1 package IO::Uncompress::Inflate ;
8 use IO::Compress::Base::Common 2.036 qw(:Status createSelfTiedObject);
9 use IO::Compress::Zlib::Constants 2.036 ;
11 use IO::Uncompress::RawInflate 2.036 ;
14 our ($VERSION, @ISA, @EXPORT_OK, %EXPORT_TAGS, $InflateError);
19 @ISA = qw( Exporter IO::Uncompress::RawInflate );
20 @EXPORT_OK = qw( $InflateError inflate ) ;
21 %EXPORT_TAGS = %IO::Uncompress::RawInflate::DEFLATE_CONSTANTS ;
22 push @{ $EXPORT_TAGS{all} }, @EXPORT_OK ;
23 Exporter::export_ok_tags('all');
29 my $obj = createSelfTiedObject($class, \$InflateError);
31 $obj->_create(undef, 0, @_);
36 my $obj = createSelfTiedObject(undef, \$InflateError);
37 return $obj->_inf(@_);
50 # gunzip always needs adler32
51 $got->value('ADLER32' => 1);
61 $self->smartReadExact(\$magic, ZLIB_HEADER_SIZE);
63 *$self->{HeaderPending} = $magic ;
65 return $self->HeaderError("Header size is " .
66 ZLIB_HEADER_SIZE . " bytes")
67 if length $magic != ZLIB_HEADER_SIZE;
69 #return $self->HeaderError("CRC mismatch.")
71 if ! $self->isZlibMagic($magic) ;
73 *$self->{Type} = 'rfc1950';
82 return $self->_readDeflateHeader($magic) ;
90 my $ADLER32 = unpack("N", $trailer) ;
91 *$self->{Info}{ADLER32} = $ADLER32;
92 return $self->TrailerError("CRC mismatch")
93 if *$self->{Strict} && $ADLER32 != *$self->{Uncomp}->adler32() ;
106 if length $buffer < ZLIB_HEADER_SIZE ;
108 my $hdr = unpack("n", $buffer) ;
109 #return 0 if $hdr % 31 != 0 ;
110 return $self->HeaderError("CRC mismatch.")
113 my ($CMF, $FLG) = unpack "C C", $buffer;
114 my $cm = bits($CMF, ZLIB_CMF_CM_OFFSET, ZLIB_CMF_CM_BITS) ;
116 # Only Deflate supported
117 return $self->HeaderError("Not Deflate (CM is $cm)")
118 if $cm != ZLIB_CMF_CM_DEFLATED ;
120 # Max window value is 7 for Deflate.
121 my $cinfo = bits($CMF, ZLIB_CMF_CINFO_OFFSET, ZLIB_CMF_CINFO_BITS) ;
122 return $self->HeaderError("CINFO > " . ZLIB_CMF_CINFO_MAX .
123 " (CINFO is $cinfo)")
124 if $cinfo > ZLIB_CMF_CINFO_MAX ;
135 ($data >> $offset ) & $mask & 0xFF ;
139 sub _readDeflateHeader
141 my ($self, $buffer) = @_ ;
144 # $self->smartReadExact(\$buffer, ZLIB_HEADER_SIZE);
146 # *$self->{HeaderPending} = $buffer ;
148 # return $self->HeaderError("Header size is " .
149 # ZLIB_HEADER_SIZE . " bytes")
150 # if length $buffer != ZLIB_HEADER_SIZE;
152 # return $self->HeaderError("CRC mismatch.")
153 # if ! isZlibMagic($buffer) ;
156 my ($CMF, $FLG) = unpack "C C", $buffer;
157 my $FDICT = bits($FLG, ZLIB_FLG_FDICT_OFFSET, ZLIB_FLG_FDICT_BITS ),
159 my $cm = bits($CMF, ZLIB_CMF_CM_OFFSET, ZLIB_CMF_CM_BITS) ;
160 $cm == ZLIB_CMF_CM_DEFLATED
161 or return $self->HeaderError("Not Deflate (CM is $cm)") ;
165 $self->smartReadExact(\$buffer, ZLIB_FDICT_SIZE)
166 or return $self->TruncatedHeader("FDICT");
168 $DICTID = unpack("N", $buffer) ;
171 *$self->{Type} = 'rfc1950';
175 'FingerprintLength' => ZLIB_HEADER_SIZE,
176 'HeaderLength' => ZLIB_HEADER_SIZE,
177 'TrailerLength' => ZLIB_TRAILER_SIZE,
181 CM => bits($CMF, ZLIB_CMF_CM_OFFSET, ZLIB_CMF_CM_BITS ),
182 CINFO => bits($CMF, ZLIB_CMF_CINFO_OFFSET, ZLIB_CMF_CINFO_BITS ),
184 FCHECK => bits($FLG, ZLIB_FLG_FCHECK_OFFSET, ZLIB_FLG_FCHECK_BITS),
185 FDICT => bits($FLG, ZLIB_FLG_FDICT_OFFSET, ZLIB_FLG_FDICT_BITS ),
186 FLEVEL => bits($FLG, ZLIB_FLG_LEVEL_OFFSET, ZLIB_FLG_LEVEL_BITS ),
202 IO::Uncompress::Inflate - Read RFC 1950 files/buffers
206 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
208 my $status = inflate $input => $output [,OPTS]
209 or die "inflate failed: $InflateError\n";
211 my $z = new IO::Uncompress::Inflate $input [OPTS]
212 or die "inflate failed: $InflateError\n";
214 $status = $z->read($buffer)
215 $status = $z->read($buffer, $length)
216 $status = $z->read($buffer, $length, $offset)
217 $line = $z->getline()
222 $status = $z->inflateSync()
224 $data = $z->trailingData()
225 $status = $z->nextStream()
226 $data = $z->getHeaderInfo()
228 $z->seek($position, $whence)
240 read($z, $buffer, $length);
241 read($z, $buffer, $length, $offset);
243 seek($z, $position, $whence)
251 This module provides a Perl interface that allows the reading of
252 files/buffers that conform to RFC 1950.
254 For writing RFC 1950 files/buffers, see the companion module IO::Compress::Deflate.
256 =head1 Functional Interface
258 A top-level function, C<inflate>, is provided to carry out
259 "one-shot" uncompression between buffers and/or files. For finer
260 control over the uncompression process, see the L</"OO Interface">
263 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
265 inflate $input => $output [,OPTS]
266 or die "inflate failed: $InflateError\n";
268 The functional interface needs Perl5.005 or better.
270 =head2 inflate $input => $output [, OPTS]
272 C<inflate> expects at least two parameters, C<$input> and C<$output>.
274 =head3 The C<$input> parameter
276 The parameter, C<$input>, is used to define the source of
279 It can take one of the following forms:
285 If the C<$input> parameter is a simple scalar, it is assumed to be a
286 filename. This file will be opened for reading and the input data
287 will be read from it.
291 If the C<$input> parameter is a filehandle, the input data will be
293 The string '-' can be used as an alias for standard input.
295 =item A scalar reference
297 If C<$input> is a scalar reference, the input data will be read
300 =item An array reference
302 If C<$input> is an array reference, each element in the array must be a
305 The input data will be read from each file in turn.
307 The complete array will be walked to ensure that it only
308 contains valid filenames before any data is uncompressed.
310 =item An Input FileGlob string
312 If C<$input> is a string that is delimited by the characters "<" and ">"
313 C<inflate> will assume that it is an I<input fileglob string>. The
314 input is the list of files that match the fileglob.
316 See L<File::GlobMapper|File::GlobMapper> for more details.
320 If the C<$input> parameter is any other type, C<undef> will be returned.
322 =head3 The C<$output> parameter
324 The parameter C<$output> is used to control the destination of the
325 uncompressed data. This parameter can take one of these forms.
331 If the C<$output> parameter is a simple scalar, it is assumed to be a
332 filename. This file will be opened for writing and the uncompressed
333 data will be written to it.
337 If the C<$output> parameter is a filehandle, the uncompressed data
338 will be written to it.
339 The string '-' can be used as an alias for standard output.
341 =item A scalar reference
343 If C<$output> is a scalar reference, the uncompressed data will be
344 stored in C<$$output>.
346 =item An Array Reference
348 If C<$output> is an array reference, the uncompressed data will be
349 pushed onto the array.
351 =item An Output FileGlob
353 If C<$output> is a string that is delimited by the characters "<" and ">"
354 C<inflate> will assume that it is an I<output fileglob string>. The
355 output is the list of files that match the fileglob.
357 When C<$output> is an fileglob string, C<$input> must also be a fileglob
358 string. Anything else is an error.
360 See L<File::GlobMapper|File::GlobMapper> for more details.
364 If the C<$output> parameter is any other type, C<undef> will be returned.
368 When C<$input> maps to multiple compressed files/buffers and C<$output> is
369 a single file/buffer, after uncompression C<$output> will contain a
370 concatenation of all the uncompressed data from each of the input
373 =head2 Optional Parameters
375 Unless specified below, the optional parameters for C<inflate>,
376 C<OPTS>, are the same as those used with the OO interface defined in the
377 L</"Constructor Options"> section below.
381 =item C<< AutoClose => 0|1 >>
383 This option applies to any input or output data streams to
384 C<inflate> that are filehandles.
386 If C<AutoClose> is specified, and the value is true, it will result in all
387 input and/or output filehandles being closed once C<inflate> has
390 This parameter defaults to 0.
392 =item C<< BinModeOut => 0|1 >>
394 When writing to a file or filehandle, set C<binmode> before writing to the
399 =item C<< Append => 0|1 >>
401 The behaviour of this option is dependent on the type of output data
408 If C<Append> is enabled, all uncompressed data will be append to the end of
409 the output buffer. Otherwise the output buffer will be cleared before any
410 uncompressed data is written to it.
414 If C<Append> is enabled, the file will be opened in append mode. Otherwise
415 the contents of the file, if any, will be truncated before any uncompressed
416 data is written to it.
420 If C<Append> is enabled, the filehandle will be positioned to the end of
421 the file via a call to C<seek> before any uncompressed data is
422 written to it. Otherwise the file pointer will not be moved.
426 When C<Append> is specified, and set to true, it will I<append> all uncompressed
427 data to the output data stream.
429 So when the output is a filehandle it will carry out a seek to the eof
430 before writing any uncompressed data. If the output is a filename, it will be opened for
431 appending. If the output is a buffer, all uncompressed data will be
432 appended to the existing buffer.
434 Conversely when C<Append> is not specified, or it is present and is set to
435 false, it will operate as follows.
437 When the output is a filename, it will truncate the contents of the file
438 before writing any uncompressed data. If the output is a filehandle
439 its position will not be changed. If the output is a buffer, it will be
440 wiped before any uncompressed data is output.
444 =item C<< MultiStream => 0|1 >>
446 If the input file/buffer contains multiple compressed data streams, this
447 option will uncompress the whole lot as a single data stream.
451 =item C<< TrailingData => $scalar >>
453 Returns the data, if any, that is present immediately after the compressed
454 data stream once uncompression is complete.
456 This option can be used when there is useful information immediately
457 following the compressed data stream, and you don't know the length of the
458 compressed data stream.
460 If the input is a buffer, C<trailingData> will return everything from the
461 end of the compressed data stream to the end of the buffer.
463 If the input is a filehandle, C<trailingData> will return the data that is
464 left in the filehandle input buffer once the end of the compressed data
465 stream has been reached. You can then use the filehandle to read the rest
468 Don't bother using C<trailingData> if the input is a filename.
470 If you know the length of the compressed data stream before you start
471 uncompressing, you can avoid having to use C<trailingData> by setting the
472 C<InputLength> option.
478 To read the contents of the file C<file1.txt.1950> and write the
479 uncompressed data to the file C<file1.txt>.
483 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
485 my $input = "file1.txt.1950";
486 my $output = "file1.txt";
487 inflate $input => $output
488 or die "inflate failed: $InflateError\n";
490 To read from an existing Perl filehandle, C<$input>, and write the
491 uncompressed data to a buffer, C<$buffer>.
495 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
498 my $input = new IO::File "<file1.txt.1950"
499 or die "Cannot open 'file1.txt.1950': $!\n" ;
501 inflate $input => \$buffer
502 or die "inflate failed: $InflateError\n";
504 To uncompress all files in the directory "/my/home" that match "*.txt.1950" and store the compressed data in the same directory
508 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
510 inflate '</my/home/*.txt.1950>' => '</my/home/#1.txt>'
511 or die "inflate failed: $InflateError\n";
513 and if you want to compress each file one at a time, this will do the trick
517 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
519 for my $input ( glob "/my/home/*.txt.1950" )
522 $output =~ s/.1950// ;
523 inflate $input => $output
524 or die "Error compressing '$input': $InflateError\n";
531 The format of the constructor for IO::Uncompress::Inflate is shown below
533 my $z = new IO::Uncompress::Inflate $input [OPTS]
534 or die "IO::Uncompress::Inflate failed: $InflateError\n";
536 Returns an C<IO::Uncompress::Inflate> object on success and undef on failure.
537 The variable C<$InflateError> will contain an error message on failure.
539 If you are running Perl 5.005 or better the object, C<$z>, returned from
540 IO::Uncompress::Inflate can be used exactly like an L<IO::File|IO::File> filehandle.
541 This means that all normal input file operations can be carried out with
542 C<$z>. For example, to read a line from a compressed file/buffer you can
543 use either of these forms
545 $line = $z->getline();
548 The mandatory parameter C<$input> is used to determine the source of the
549 compressed data. This parameter can take one of three forms.
555 If the C<$input> parameter is a scalar, it is assumed to be a filename. This
556 file will be opened for reading and the compressed data will be read from it.
560 If the C<$input> parameter is a filehandle, the compressed data will be
562 The string '-' can be used as an alias for standard input.
564 =item A scalar reference
566 If C<$input> is a scalar reference, the compressed data will be read from
571 =head2 Constructor Options
573 The option names defined below are case insensitive and can be optionally
574 prefixed by a '-'. So all of the following are valid
581 OPTS is a combination of the following options:
585 =item C<< AutoClose => 0|1 >>
587 This option is only valid when the C<$input> parameter is a filehandle. If
588 specified, and the value is true, it will result in the file being closed once
589 either the C<close> method is called or the IO::Uncompress::Inflate object is
592 This parameter defaults to 0.
594 =item C<< MultiStream => 0|1 >>
596 Allows multiple concatenated compressed streams to be treated as a single
597 compressed stream. Decompression will stop once either the end of the
598 file/buffer is reached, an error is encountered (premature eof, corrupt
599 compressed data) or the end of a stream is not immediately followed by the
600 start of another stream.
602 This parameter defaults to 0.
604 =item C<< Prime => $string >>
606 This option will uncompress the contents of C<$string> before processing the
609 This option can be useful when the compressed data is embedded in another
610 file/data structure and it is not possible to work out where the compressed
611 data begins without having to read the first few bytes. If this is the
612 case, the uncompression can be I<primed> with these bytes using this
615 =item C<< Transparent => 0|1 >>
617 If this option is set and the input file/buffer is not compressed data,
618 the module will allow reading of it anyway.
620 In addition, if the input file/buffer does contain compressed data and
621 there is non-compressed data immediately following it, setting this option
622 will make this module treat the whole file/buffer as a single data stream.
624 This option defaults to 1.
626 =item C<< BlockSize => $num >>
628 When reading the compressed input data, IO::Uncompress::Inflate will read it in
629 blocks of C<$num> bytes.
631 This option defaults to 4096.
633 =item C<< InputLength => $size >>
635 When present this option will limit the number of compressed bytes read
636 from the input file/buffer to C<$size>. This option can be used in the
637 situation where there is useful data directly after the compressed data
638 stream and you know beforehand the exact length of the compressed data
641 This option is mostly used when reading from a filehandle, in which case
642 the file pointer will be left pointing to the first byte directly after the
643 compressed data stream.
645 This option defaults to off.
647 =item C<< Append => 0|1 >>
649 This option controls what the C<read> method does with uncompressed data.
651 If set to 1, all uncompressed data will be appended to the output parameter
652 of the C<read> method.
654 If set to 0, the contents of the output parameter of the C<read> method
655 will be overwritten by the uncompressed data.
659 =item C<< Strict => 0|1 >>
661 This option controls whether the extra checks defined below are used when
662 carrying out the decompression. When Strict is on, the extra tests are
663 carried out, when Strict is off they are not.
665 The default for this option is off.
671 The ADLER32 checksum field must be present.
675 The value of the ADLER32 field read must match the adler32 value of the
676 uncompressed data actually contained in the file.
692 $status = $z->read($buffer)
694 Reads a block of compressed data (the size the the compressed block is
695 determined by the C<Buffer> option in the constructor), uncompresses it and
696 writes any uncompressed data into C<$buffer>. If the C<Append> parameter is
697 set in the constructor, the uncompressed data will be appended to the
698 C<$buffer> parameter. Otherwise C<$buffer> will be overwritten.
700 Returns the number of uncompressed bytes written to C<$buffer>, zero if eof
701 or a negative number on error.
707 $status = $z->read($buffer, $length)
708 $status = $z->read($buffer, $length, $offset)
710 $status = read($z, $buffer, $length)
711 $status = read($z, $buffer, $length, $offset)
713 Attempt to read C<$length> bytes of uncompressed data into C<$buffer>.
715 The main difference between this form of the C<read> method and the
716 previous one, is that this one will attempt to return I<exactly> C<$length>
717 bytes. The only circumstances that this function will not is if end-of-file
718 or an IO error is encountered.
720 Returns the number of uncompressed bytes written to C<$buffer>, zero if eof
721 or a negative number on error.
727 $line = $z->getline()
732 This method fully supports the use of of the variable C<$/> (or
733 C<$INPUT_RECORD_SEPARATOR> or C<$RS> when C<English> is in use) to
734 determine what constitutes an end of line. Paragraph mode, record mode and
735 file slurp mode are all supported.
743 Read a single character.
749 $char = $z->ungetc($string)
755 $status = $z->inflateSync()
763 $hdr = $z->getHeaderInfo();
764 @hdrs = $z->getHeaderInfo();
766 This method returns either a hash reference (in scalar context) or a list
767 or hash references (in array context) that contains information about each
768 of the header fields in the compressed data stream(s).
777 Returns the uncompressed file offset.
786 Returns true if the end of the compressed input stream has been reached.
790 $z->seek($position, $whence);
791 seek($z, $position, $whence);
793 Provides a sub-set of the C<seek> functionality, with the restriction
794 that it is only legal to seek forward in the input file/buffer.
795 It is a fatal error to attempt to seek backward.
797 The C<$whence> parameter takes one the usual values, namely SEEK_SET,
798 SEEK_CUR or SEEK_END.
800 Returns 1 on success, 0 on failure.
809 This is a noop provided for completeness.
815 Returns true if the object currently refers to a opened file/buffer.
819 my $prev = $z->autoflush()
820 my $prev = $z->autoflush(EXPR)
822 If the C<$z> object is associated with a file or a filehandle, this method
823 returns the current autoflush setting for the underlying filehandle. If
824 C<EXPR> is present, and is non-zero, it will enable flushing after every
825 write/print operation.
827 If C<$z> is associated with a buffer, this method has no effect and always
830 B<Note> that the special variable C<$|> B<cannot> be used to set or
831 retrieve the autoflush setting.
833 =head2 input_line_number
835 $z->input_line_number()
836 $z->input_line_number(EXPR)
838 Returns the current uncompressed line number. If C<EXPR> is present it has
839 the effect of setting the line number. Note that setting the line number
840 does not change the current position within the file/buffer being read.
842 The contents of C<$/> are used to to determine what constitutes a line
850 If the C<$z> object is associated with a file or a filehandle, C<fileno>
851 will return the underlying file descriptor. Once the C<close> method is
852 called C<fileno> will return C<undef>.
854 If the C<$z> object is associated with a buffer, this method will return
862 Closes the output file/buffer.
864 For most versions of Perl this method will be automatically invoked if
865 the IO::Uncompress::Inflate object is destroyed (either explicitly or by the
866 variable with the reference to the object going out of scope). The
867 exceptions are Perl versions 5.005 through 5.00504 and 5.8.0. In
868 these cases, the C<close> method will be called automatically, but
869 not until global destruction of all live objects when the program is
872 Therefore, if you want your scripts to be able to run on all versions
873 of Perl, you should call C<close> explicitly and not rely on automatic
876 Returns true on success, otherwise 0.
878 If the C<AutoClose> option has been enabled when the IO::Uncompress::Inflate
879 object was created, and the object is associated with a file, the
880 underlying file will also be closed.
886 my $status = $z->nextStream();
888 Skips to the next compressed data stream in the input file/buffer. If a new
889 compressed data stream is found, the eof marker will be cleared and C<$.>
892 Returns 1 if a new stream was found, 0 if none was found, and -1 if an
893 error was encountered.
899 my $data = $z->trailingData();
901 Returns the data, if any, that is present immediately after the compressed
902 data stream once uncompression is complete. It only makes sense to call
903 this method once the end of the compressed data stream has been
906 This option can be used when there is useful information immediately
907 following the compressed data stream, and you don't know the length of the
908 compressed data stream.
910 If the input is a buffer, C<trailingData> will return everything from the
911 end of the compressed data stream to the end of the buffer.
913 If the input is a filehandle, C<trailingData> will return the data that is
914 left in the filehandle input buffer once the end of the compressed data
915 stream has been reached. You can then use the filehandle to read the rest
918 Don't bother using C<trailingData> if the input is a filename.
920 If you know the length of the compressed data stream before you start
921 uncompressing, you can avoid having to use C<trailingData> by setting the
922 C<InputLength> option in the constructor.
926 No symbolic constants are required by this IO::Uncompress::Inflate at present.
932 Imports C<inflate> and C<$InflateError>.
935 use IO::Uncompress::Inflate qw(inflate $InflateError) ;
941 =head2 Working with Net::FTP
943 See L<IO::Uncompress::Inflate::FAQ|IO::Uncompress::Inflate::FAQ/"Compressed files and Net::FTP">
947 L<Compress::Zlib>, L<IO::Compress::Gzip>, L<IO::Uncompress::Gunzip>, L<IO::Compress::Deflate>, L<IO::Compress::RawDeflate>, L<IO::Uncompress::RawInflate>, L<IO::Compress::Bzip2>, L<IO::Uncompress::Bunzip2>, L<IO::Compress::Lzma>, L<IO::Uncompress::UnLzma>, L<IO::Compress::Xz>, L<IO::Uncompress::UnXz>, L<IO::Compress::Lzop>, L<IO::Uncompress::UnLzop>, L<IO::Compress::Lzf>, L<IO::Uncompress::UnLzf>, L<IO::Uncompress::AnyInflate>, L<IO::Uncompress::AnyUncompress>
949 L<Compress::Zlib::FAQ|Compress::Zlib::FAQ>
951 L<File::GlobMapper|File::GlobMapper>, L<Archive::Zip|Archive::Zip>,
952 L<Archive::Tar|Archive::Tar>,
955 For RFC 1950, 1951 and 1952 see
956 F<http://www.faqs.org/rfcs/rfc1950.html>,
957 F<http://www.faqs.org/rfcs/rfc1951.html> and
958 F<http://www.faqs.org/rfcs/rfc1952.html>
960 The I<zlib> compression library was written by Jean-loup Gailly
961 F<gzip@prep.ai.mit.edu> and Mark Adler F<madler@alumni.caltech.edu>.
963 The primary site for the I<zlib> compression library is
964 F<http://www.zlib.org>.
966 The primary site for gzip is F<http://www.gzip.org>.
970 This module was written by Paul Marquess, F<pmqs@cpan.org>.
972 =head1 MODIFICATION HISTORY
974 See the Changes file.
976 =head1 COPYRIGHT AND LICENSE
978 Copyright (c) 2005-2011 Paul Marquess. All rights reserved.
980 This program is free software; you can redistribute it and/or
981 modify it under the same terms as Perl itself.