MAN page from Trustix bzip2-1.0.2-9tr.i586.rpm
Section: User Commands (1)Index
bzip2, bunzip2 - a block-sorting file compressor, v1.0.2
bzcat - decompresses files to stdout
bzip2recover - recovers data from damaged bzip2 files
][ filenames ...
][ filenames ...
compresses files using the Burrows-Wheeler block sortingtext compression algorithm, and Huffman coding. Compression isgenerally considerably better than that achieved by more conventionalLZ77/LZ78-based compressors, and approaches the performance of the PPMfamily of statistical compressors.
The command-line options are deliberately very similar to those of GNU gzip, but they are not identical.
bzip2expects a list of file names to accompany thecommand-line flags. Each file is replaced by a compressed version ofitself, with the name "original_name.bz2". Each compressed filehas the same modification date, permissions, and, when possible,ownership as the corresponding original, so that these properties canbe correctly restored at decompression time. File name handling isnaive in the sense that there is no mechanism for preserving originalfile names, permissions, ownerships or dates in filesystems which lackthese concepts, or have serious file name length restrictions, such asMS-DOS.
bzip2andbunzip2will by default not overwrite existingfiles. If you want this to happen, specify the -f flag.
If no file names are specified,bzip2compresses from standardinput to standard output. In this case,bzip2will decline towrite compressed output to a terminal, as this would be entirelyincomprehensible and therefore pointless.
bunzip2(orbzip2 -d) decompresses allspecified files. Files which were not created by bzip2will be detected and ignored, and a warning issued. bzip2attempts to guess the filename for the decompressed file from that of the compressed file as follows:
filename.bz2 becomes filename
filename.bz becomes filename
filename.tbz2 becomes filename.tar
filename.tbz becomes filename.tar
anyothername becomes anyothername.out
If the file does not end in one of the recognised endings, .bz2, .bz, .tbz2or.tbz, bzip2 complains that it cannotguess the name of the original file, and uses the original namewith.outappended.
As with compression, supplying nofilenames causes decompression from standard input to standard output.
bunzip2 will correctly decompress a file which is theconcatenation of two or more compressed files. The result is theconcatenation of the corresponding uncompressed files. Integritytesting (-t) of concatenated compressed files is also supported.
You can also compress or decompress files to the standard output bygiving the -c flag. Multiple files may be compressed anddecompressed like this. The resulting outputs are fed sequentially tostdout. Compression of multiple files in this manner generates a streamcontaining multiple compressed file representations. Such a streamcan be decompressed correctly only bybzip2 version 0.9.0 orlater. Earlier versions ofbzip2will stop after decompressingthe first file in the stream.
bzcat(orbzip2 -dc) decompresses all specified files tothe standard output.
bzip2will read arguments from the environment variablesBZIP2andBZIP,in that order, and will process thembefore any arguments read from the command line. This gives a convenient way to supply default arguments.
Compression is always performed, even if the compressed file is slightlylarger than the original. Files of less than about one hundred bytestend to get larger, since the compression mechanism has a constantoverhead in the region of 50 bytes. Random data (including the outputof most file compressors) is coded at about 8.05 bits per byte, givingan expansion of around 0.5%.
As a self-check for your protection, bzip2uses 32-bit CRCs tomake sure that the decompressed version of a file is identical to theoriginal. This guards against corruption of the compressed data, andagainst undetected bugs inbzip2(hopefully very unlikely). Thechances of data corruption going undetected is microscopic, about onechance in four billion for each file processed. Be aware, though, thatthe check occurs upon decompression, so it can only tell you thatsomething is wrong. It can't help you recover the original uncompresseddata. You can use bzip2recoverto try to recover data fromdamaged files.
Return values: 0 for a normal exit, 1 for environmental problems (filenot found, invalid flags, I/O errors, &c), 2 to indicate a corruptcompressed file, 3 for an internal consistency error (eg, bug) whichcausedbzip2to panic.
- -c --stdout
- Compress or decompress to standard output.
- -d --decompress
- Force decompression. bzip2, bunzip2 andbzcat arereally the same program, and the decision about what actions to take isdone on the basis of which name is used. This flag overrides thatmechanism, and forces bzip2to decompress.
- -z --compress
- The complement to -d: forces compression, regardless of theinvocation name.
- -t --test
- Check integrity of the specified file(s), but don't decompress them.This really performs a trial decompression and throws away the result.
- -f --force
- Force overwrite of output files. Normally,bzip2 will not overwriteexisting output files. Also forces bzip2 to break hard linksto files, which it otherwise wouldn't do.
bzip2 normally declines to decompress files which don't have thecorrect magic header bytes. If forced (-f), however, it will passsuch files through unmodified. This is how GNU gzip behaves.
- -k --keep
- Keep (don't delete) input files during compressionor decompression.
- -s --small
- Reduce memory usage, for compression, decompression and testing. Filesare decompressed and tested using a modified algorithm which onlyrequires 2.5 bytes per block byte. This means any file can bedecompressed in 2300k of memory, albeit at about half the normal speed.
During compression, -s selects a block size of 200k, which limitsmemory use to around the same figure, at the expense of your compressionratio. In short, if your machine is low on memory (8 megabytes orless), use -s for everything. See MEMORY MANAGEMENT below.
- -q --quiet
- Suppress non-essential warning messages. Messages pertaining toI/O errors and other critical events will not be suppressed.
- -v --verbose
- Verbose mode -- show the compression ratio for each file processed.Further -v's increase the verbosity level, spewing out lots ofinformation which is primarily of interest for diagnostic purposes.
- -L --license -V --version
- Display the software version, license terms and conditions.
- -1 (or --fast) to -9 (or --best)
- Set the block size to 100 k, 200 k .. 900 k when compressing. Has noeffect when decompressing. See MEMORY MANAGEMENT below.The --fast and --best aliases are primarily for GNU gzip compatibility. In particular, --fast doesn't make thingssignificantly faster. And --best merely selects the default behaviour.
- Treats all subsequent arguments as file names, even if they startwith a dash. This is so you can handle files with names beginningwith a dash, for example: bzip2 -- -myfilename.
- --repetitive-fast --repetitive-best
- These flags are redundant in versions 0.9.5 and above. They providedsome coarse control over the behaviour of the sorting algorithm inearlier versions, which was sometimes useful. 0.9.5 and above have animproved algorithm which renders these flags irrelevant.
compresses large files in blocks. The block size affectsboth the compression ratio achieved, and the amount of memory needed forcompression and decompression. The flags -1 through -9specify the block size to be 100,000 bytes through 900,000 bytes (thedefault) respectively. At decompression time, the block size used forcompression is read from the header of the compressed file, andbunzip2
then allocates itself just enough memory to decompressthe file. Since block sizes are stored in compressed files, it followsthat the flags -1 to -9 are irrelevant to and so ignoredduring decompression.
Compression and decompression requirements, in bytes, can be estimated as:
Compression: 400k + ( 8 x block size )
Decompression: 100k + ( 4 x block size ), or
100k + ( 2.5 x block size )
Larger block sizes give rapidly diminishing marginal returns. Most ofthe compression comes from the first two or three hundred k of blocksize, a fact worth bearing in mind when usingbzip2on small machines.It is also important to appreciate that the decompression memoryrequirement is set at compression time by the choice of block size.
For files compressed with the default 900k block size,bunzip2will require about 3700 kbytes to decompress. To support decompressionof any file on a 4 megabyte machine, bunzip2has an option todecompress using approximately half this amount of memory, about 2300kbytes. Decompression speed is also halved, so you should use thisoption only where necessary. The relevant flag is -s.
In general, try and use the largest block size memory constraints allow,since that maximises the compression achieved. Compression anddecompression speed are virtually unaffected by block size.
Another significant point applies to files which fit in a single block-- that means most files you'd encounter using a large block size. Theamount of real memory touched is proportional to the size of the file,since the file is smaller than a block. For example, compressing a file20,000 bytes long with the flag -9 will cause the compressor toallocate around 7600k of memory, but only touch 400k + 20000 * 8 = 560kbytes of it. Similarly, the decompressor will allocate 3700k but onlytouch 100k + 20000 * 4 = 180 kbytes.
Here is a table which summarises the maximum memory usage for differentblock sizes. Also recorded is the total compressed size for 14 files ofthe Calgary Text Compression Corpus totalling 3,141,622 bytes. Thiscolumn gives some feel for how compression varies with block size.These figures tend to understate the advantage of larger block sizes forlarger files, since the Corpus is dominated by smaller files.
Compress Decompress Decompress Corpus
Flag usage usage -s usage Size
-1 1200k 500k 350k 914704
-2 2000k 900k 600k 877703
-3 2800k 1300k 850k 860338
-4 3600k 1700k 1100k 846899
-5 4400k 2100k 1350k 845160
-6 5200k 2500k 1600k 838626
-7 6100k 2900k 1850k 834096
-8 6800k 3300k 2100k 828642
-9 7600k 3700k 2350k 828642
RECOVERING DATA FROM DAMAGED FILESbzip2
compresses files in blocks, usually 900kbytes long. Eachblock is handled independently. If a media or transmission error causesa multi-block .bz2file to become damaged, it may be possible torecover data from the undamaged blocks in the file.
The compressed representation of each block is delimited by a 48-bitpattern, which makes it possible to find the block boundaries withreasonable certainty. Each block also carries its own 32-bit CRC, sodamaged blocks can be distinguished from undamaged ones.
bzip2recoveris a simple program whose purpose is to search forblocks in .bz2 files, and write each block out into its own .bz2 file. You can then usebzip2 -tto test theintegrity of the resulting files, and decompress those which areundamaged.
bzip2recovertakes a single argument, the name of the damaged file, and writes a number of files "rec00001file.bz2","rec00002file.bz2", etc, containing the extracted blocks.The output filenames are designed so that the use ofwildcards in subsequent processing -- for example, "bzip2 -dc rec*file.bz2 > recovered_data" -- processes the files inthe correct order.
bzip2recovershould be of most use dealing with large .bz2files, as these will contain many blocks. It is clearlyfutile to use it on damaged single-block files, since adamaged block cannot be recovered. If you wish to minimise any potential data loss through media or transmission errors, you might consider compressing with a smallerblock size.
The sorting phase of compression gathers together similar strings in thefile. Because of this, files containing very long runs of repeatedsymbols, like "aabaabaabaab ..." (repeated several hundred times) maycompress more slowly than normal. Versions 0.9.5 and above fare muchbetter than previous versions in this respect. The ratio betweenworst-case and average-case compression time is in the region of 10:1.For previous versions, this figure was more like 100:1. You can use the-vvvv option to monitor progress in great detail, if you want.
Decompression speed is unaffected by these phenomena.
bzip2usually allocates several megabytes of memory to operatein, and then charges all over it in a fairly random fashion. This meansthat performance, both for compressing and decompressing, is largelydetermined by the speed at which your machine can service cache misses.Because of this, small changes to the code to reduce the miss rate havebeen observed to give disproportionately large performance improvements.I imagine bzip2will perform best on machines with very large caches.
I/O error messages are not as helpful as they could be.bzip2
tries hard to detect I/O errors and exit cleanly, but the details ofwhat the problem is sometimes seem rather misleading.
This manual page pertains to version 1.0.2 ofbzip2. Compressed data created by this version is entirely forwards andbackwards compatible with the previous public releases, versions0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1, but with the followingexception: 0.9.0 and above can correctly decompress multipleconcatenated compressed files. 0.1pl2 cannot do this; it will stopafter decompressing just the first file in the stream.
bzip2recoverversions prior to this one, 1.0.2, used 32-bit integers to representbit positions in compressed files, so it could not handle compressedfiles more than 512 megabytes long. Version 1.0.2 and above uses64-bit ints on some platforms which support them (GNU supportedtargets, and Windows). To establish whether or not bzip2recover wasbuilt with such a limitation, run it without arguments. In any eventyou can build yourself an unlimited version if you can recompile itwith MaybeUInt64 set to be an unsigned 64-bit integer.
Julian Seward, jsewardAATTacm.org.
The ideas embodied inbzip2are due to (at least) the followingpeople: Michael Burrows and David Wheeler (for the block sortingtransformation), David Wheeler (again, for the Huffman coder), PeterFenwick (for the structured coding model in the originalbzip,and many refinements), and Alistair Moffat, Radford Neal and Ian Witten(for the arithmetic coder in the originalbzip). I am muchindebted for their help, support and advice. See the manual in thesource distribution for pointers to sources of documentation. Christianvon Roques encouraged me to look for faster sorting algorithms, so as tospeed up compression. Bela Lubkin encouraged me to improve theworst-case compression performance. The bz* scripts are derived from those of GNU gzip.Many people sent patches, helpedwith portability problems, lent machines, gave advice and were generallyhelpful.
- MEMORY MANAGEMENT
- RECOVERING DATA FROM DAMAGED FILES
- PERFORMANCE NOTES
This document was created byman2html,using the manual pages.