ERROR() |
475 | if length($$string) < GZIP_MIN_HEADER_SIZE ; |
476 | |
477 | my ($magic1, $magic2, $method, $flags, $time, $xflags, $oscode) = |
478 | unpack ('CCCCVCC', $$string); |
479 | |
480 | return Z_DATA_ERROR() |
481 | unless $magic1 == GZIP_ID1 and $magic2 == GZIP_ID2 and |
482 | $method == Z_DEFLATED() and !($flags & GZIP_FLG_RESERVED) ; |
483 | substr($$string, 0, GZIP_MIN_HEADER_SIZE) = '' ; |
484 | |
485 | # skip extra field |
486 | if ($flags & GZIP_FLG_FEXTRA) |
487 | { |
488 | return Z_DATA_ERROR() |
489 | if length($$string) < GZIP_FEXTRA_HEADER_SIZE ; |
490 | |
491 | my ($extra_len) = unpack ('v', $$string); |
492 | $extra_len += GZIP_FEXTRA_HEADER_SIZE; |
493 | return Z_DATA_ERROR() |
494 | if length($$string) < $extra_len ; |
495 | |
496 | substr($$string, 0, $extra_len) = ''; |
497 | } |
498 | |
499 | # skip orig name |
500 | if ($flags & GZIP_FLG_FNAME) |
501 | { |
502 | my $name_end = index ($$string, GZIP_NULL_BYTE); |
503 | return Z_DATA_ERROR() |
504 | if $name_end == -1 ; |
505 | substr($$string, 0, $name_end + 1) = ''; |
506 | } |
507 | |
508 | # skip comment |
509 | if ($flags & GZIP_FLG_FCOMMENT) |
510 | { |
511 | my $comment_end = index ($$string, GZIP_NULL_BYTE); |
512 | return Z_DATA_ERROR() |
513 | if $comment_end == -1 ; |
514 | substr($$string, 0, $comment_end + 1) = ''; |
515 | } |
516 | |
517 | # skip header crc |
518 | if ($flags & GZIP_FLG_FHCRC) |
519 | { |
520 | return Z_DATA_ERROR() |
521 | if length ($$string) < GZIP_FHCRC_SIZE ; |
522 | substr($$string, 0, GZIP_FHCRC_SIZE) = ''; |
523 | } |
524 | |
525 | return Z_OK(); |
526 | } |
527 | |
528 | |
529 | sub memGunzip($) |
530 | { |
531 | # if the buffer isn't a reference, make it one |
532 | my $string = (ref $_[0] ? $_[0] : \$_[0]); |
533 | |
534 | _removeGzipHeader($string) == Z_OK() |
535 | or return undef; |
536 | |
537 | my $bufsize = length $$string > 4096 ? length $$string : 4096 ; |
538 | my $x = new Compress::Raw::Zlib::Inflate({-WindowBits => - MAX_WBITS(), |
539 | -Bufsize => $bufsize}) |
540 | |
541 | or return undef; |
542 | |
543 | my $output = "" ; |
544 | my $status = $x->inflate($string, $output); |
545 | return undef |
546 | unless $status == Z_STREAM_END(); |
547 | |
548 | if (length $$string >= 8) |
549 | { |
550 | my ($crc, $len) = unpack ("VV", substr($$string, 0, 8)); |
551 | substr($$string, 0, 8) = ''; |
552 | return undef |
553 | unless $len == length($output) and |
554 | $crc == crc32($output); |
555 | } |
556 | else |
557 | { |
558 | $$string = ''; |
559 | } |
560 | return $output; |
561 | } |
562 | |
563 | # Autoload methods go after __END__, and are processed by the autosplit program. |
564 | |
565 | 1; |
566 | __END__ |
567 | |
568 | |
569 | =head1 NAME |
570 | |
571 | Compress::Zlib - Interface to zlib compression library |
572 | |
573 | =head1 SYNOPSIS |
574 | |
575 | use Compress::Zlib ; |
576 | |
577 | ($d, $status) = deflateInit( [OPT] ) ; |
578 | $status = $d->deflate($input, $output) ; |
579 | $status = $d->flush($output [, $flush_type]) ; |
580 | $d->deflateParams(OPTS) ; |
581 | $d->deflateTune(OPTS) ; |
582 | $d->dict_adler() ; |
583 | $d->crc32() ; |
584 | $d->adler32() ; |
585 | $d->total_in() ; |
586 | $d->total_out() ; |
587 | $d->msg() ; |
588 | $d->get_Strategy(); |
589 | $d->get_Level(); |
590 | $d->get_BufSize(); |
591 | |
592 | ($i, $status) = inflateInit( [OPT] ) ; |
593 | $status = $i->inflate($input, $output [, $eof]) ; |
594 | $status = $i->inflateSync($input) ; |
595 | $i->dict_adler() ; |
596 | $d->crc32() ; |
597 | $d->adler32() ; |
598 | $i->total_in() ; |
599 | $i->total_out() ; |
600 | $i->msg() ; |
601 | $d->get_BufSize(); |
602 | |
603 | $dest = compress($source) ; |
604 | $dest = uncompress($source) ; |
605 | |
606 | $gz = gzopen($filename or filehandle, $mode) ; |
607 | $bytesread = $gz->gzread($buffer [,$size]) ; |
608 | $bytesread = $gz->gzreadline($line) ; |
609 | $byteswritten = $gz->gzwrite($buffer) ; |
610 | $status = $gz->gzflush($flush) ; |
611 | $offset = $gz->gztell() ; |
612 | $status = $gz->gzseek($offset, $whence) ; |
613 | $status = $gz->gzclose() ; |
614 | $status = $gz->gzeof() ; |
615 | $status = $gz->gzsetparams($level, $strategy) ; |
616 | $errstring = $gz->gzerror() ; |
617 | $gzerrno |
618 | |
619 | $dest = Compress::Zlib::memGzip($buffer) ; |
620 | $dest = Compress::Zlib::memGunzip($buffer) ; |
621 | |
622 | $crc = adler32($buffer [,$crc]) ; |
623 | $crc = crc32($buffer [,$crc]) ; |
624 | |
625 | $crc = adler32_combine($crc1, $crc2, $len2)l |
626 | $crc = crc32_combine($adler1, $adler2, $len2) |
627 | |
628 | ZLIB_VERSION |
629 | ZLIB_VERNUM |
630 | |
631 | |
632 | |
633 | =head1 DESCRIPTION |
634 | |
635 | The I<Compress::Zlib> module provides a Perl interface to the I<zlib> |
636 | compression library (see L</AUTHOR> for details about where to get |
637 | I<zlib>). |
638 | |
639 | The C<Compress::Zlib> module can be split into two general areas of |
640 | functionality, namely a simple read/write interface to I<gzip> files |
641 | and a low-level in-memory compression/decompression interface. |
642 | |
643 | Each of these areas will be discussed separately below. |
644 | |
645 | =head2 Notes for users of Compress::Zlib version 1 |
646 | |
647 | Version 2 of this module is a total rewrite. The net result of this is that |
648 | C<Compress::Zlib> does not now access the zlib library directly. |
649 | |
650 | It now uses the C<IO::Compress::Gzip> and C<IO::Uncompress::Gunzip> modules |
651 | for reading/writing gzip files, and the C<Compress::Raw::Zlib> module for |
652 | low-level zlib access. |
653 | |
654 | If you are writing new code, your first port of call should be to use one |
655 | these new modules. |
656 | |
657 | =head1 GZIP INTERFACE |
658 | |
659 | A number of functions are supplied in I<zlib> for reading and writing |
660 | I<gzip> files that conform to RFC 1952. This module provides an interface |
661 | to most of them. |
662 | |
663 | If you are upgrading from C<Compress::Zlib> 1.x, the following |
664 | enhancements/changes have been made to the C<gzopen> interface: |
665 | |
666 | =over 5 |
667 | |
668 | =item 1 |
669 | |
670 | If you want to to open either STDIN or STDOUT with C<gzopen>, you can now |
671 | optionally use the special filename "C<->" as a synonym for C<\*STDIN> and |
672 | C<\*STDOUT>. |
673 | |
674 | =item 2 |
675 | |
676 | In C<Compress::Zlib> version 1.x, C<gzopen> used the zlib library to open |
677 | the underlying file. This made things especially tricky when a Perl |
678 | filehandle was passed to C<gzopen>. Behind the scenes the numeric C file |
679 | descriptor had to be extracted from the Perl filehandle and this passed to |
680 | the zlib library. |
681 | |
682 | Apart from being non-portable to some operating systems, this made it |
683 | difficult to use C<gzopen> in situations where you wanted to extract/create |
684 | a gzip data stream that is embedded in a larger file, without having to |
685 | resort to opening and closing the file multiple times. |
686 | |
687 | It also made it impossible to pass a perl filehandle that wasn't associated |
688 | with a real filesystem file, like, say, an C<IO::String>. |
689 | |
690 | In C<Compress::Zlib> version 2.x, the C<gzopen> interface has been |
691 | completely rewritten to use the L<IO::Compress::Gzip|IO::Compress::Gzip> |
692 | for writing gzip files and L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip> |
693 | for reading gzip files. None of the limitations mentioned above apply. |
694 | |
695 | =item 3 |
696 | |
697 | Addition of C<gzseek> to provide a restricted C<seek> interface. |
698 | |
699 | =item 4. |
700 | |
701 | Added C<gztell>. |
702 | |
703 | =back |
704 | |
705 | A more complete and flexible interface for reading/writing gzip |
706 | files/buffers is included with the module C<IO-Compress-ZLib>. See |
707 | L<IO::Compress::Gzip|IO::Compress::Gzip> and |
708 | L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip> for more details. |
709 | |
710 | =over 5 |
711 | |
712 | =item B<$gz = gzopen($filename, $mode)> |
713 | |
714 | =item B<$gz = gzopen($filehandle, $mode)> |
715 | |
716 | This function opens either the I<gzip> file C<$filename> for reading or |
717 | writing or attaches to the opened filehandle, C<$filehandle>. |
718 | It returns an object on success and C<undef> on failure. |
719 | |
720 | When writing a gzip file this interface will always create the smallest |
721 | possible gzip header (exactly 10 bytes). If you want greater control over |
722 | the information stored in the gzip header (like the original filename or a |
723 | comment) use L<IO::Compress::Gzip|IO::Compress::Gzip> instead. |
724 | |
725 | The second parameter, C<$mode>, is used to specify whether the file is |
726 | opened for reading or writing and to optionally specify a compression |
727 | level and compression strategy when writing. The format of the C<$mode> |
728 | parameter is similar to the mode parameter to the 'C' function C<fopen>, |
729 | so "rb" is used to open for reading and "wb" for writing. |
730 | |
731 | To specify a compression level when writing, append a digit between 0 |
732 | and 9 to the mode string -- 0 means no compression and 9 means maximum |
733 | compression. |
734 | If no compression level is specified Z_DEFAULT_COMPRESSION is used. |
735 | |
736 | To specify the compression strategy when writing, append 'f' for filtered |
737 | data, 'h' for Huffman only compression, or 'R' for run-length encoding. |
738 | If no strategy is specified Z_DEFAULT_STRATEGY is used. |
739 | |
740 | So, for example, "wb9" means open for writing with the maximum compression |
741 | using the default strategy and "wb4R" means open for writing with compression |
742 | level 4 and run-length encoding. |
743 | |
744 | Refer to the I<zlib> documentation for the exact format of the C<$mode> |
745 | parameter. |
746 | |
747 | |
748 | =item B<$bytesread = $gz-E<gt>gzread($buffer [, $size]) ;> |
749 | |
750 | Reads C<$size> bytes from the compressed file into C<$buffer>. If |
751 | C<$size> is not specified, it will default to 4096. If the scalar |
752 | C<$buffer> is not large enough, it will be extended automatically. |
753 | |
754 | Returns the number of bytes actually read. On EOF it returns 0 and in |
755 | the case of an error, -1. |
756 | |
757 | =item B<$bytesread = $gz-E<gt>gzreadline($line) ;> |
758 | |
759 | Reads the next line from the compressed file into C<$line>. |
760 | |
761 | Returns the number of bytes actually read. On EOF it returns 0 and in |
762 | the case of an error, -1. |
763 | |
764 | It is legal to intermix calls to C<gzread> and C<gzreadline>. |
765 | |
766 | In addition, C<gzreadline> fully supports the use of of the variable C<$/> |
767 | (C<$INPUT_RECORD_SEPARATOR> or C<$RS> when C<English> is in use) to |
768 | determine what constitutes an end of line. Both paragraph mode and file |
769 | slurp mode are supported. |
770 | |
771 | |
772 | =item B<$byteswritten = $gz-E<gt>gzwrite($buffer) ;> |
773 | |
774 | Writes the contents of C<$buffer> to the compressed file. Returns the |
775 | number of bytes actually written, or 0 on error. |
776 | |
777 | =item B<$status = $gz-E<gt>gzflush($flush_type) ;> |
778 | |
779 | Flushes all pending output into the compressed file. |
780 | |
781 | This method takes an optional parameter, C<$flush_type>, that controls |
782 | how the flushing will be carried out. By default the C<$flush_type> |
783 | used is C<Z_FINISH>. Other valid values for C<$flush_type> are |
784 | C<Z_NO_FLUSH>, C<Z_SYNC_FLUSH>, C<Z_FULL_FLUSH> and C<Z_BLOCK>. It is |
785 | strongly recommended that you only set the C<flush_type> parameter if |
786 | you fully understand the implications of what it does - overuse of C<flush> |
787 | can seriously degrade the level of compression achieved. See the C<zlib> |
788 | documentation for details. |
789 | |
790 | Returns 1 on success, 0 on failure. |
791 | |
792 | |
793 | =item B<$offset = $gz-E<gt>gztell() ;> |
794 | |
795 | Returns the uncompressed file offset. |
796 | |
797 | =item B<$status = $gz-E<gt>gzseek($offset, $whence) ;> |
798 | |
799 | Provides a sub-set of the C<seek> functionality, with the restriction |
800 | that it is only legal to seek forward in the compressed file. |
801 | It is a fatal error to attempt to seek backward. |
802 | |
803 | When opened for writing, empty parts of the file will have NULL (0x00) |
804 | bytes written to them. |
805 | |
806 | The C<$whence> parameter should be one of SEEK_SET, SEEK_CUR or SEEK_END. |
807 | |
808 | Returns 1 on success, 0 on failure. |
809 | |
810 | =item B<$gz-E<gt>gzclose> |
811 | |
812 | Closes the compressed file. Any pending data is flushed to the file |
813 | before it is closed. |
814 | |
815 | Returns 1 on success, 0 on failure. |
816 | |
817 | =item B<$gz-E<gt>gzsetparams($level, $strategy> |
818 | |
819 | Change settings for the deflate stream C<$gz>. |
820 | |
821 | The list of the valid options is shown below. Options not specified |
822 | will remain unchanged. |
823 | |
824 | Note: This method is only available if you are running zlib 1.0.6 or better. |
825 | |
826 | =over 5 |
827 | |
828 | =item B<$level> |
829 | |
830 | Defines the compression level. Valid values are 0 through 9, |
831 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
832 | C<Z_DEFAULT_COMPRESSION>. |
833 | |
834 | =item B<$strategy> |
835 | |
836 | Defines the strategy used to tune the compression. The valid values are |
837 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
838 | |
839 | =back |
840 | |
841 | =item B<$gz-E<gt>gzerror> |
842 | |
843 | Returns the I<zlib> error message or number for the last operation |
844 | associated with C<$gz>. The return value will be the I<zlib> error |
845 | number when used in a numeric context and the I<zlib> error message |
846 | when used in a string context. The I<zlib> error number constants, |
847 | shown below, are available for use. |
848 | |
849 | Z_OK |
850 | Z_STREAM_END |
851 | Z_ERRNO |
852 | Z_STREAM_ERROR |
853 | Z_DATA_ERROR |
854 | Z_MEM_ERROR |
855 | Z_BUF_ERROR |
856 | |
857 | =item B<$gzerrno> |
858 | |
859 | The C<$gzerrno> scalar holds the error code associated with the most |
860 | recent I<gzip> routine. Note that unlike C<gzerror()>, the error is |
861 | I<not> associated with a particular file. |
862 | |
863 | As with C<gzerror()> it returns an error number in numeric context and |
864 | an error message in string context. Unlike C<gzerror()> though, the |
865 | error message will correspond to the I<zlib> message when the error is |
866 | associated with I<zlib> itself, or the UNIX error message when it is |
867 | not (i.e. I<zlib> returned C<Z_ERRORNO>). |
868 | |
869 | As there is an overlap between the error numbers used by I<zlib> and |
870 | UNIX, C<$gzerrno> should only be used to check for the presence of |
871 | I<an> error in numeric context. Use C<gzerror()> to check for specific |
872 | I<zlib> errors. The I<gzcat> example below shows how the variable can |
873 | be used safely. |
874 | |
875 | =back |
876 | |
877 | |
878 | =head2 Examples |
879 | |
880 | Here is an example script which uses the interface. It implements a |
881 | I<gzcat> function. |
882 | |
883 | use strict ; |
884 | use warnings ; |
885 | |
886 | use Compress::Zlib ; |
887 | |
888 | # use stdin if no files supplied |
889 | @ARGV = '-' unless @ARGV ; |
890 | |
891 | foreach my $file (@ARGV) { |
892 | my $buffer ; |
893 | |
894 | my $gz = gzopen($file, "rb") |
895 | or die "Cannot open $file: $gzerrno\n" ; |
896 | |
897 | print $buffer while $gz->gzread($buffer) > 0 ; |
898 | |
899 | die "Error reading from $file: $gzerrno" . ($gzerrno+0) . "\n" |
900 | if $gzerrno != Z_STREAM_END ; |
901 | |
902 | $gz->gzclose() ; |
903 | } |
904 | |
905 | Below is a script which makes use of C<gzreadline>. It implements a |
906 | very simple I<grep> like script. |
907 | |
908 | use strict ; |
909 | use warnings ; |
910 | |
911 | use Compress::Zlib ; |
912 | |
913 | die "Usage: gzgrep pattern [file...]\n" |
914 | unless @ARGV >= 1; |
915 | |
916 | my $pattern = shift ; |
917 | |
918 | # use stdin if no files supplied |
919 | @ARGV = '-' unless @ARGV ; |
920 | |
921 | foreach my $file (@ARGV) { |
922 | my $gz = gzopen($file, "rb") |
923 | or die "Cannot open $file: $gzerrno\n" ; |
924 | |
925 | while ($gz->gzreadline($_) > 0) { |
926 | print if /$pattern/ ; |
927 | } |
928 | |
929 | die "Error reading from $file: $gzerrno\n" |
930 | if $gzerrno != Z_STREAM_END ; |
931 | |
932 | $gz->gzclose() ; |
933 | } |
934 | |
935 | This script, I<gzstream>, does the opposite of the I<gzcat> script |
936 | above. It reads from standard input and writes a gzip data stream to |
937 | standard output. |
938 | |
939 | use strict ; |
940 | use warnings ; |
941 | |
942 | use Compress::Zlib ; |
943 | |
944 | binmode STDOUT; # gzopen only sets it on the fd |
945 | |
946 | my $gz = gzopen(\*STDOUT, "wb") |
947 | or die "Cannot open stdout: $gzerrno\n" ; |
948 | |
949 | while (<>) { |
950 | $gz->gzwrite($_) |
951 | or die "error writing: $gzerrno\n" ; |
952 | } |
953 | |
954 | $gz->gzclose ; |
955 | |
956 | =head2 Compress::Zlib::memGzip |
957 | |
958 | This function is used to create an in-memory gzip file with the minimum |
959 | possible gzip header (exactly 10 bytes). |
960 | |
961 | $dest = Compress::Zlib::memGzip($buffer) ; |
962 | |
963 | If successful, it returns the in-memory gzip file, otherwise it returns |
964 | undef. |
965 | |
966 | The C<$buffer> parameter can either be a scalar or a scalar reference. |
967 | |
968 | See L<IO::Compress::Gzip|IO::Compress::Gzip> for an alternative way to |
969 | carry out in-memory gzip compression. |
970 | |
971 | =head2 Compress::Zlib::memGunzip |
972 | |
973 | This function is used to uncompress an in-memory gzip file. |
974 | |
975 | $dest = Compress::Zlib::memGunzip($buffer) ; |
976 | |
977 | If successful, it returns the uncompressed gzip file, otherwise it |
978 | returns undef. |
979 | |
980 | The C<$buffer> parameter can either be a scalar or a scalar reference. The |
981 | contents of the C<$buffer> parameter are destroyed after calling this function. |
982 | |
983 | See L<IO::Uncompress::Gunzip|IO::Uncompress::Gunzip> for an alternative way |
984 | to carry out in-memory gzip uncompression. |
985 | |
986 | =head1 COMPRESS/UNCOMPRESS |
987 | |
988 | Two functions are provided to perform in-memory compression/uncompression of |
989 | RFC 1950 data streams. They are called C<compress> and C<uncompress>. |
990 | |
991 | =over 5 |
992 | |
993 | =item B<$dest = compress($source [, $level] ) ;> |
994 | |
995 | Compresses C<$source>. If successful it returns the compressed |
996 | data. Otherwise it returns I<undef>. |
997 | |
998 | The source buffer, C<$source>, can either be a scalar or a scalar |
999 | reference. |
1000 | |
1001 | The C<$level> parameter defines the compression level. Valid values are |
1002 | 0 through 9, C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, |
1003 | C<Z_BEST_COMPRESSION>, and C<Z_DEFAULT_COMPRESSION>. |
1004 | If C<$level> is not specified C<Z_DEFAULT_COMPRESSION> will be used. |
1005 | |
1006 | |
1007 | =item B<$dest = uncompress($source) ;> |
1008 | |
1009 | Uncompresses C<$source>. If successful it returns the uncompressed |
1010 | data. Otherwise it returns I<undef>. |
1011 | |
1012 | The source buffer can either be a scalar or a scalar reference. |
1013 | |
1014 | =back |
1015 | |
1016 | Please note: the two functions defined above are I<not> compatible with |
1017 | the Unix commands of the same name. |
1018 | |
1019 | See L<IO::Deflate|IO::Deflate> and L<IO::Inflate|IO::Inflate> included with |
1020 | this distribution for an alternative interface for reading/writing RFC 1950 |
1021 | files/buffers. |
1022 | |
1023 | |
1024 | =head1 Deflate Interface |
1025 | |
1026 | This section defines an interface that allows in-memory compression using |
1027 | the I<deflate> interface provided by zlib. |
1028 | |
1029 | Here is a definition of the interface available: |
1030 | |
1031 | |
1032 | =head2 B<($d, $status) = deflateInit( [OPT] )> |
1033 | |
1034 | Initialises a deflation stream. |
1035 | |
1036 | It combines the features of the I<zlib> functions C<deflateInit>, |
1037 | C<deflateInit2> and C<deflateSetDictionary>. |
1038 | |
1039 | If successful, it will return the initialised deflation stream, C<$d> |
1040 | and C<$status> of C<Z_OK> in a list context. In scalar context it |
1041 | returns the deflation stream, C<$d>, only. |
1042 | |
1043 | If not successful, the returned deflation stream (C<$d>) will be |
1044 | I<undef> and C<$status> will hold the exact I<zlib> error code. |
1045 | |
1046 | The function optionally takes a number of named options specified as |
1047 | C<-Name=E<gt>value> pairs. This allows individual options to be |
1048 | tailored without having to specify them all in the parameter list. |
1049 | |
1050 | For backward compatibility, it is also possible to pass the parameters |
1051 | as a reference to a hash containing the name=>value pairs. |
1052 | |
1053 | The function takes one optional parameter, a reference to a hash. The |
1054 | contents of the hash allow the deflation interface to be tailored. |
1055 | |
1056 | Here is a list of the valid options: |
1057 | |
1058 | =over 5 |
1059 | |
1060 | =item B<-Level> |
1061 | |
1062 | Defines the compression level. Valid values are 0 through 9, |
1063 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
1064 | C<Z_DEFAULT_COMPRESSION>. |
1065 | |
1066 | The default is C<-Level =E<gt>Z_DEFAULT_COMPRESSION>. |
1067 | |
1068 | =item B<-Method> |
1069 | |
1070 | Defines the compression method. The only valid value at present (and |
1071 | the default) is C<-Method =E<gt>Z_DEFLATED>. |
1072 | |
1073 | =item B<-WindowBits> |
1074 | |
1075 | To create an RFC 1950 data stream, set C<WindowBits> to a positive number. |
1076 | |
1077 | To create an RFC 1951 data stream, set C<WindowBits> to C<-MAX_WBITS>. |
1078 | |
1079 | For a full definition of the meaning and valid values for C<WindowBits> refer |
1080 | to the I<zlib> documentation for I<deflateInit2>. |
1081 | |
1082 | Defaults to C<-WindowBits =E<gt>MAX_WBITS>. |
1083 | |
1084 | =item B<-MemLevel> |
1085 | |
1086 | For a definition of the meaning and valid values for C<MemLevel> |
1087 | refer to the I<zlib> documentation for I<deflateInit2>. |
1088 | |
1089 | Defaults to C<-MemLevel =E<gt>MAX_MEM_LEVEL>. |
1090 | |
1091 | =item B<-Strategy> |
1092 | |
1093 | Defines the strategy used to tune the compression. The valid values are |
1094 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
1095 | |
1096 | The default is C<-Strategy =E<gt>Z_DEFAULT_STRATEGY>. |
1097 | |
1098 | =item B<-Dictionary> |
1099 | |
1100 | When a dictionary is specified I<Compress::Zlib> will automatically |
1101 | call C<deflateSetDictionary> directly after calling C<deflateInit>. The |
1102 | Adler32 value for the dictionary can be obtained by calling the method |
1103 | C<$d->dict_adler()>. |
1104 | |
1105 | The default is no dictionary. |
1106 | |
1107 | =item B<-Bufsize> |
1108 | |
1109 | Sets the initial size for the deflation buffer. If the buffer has to be |
1110 | reallocated to increase the size, it will grow in increments of |
1111 | C<Bufsize>. |
1112 | |
1113 | The default is 4096. |
1114 | |
1115 | =back |
1116 | |
1117 | Here is an example of using the C<deflateInit> optional parameter list |
1118 | to override the default buffer size and compression level. All other |
1119 | options will take their default values. |
1120 | |
1121 | deflateInit( -Bufsize => 300, |
1122 | -Level => Z_BEST_SPEED ) ; |
1123 | |
1124 | |
1125 | =head2 B<($out, $status) = $d-E<gt>deflate($buffer)> |
1126 | |
1127 | |
1128 | Deflates the contents of C<$buffer>. The buffer can either be a scalar |
1129 | or a scalar reference. When finished, C<$buffer> will be |
1130 | completely processed (assuming there were no errors). If the deflation |
1131 | was successful it returns the deflated output, C<$out>, and a status |
1132 | value, C<$status>, of C<Z_OK>. |
1133 | |
1134 | On error, C<$out> will be I<undef> and C<$status> will contain the |
1135 | I<zlib> error code. |
1136 | |
1137 | In a scalar context C<deflate> will return C<$out> only. |
1138 | |
1139 | As with the I<deflate> function in I<zlib>, it is not necessarily the |
1140 | case that any output will be produced by this method. So don't rely on |
1141 | the fact that C<$out> is empty for an error test. |
1142 | |
1143 | |
1144 | =head2 B<($out, $status) = $d-E<gt>flush([flush_type])> |
1145 | |
1146 | Typically used to finish the deflation. Any pending output will be |
1147 | returned via C<$out>. |
1148 | C<$status> will have a value C<Z_OK> if successful. |
1149 | |
1150 | In a scalar context C<flush> will return C<$out> only. |
1151 | |
1152 | Note that flushing can seriously degrade the compression ratio, so it |
1153 | should only be used to terminate a decompression (using C<Z_FINISH>) or |
1154 | when you want to create a I<full flush point> (using C<Z_FULL_FLUSH>). |
1155 | |
1156 | By default the C<flush_type> used is C<Z_FINISH>. Other valid values |
1157 | for C<flush_type> are C<Z_NO_FLUSH>, C<Z_PARTIAL_FLUSH>, C<Z_SYNC_FLUSH> |
1158 | and C<Z_FULL_FLUSH>. It is strongly recommended that you only set the |
1159 | C<flush_type> parameter if you fully understand the implications of |
1160 | what it does. See the C<zlib> documentation for details. |
1161 | |
1162 | =head2 B<$status = $d-E<gt>deflateParams([OPT])> |
1163 | |
1164 | Change settings for the deflate stream C<$d>. |
1165 | |
1166 | The list of the valid options is shown below. Options not specified |
1167 | will remain unchanged. |
1168 | |
1169 | =over 5 |
1170 | |
1171 | =item B<-Level> |
1172 | |
1173 | Defines the compression level. Valid values are 0 through 9, |
1174 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
1175 | C<Z_DEFAULT_COMPRESSION>. |
1176 | |
1177 | =item B<-Strategy> |
1178 | |
1179 | Defines the strategy used to tune the compression. The valid values are |
1180 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
1181 | |
1182 | =back |
1183 | |
1184 | =head2 B<$d-E<gt>dict_adler()> |
1185 | |
1186 | Returns the adler32 value for the dictionary. |
1187 | |
1188 | =head2 B<$d-E<gt>msg()> |
1189 | |
1190 | Returns the last error message generated by zlib. |
1191 | |
1192 | =head2 B<$d-E<gt>total_in()> |
1193 | |
1194 | Returns the total number of bytes uncompressed bytes input to deflate. |
1195 | |
1196 | =head2 B<$d-E<gt>total_out()> |
1197 | |
1198 | Returns the total number of compressed bytes output from deflate. |
1199 | |
1200 | =head2 Example |
1201 | |
1202 | |
1203 | Here is a trivial example of using C<deflate>. It simply reads standard |
1204 | input, deflates it and writes it to standard output. |
1205 | |
1206 | use strict ; |
1207 | use warnings ; |
1208 | |
1209 | use Compress::Zlib ; |
1210 | |
1211 | binmode STDIN; |
1212 | binmode STDOUT; |
1213 | my $x = deflateInit() |
1214 | or die "Cannot create a deflation stream\n" ; |
1215 | |
1216 | my ($output, $status) ; |
1217 | while (<>) |
1218 | { |
1219 | ($output, $status) = $x->deflate($_) ; |
1220 | |
1221 | $status == Z_OK |
1222 | or die "deflation failed\n" ; |
1223 | |
1224 | print $output ; |
1225 | } |
1226 | |
1227 | ($output, $status) = $x->flush() ; |
1228 | |
1229 | $status == Z_OK |
1230 | or die "deflation failed\n" ; |
1231 | |
1232 | print $output ; |
1233 | |
1234 | =head1 Inflate Interface |
1235 | |
1236 | This section defines the interface available that allows in-memory |
1237 | uncompression using the I<deflate> interface provided by zlib. |
1238 | |
1239 | Here is a definition of the interface: |
1240 | |
1241 | |
1242 | =head2 B<($i, $status) = inflateInit()> |
1243 | |
1244 | Initialises an inflation stream. |
1245 | |
1246 | In a list context it returns the inflation stream, C<$i>, and the |
1247 | I<zlib> status code in C<$status>. In a scalar context it returns the |
1248 | inflation stream only. |
1249 | |
1250 | If successful, C<$i> will hold the inflation stream and C<$status> will |
1251 | be C<Z_OK>. |
1252 | |
1253 | If not successful, C<$i> will be I<undef> and C<$status> will hold the |
1254 | I<zlib> error code. |
1255 | |
1256 | The function optionally takes a number of named options specified as |
1257 | C<-Name=E<gt>value> pairs. This allows individual options to be |
1258 | tailored without having to specify them all in the parameter list. |
1259 | |
1260 | For backward compatibility, it is also possible to pass the parameters |
1261 | as a reference to a hash containing the name=>value pairs. |
1262 | |
1263 | The function takes one optional parameter, a reference to a hash. The |
1264 | contents of the hash allow the deflation interface to be tailored. |
1265 | |
1266 | Here is a list of the valid options: |
1267 | |
1268 | =over 5 |
1269 | |
1270 | =item B<-WindowBits> |
1271 | |
1272 | To uncompress an RFC 1950 data stream, set C<WindowBits> to a positive number. |
1273 | |
1274 | To uncompress an RFC 1951 data stream, set C<WindowBits> to C<-MAX_WBITS>. |
1275 | |
1276 | For a full definition of the meaning and valid values for C<WindowBits> refer |
1277 | to the I<zlib> documentation for I<inflateInit2>. |
1278 | |
1279 | Defaults to C<-WindowBits =E<gt>MAX_WBITS>. |
1280 | |
1281 | =item B<-Bufsize> |
1282 | |
1283 | Sets the initial size for the inflation buffer. If the buffer has to be |
1284 | reallocated to increase the size, it will grow in increments of |
1285 | C<Bufsize>. |
1286 | |
1287 | Default is 4096. |
1288 | |
1289 | =item B<-Dictionary> |
1290 | |
1291 | The default is no dictionary. |
1292 | |
1293 | =back |
1294 | |
1295 | Here is an example of using the C<inflateInit> optional parameter to |
1296 | override the default buffer size. |
1297 | |
1298 | inflateInit( -Bufsize => 300 ) ; |
1299 | |
1300 | =head2 B<($out, $status) = $i-E<gt>inflate($buffer)> |
1301 | |
1302 | Inflates the complete contents of C<$buffer>. The buffer can either be |
1303 | a scalar or a scalar reference. |
1304 | |
1305 | Returns C<Z_OK> if successful and C<Z_STREAM_END> if the end of the |
1306 | compressed data has been successfully reached. |
1307 | If not successful, C<$out> will be I<undef> and C<$status> will hold |
1308 | the I<zlib> error code. |
1309 | |
1310 | The C<$buffer> parameter is modified by C<inflate>. On completion it |
1311 | will contain what remains of the input buffer after inflation. This |
1312 | means that C<$buffer> will be an empty string when the return status is |
1313 | C<Z_OK>. When the return status is C<Z_STREAM_END> the C<$buffer> |
1314 | parameter will contains what (if anything) was stored in the input |
1315 | buffer after the deflated data stream. |
1316 | |
1317 | This feature is useful when processing a file format that encapsulates |
1318 | a compressed data stream (e.g. gzip, zip). |
1319 | |
1320 | =head2 B<$status = $i-E<gt>inflateSync($buffer)> |
1321 | |
1322 | Scans C<$buffer> until it reaches either a I<full flush point> or the |
1323 | end of the buffer. |
1324 | |
1325 | If a I<full flush point> is found, C<Z_OK> is returned and C<$buffer> |
1326 | will be have all data up to the flush point removed. This can then be |
1327 | passed to the C<deflate> method. |
1328 | |
1329 | Any other return code means that a flush point was not found. If more |
1330 | data is available, C<inflateSync> can be called repeatedly with more |
1331 | compressed data until the flush point is found. |
1332 | |
1333 | |
1334 | =head2 B<$i-E<gt>dict_adler()> |
1335 | |
1336 | Returns the adler32 value for the dictionary. |
1337 | |
1338 | =head2 B<$i-E<gt>msg()> |
1339 | |
1340 | Returns the last error message generated by zlib. |
1341 | |
1342 | =head2 B<$i-E<gt>total_in()> |
1343 | |
1344 | Returns the total number of bytes compressed bytes input to inflate. |
1345 | |
1346 | =head2 B<$i-E<gt>total_out()> |
1347 | |
1348 | Returns the total number of uncompressed bytes output from inflate. |
1349 | |
1350 | =head2 Example |
1351 | |
1352 | Here is an example of using C<inflate>. |
1353 | |
1354 | use strict ; |
1355 | use warnings ; |
1356 | |
1357 | use Compress::Zlib ; |
1358 | |
1359 | my $x = inflateInit() |
1360 | or die "Cannot create a inflation stream\n" ; |
1361 | |
1362 | my $input = '' ; |
1363 | binmode STDIN; |
1364 | binmode STDOUT; |
1365 | |
1366 | my ($output, $status) ; |
1367 | while (read(STDIN, $input, 4096)) |
1368 | { |
1369 | ($output, $status) = $x->inflate(\$input) ; |
1370 | |
1371 | print $output |
1372 | if $status == Z_OK or $status == Z_STREAM_END ; |
1373 | |
1374 | last if $status != Z_OK ; |
1375 | } |
1376 | |
1377 | die "inflation failed\n" |
1378 | unless $status == Z_STREAM_END ; |
1379 | |
1380 | =head1 CHECKSUM FUNCTIONS |
1381 | |
1382 | Two functions are provided by I<zlib> to calculate checksums. For the |
1383 | Perl interface, the order of the two parameters in both functions has |
1384 | been reversed. This allows both running checksums and one off |
1385 | calculations to be done. |
1386 | |
1387 | $crc = adler32($buffer [,$crc]) ; |
1388 | $crc = crc32($buffer [,$crc]) ; |
1389 | |
1390 | The buffer parameters can either be a scalar or a scalar reference. |
1391 | |
1392 | If the $crc parameters is C<undef>, the crc value will be reset. |
1393 | |
1394 | If you have built this module with zlib 1.2.3 or better, two more |
1395 | CRC-related functions are available. |
1396 | |
1397 | $crc = adler32_combine($crc1, $crc2, $len2)l |
1398 | $crc = crc32_combine($adler1, $adler2, $len2) |
1399 | |
1400 | These functions allow checksums to be merged. |
1401 | |
1402 | =head1 ACCESSING ZIP FILES |
1403 | |
1404 | Although it is possible (with some effort on your part) to use this |
1405 | module to access .zip files, there is a module on CPAN that will do all |
1406 | the hard work for you. Check out the C<Archive::Zip> module on CPAN at |
1407 | |
1408 | http://www.cpan.org/modules/by-module/Archive/Archive-Zip-*.tar.gz |
1409 | |
1410 | |
1411 | =head1 CONSTANTS |
1412 | |
1413 | All the I<zlib> constants are automatically imported when you make use |
1414 | of I<Compress::Zlib>. |
1415 | |
1416 | |
1417 | =head1 SEE ALSO |
1418 | |
1419 | 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::Uncompress::AnyInflate>, L<IO::Uncompress::AnyUncompress> |
1420 | |
1421 | L<Compress::Zlib::FAQ|Compress::Zlib::FAQ> |
1422 | |
1423 | L<File::GlobMapper|File::GlobMapper>, L<Archive::Zip|Archive::Zip>, |
1424 | L<Archive::Tar|Archive::Tar>, |
1425 | L<IO::Zlib|IO::Zlib> |
1426 | |
1427 | |
1428 | For RFC 1950, 1951 and 1952 see |
1429 | F<http://www.faqs.org/rfcs/rfc1950.html>, |
1430 | F<http://www.faqs.org/rfcs/rfc1951.html> and |
1431 | F<http://www.faqs.org/rfcs/rfc1952.html> |
1432 | |
1433 | The I<zlib> compression library was written by Jean-loup Gailly |
1434 | F<gzip@prep.ai.mit.edu> and Mark Adler F<madler@alumni.caltech.edu>. |
1435 | |
1436 | The primary site for the I<zlib> compression library is |
1437 | F<http://www.zlib.org>. |
1438 | |
1439 | The primary site for gzip is F<http://www.gzip.org>. |
1440 | |
1441 | |
1442 | |
1443 | |
25f0751f |
1444 | =head1 AUTHOR |
1445 | |
cb7abd7f |
1446 | This module was written by Paul Marquess, F<pmqs@cpan.org>. |
25f0751f |
1447 | |
1448 | |
1449 | |
1450 | =head1 MODIFICATION HISTORY |
1451 | |
1452 | See the Changes file. |
1453 | |
1454 | =head1 COPYRIGHT AND LICENSE |
25f0751f |
1455 | |
1456 | Copyright (c) 1995-2006 Paul Marquess. All rights reserved. |
1457 | |
1458 | This program is free software; you can redistribute it and/or |
1459 | modify it under the same terms as Perl itself. |
1460 | |
1461 | |
1462 | |