Compress::Raw::Zlib - Low-Level Interface to zlib compression library
use Compress::Raw::Zlib ;
($d, $status) = new Compress::Raw::Zlib::Deflate( [OPT] ) ;
$status = $d->deflate($input, $output) ;
$status = $d->flush($output [, $flush_type]) ;
$d->deflateReset() ;
$d->deflateParams(OPTS) ;
$d->deflateTune(OPTS) ;
$d->dict_adler() ;
$d->crc32() ;
$d->adler32() ;
$d->total_in() ;
$d->total_out() ;
$d->msg() ;
$d->get_Strategy();
$d->get_Level();
$d->get_BufSize();
($i, $status) = new Compress::Raw::Zlib::Inflate( [OPT] ) ;
$status = $i->inflate($input, $output [, $eof]) ;
$status = $i->inflateSync($input) ;
$i->inflateReset() ;
$i->dict_adler() ;
$d->crc32() ;
$d->adler32() ;
$i->total_in() ;
$i->total_out() ;
$i->msg() ;
$d->get_BufSize();
$crc = adler32($buffer [,$crc]) ;
$crc = crc32($buffer [,$crc]) ;
$crc = crc32_combine($crc1, $crc2, $len2);
$adler = adler32_combine($adler1, $adler2, $len2);
my $version = Compress::Raw::Zlib::zlib_version();
my $flags = Compress::Raw::Zlib::zlibCompileFlags();
The Compress::Raw::Zlib module provides a Perl interface to the zlib compression library (see "AUTHOR" for details about where to get zlib).
This section defines an interface that allows in-memory compression using the deflate interface provided by zlib.
Here is a definition of the interface available:
Initialises a deflation object.
If you are familiar with the zlib library, it combines the features of the zlib functions deflateInit
, deflateInit2
and deflateSetDictionary
.
If successful, it will return the initialised deflation object, $d
and a $status
of Z_OK
in a list context. In scalar context it returns the deflation object, $d
, only.
If not successful, the returned deflation object, $d
, will be undef and $status
will hold the a zlib error code.
The function optionally takes a number of named options specified as Name => value
pairs. This allows individual options to be tailored without having to specify them all in the parameter list.
For backward compatibility, it is also possible to pass the parameters as a reference to a hash containing the name=>value pairs.
Below is a list of the valid options:
Defines the compression level. Valid values are 0 through 9, Z_NO_COMPRESSION
, Z_BEST_SPEED
, Z_BEST_COMPRESSION
, and Z_DEFAULT_COMPRESSION
.
The default is Z_DEFAULT_COMPRESSION
.
Defines the compression method. The only valid value at present (and the default) is Z_DEFLATED
.
To compress an RFC 1950 data stream, set WindowBits
to a positive number between 8 and 15.
To compress an RFC 1951 data stream, set WindowBits
to -MAX_WBITS
.
To compress an RFC 1952 data stream (i.e. gzip), set WindowBits
to WANT_GZIP
.
For a definition of the meaning and valid values for WindowBits
refer to the zlib documentation for deflateInit2.
Defaults to MAX_WBITS
.
For a definition of the meaning and valid values for MemLevel
refer to the zlib documentation for deflateInit2.
Defaults to MAX_MEM_LEVEL.
Defines the strategy used to tune the compression. The valid values are Z_DEFAULT_STRATEGY
, Z_FILTERED
, Z_RLE
, Z_FIXED
and Z_HUFFMAN_ONLY
.
The default is Z_DEFAULT_STRATEGY
.
When a dictionary is specified Compress::Raw::Zlib will automatically call deflateSetDictionary
directly after calling deflateInit
. The Adler32 value for the dictionary can be obtained by calling the method $d->dict_adler()
.
The default is no dictionary.
Sets the initial size for the output buffer used by the $d->deflate
and $d->flush
methods. If the buffer has to be reallocated to increase the size, it will grow in increments of Bufsize
.
The default buffer size is 4096.
This option controls how data is written to the output buffer by the $d->deflate
and $d->flush
methods.
If the AppendOutput
option is set to false, the output buffers in the $d->deflate
and $d->flush
methods will be truncated before uncompressed data is written to them.
If the option is set to true, uncompressed data will be appended to the output buffer in the $d->deflate
and $d->flush
methods.
This option defaults to false.
If set to true, a crc32 checksum of the uncompressed data will be calculated. Use the $d->crc32
method to retrieve this value.
This option defaults to false.
If set to true, an adler32 checksum of the uncompressed data will be calculated. Use the $d->adler32
method to retrieve this value.
This option defaults to false.
Here is an example of using the Compress::Raw::Zlib::Deflate
optional parameter list to override the default buffer size and compression level. All other options will take their default values.
my $d = new Compress::Raw::Zlib::Deflate ( -Bufsize => 300,
-Level => Z_BEST_SPEED ) ;
Deflates the contents of $input
and writes the compressed data to $output
.
The $input
and $output
parameters can be either scalars or scalar references.
When finished, $input
will be completely processed (assuming there were no errors). If the deflation was successful it writes the deflated data to $output
and returns a status value of Z_OK
.
On error, it returns a zlib error code.
If the AppendOutput
option is set to true in the constructor for the $d
object, the compressed data will be appended to $output
. If it is false, $output
will be truncated before any compressed data is written to it.
Note: This method will not necessarily write compressed data to $output
every time it is called. So do not assume that there has been an error if the contents of $output
is empty on returning from this method. As long as the return code from the method is Z_OK
, the deflate has succeeded.
Typically used to finish the deflation. Any pending output will be written to $output
.
Returns Z_OK
if successful.
Note that flushing can seriously degrade the compression ratio, so it should only be used to terminate a decompression (using Z_FINISH
) or when you want to create a full flush point (using Z_FULL_FLUSH
).
By default the flush_type
used is Z_FINISH
. Other valid values for flush_type
are Z_NO_FLUSH
, Z_PARTIAL_FLUSH
, Z_SYNC_FLUSH
and Z_FULL_FLUSH
. It is strongly recommended that you only set the flush_type
parameter if you fully understand the implications of what it does. See the zlib
documentation for details.
If the AppendOutput
option is set to true in the constructor for the $d
object, the compressed data will be appended to $output
. If it is false, $output
will be truncated before any compressed data is written to it.
This method will reset the deflation object $d
. It can be used when you are compressing multiple data streams and want to use the same object to compress each of them. It should only be used once the previous data stream has been flushed successfully, i.e. a call to $d->flush(Z_FINISH)
has returned Z_OK
.
Returns Z_OK
if successful.
Change settings for the deflate object $d
.
The list of the valid options is shown below. Options not specified will remain unchanged.
Defines the compression level. Valid values are 0 through 9, Z_NO_COMPRESSION
, Z_BEST_SPEED
, Z_BEST_COMPRESSION
, and Z_DEFAULT_COMPRESSION
.
Defines the strategy used to tune the compression. The valid values are Z_DEFAULT_STRATEGY
, Z_FILTERED
and Z_HUFFMAN_ONLY
.
Sets the initial size for the output buffer used by the $d->deflate
and $d->flush
methods. If the buffer has to be reallocated to increase the size, it will grow in increments of Bufsize
.
Tune the internal settings for the deflate object $d
. This option is only available if you are running zlib 1.2.2.3 or better.
Refer to the documentation in zlib.h for instructions on how to fly deflateTune
.
Returns the adler32 value for the dictionary.
Returns the crc32 value for the uncompressed data to date.
If the CRC32
option is not enabled in the constructor for this object, this method will always return 0;
Returns the adler32 value for the uncompressed data to date.
Returns the last error message generated by zlib.
Returns the total number of bytes uncompressed bytes input to deflate.
Returns the total number of compressed bytes output from deflate.
Returns the deflation strategy currently used. Valid values are Z_DEFAULT_STRATEGY
, Z_FILTERED
and Z_HUFFMAN_ONLY
.
Returns the compression level being used.
Returns the buffer size used to carry out the compression.
Here is a trivial example of using deflate
. It simply reads standard input, deflates it and writes it to standard output.
use strict ;
use warnings ;
use Compress::Raw::Zlib ;
binmode STDIN;
binmode STDOUT;
my $x = new Compress::Raw::Zlib::Deflate
or die "Cannot create a deflation stream\n" ;
my ($output, $status) ;
while (<>)
{
$status = $x->deflate($_, $output) ;
$status == Z_OK
or die "deflation failed\n" ;
print $output ;
}
$status = $x->flush($output) ;
$status == Z_OK
or die "deflation failed\n" ;
print $output ;
This section defines an interface that allows in-memory uncompression using the inflate interface provided by zlib.
Here is a definition of the interface:
Initialises an inflation object.
In a list context it returns the inflation object, $i
, and the zlib status code ($status
). In a scalar context it returns the inflation object only.
If successful, $i
will hold the inflation object and $status
will be Z_OK
.
If not successful, $i
will be undef and $status
will hold the zlib error code.
The function optionally takes a number of named options specified as -Name => value
pairs. This allows individual options to be tailored without having to specify them all in the parameter list.
For backward compatibility, it is also possible to pass the parameters as a reference to a hash containing the name=>value
pairs.
Here is a list of the valid options:
To uncompress an RFC 1950 data stream, set WindowBits
to a positive number between 8 and 15.
To uncompress an RFC 1951 data stream, set WindowBits
to -MAX_WBITS
.
To uncompress an RFC 1952 data stream (i.e. gzip), set WindowBits
to WANT_GZIP
.
To auto-detect and uncompress an RFC 1950 or RFC 1952 data stream (i.e. gzip), set WindowBits
to WANT_GZIP_OR_ZLIB
.
For a full definition of the meaning and valid values for WindowBits
refer to the zlib documentation for inflateInit2.
Defaults to MAX_WBITS
.
Sets the initial size for the output buffer used by the $i->inflate
method. If the output buffer in this method has to be reallocated to increase the size, it will grow in increments of Bufsize
.
Default is 4096.
The default is no dictionary.
This option controls how data is written to the output buffer by the $i->inflate
method.
If the option is set to false, the output buffer in the $i->inflate
method will be truncated before uncompressed data is written to it.
If the option is set to true, uncompressed data will be appended to the output buffer by the $i->inflate
method.
This option defaults to false.
If set to true, a crc32 checksum of the uncompressed data will be calculated. Use the $i->crc32
method to retrieve this value.
This option defaults to false.
If set to true, an adler32 checksum of the uncompressed data will be calculated. Use the $i->adler32
method to retrieve this value.
This option defaults to false.
If set to true, this option will remove compressed data from the input buffer of the $i->inflate
method as the inflate progresses.
This option can be useful when you are processing compressed data that is embedded in another file/buffer. In this case the data that immediately follows the compressed stream will be left in the input buffer.
This option defaults to true.
The LimitOutput
option changes the behavior of the $i->inflate
method so that the amount of memory used by the output buffer can be limited.
When LimitOutput
is used the size of the output buffer used will either be the value of the Bufsize
option or the amount of memory already allocated to $output
, whichever is larger. Predicting the output size available is tricky, so don't rely on getting an exact output buffer size.
When LimitOutout
is not specified $i->inflate
will use as much memory as it takes to write all the uncompressed data it creates by uncompressing the input buffer.
If LimitOutput
is enabled, the ConsumeInput
option will also be enabled.
This option defaults to false.
See "The LimitOutput option" for a discussion on why LimitOutput
is needed and how to use it.
Here is an example of using an optional parameter to override the default buffer size.
my ($i, $status) = new Compress::Raw::Zlib::Inflate( -Bufsize => 300 ) ;
Inflates the complete contents of $input
and writes the uncompressed data to $output
. The $input
and $output
parameters can either be scalars or scalar references.
Returns Z_OK
if successful and Z_STREAM_END
if the end of the compressed data has been successfully reached.
If not successful $status
will hold the zlib error code.
If the ConsumeInput
option has been set to true when the Compress::Raw::Zlib::Inflate
object is created, the $input
parameter is modified by inflate
. On completion it will contain what remains of the input buffer after inflation. In practice, this means that when the return status is Z_OK
the $input
parameter will contain an empty string, and when the return status is Z_STREAM_END
the $input
parameter will contains what (if anything) was stored in the input buffer after the deflated data stream.
This feature is useful when processing a file format that encapsulates a compressed data stream (e.g. gzip, zip) and there is useful data immediately after the deflation stream.
If the AppendOutput
option is set to true in the constructor for this object, the uncompressed data will be appended to $output
. If it is false, $output
will be truncated before any uncompressed data is written to it.
The $eof
parameter needs a bit of explanation.
Prior to version 1.2.0, zlib assumed that there was at least one trailing byte immediately after the compressed data stream when it was carrying out decompression. This normally isn't a problem because the majority of zlib applications guarantee that there will be data directly after the compressed data stream. For example, both gzip (RFC 1950) and zip both define trailing data that follows the compressed data stream.
The $eof
parameter only needs to be used if all of the following conditions apply
You are either using a copy of zlib that is older than version 1.2.0 or you want your application code to be able to run with as many different versions of zlib as possible.
You have set the WindowBits
parameter to -MAX_WBITS
in the constructor for this object, i.e. you are uncompressing a raw deflated data stream (RFC 1951).
There is no data immediately after the compressed data stream.
If all of these are the case, then you need to set the $eof
parameter to true on the final call (and only the final call) to $i->inflate
.
If you have built this module with zlib >= 1.2.0, the $eof
parameter is ignored. You can still set it if you want, but it won't be used behind the scenes.
This method can be used to attempt to recover good data from a compressed data stream that is partially corrupt. It scans $input
until it reaches either a full flush point or the end of the buffer.
If a full flush point is found, Z_OK
is returned and $input
will be have all data up to the flush point removed. This data can then be passed to the $i->inflate
method to be uncompressed.
Any other return code means that a flush point was not found. If more data is available, inflateSync
can be called repeatedly with more compressed data until the flush point is found.
Note full flush points are not present by default in compressed data streams. They must have been added explicitly when the data stream was created by calling Compress::Deflate::flush
with Z_FULL_FLUSH
.
This method will reset the inflation object $i
. It can be used when you are uncompressing multiple data streams and want to use the same object to uncompress each of them.
Returns Z_OK
if successful.
Returns the adler32 value for the dictionary.
Returns the crc32 value for the uncompressed data to date.
If the CRC32
option is not enabled in the constructor for this object, this method will always return 0;
Returns the adler32 value for the uncompressed data to date.
If the ADLER32
option is not enabled in the constructor for this object, this method will always return 0;
Returns the last error message generated by zlib.
Returns the total number of bytes compressed bytes input to inflate.
Returns the total number of uncompressed bytes output from inflate.
Returns the buffer size used to carry out the decompression.
Here is an example of using inflate
.
use strict ;
use warnings ;
use Compress::Raw::Zlib;
my $x = new Compress::Raw::Zlib::Inflate()
or die "Cannot create a inflation stream\n" ;
my $input = '' ;
binmode STDIN;
binmode STDOUT;
my ($output, $status) ;
while (read(STDIN, $input, 4096))
{
$status = $x->inflate($input, $output) ;
print $output ;
last if $status != Z_OK ;
}
die "inflation failed\n"
unless $status == Z_STREAM_END ;
The next example show how to use the LimitOutput
option. Notice the use of two nested loops in this case. The outer loop reads the data from the input source - STDIN and the inner loop repeatedly calls inflate
until $input
is exhausted, we get an error, or the end of the stream is reached. One point worth remembering is by using the LimitOutput
option you also get ConsumeInput
set as well - this makes the code below much simpler.
use strict ;
use warnings ;
use Compress::Raw::Zlib;
my $x = new Compress::Raw::Zlib::Inflate(LimitOutput => 1)
or die "Cannot create a inflation stream\n" ;
my $input = '' ;
binmode STDIN;
binmode STDOUT;
my ($output, $status) ;
OUTER:
while (read(STDIN, $input, 4096))
{
do
{
$status = $x->inflate($input, $output) ;
print $output ;
last OUTER
unless $status == Z_OK || $status == Z_BUF_ERROR ;
}
while ($status == Z_OK && length $input);
}
die "inflation failed\n"
unless $status == Z_STREAM_END ;
Two functions are provided by zlib to calculate checksums. For the Perl interface, the order of the two parameters in both functions has been reversed. This allows both running checksums and one off calculations to be done.
$crc = adler32($buffer [,$crc]) ;
$crc = crc32($buffer [,$crc]) ;
The buffer parameters can either be a scalar or a scalar reference.
If the $crc parameters is undef
, the crc value will be reset.
If you have built this module with zlib 1.2.3 or better, two more CRC-related functions are available.
$crc = crc32_combine($crc1, $crc2, $len2);
$adler = adler32_combine($adler1, $adler2, $len2);
These functions allow checksums to be merged. Refer to the zlib documentation for more details.
Returns the version of the zlib library.
Returns the flags indicating compile-time options that were used to build the zlib library. See the zlib documentation for a description of the flags returned by zlibCompileFlags
.
Note that when the zlib sources are built along with this module the sprintf
flags (bits 24, 25 and 26) should be ignored.
If you are using zlib 1.2.0 or older, zlibCompileFlags
will return 0.
By default $i->inflate($input, $output)
will uncompress all data in $input
and write all of the uncompressed data it has generated to $output
. This makes the interface to inflate
much simpler - if the method has uncompressed $input
successfully all compressed data in $input
will have been dealt with. So if you are reading from an input source and uncompressing as you go the code will look something like this
use strict ;
use warnings ;
use Compress::Raw::Zlib;
my $x = new Compress::Raw::Zlib::Inflate()
or die "Cannot create a inflation stream\n" ;
my $input = '' ;
my ($output, $status) ;
while (read(STDIN, $input, 4096))
{
$status = $x->inflate($input, $output) ;
print $output ;
last if $status != Z_OK ;
}
die "inflation failed\n"
unless $status == Z_STREAM_END ;
The points to note are
The main processing loop in the code handles reading of compressed data from STDIN.
The status code returned from inflate
will only trigger termination of the main processing loop if it isn't Z_OK
. When LimitOutput
has not been used the Z_OK
status means that the end of the compressed data stream has been reached or there has been an error in uncompression.
After the call to inflate
all of the uncompressed data in $input
will have been processed. This means the subsequent call to read
can overwrite it's contents without any problem.
For most use-cases the behavior described above is acceptable (this module and it's predecessor, Compress::Zlib
, have used it for over 10 years without an issue), but in a few very specific use-cases the amount of memory required for $output
can prohibitively large. For example, if the compressed data stream contains the same pattern repeated thousands of times, a relatively small compressed data stream can uncompress into hundreds of megabytes. Remember inflate
will keep allocating memory until all the uncompressed data has been written to the output buffer - the size of $output
is unbounded.
The LimitOutput
option is designed to help with this use-case.
The main difference in your code when using LimitOutput
is having to deal with cases where the $input
parameter still contains some uncompressed data that inflate
hasn't processed yet. The status code returned from inflate
will be Z_OK
if uncompression took place and Z_BUF_ERROR
if the output buffer is full.
Below is typical code that shows how to use LimitOutput
.
use strict ;
use warnings ;
use Compress::Raw::Zlib;
my $x = new Compress::Raw::Zlib::Inflate(LimitOutput => 1)
or die "Cannot create a inflation stream\n" ;
my $input = '' ;
binmode STDIN;
binmode STDOUT;
my ($output, $status) ;
OUTER:
while (read(STDIN, $input, 4096))
{
do
{
$status = $x->inflate($input, $output) ;
print $output ;
last OUTER
unless $status == Z_OK || $status == Z_BUF_ERROR ;
}
while ($status == Z_OK && length $input);
}
die "inflation failed\n"
unless $status == Z_STREAM_END ;
Points to note this time:
There are now two nested loops in the code: the outer loop for reading the compressed data from STDIN, as before; and the inner loop to carry out the uncompression.
There are two exit points from the inner uncompression loop.
Firstly when inflate
has returned a status other than Z_OK
or Z_BUF_ERROR
. This means that either the end of the compressed data stream has been reached (Z_STREAM_END
) or there is an error in the compressed data. In either of these cases there is no point in continuing with reading the compressed data, so both loops are terminated.
The second exit point tests if there is any data left in the input buffer, $input
- remember that the ConsumeInput
option is automatically enabled when LimitOutput
is used. When the input buffer has been exhausted, the outer loop can run again and overwrite a now empty $input
.
Although it is possible (with some effort on your part) to use this module to access .zip files, there are other perl modules available that will do all the hard work for you. Check out Archive::Zip
, Archive::Zip::SimpleZip
, IO::Compress::Zip
and IO::Uncompress::Unzip
.
This module is not compatible with Unix compress
.
If you have the uncompress
program available, you can use this to read compressed files
open F, "uncompress -c $filename |";
while (<F>)
{
...
Alternatively, if you have the gunzip
program available, you can use this to read compressed files
open F, "gunzip -c $filename |";
while (<F>)
{
...
and this to write compress files, if you have the compress
program available
open F, "| compress -c $filename ";
print F "data";
...
close F ;
See previous FAQ item.
If the Archive::Tar
module is installed and either the uncompress
or gunzip
programs are available, you can use one of these workarounds to read .tar.Z
files.
Firstly with uncompress
use strict;
use warnings;
use Archive::Tar;
open F, "uncompress -c $filename |";
my $tar = Archive::Tar->new(*F);
...
and this with gunzip
use strict;
use warnings;
use Archive::Tar;
open F, "gunzip -c $filename |";
my $tar = Archive::Tar->new(*F);
...
Similarly, if the compress
program is available, you can use this to write a .tar.Z
file
use strict;
use warnings;
use Archive::Tar;
use IO::File;
my $fh = new IO::File "| compress -c >$filename";
my $tar = Archive::Tar->new();
...
$tar->write($fh);
$fh->close ;
By default Compress::Raw::Zlib
will build with a private copy of version 1.2.5 of the zlib library. (See the README file for details of how to override this behaviour)
If you decide to use a different version of the zlib library, you need to be aware of the following issues
First off, you must have zlib 1.0.5 or better.
You need to have zlib 1.2.1 or better if you want to use the -Merge
option with IO::Compress::Gzip
, IO::Compress::Deflate
and IO::Compress::RawDeflate
.
All the zlib constants are automatically imported when you make use of Compress::Raw::Zlib.
Compress::Zlib, IO::Compress::Gzip, IO::Uncompress::Gunzip, IO::Compress::Deflate, IO::Uncompress::Inflate, IO::Compress::RawDeflate, IO::Uncompress::RawInflate, IO::Compress::Bzip2, IO::Uncompress::Bunzip2, IO::Compress::Lzma, IO::Uncompress::UnLzma, IO::Compress::Xz, IO::Uncompress::UnXz, IO::Compress::Lzop, IO::Uncompress::UnLzop, IO::Compress::Lzf, IO::Uncompress::UnLzf, IO::Uncompress::AnyInflate, IO::Uncompress::AnyUncompress
File::GlobMapper, Archive::Zip, Archive::Tar, IO::Zlib
For RFC 1950, 1951 and 1952 see http://www.faqs.org/rfcs/rfc1950.html, http://www.faqs.org/rfcs/rfc1951.html and http://www.faqs.org/rfcs/rfc1952.html
The zlib compression library was written by Jean-loup Gailly gzip@prep.ai.mit.edu
and Mark Adler madler@alumni.caltech.edu
.
The primary site for the zlib compression library is http://www.zlib.org.
The primary site for gzip is http://www.gzip.org.
This module was written by Paul Marquess, pmqs@cpan.org
.
See the Changes file.
Copyright (c) 2005-2017 Paul Marquess. All rights reserved.
This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.