|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
NAME
|
|
Packit |
71fd91 |
bzip2, bunzip2 - a block-sorting file compressor, v1.0.6
|
|
Packit |
71fd91 |
bzcat - decompresses files to stdout
|
|
Packit |
71fd91 |
bzip2recover - recovers data from damaged bzip2 files
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
SYNOPSIS
|
|
Packit |
71fd91 |
bzip2 [ -cdfkqstvzVL123456789 ] [ filenames ... ]
|
|
Packit |
71fd91 |
bunzip2 [ -fkvsVL ] [ filenames ... ]
|
|
Packit |
71fd91 |
bzcat [ -s ] [ filenames ... ]
|
|
Packit |
71fd91 |
bzip2recover filename
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
DESCRIPTION
|
|
Packit |
71fd91 |
bzip2 compresses files using the Burrows-Wheeler block
|
|
Packit |
71fd91 |
sorting text compression algorithm, and Huffman coding.
|
|
Packit |
71fd91 |
Compression is generally considerably better than that
|
|
Packit |
71fd91 |
achieved by more conventional LZ77/LZ78-based compressors,
|
|
Packit |
71fd91 |
and approaches the performance of the PPM family of sta-
|
|
Packit |
71fd91 |
tistical compressors.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
The command-line options are deliberately very similar to
|
|
Packit |
71fd91 |
those of GNU gzip, but they are not identical.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2 expects a list of file names to accompany the com-
|
|
Packit |
71fd91 |
mand-line flags. Each file is replaced by a compressed
|
|
Packit |
71fd91 |
version of itself, with the name "original_name.bz2".
|
|
Packit |
71fd91 |
Each compressed file has the same modification date, per-
|
|
Packit |
71fd91 |
missions, and, when possible, ownership as the correspond-
|
|
Packit |
71fd91 |
ing original, so that these properties can be correctly
|
|
Packit |
71fd91 |
restored at decompression time. File name handling is
|
|
Packit |
71fd91 |
naive in the sense that there is no mechanism for preserv-
|
|
Packit |
71fd91 |
ing original file names, permissions, ownerships or dates
|
|
Packit |
71fd91 |
in filesystems which lack these concepts, or have serious
|
|
Packit |
71fd91 |
file name length restrictions, such as MS-DOS.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2 and bunzip2 will by default not overwrite existing
|
|
Packit |
71fd91 |
files. If you want this to happen, specify the -f flag.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
If no file names are specified, bzip2 compresses from
|
|
Packit |
71fd91 |
standard input to standard output. In this case, bzip2
|
|
Packit |
71fd91 |
will decline to write compressed output to a terminal, as
|
|
Packit |
71fd91 |
this would be entirely incomprehensible and therefore
|
|
Packit |
71fd91 |
pointless.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bunzip2 (or bzip2 -d) decompresses all specified files.
|
|
Packit |
71fd91 |
Files which were not created by bzip2 will be detected and
|
|
Packit |
71fd91 |
ignored, and a warning issued. bzip2 attempts to guess
|
|
Packit |
71fd91 |
the filename for the decompressed file from that of the
|
|
Packit |
71fd91 |
compressed file as follows:
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
filename.bz2 becomes filename
|
|
Packit |
71fd91 |
filename.bz becomes filename
|
|
Packit |
71fd91 |
filename.tbz2 becomes filename.tar
|
|
Packit |
71fd91 |
filename.tbz becomes filename.tar
|
|
Packit |
71fd91 |
anyothername becomes anyothername.out
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
If the file does not end in one of the recognised endings,
|
|
Packit |
71fd91 |
.bz2, .bz, .tbz2 or .tbz, bzip2 complains that it cannot
|
|
Packit |
71fd91 |
guess the name of the original file, and uses the original
|
|
Packit |
71fd91 |
name with .out appended.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
As with compression, supplying no filenames causes decom-
|
|
Packit |
71fd91 |
pression from standard input to standard output.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bunzip2 will correctly decompress a file which is the con-
|
|
Packit |
71fd91 |
catenation of two or more compressed files. The result is
|
|
Packit |
71fd91 |
the concatenation of the corresponding uncompressed files.
|
|
Packit |
71fd91 |
Integrity testing (-t) of concatenated compressed files is
|
|
Packit |
71fd91 |
also supported.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
You can also compress or decompress files to the standard
|
|
Packit |
71fd91 |
output by giving the -c flag. Multiple files may be com-
|
|
Packit |
71fd91 |
pressed and decompressed like this. The resulting outputs
|
|
Packit |
71fd91 |
are fed sequentially to stdout. Compression of multiple
|
|
Packit |
71fd91 |
files in this manner generates a stream containing multi-
|
|
Packit |
71fd91 |
ple compressed file representations. Such a stream can be
|
|
Packit |
71fd91 |
decompressed correctly only by bzip2 version 0.9.0 or
|
|
Packit |
71fd91 |
later. Earlier versions of bzip2 will stop after decom-
|
|
Packit |
71fd91 |
pressing the first file in the stream.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzcat (or bzip2 -dc) decompresses all specified files to
|
|
Packit |
71fd91 |
the standard output.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2 will read arguments from the environment variables
|
|
Packit |
71fd91 |
BZIP2 and BZIP, in that order, and will process them
|
|
Packit |
71fd91 |
before any arguments read from the command line. This
|
|
Packit |
71fd91 |
gives a convenient way to supply default arguments.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Compression is always performed, even if the compressed
|
|
Packit |
71fd91 |
file is slightly larger than the original. Files of less
|
|
Packit |
71fd91 |
than about one hundred bytes tend to get larger, since the
|
|
Packit |
71fd91 |
compression mechanism has a constant overhead in the
|
|
Packit |
71fd91 |
region of 50 bytes. Random data (including the output of
|
|
Packit |
71fd91 |
most file compressors) is coded at about 8.05 bits per
|
|
Packit |
71fd91 |
byte, giving an expansion of around 0.5%.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
As a self-check for your protection, bzip2 uses 32-bit
|
|
Packit |
71fd91 |
CRCs to make sure that the decompressed version of a file
|
|
Packit |
71fd91 |
is identical to the original. This guards against corrup-
|
|
Packit |
71fd91 |
tion of the compressed data, and against undetected bugs
|
|
Packit |
71fd91 |
in bzip2 (hopefully very unlikely). The chances of data
|
|
Packit |
71fd91 |
corruption going undetected is microscopic, about one
|
|
Packit |
71fd91 |
chance in four billion for each file processed. Be aware,
|
|
Packit |
71fd91 |
though, that the check occurs upon decompression, so it
|
|
Packit |
71fd91 |
can only tell you that something is wrong. It can't help
|
|
Packit |
71fd91 |
you recover the original uncompressed data. You can use
|
|
Packit |
71fd91 |
bzip2recover to try to recover data from damaged files.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Return values: 0 for a normal exit, 1 for environmental
|
|
Packit |
71fd91 |
problems (file not found, invalid flags, I/O errors, &c),
|
|
Packit |
71fd91 |
2 to indicate a corrupt compressed file, 3 for an internal
|
|
Packit |
71fd91 |
consistency error (eg, bug) which caused bzip2 to panic.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
OPTIONS
|
|
Packit |
71fd91 |
-c --stdout
|
|
Packit |
71fd91 |
Compress or decompress to standard output.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-d --decompress
|
|
Packit |
71fd91 |
Force decompression. bzip2, bunzip2 and bzcat are
|
|
Packit |
71fd91 |
really the same program, and the decision about
|
|
Packit |
71fd91 |
what actions to take is done on the basis of which
|
|
Packit |
71fd91 |
name is used. This flag overrides that mechanism,
|
|
Packit |
71fd91 |
and forces bzip2 to decompress.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-z --compress
|
|
Packit |
71fd91 |
The complement to -d: forces compression,
|
|
Packit |
71fd91 |
regardless of the invocation name.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-t --test
|
|
Packit |
71fd91 |
Check integrity of the specified file(s), but don't
|
|
Packit |
71fd91 |
decompress them. This really performs a trial
|
|
Packit |
71fd91 |
decompression and throws away the result.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-f --force
|
|
Packit |
71fd91 |
Force overwrite of output files. Normally, bzip2
|
|
Packit |
71fd91 |
will not overwrite existing output files. Also
|
|
Packit |
71fd91 |
forces bzip2 to break hard links to files, which it
|
|
Packit |
71fd91 |
otherwise wouldn't do.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2 normally declines to decompress files which
|
|
Packit |
71fd91 |
don't have the correct magic header bytes. If
|
|
Packit |
71fd91 |
forced (-f), however, it will pass such files
|
|
Packit |
71fd91 |
through unmodified. This is how GNU gzip behaves.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-k --keep
|
|
Packit |
71fd91 |
Keep (don't delete) input files during compression
|
|
Packit |
71fd91 |
or decompression.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-s --small
|
|
Packit |
71fd91 |
Reduce memory usage, for compression, decompression
|
|
Packit |
71fd91 |
and testing. Files are decompressed and tested
|
|
Packit |
71fd91 |
using a modified algorithm which only requires 2.5
|
|
Packit |
71fd91 |
bytes per block byte. This means any file can be
|
|
Packit |
71fd91 |
decompressed in 2300k of memory, albeit at about
|
|
Packit |
71fd91 |
half the normal speed.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
During compression, -s selects a block size of
|
|
Packit |
71fd91 |
200k, which limits memory use to around the same
|
|
Packit |
71fd91 |
figure, at the expense of your compression ratio.
|
|
Packit |
71fd91 |
In short, if your machine is low on memory (8
|
|
Packit |
71fd91 |
megabytes or less), use -s for everything. See
|
|
Packit |
71fd91 |
MEMORY MANAGEMENT below.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-q --quiet
|
|
Packit |
71fd91 |
Suppress non-essential warning messages. Messages
|
|
Packit |
71fd91 |
pertaining to I/O errors and other critical events
|
|
Packit |
71fd91 |
will not be suppressed.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-v --verbose
|
|
Packit |
71fd91 |
Verbose mode -- show the compression ratio for each
|
|
Packit |
71fd91 |
file processed. Further -v's increase the ver-
|
|
Packit |
71fd91 |
bosity level, spewing out lots of information which
|
|
Packit |
71fd91 |
is primarily of interest for diagnostic purposes.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-L --license -V --version
|
|
Packit |
71fd91 |
Display the software version, license terms and
|
|
Packit |
71fd91 |
conditions.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-1 (or --fast) to -9 (or --best)
|
|
Packit |
71fd91 |
Set the block size to 100 k, 200 k .. 900 k when
|
|
Packit |
71fd91 |
compressing. Has no effect when decompressing.
|
|
Packit |
71fd91 |
See MEMORY MANAGEMENT below. The --fast and --best
|
|
Packit |
71fd91 |
aliases are primarily for GNU gzip compatibility.
|
|
Packit |
71fd91 |
In particular, --fast doesn't make things signifi-
|
|
Packit |
71fd91 |
cantly faster. And --best merely selects the
|
|
Packit |
71fd91 |
default behaviour.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-- Treats all subsequent arguments as file names, even
|
|
Packit |
71fd91 |
if they start with a dash. This is so you can han-
|
|
Packit |
71fd91 |
dle files with names beginning with a dash, for
|
|
Packit |
71fd91 |
example: bzip2 -- -myfilename.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
--repetitive-fast --repetitive-best
|
|
Packit |
71fd91 |
These flags are redundant in versions 0.9.5 and
|
|
Packit |
71fd91 |
above. They provided some coarse control over the
|
|
Packit |
71fd91 |
behaviour of the sorting algorithm in earlier ver-
|
|
Packit |
71fd91 |
sions, which was sometimes useful. 0.9.5 and above
|
|
Packit |
71fd91 |
have an improved algorithm which renders these
|
|
Packit |
71fd91 |
flags irrelevant.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
MEMORY MANAGEMENT
|
|
Packit |
71fd91 |
bzip2 compresses large files in blocks. The block size
|
|
Packit |
71fd91 |
affects both the compression ratio achieved, and the
|
|
Packit |
71fd91 |
amount of memory needed for compression and decompression.
|
|
Packit |
71fd91 |
The flags -1 through -9 specify the block size to be
|
|
Packit |
71fd91 |
100,000 bytes through 900,000 bytes (the default) respec-
|
|
Packit |
71fd91 |
tively. At decompression time, the block size used for
|
|
Packit |
71fd91 |
compression is read from the header of the compressed
|
|
Packit |
71fd91 |
file, and bunzip2 then allocates itself just enough memory
|
|
Packit |
71fd91 |
to decompress the file. Since block sizes are stored in
|
|
Packit |
71fd91 |
compressed files, it follows that the flags -1 to -9 are
|
|
Packit |
71fd91 |
irrelevant to and so ignored during decompression.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Compression and decompression requirements, in bytes, can
|
|
Packit |
71fd91 |
be estimated as:
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Compression: 400k + ( 8 x block size )
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Decompression: 100k + ( 4 x block size ), or
|
|
Packit |
71fd91 |
100k + ( 2.5 x block size )
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Larger block sizes give rapidly diminishing marginal
|
|
Packit |
71fd91 |
returns. Most of the compression comes from the first two
|
|
Packit |
71fd91 |
or three hundred k of block size, a fact worth bearing in
|
|
Packit |
71fd91 |
mind when using bzip2 on small machines. It is also
|
|
Packit |
71fd91 |
important to appreciate that the decompression memory
|
|
Packit |
71fd91 |
requirement is set at compression time by the choice of
|
|
Packit |
71fd91 |
block size.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
For files compressed with the default 900k block size,
|
|
Packit |
71fd91 |
bunzip2 will require about 3700 kbytes to decompress. To
|
|
Packit |
71fd91 |
support decompression of any file on a 4 megabyte machine,
|
|
Packit |
71fd91 |
bunzip2 has an option to decompress using approximately
|
|
Packit |
71fd91 |
half this amount of memory, about 2300 kbytes. Decompres-
|
|
Packit |
71fd91 |
sion speed is also halved, so you should use this option
|
|
Packit |
71fd91 |
only where necessary. The relevant flag is -s.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
In general, try and use the largest block size memory con-
|
|
Packit |
71fd91 |
straints allow, since that maximises the compression
|
|
Packit |
71fd91 |
achieved. Compression and decompression speed are virtu-
|
|
Packit |
71fd91 |
ally unaffected by block size.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Another significant point applies to files which fit in a
|
|
Packit |
71fd91 |
single block -- that means most files you'd encounter
|
|
Packit |
71fd91 |
using a large block size. The amount of real memory
|
|
Packit |
71fd91 |
touched is proportional to the size of the file, since the
|
|
Packit |
71fd91 |
file is smaller than a block. For example, compressing a
|
|
Packit |
71fd91 |
file 20,000 bytes long with the flag -9 will cause the
|
|
Packit |
71fd91 |
compressor to allocate around 7600k of memory, but only
|
|
Packit |
71fd91 |
touch 400k + 20000 * 8 = 560 kbytes of it. Similarly, the
|
|
Packit |
71fd91 |
decompressor will allocate 3700k but only touch 100k +
|
|
Packit |
71fd91 |
20000 * 4 = 180 kbytes.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Here is a table which summarises the maximum memory usage
|
|
Packit |
71fd91 |
for different block sizes. Also recorded is the total
|
|
Packit |
71fd91 |
compressed size for 14 files of the Calgary Text Compres-
|
|
Packit |
71fd91 |
sion Corpus totalling 3,141,622 bytes. This column gives
|
|
Packit |
71fd91 |
some feel for how compression varies with block size.
|
|
Packit |
71fd91 |
These figures tend to understate the advantage of larger
|
|
Packit |
71fd91 |
block sizes for larger files, since the Corpus is domi-
|
|
Packit |
71fd91 |
nated by smaller files.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Compress Decompress Decompress Corpus
|
|
Packit |
71fd91 |
Flag usage usage -s usage Size
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
-1 1200k 500k 350k 914704
|
|
Packit |
71fd91 |
-2 2000k 900k 600k 877703
|
|
Packit |
71fd91 |
-3 2800k 1300k 850k 860338
|
|
Packit |
71fd91 |
-4 3600k 1700k 1100k 846899
|
|
Packit |
71fd91 |
-5 4400k 2100k 1350k 845160
|
|
Packit |
71fd91 |
-6 5200k 2500k 1600k 838626
|
|
Packit |
71fd91 |
-7 6100k 2900k 1850k 834096
|
|
Packit |
71fd91 |
-8 6800k 3300k 2100k 828642
|
|
Packit |
71fd91 |
-9 7600k 3700k 2350k 828642
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
RECOVERING DATA FROM DAMAGED FILES
|
|
Packit |
71fd91 |
bzip2 compresses files in blocks, usually 900kbytes long.
|
|
Packit |
71fd91 |
Each block is handled independently. If a media or trans-
|
|
Packit |
71fd91 |
mission error causes a multi-block .bz2 file to become
|
|
Packit |
71fd91 |
damaged, it may be possible to recover data from the
|
|
Packit |
71fd91 |
undamaged blocks in the file.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
The compressed representation of each block is delimited
|
|
Packit |
71fd91 |
by a 48-bit pattern, which makes it possible to find the
|
|
Packit |
71fd91 |
block boundaries with reasonable certainty. Each block
|
|
Packit |
71fd91 |
also carries its own 32-bit CRC, so damaged blocks can be
|
|
Packit |
71fd91 |
distinguished from undamaged ones.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2recover is a simple program whose purpose is to
|
|
Packit |
71fd91 |
search for blocks in .bz2 files, and write each block out
|
|
Packit |
71fd91 |
into its own .bz2 file. You can then use bzip2 -t to test
|
|
Packit |
71fd91 |
the integrity of the resulting files, and decompress those
|
|
Packit |
71fd91 |
which are undamaged.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2recover takes a single argument, the name of the dam-
|
|
Packit |
71fd91 |
aged file, and writes a number of files
|
|
Packit |
71fd91 |
"rec00001file.bz2", "rec00002file.bz2", etc, containing
|
|
Packit |
71fd91 |
the extracted blocks. The output filenames are
|
|
Packit |
71fd91 |
designed so that the use of wildcards in subsequent pro-
|
|
Packit |
71fd91 |
cessing -- for example, "bzip2 -dc rec*file.bz2 > recov-
|
|
Packit |
71fd91 |
ered_data" -- processes the files in the correct order.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2recover should be of most use dealing with large .bz2
|
|
Packit |
71fd91 |
files, as these will contain many blocks. It is clearly
|
|
Packit |
71fd91 |
futile to use it on damaged single-block files, since a
|
|
Packit |
71fd91 |
damaged block cannot be recovered. If you wish to min-
|
|
Packit |
71fd91 |
imise any potential data loss through media or transmis-
|
|
Packit |
71fd91 |
sion errors, you might consider compressing with a smaller
|
|
Packit |
71fd91 |
block size.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
PERFORMANCE NOTES
|
|
Packit |
71fd91 |
The sorting phase of compression gathers together similar
|
|
Packit |
71fd91 |
strings in the file. Because of this, files containing
|
|
Packit |
71fd91 |
very long runs of repeated symbols, like "aabaabaabaab
|
|
Packit |
71fd91 |
..." (repeated several hundred times) may compress more
|
|
Packit |
71fd91 |
slowly than normal. Versions 0.9.5 and above fare much
|
|
Packit |
71fd91 |
better than previous versions in this respect. The ratio
|
|
Packit |
71fd91 |
between worst-case and average-case compression time is in
|
|
Packit |
71fd91 |
the region of 10:1. For previous versions, this figure
|
|
Packit |
71fd91 |
was more like 100:1. You can use the -vvvv option to mon-
|
|
Packit |
71fd91 |
itor progress in great detail, if you want.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
Decompression speed is unaffected by these phenomena.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2 usually allocates several megabytes of memory to
|
|
Packit |
71fd91 |
operate in, and then charges all over it in a fairly ran-
|
|
Packit |
71fd91 |
dom fashion. This means that performance, both for com-
|
|
Packit |
71fd91 |
pressing and decompressing, is largely determined by the
|
|
Packit |
71fd91 |
speed at which your machine can service cache misses.
|
|
Packit |
71fd91 |
Because of this, small changes to the code to reduce the
|
|
Packit |
71fd91 |
miss rate have been observed to give disproportionately
|
|
Packit |
71fd91 |
large performance improvements. I imagine bzip2 will per-
|
|
Packit |
71fd91 |
form best on machines with very large caches.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
CAVEATS
|
|
Packit |
71fd91 |
I/O error messages are not as helpful as they could be.
|
|
Packit |
71fd91 |
bzip2 tries hard to detect I/O errors and exit cleanly,
|
|
Packit |
71fd91 |
but the details of what the problem is sometimes seem
|
|
Packit |
71fd91 |
rather misleading.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
This manual page pertains to version 1.0.6 of bzip2. Com-
|
|
Packit |
71fd91 |
pressed data created by this version is entirely forwards
|
|
Packit |
71fd91 |
and backwards compatible with the previous public
|
|
Packit |
71fd91 |
releases, versions 0.1pl2, 0.9.0, 0.9.5, 1.0.0, 1.0.1,
|
|
Packit |
71fd91 |
1.0.2 and above, but with the following exception: 0.9.0
|
|
Packit |
71fd91 |
and above can correctly decompress multiple concatenated
|
|
Packit |
71fd91 |
compressed files. 0.1pl2 cannot do this; it will stop
|
|
Packit |
71fd91 |
after decompressing just the first file in the stream.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
bzip2recover versions prior to 1.0.2 used 32-bit integers
|
|
Packit |
71fd91 |
to represent bit positions in compressed files, so they
|
|
Packit |
71fd91 |
could not handle compressed files more than 512 megabytes
|
|
Packit |
71fd91 |
long. Versions 1.0.2 and above use 64-bit ints on some
|
|
Packit |
71fd91 |
platforms which support them (GNU supported targets, and
|
|
Packit |
71fd91 |
Windows). To establish whether or not bzip2recover was
|
|
Packit |
71fd91 |
built with such a limitation, run it without arguments.
|
|
Packit |
71fd91 |
In any event you can build yourself an unlimited version
|
|
Packit |
71fd91 |
if you can recompile it with MaybeUInt64 set to be an
|
|
Packit |
71fd91 |
unsigned 64-bit integer.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
AUTHOR
|
|
Packit |
71fd91 |
Julian Seward, jsewardbzip.org.
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
http://www.bzip.org
|
|
Packit |
71fd91 |
|
|
Packit |
71fd91 |
The ideas embodied in bzip2 are due to (at least) the fol-
|
|
Packit |
71fd91 |
lowing people: Michael Burrows and David Wheeler (for the
|
|
Packit |
71fd91 |
block sorting transformation), David Wheeler (again, for
|
|
Packit |
71fd91 |
the Huffman coder), Peter Fenwick (for the structured cod-
|
|
Packit |
71fd91 |
ing model in the original bzip, and many refinements), and
|
|
Packit |
71fd91 |
Alistair Moffat, Radford Neal and Ian Witten (for the
|
|
Packit |
71fd91 |
arithmetic coder in the original bzip). I am much
|
|
Packit |
71fd91 |
indebted for their help, support and advice. See the man-
|
|
Packit |
71fd91 |
ual in the source distribution for pointers to sources of
|
|
Packit |
71fd91 |
documentation. Christian von Roques encouraged me to look
|
|
Packit |
71fd91 |
for faster sorting algorithms, so as to speed up compres-
|
|
Packit |
71fd91 |
sion. Bela Lubkin encouraged me to improve the worst-case
|
|
Packit |
71fd91 |
compression performance. Donna Robinson XMLised the docu-
|
|
Packit |
71fd91 |
mentation. The bz* scripts are derived from those of GNU
|
|
Packit |
71fd91 |
gzip. Many people sent patches, helped with portability
|
|
Packit |
71fd91 |
problems, lent machines, gave advice and were generally
|
|
Packit |
71fd91 |
helpful.
|
|
Packit |
71fd91 |
|