Commit | Line | Data |
f4c6fd49 |
1 | # File : Zlib.pm |
2 | # Author : Paul Marquess |
3 | # Created : 30 January 2005 |
4 | # Version : 1.34 |
5 | # |
6 | # Copyright (c) 1995-2005 Paul Marquess. All rights reserved. |
7 | # This program is free software; you can redistribute it and/or |
8 | # modify it under the same terms as Perl itself. |
9 | # |
10 | |
11 | package Compress::Zlib; |
12 | |
13 | require 5.004 ; |
14 | require Exporter; |
15 | require DynaLoader; |
16 | use AutoLoader; |
17 | use Carp ; |
18 | use IO::Handle ; |
19 | |
20 | use strict ; |
8aa25532 |
21 | use warnings ; |
22 | our ($VERSION, @ISA, @EXPORT, $AUTOLOAD); |
23 | our ($deflateDefault, $deflateParamsDefault, $inflateDefault); |
f4c6fd49 |
24 | |
213b85e9 |
25 | $VERSION = "1.34_01" ; |
f4c6fd49 |
26 | |
27 | @ISA = qw(Exporter DynaLoader); |
28 | # Items to export into callers namespace by default. Note: do not export |
29 | # names by default without a very good reason. Use EXPORT_OK instead. |
30 | # Do not simply export all your public functions/methods/constants. |
31 | @EXPORT = qw( |
32 | deflateInit |
33 | inflateInit |
34 | |
35 | compress |
36 | uncompress |
37 | |
38 | gzip gunzip |
39 | |
40 | gzopen |
41 | $gzerrno |
42 | |
43 | adler32 |
44 | crc32 |
45 | |
46 | ZLIB_VERSION |
47 | |
48 | DEF_WBITS |
49 | OS_CODE |
50 | |
51 | MAX_MEM_LEVEL |
52 | MAX_WBITS |
53 | |
54 | Z_ASCII |
55 | Z_BEST_COMPRESSION |
56 | Z_BEST_SPEED |
57 | Z_BINARY |
58 | Z_BUF_ERROR |
59 | Z_DATA_ERROR |
60 | Z_DEFAULT_COMPRESSION |
61 | Z_DEFAULT_STRATEGY |
62 | Z_DEFLATED |
63 | Z_ERRNO |
64 | Z_FILTERED |
65 | Z_FINISH |
66 | Z_FULL_FLUSH |
67 | Z_HUFFMAN_ONLY |
68 | Z_MEM_ERROR |
69 | Z_NEED_DICT |
70 | Z_NO_COMPRESSION |
71 | Z_NO_FLUSH |
72 | Z_NULL |
73 | Z_OK |
74 | Z_PARTIAL_FLUSH |
75 | Z_STREAM_END |
76 | Z_STREAM_ERROR |
77 | Z_SYNC_FLUSH |
78 | Z_UNKNOWN |
79 | Z_VERSION_ERROR |
80 | ); |
81 | |
82 | |
83 | |
84 | sub AUTOLOAD { |
85 | my($constname); |
86 | ($constname = $AUTOLOAD) =~ s/.*:://; |
87 | my ($error, $val) = constant($constname); |
88 | Carp::croak $error if $error; |
89 | no strict 'refs'; |
90 | *{$AUTOLOAD} = sub { $val }; |
91 | goto &{$AUTOLOAD}; |
92 | } |
93 | |
94 | bootstrap Compress::Zlib $VERSION ; |
95 | |
96 | # Preloaded methods go here. |
97 | |
98 | sub isaFilehandle($) |
99 | { |
100 | my $fh = shift ; |
101 | |
102 | return ((UNIVERSAL::isa($fh,'GLOB') or UNIVERSAL::isa(\$fh,'GLOB')) |
103 | and defined fileno($fh) ) |
104 | |
105 | } |
106 | |
107 | sub isaFilename($) |
108 | { |
109 | my $name = shift ; |
110 | |
111 | return (! ref $name and UNIVERSAL::isa(\$name, 'SCALAR')) ; |
112 | } |
113 | |
114 | sub gzopen($$) |
115 | { |
116 | my ($file, $mode) = @_ ; |
117 | |
118 | if (isaFilehandle $file) { |
119 | IO::Handle::flush($file) ; |
120 | my $offset = tell($file) ; |
121 | gzdopen_(fileno($file), $mode, $offset) ; |
122 | } |
123 | elsif (isaFilename $file) { |
124 | gzopen_($file, $mode) |
125 | } |
126 | else { |
127 | croak "gzopen: file parameter is not a filehandle or filename" |
128 | } |
129 | } |
130 | |
131 | sub ParseParameters($@) |
132 | { |
133 | my ($default, @rest) = @_ ; |
134 | my (%got) = %$default ; |
135 | my (@Bad) ; |
136 | my ($key, $value) ; |
137 | my $sub = (caller(1))[3] ; |
138 | my %options = () ; |
139 | |
140 | # allow the options to be passed as a hash reference or |
141 | # as the complete hash. |
142 | if (@rest == 1) { |
143 | |
144 | croak "$sub: parameter is not a reference to a hash" |
145 | if ref $rest[0] ne "HASH" ; |
146 | |
147 | %options = %{ $rest[0] } ; |
148 | } |
149 | elsif (@rest >= 2) { |
150 | my $count = @rest; |
151 | croak "$sub: Expected even number of parameters, got $count" |
152 | if @rest % 2 != 0 ; |
153 | %options = @rest ; |
154 | } |
155 | |
156 | while (($key, $value) = each %options) |
157 | { |
158 | $key =~ s/^-// ; |
159 | |
160 | if (exists $default->{$key}) |
161 | { $got{$key} = $value } |
162 | else |
163 | { push (@Bad, $key) } |
164 | } |
165 | |
166 | if (@Bad) { |
167 | my ($bad) = join(", ", @Bad) ; |
168 | croak "unknown key value(s) @Bad" ; |
169 | } |
170 | |
171 | return \%got ; |
172 | } |
173 | |
174 | $deflateDefault = { |
175 | 'Level' => Z_DEFAULT_COMPRESSION(), |
176 | 'Method' => Z_DEFLATED(), |
177 | 'WindowBits' => MAX_WBITS(), |
178 | 'MemLevel' => MAX_MEM_LEVEL(), |
179 | 'Strategy' => Z_DEFAULT_STRATEGY(), |
180 | 'Bufsize' => 4096, |
181 | 'Dictionary' => "", |
182 | } ; |
183 | |
184 | $deflateParamsDefault = { |
185 | 'Level' => undef, |
186 | 'Strategy' => undef, |
187 | 'Bufsize' => undef, |
188 | } ; |
189 | |
190 | $inflateDefault = { |
191 | 'WindowBits' => MAX_WBITS(), |
192 | 'Bufsize' => 4096, |
193 | 'Dictionary' => "", |
194 | } ; |
195 | |
196 | |
197 | sub deflateInit(@) |
198 | { |
199 | my ($got) = ParseParameters($deflateDefault, @_) ; |
8aa25532 |
200 | no warnings; |
f4c6fd49 |
201 | croak "deflateInit: Bufsize must be >= 1, you specified $got->{Bufsize}" |
202 | unless $got->{Bufsize} >= 1; |
203 | _deflateInit($got->{Level}, $got->{Method}, $got->{WindowBits}, |
204 | $got->{MemLevel}, $got->{Strategy}, $got->{Bufsize}, |
205 | $got->{Dictionary}) ; |
206 | |
207 | } |
208 | |
209 | sub inflateInit(@) |
210 | { |
211 | my ($got) = ParseParameters($inflateDefault, @_) ; |
8aa25532 |
212 | no warnings; |
f4c6fd49 |
213 | croak "inflateInit: Bufsize must be >= 1, you specified $got->{Bufsize}" |
214 | unless $got->{Bufsize} >= 1; |
215 | _inflateInit($got->{WindowBits}, $got->{Bufsize}, $got->{Dictionary}); |
216 | |
217 | } |
218 | |
219 | sub Compress::Zlib::deflateStream::deflateParams |
220 | { |
221 | my $self = shift ; |
222 | my ($got) = ParseParameters($deflateParamsDefault, @_) ; |
223 | croak "deflateParams needs Level and/or Strategy" |
224 | unless defined $got->{Level} || defined $got->{Strategy}; |
8aa25532 |
225 | no warnings; |
f4c6fd49 |
226 | croak "deflateParams: Bufsize must be >= 1, you specified $got->{Bufsize}" |
227 | unless !defined $got->{Bufsize} || $got->{Bufsize} >= 1; |
228 | |
229 | my $flags = 0; |
230 | if (defined $got->{Level}) |
231 | { $flags |= 1 } |
232 | else |
233 | { $got->{Level} = 0 } |
234 | |
235 | if (defined $got->{Strategy}) |
236 | { $flags |= 2 } |
237 | else |
238 | { $got->{Strategy} = 0 } |
239 | |
240 | $got->{Bufsize} = 0 |
241 | if !defined $got->{Bufsize}; |
242 | |
243 | $self->_deflateParams($flags, $got->{Level}, $got->{Strategy}, |
244 | $got->{Bufsize}); |
245 | |
246 | } |
247 | |
248 | sub compress($;$) |
249 | { |
250 | my ($x, $output, $out, $err, $in) ; |
251 | |
252 | if (ref $_[0] ) { |
253 | $in = $_[0] ; |
254 | croak "not a scalar reference" unless ref $in eq 'SCALAR' ; |
255 | } |
256 | else { |
257 | $in = \$_[0] ; |
258 | } |
259 | |
260 | my $level = (@_ == 2 ? $_[1] : Z_DEFAULT_COMPRESSION() ); |
261 | |
262 | |
263 | if ( (($x, $err) = deflateInit(Level => $level))[1] == Z_OK()) { |
264 | |
265 | ($output, $err) = $x->deflate($in) ; |
266 | return undef unless $err == Z_OK() ; |
267 | |
268 | ($out, $err) = $x->flush() ; |
269 | return undef unless $err == Z_OK() ; |
270 | |
271 | return ($output . $out) ; |
272 | |
273 | } |
274 | |
275 | return undef ; |
276 | } |
277 | |
278 | |
279 | sub uncompress($) |
280 | { |
281 | my ($x, $output, $err, $in) ; |
282 | |
283 | if (ref $_[0] ) { |
284 | $in = $_[0] ; |
285 | croak "not a scalar reference" unless ref $in eq 'SCALAR' ; |
286 | } |
287 | else { |
288 | $in = \$_[0] ; |
289 | } |
290 | |
291 | if ( (($x, $err) = inflateInit())[1] == Z_OK()) { |
292 | |
293 | ($output, $err) = $x->__unc_inflate($in) ; |
294 | return undef unless $err == Z_STREAM_END() ; |
295 | |
296 | return $output ; |
297 | } |
298 | |
299 | return undef ; |
300 | } |
301 | |
302 | |
303 | # Constants |
304 | use constant MAGIC1 => 0x1f ; |
305 | use constant MAGIC2 => 0x8b ; |
306 | use constant OSCODE => 3 ; |
307 | |
308 | use constant FTEXT => 1 ; |
309 | use constant FHCRC => 2 ; |
310 | use constant FEXTRA => 4 ; |
311 | use constant FNAME => 8 ; |
312 | use constant FCOMMENT => 16 ; |
313 | use constant NULL => pack("C", 0) ; |
314 | use constant RESERVED => 0xE0 ; |
315 | |
316 | use constant MIN_HDR_SIZE => 10 ; # minimum gzip header size |
317 | |
318 | sub memGzip($) |
319 | { |
320 | my $x = deflateInit( |
321 | -Level => Z_BEST_COMPRESSION(), |
322 | -WindowBits => - MAX_WBITS(), |
323 | ) |
324 | or return undef ; |
325 | |
326 | # write a minimal gzip header |
327 | my(@m); |
328 | push @m, pack("C" . MIN_HDR_SIZE, |
329 | MAGIC1, MAGIC2, Z_DEFLATED(), 0,0,0,0,0,0, OSCODE) ; |
330 | |
331 | # if the deflation buffer isn't a reference, make it one |
332 | my $string = (ref $_[0] ? $_[0] : \$_[0]) ; |
333 | |
334 | my ($output, $status) = $x->deflate($string) ; |
335 | push @m, $output ; |
336 | $status == Z_OK() |
337 | or return undef ; |
338 | |
339 | ($output, $status) = $x->flush() ; |
340 | push @m, $output ; |
341 | $status == Z_OK() |
342 | or return undef ; |
343 | |
344 | push @m, pack("V V", crc32($string), $x->total_in()); |
345 | |
346 | return join "", @m; |
347 | } |
348 | |
349 | sub _removeGzipHeader($) |
350 | { |
351 | my $string = shift ; |
352 | |
353 | return Z_DATA_ERROR() |
354 | if length($$string) < MIN_HDR_SIZE ; |
355 | |
356 | my ($magic1, $magic2, $method, $flags, $time, $xflags, $oscode) = |
357 | unpack ('CCCCVCC', $$string); |
358 | |
359 | return Z_DATA_ERROR() |
360 | unless $magic1 == MAGIC1 and $magic2 == MAGIC2 and |
361 | $method == Z_DEFLATED() and !($flags & RESERVED()) ; |
362 | substr($$string, 0, MIN_HDR_SIZE) = '' ; |
363 | |
364 | # skip extra field |
365 | if ($flags & FEXTRA) |
366 | { |
367 | return Z_DATA_ERROR() |
368 | if length($$string) < 2 ; |
369 | |
370 | my ($extra_len) = unpack ('v', $$string); |
371 | $extra_len += 2; |
372 | return Z_DATA_ERROR() |
373 | if length($$string) < $extra_len ; |
374 | |
375 | substr($$string, 0, $extra_len) = ''; |
376 | } |
377 | |
378 | # skip orig name |
379 | if ($flags & FNAME) |
380 | { |
381 | my $name_end = index ($$string, NULL); |
382 | return Z_DATA_ERROR() |
383 | if $name_end == -1 ; |
384 | substr($$string, 0, $name_end + 1) = ''; |
385 | } |
386 | |
387 | # skip comment |
388 | if ($flags & FCOMMENT) |
389 | { |
390 | my $comment_end = index ($$string, NULL); |
391 | return Z_DATA_ERROR() |
392 | if $comment_end == -1 ; |
393 | substr($$string, 0, $comment_end + 1) = ''; |
394 | } |
395 | |
396 | # skip header crc |
397 | if ($flags & FHCRC) |
398 | { |
399 | return Z_DATA_ERROR() |
400 | if length ($$string) < 2 ; |
401 | substr($$string, 0, 2) = ''; |
402 | } |
403 | |
404 | return Z_OK(); |
405 | } |
406 | |
407 | |
408 | sub memGunzip($) |
409 | { |
410 | # if the buffer isn't a reference, make it one |
411 | my $string = (ref $_[0] ? $_[0] : \$_[0]); |
412 | |
413 | _removeGzipHeader($string) == Z_OK() |
414 | or return undef; |
415 | |
416 | my $bufsize = length $$string > 4096 ? length $$string : 4096 ; |
417 | my $x = inflateInit( -WindowBits => - MAX_WBITS(), |
418 | -Bufsize => $bufsize) |
419 | or return undef; |
420 | my ($output, $status) = $x->inflate($string); |
421 | return undef |
422 | unless $status == Z_STREAM_END(); |
423 | |
424 | if (length $$string >= 8) |
425 | { |
426 | my ($crc, $len) = unpack ("VV", substr($$string, 0, 8)); |
427 | substr($$string, 0, 8) = ''; |
428 | return undef |
429 | unless $len == length($output) and |
430 | $crc == crc32($output); |
431 | } |
432 | else |
433 | { |
434 | $$string = ''; |
435 | } |
436 | |
437 | return $output; |
438 | } |
439 | |
440 | # Autoload methods go after __END__, and are processed by the autosplit program. |
441 | |
442 | 1; |
443 | __END__ |
444 | |
445 | =cut |
446 | |
447 | =head1 NAME |
448 | |
449 | Compress::Zlib - Interface to zlib compression library |
450 | |
451 | =head1 SYNOPSIS |
452 | |
453 | use Compress::Zlib ; |
454 | |
455 | ($d, $status) = deflateInit( [OPT] ) ; |
456 | ($out, $status) = $d->deflate($buffer) ; |
457 | $status = $d->deflateParams([OPT]) ; |
458 | ($out, $status) = $d->flush() ; |
459 | $d->dict_adler() ; |
460 | $d->total_in() ; |
461 | $d->total_out() ; |
462 | $d->msg() ; |
463 | |
464 | ($i, $status) = inflateInit( [OPT] ) ; |
465 | ($out, $status) = $i->inflate($buffer) ; |
466 | $status = $i->inflateSync($buffer) ; |
467 | $i->dict_adler() ; |
468 | $i->total_in() ; |
469 | $i->total_out() ; |
470 | $i->msg() ; |
471 | |
472 | $dest = compress($source, [$level]) ; |
473 | $dest = uncompress($source) ; |
474 | |
475 | $gz = gzopen($filename or filehandle, $mode) ; |
476 | $bytesread = $gz->gzread($buffer [,$size]) ; |
477 | $bytesread = $gz->gzreadline($line) ; |
478 | $byteswritten = $gz->gzwrite($buffer) ; |
479 | $status = $gz->gzflush($flush) ; |
480 | $status = $gz->gzclose() ; |
481 | $status = $gz->gzeof() ; |
482 | $status = $gz->gzsetparams($level, $strategy) ; |
483 | $errstring = $gz->gzerror() ; |
484 | $gzerrno |
485 | |
486 | $dest = Compress::Zlib::memGzip($buffer) ; |
487 | $dest = Compress::Zlib::memGunzip($buffer) ; |
488 | |
489 | $crc = adler32($buffer [,$crc]) ; |
490 | $crc = crc32($buffer [,$crc]) ; |
491 | |
492 | ZLIB_VERSION |
493 | |
494 | =head1 DESCRIPTION |
495 | |
496 | The I<Compress::Zlib> module provides a Perl interface to the I<zlib> |
497 | compression library (see L</AUTHOR> for details about where to get |
498 | I<zlib>). Most of the functionality provided by I<zlib> is available |
499 | in I<Compress::Zlib>. |
500 | |
501 | The module can be split into two general areas of functionality, namely |
502 | in-memory compression/decompression and read/write access to I<gzip> |
503 | files. Each of these areas will be discussed separately below. |
504 | |
505 | =head1 DEFLATE |
506 | |
507 | The interface I<Compress::Zlib> provides to the in-memory I<deflate> |
508 | (and I<inflate>) functions has been modified to fit into a Perl model. |
509 | |
510 | The main difference is that for both inflation and deflation, the Perl |
511 | interface will I<always> consume the complete input buffer before |
512 | returning. Also the output buffer returned will be automatically grown |
513 | to fit the amount of output available. |
514 | |
515 | Here is a definition of the interface available: |
516 | |
517 | |
518 | =head2 B<($d, $status) = deflateInit( [OPT] )> |
519 | |
520 | Initialises a deflation stream. |
521 | |
522 | It combines the features of the I<zlib> functions B<deflateInit>, |
523 | B<deflateInit2> and B<deflateSetDictionary>. |
524 | |
525 | If successful, it will return the initialised deflation stream, B<$d> |
526 | and B<$status> of C<Z_OK> in a list context. In scalar context it |
527 | returns the deflation stream, B<$d>, only. |
528 | |
529 | If not successful, the returned deflation stream (B<$d>) will be |
530 | I<undef> and B<$status> will hold the exact I<zlib> error code. |
531 | |
532 | The function optionally takes a number of named options specified as |
533 | C<-Name=E<gt>value> pairs. This allows individual options to be |
534 | tailored without having to specify them all in the parameter list. |
535 | |
536 | For backward compatibility, it is also possible to pass the parameters |
537 | as a reference to a hash containing the name=>value pairs. |
538 | |
539 | The function takes one optional parameter, a reference to a hash. The |
540 | contents of the hash allow the deflation interface to be tailored. |
541 | |
542 | Here is a list of the valid options: |
543 | |
544 | =over 5 |
545 | |
546 | =item B<-Level> |
547 | |
548 | Defines the compression level. Valid values are 0 through 9, |
549 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
550 | C<Z_DEFAULT_COMPRESSION>. |
551 | |
552 | The default is C<-Level =E<gt>Z_DEFAULT_COMPRESSION>. |
553 | |
554 | =item B<-Method> |
555 | |
556 | Defines the compression method. The only valid value at present (and |
557 | the default) is C<-Method =E<gt>Z_DEFLATED>. |
558 | |
559 | =item B<-WindowBits> |
560 | |
561 | For a definition of the meaning and valid values for B<WindowBits> |
562 | refer to the I<zlib> documentation for I<deflateInit2>. |
563 | |
564 | Defaults to C<-WindowBits =E<gt>MAX_WBITS>. |
565 | |
566 | =item B<-MemLevel> |
567 | |
568 | For a definition of the meaning and valid values for B<MemLevel> |
569 | refer to the I<zlib> documentation for I<deflateInit2>. |
570 | |
571 | Defaults to C<-MemLevel =E<gt>MAX_MEM_LEVEL>. |
572 | |
573 | =item B<-Strategy> |
574 | |
575 | Defines the strategy used to tune the compression. The valid values are |
576 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
577 | |
578 | The default is C<-Strategy =E<gt>Z_DEFAULT_STRATEGY>. |
579 | |
580 | =item B<-Dictionary> |
581 | |
582 | When a dictionary is specified I<Compress::Zlib> will automatically |
583 | call B<deflateSetDictionary> directly after calling B<deflateInit>. The |
584 | Adler32 value for the dictionary can be obtained by calling the method |
585 | C<$d->dict_adler()>. |
586 | |
587 | The default is no dictionary. |
588 | |
589 | =item B<-Bufsize> |
590 | |
591 | Sets the initial size for the deflation buffer. If the buffer has to be |
592 | reallocated to increase the size, it will grow in increments of |
593 | B<Bufsize>. |
594 | |
595 | The default is 4096. |
596 | |
597 | =back |
598 | |
599 | Here is an example of using the B<deflateInit> optional parameter list |
600 | to override the default buffer size and compression level. All other |
601 | options will take their default values. |
602 | |
603 | deflateInit( -Bufsize => 300, |
604 | -Level => Z_BEST_SPEED ) ; |
605 | |
606 | |
607 | =head2 B<($out, $status) = $d-E<gt>deflate($buffer)> |
608 | |
609 | |
610 | Deflates the contents of B<$buffer>. The buffer can either be a scalar |
611 | or a scalar reference. When finished, B<$buffer> will be |
612 | completely processed (assuming there were no errors). If the deflation |
613 | was successful it returns the deflated output, B<$out>, and a status |
614 | value, B<$status>, of C<Z_OK>. |
615 | |
616 | On error, B<$out> will be I<undef> and B<$status> will contain the |
617 | I<zlib> error code. |
618 | |
619 | In a scalar context B<deflate> will return B<$out> only. |
620 | |
621 | As with the I<deflate> function in I<zlib>, it is not necessarily the |
622 | case that any output will be produced by this method. So don't rely on |
623 | the fact that B<$out> is empty for an error test. |
624 | |
625 | |
626 | =head2 B<($out, $status) = $d-E<gt>flush([flush_type])> |
627 | |
628 | Typically used to finish the deflation. Any pending output will be |
629 | returned via B<$out>. |
630 | B<$status> will have a value C<Z_OK> if successful. |
631 | |
632 | In a scalar context B<flush> will return B<$out> only. |
633 | |
634 | Note that flushing can seriously degrade the compression ratio, so it |
635 | should only be used to terminate a decompression (using C<Z_FINISH>) or |
636 | when you want to create a I<full flush point> (using C<Z_FULL_FLUSH>). |
637 | |
638 | By default the C<flush_type> used is C<Z_FINISH>. Other valid values |
639 | for C<flush_type> are C<Z_NO_FLUSH>, C<Z_PARTIAL_FLUSH>, C<Z_SYNC_FLUSH> |
640 | and C<Z_FULL_FLUSH>. It is strongly recommended that you only set the |
641 | C<flush_type> parameter if you fully understand the implications of |
642 | what it does. See the C<zlib> documentation for details. |
643 | |
644 | =head2 B<$status = $d-E<gt>deflateParams([OPT])> |
645 | |
646 | Change settings for the deflate stream C<$d>. |
647 | |
648 | The list of the valid options is shown below. Options not specified |
649 | will remain unchanged. |
650 | |
651 | =over 5 |
652 | |
653 | =item B<-Level> |
654 | |
655 | Defines the compression level. Valid values are 0 through 9, |
656 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
657 | C<Z_DEFAULT_COMPRESSION>. |
658 | |
659 | =item B<-Strategy> |
660 | |
661 | Defines the strategy used to tune the compression. The valid values are |
662 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
663 | |
664 | =back |
665 | |
666 | =head2 B<$d-E<gt>dict_adler()> |
667 | |
668 | Returns the adler32 value for the dictionary. |
669 | |
670 | =head2 B<$d-E<gt>msg()> |
671 | |
672 | Returns the last error message generated by zlib. |
673 | |
674 | =head2 B<$d-E<gt>total_in()> |
675 | |
676 | Returns the total number of bytes uncompressed bytes input to deflate. |
677 | |
678 | =head2 B<$d-E<gt>total_out()> |
679 | |
680 | Returns the total number of compressed bytes output from deflate. |
681 | |
682 | =head2 Example |
683 | |
684 | |
685 | Here is a trivial example of using B<deflate>. It simply reads standard |
686 | input, deflates it and writes it to standard output. |
687 | |
688 | use strict ; |
689 | use warnings ; |
690 | |
691 | use Compress::Zlib ; |
692 | |
693 | binmode STDIN; |
694 | binmode STDOUT; |
695 | my $x = deflateInit() |
696 | or die "Cannot create a deflation stream\n" ; |
697 | |
698 | my ($output, $status) ; |
699 | while (<>) |
700 | { |
701 | ($output, $status) = $x->deflate($_) ; |
702 | |
703 | $status == Z_OK |
704 | or die "deflation failed\n" ; |
705 | |
706 | print $output ; |
707 | } |
708 | |
709 | ($output, $status) = $x->flush() ; |
710 | |
711 | $status == Z_OK |
712 | or die "deflation failed\n" ; |
713 | |
714 | print $output ; |
715 | |
716 | =head1 INFLATE |
717 | |
718 | Here is a definition of the interface: |
719 | |
720 | |
721 | =head2 B<($i, $status) = inflateInit()> |
722 | |
723 | Initialises an inflation stream. |
724 | |
725 | In a list context it returns the inflation stream, B<$i>, and the |
726 | I<zlib> status code (B<$status>). In a scalar context it returns the |
727 | inflation stream only. |
728 | |
729 | If successful, B<$i> will hold the inflation stream and B<$status> will |
730 | be C<Z_OK>. |
731 | |
732 | If not successful, B<$i> will be I<undef> and B<$status> will hold the |
733 | I<zlib> error code. |
734 | |
735 | The function optionally takes a number of named options specified as |
736 | C<-Name=E<gt>value> pairs. This allows individual options to be |
737 | tailored without having to specify them all in the parameter list. |
738 | |
739 | For backward compatibility, it is also possible to pass the parameters |
740 | as a reference to a hash containing the name=>value pairs. |
741 | |
742 | The function takes one optional parameter, a reference to a hash. The |
743 | contents of the hash allow the deflation interface to be tailored. |
744 | |
745 | Here is a list of the valid options: |
746 | |
747 | =over 5 |
748 | |
749 | =item B<-WindowBits> |
750 | |
751 | For a definition of the meaning and valid values for B<WindowBits> |
752 | refer to the I<zlib> documentation for I<inflateInit2>. |
753 | |
754 | Defaults to C<-WindowBits =E<gt>MAX_WBITS>. |
755 | |
756 | =item B<-Bufsize> |
757 | |
758 | Sets the initial size for the inflation buffer. If the buffer has to be |
759 | reallocated to increase the size, it will grow in increments of |
760 | B<Bufsize>. |
761 | |
762 | Default is 4096. |
763 | |
764 | =item B<-Dictionary> |
765 | |
766 | The default is no dictionary. |
767 | |
768 | =back |
769 | |
770 | Here is an example of using the B<inflateInit> optional parameter to |
771 | override the default buffer size. |
772 | |
773 | inflateInit( -Bufsize => 300 ) ; |
774 | |
775 | =head2 B<($out, $status) = $i-E<gt>inflate($buffer)> |
776 | |
777 | Inflates the complete contents of B<$buffer>. The buffer can either be |
778 | a scalar or a scalar reference. |
779 | |
780 | Returns C<Z_OK> if successful and C<Z_STREAM_END> if the end of the |
781 | compressed data has been successfully reached. |
782 | If not successful, B<$out> will be I<undef> and B<$status> will hold |
783 | the I<zlib> error code. |
784 | |
785 | The C<$buffer> parameter is modified by C<inflate>. On completion it |
786 | will contain what remains of the input buffer after inflation. This |
787 | means that C<$buffer> will be an empty string when the return status is |
788 | C<Z_OK>. When the return status is C<Z_STREAM_END> the C<$buffer> |
789 | parameter will contains what (if anything) was stored in the input |
790 | buffer after the deflated data stream. |
791 | |
792 | This feature is useful when processing a file format that encapsulates |
793 | a compressed data stream (e.g. gzip, zip). |
794 | |
795 | =head2 B<$status = $i-E<gt>inflateSync($buffer)> |
796 | |
797 | Scans C<$buffer> until it reaches either a I<full flush point> or the |
798 | end of the buffer. |
799 | |
800 | If a I<full flush point> is found, C<Z_OK> is returned and C<$buffer> |
801 | will be have all data up to the flush point removed. This can then be |
802 | passed to the C<deflate> method. |
803 | |
804 | Any other return code means that a flush point was not found. If more |
805 | data is available, C<inflateSync> can be called repeatedly with more |
806 | compressed data until the flush point is found. |
807 | |
808 | |
809 | =head2 B<$i-E<gt>dict_adler()> |
810 | |
811 | Returns the adler32 value for the dictionary. |
812 | |
813 | =head2 B<$i-E<gt>msg()> |
814 | |
815 | Returns the last error message generated by zlib. |
816 | |
817 | =head2 B<$i-E<gt>total_in()> |
818 | |
819 | Returns the total number of bytes compressed bytes input to inflate. |
820 | |
821 | =head2 B<$i-E<gt>total_out()> |
822 | |
823 | Returns the total number of uncompressed bytes output from inflate. |
824 | |
825 | =head2 Example |
826 | |
827 | Here is an example of using B<inflate>. |
828 | |
829 | use strict ; |
830 | use warnings ; |
831 | |
832 | use Compress::Zlib ; |
833 | |
834 | my $x = inflateInit() |
835 | or die "Cannot create a inflation stream\n" ; |
836 | |
837 | my $input = '' ; |
838 | binmode STDIN; |
839 | binmode STDOUT; |
840 | |
841 | my ($output, $status) ; |
842 | while (read(STDIN, $input, 4096)) |
843 | { |
844 | ($output, $status) = $x->inflate(\$input) ; |
845 | |
846 | print $output |
847 | if $status == Z_OK or $status == Z_STREAM_END ; |
848 | |
849 | last if $status != Z_OK ; |
850 | } |
851 | |
852 | die "inflation failed\n" |
853 | unless $status == Z_STREAM_END ; |
854 | |
855 | =head1 COMPRESS/UNCOMPRESS |
856 | |
857 | Two high-level functions are provided by I<zlib> to perform in-memory |
858 | compression. They are B<compress> and B<uncompress>. Two Perl subs are |
859 | provided which provide similar functionality. |
860 | |
861 | =over 5 |
862 | |
863 | =item B<$dest = compress($source [, $level] ) ;> |
864 | |
865 | Compresses B<$source>. If successful it returns the |
866 | compressed data. Otherwise it returns I<undef>. |
867 | |
868 | The source buffer can either be a scalar or a scalar reference. |
869 | |
870 | The B<$level> paramter defines the compression level. Valid values are |
871 | 0 through 9, C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, |
872 | C<Z_BEST_COMPRESSION>, and C<Z_DEFAULT_COMPRESSION>. |
873 | If B<$level> is not specified C<Z_DEFAULT_COMPRESSION> will be used. |
874 | |
875 | |
876 | =item B<$dest = uncompress($source) ;> |
877 | |
878 | Uncompresses B<$source>. If successful it returns the uncompressed |
879 | data. Otherwise it returns I<undef>. |
880 | |
881 | The source buffer can either be a scalar or a scalar reference. |
882 | |
883 | =back |
884 | |
885 | =head1 GZIP INTERFACE |
886 | |
887 | A number of functions are supplied in I<zlib> for reading and writing |
888 | I<gzip> files. This module provides an interface to most of them. In |
889 | general the interface provided by this module operates identically to |
890 | the functions provided by I<zlib>. Any differences are explained |
891 | below. |
892 | |
893 | =over 5 |
894 | |
895 | =item B<$gz = gzopen(filename or filehandle, mode)> |
896 | |
897 | This function operates identically to the I<zlib> equivalent except |
898 | that it returns an object which is used to access the other I<gzip> |
899 | methods. |
900 | |
901 | As with the I<zlib> equivalent, the B<mode> parameter is used to |
902 | specify both whether the file is opened for reading or writing and to |
903 | optionally specify a a compression level. Refer to the I<zlib> |
904 | documentation for the exact format of the B<mode> parameter. |
905 | |
906 | If a reference to an open filehandle is passed in place of the |
907 | filename, gzdopen will be called behind the scenes. The third example |
908 | at the end of this section, I<gzstream>, uses this feature. |
909 | |
910 | =item B<$bytesread = $gz-E<gt>gzread($buffer [, $size]) ;> |
911 | |
912 | Reads B<$size> bytes from the compressed file into B<$buffer>. If |
913 | B<$size> is not specified, it will default to 4096. If the scalar |
914 | B<$buffer> is not large enough, it will be extended automatically. |
915 | |
916 | Returns the number of bytes actually read. On EOF it returns 0 and in |
917 | the case of an error, -1. |
918 | |
919 | =item B<$bytesread = $gz-E<gt>gzreadline($line) ;> |
920 | |
921 | Reads the next line from the compressed file into B<$line>. |
922 | |
923 | Returns the number of bytes actually read. On EOF it returns 0 and in |
924 | the case of an error, -1. |
925 | |
926 | It is legal to intermix calls to B<gzread> and B<gzreadline>. |
927 | |
928 | At this time B<gzreadline> ignores the variable C<$/> |
929 | (C<$INPUT_RECORD_SEPARATOR> or C<$RS> when C<English> is in use). The |
930 | end of a line is denoted by the C character C<'\n'>. |
931 | |
932 | =item B<$byteswritten = $gz-E<gt>gzwrite($buffer) ;> |
933 | |
934 | Writes the contents of B<$buffer> to the compressed file. Returns the |
935 | number of bytes actually written, or 0 on error. |
936 | |
937 | =item B<$status = $gz-E<gt>gzflush($flush) ;> |
938 | |
939 | Flushes all pending output to the compressed file. |
940 | Works identically to the I<zlib> function it interfaces to. Note that |
941 | the use of B<gzflush> can degrade compression. |
942 | |
943 | Returns C<Z_OK> if B<$flush> is C<Z_FINISH> and all output could be |
944 | flushed. Otherwise the zlib error code is returned. |
945 | |
946 | Refer to the I<zlib> documentation for the valid values of B<$flush>. |
947 | |
948 | =item B<$status = $gz-E<gt>gzeof() ;> |
949 | |
950 | Returns 1 if the end of file has been detected while reading the input |
951 | file, otherwise returns 0. |
952 | |
953 | =item B<$gz-E<gt>gzclose> |
954 | |
955 | Closes the compressed file. Any pending data is flushed to the file |
956 | before it is closed. |
957 | |
958 | =item B<$gz-E<gt>gzsetparams($level, $strategy> |
959 | |
960 | Change settings for the deflate stream C<$gz>. |
961 | |
962 | The list of the valid options is shown below. Options not specified |
963 | will remain unchanged. |
964 | |
965 | Note: This method is only available if you are running zlib 1.0.6 or better. |
966 | |
967 | =over 5 |
968 | |
969 | =item B<$level> |
970 | |
971 | Defines the compression level. Valid values are 0 through 9, |
972 | C<Z_NO_COMPRESSION>, C<Z_BEST_SPEED>, C<Z_BEST_COMPRESSION>, and |
973 | C<Z_DEFAULT_COMPRESSION>. |
974 | |
975 | =item B<$strategy> |
976 | |
977 | Defines the strategy used to tune the compression. The valid values are |
978 | C<Z_DEFAULT_STRATEGY>, C<Z_FILTERED> and C<Z_HUFFMAN_ONLY>. |
979 | |
980 | =back |
981 | |
982 | =item B<$gz-E<gt>gzerror> |
983 | |
984 | Returns the I<zlib> error message or number for the last operation |
985 | associated with B<$gz>. The return value will be the I<zlib> error |
986 | number when used in a numeric context and the I<zlib> error message |
987 | when used in a string context. The I<zlib> error number constants, |
988 | shown below, are available for use. |
989 | |
990 | Z_OK |
991 | Z_STREAM_END |
992 | Z_ERRNO |
993 | Z_STREAM_ERROR |
994 | Z_DATA_ERROR |
995 | Z_MEM_ERROR |
996 | Z_BUF_ERROR |
997 | |
998 | =item B<$gzerrno> |
999 | |
1000 | The B<$gzerrno> scalar holds the error code associated with the most |
1001 | recent I<gzip> routine. Note that unlike B<gzerror()>, the error is |
1002 | I<not> associated with a particular file. |
1003 | |
1004 | As with B<gzerror()> it returns an error number in numeric context and |
1005 | an error message in string context. Unlike B<gzerror()> though, the |
1006 | error message will correspond to the I<zlib> message when the error is |
1007 | associated with I<zlib> itself, or the UNIX error message when it is |
1008 | not (i.e. I<zlib> returned C<Z_ERRORNO>). |
1009 | |
1010 | As there is an overlap between the error numbers used by I<zlib> and |
1011 | UNIX, B<$gzerrno> should only be used to check for the presence of |
1012 | I<an> error in numeric context. Use B<gzerror()> to check for specific |
1013 | I<zlib> errors. The I<gzcat> example below shows how the variable can |
1014 | be used safely. |
1015 | |
1016 | =back |
1017 | |
1018 | |
1019 | =head2 Examples |
1020 | |
1021 | Here is an example script which uses the interface. It implements a |
1022 | I<gzcat> function. |
1023 | |
1024 | use strict ; |
1025 | use warnings ; |
1026 | |
1027 | use Compress::Zlib ; |
1028 | |
1029 | die "Usage: gzcat file...\n" |
1030 | unless @ARGV ; |
1031 | |
1032 | my $file ; |
1033 | |
1034 | foreach $file (@ARGV) { |
1035 | my $buffer ; |
1036 | |
1037 | my $gz = gzopen($file, "rb") |
1038 | or die "Cannot open $file: $gzerrno\n" ; |
1039 | |
1040 | print $buffer while $gz->gzread($buffer) > 0 ; |
1041 | |
1042 | die "Error reading from $file: $gzerrno" . ($gzerrno+0) . "\n" |
1043 | if $gzerrno != Z_STREAM_END ; |
1044 | |
1045 | $gz->gzclose() ; |
1046 | } |
1047 | |
1048 | Below is a script which makes use of B<gzreadline>. It implements a |
1049 | very simple I<grep> like script. |
1050 | |
1051 | use strict ; |
1052 | use warnings ; |
1053 | |
1054 | use Compress::Zlib ; |
1055 | |
1056 | die "Usage: gzgrep pattern file...\n" |
1057 | unless @ARGV >= 2; |
1058 | |
1059 | my $pattern = shift ; |
1060 | |
1061 | my $file ; |
1062 | |
1063 | foreach $file (@ARGV) { |
1064 | my $gz = gzopen($file, "rb") |
1065 | or die "Cannot open $file: $gzerrno\n" ; |
1066 | |
1067 | while ($gz->gzreadline($_) > 0) { |
1068 | print if /$pattern/ ; |
1069 | } |
1070 | |
1071 | die "Error reading from $file: $gzerrno\n" |
1072 | if $gzerrno != Z_STREAM_END ; |
1073 | |
1074 | $gz->gzclose() ; |
1075 | } |
1076 | |
1077 | This script, I<gzstream>, does the opposite of the I<gzcat> script |
1078 | above. It reads from standard input and writes a gzip file to standard |
1079 | output. |
1080 | |
1081 | use strict ; |
1082 | use warnings ; |
1083 | |
1084 | use Compress::Zlib ; |
1085 | |
1086 | binmode STDOUT; # gzopen only sets it on the fd |
1087 | |
1088 | my $gz = gzopen(\*STDOUT, "wb") |
1089 | or die "Cannot open stdout: $gzerrno\n" ; |
1090 | |
1091 | while (<>) { |
1092 | $gz->gzwrite($_) |
1093 | or die "error writing: $gzerrno\n" ; |
1094 | } |
1095 | |
1096 | $gz->gzclose ; |
1097 | |
1098 | =head2 Compress::Zlib::memGzip |
1099 | |
1100 | This function is used to create an in-memory gzip file. |
1101 | It creates a minimal gzip header. |
1102 | |
1103 | $dest = Compress::Zlib::memGzip($buffer) ; |
1104 | |
1105 | If successful, it returns the in-memory gzip file, otherwise it returns |
1106 | undef. |
1107 | |
1108 | The buffer parameter can either be a scalar or a scalar reference. |
1109 | |
1110 | =head2 Compress::Zlib::memGunzip |
1111 | |
1112 | This function is used to uncompress an in-memory gzip file. |
1113 | |
1114 | $dest = Compress::Zlib::memGunzip($buffer) ; |
1115 | |
1116 | If successful, it returns the uncompressed gzip file, otherwise it |
1117 | returns undef. |
1118 | |
1119 | The buffer parameter can either be a scalar or a scalar reference. The |
1120 | contents of the buffer parameter are destroyed after calling this |
1121 | function. |
1122 | |
1123 | =head1 CHECKSUM FUNCTIONS |
1124 | |
1125 | Two functions are provided by I<zlib> to calculate a checksum. For the |
1126 | Perl interface, the order of the two parameters in both functions has |
1127 | been reversed. This allows both running checksums and one off |
1128 | calculations to be done. |
1129 | |
1130 | $crc = adler32($buffer [,$crc]) ; |
1131 | $crc = crc32($buffer [,$crc]) ; |
1132 | |
1133 | The buffer parameters can either be a scalar or a scalar reference. |
1134 | |
1135 | If the $crc parameters is C<undef>, the crc value will be reset. |
1136 | |
1137 | =head1 ACCESSING ZIP FILES |
1138 | |
1139 | Although it is possible to use this module to access .zip files, there |
1140 | is a module on CPAN that will do all the hard work for you. Check out |
1141 | |
1142 | http://www.cpan.org/modules/by-module/Archive/Archive-Zip-*.tar.gz |
1143 | |
1144 | Assuming you don't want to use this module to access zip files there |
1145 | are a number of undocumented features in the zlib library you need to |
1146 | be aware of. |
1147 | |
1148 | =over 5 |
1149 | |
1150 | =item 1. |
1151 | |
1152 | When calling B<inflateInit> or B<deflateInit> the B<WindowBits> parameter |
1153 | must be set to C<-MAX_WBITS>. This disables the creation of the zlib |
1154 | header. |
1155 | |
1156 | =item 2. |
1157 | |
1158 | The zlib function B<inflate>, and so the B<inflate> method supplied in |
1159 | this module, assume that there is at least one trailing byte after the |
1160 | compressed data stream. Normally this isn't a problem because both |
1161 | the gzip and zip file formats will guarantee that there is data directly |
1162 | after the compressed data stream. |
1163 | |
1164 | =back |
1165 | |
1166 | =head1 CONSTANTS |
1167 | |
1168 | All the I<zlib> constants are automatically imported when you make use |
1169 | of I<Compress::Zlib>. |
1170 | |
1171 | =head1 AUTHOR |
1172 | |
1173 | The I<Compress::Zlib> module was written by Paul Marquess, |
1174 | F<pmqs@cpan.org>. The latest copy of the module can be |
1175 | found on CPAN in F<modules/by-module/Compress/Compress-Zlib-x.x.tar.gz>. |
1176 | |
1177 | The primary site for the I<zlib> compression library is |
1178 | F<http://www.zlib.org>. |
1179 | |
1180 | =head1 MODIFICATION HISTORY |
1181 | |
1182 | See the Changes file. |