Compress reduces the size of the named files using adaptive Lempel-Ziv coding. Whenever possible, each file is replaced by one with the extension .Z, while keeping the same ownership modes, access and modification times. If no files are specified, the standard input is compressed to the standard output. Compress will only attempt to compress regular files. In particular, it will ignore symbolic links. If a file has multiple hard links, compress will refuse to compress it unless the -f flag is given.
If -f is not given and compress is run in the foreground, the user is prompted as to whether an existing file should be overwritten.
Compressed files can be restored to their original form using uncompress or zcat.
uncompress takes a list of files on its command line and replaces each file whose name ends with .Z and which begins with the correct magic number with an uncompressed file without the .Z. The uncompressed file will have the mode, ownership and timestamps of the compressed file.
The -c option makes compress/uncompress write to the standard output; no files are changed.
zcat is identical to uncompress -c. zcat uncompresses either a list of files on the command line or its standard input and writes the uncompressed data on standard output. zcat will uncompress files that have the correct magic number whether they have a .Z suffix or not.
If the -r flag is specified, compress will operate recursively. If any of the file names specified on the command line are directories, compress will descend into the directory and compress all the files it finds there.
The -V flag tells each of these programs to print its version and patchlevel, along with any preprocessor flags specified during compilation, on stderr before doing any compression or uncompression.
Compress uses the modified Lempel-Ziv algorithm popularized in "A Technique for High Performance Data Compression", Terry A. Welch, IEEE Computer, vol. 17, no. 6 (June 1984), pp. 8-19. Common substrings in the file are first replaced by 9-bit codes 257 and up. When code 512 is reached, the algorithm switches to 10-bit codes and continues to use more bits until the limit specified by the -b flag is reached (default 16). Bits must be between 9 and 16. The default can be changed in the source to allow compress to be run on a smaller machine.
After the bits limit is attained, compress periodically checks the compression ratio. If it is increasing, compress continues to use the existing code dictionary. However, if the compression ratio decreases, compress discards the table of substrings and rebuilds it from scratch. This allows the algorithm to adapt to the next "block" of the file.
Note that the -b flag is omitted for uncompress, since the bits parameter specified during compression is encoded within the output, along with a magic number to ensure that neither decompression of random data nor recompression of compressed data is attempted.
The amount of compression obtained depends on the size of the input, the number of bits per code, and the distribution of common substrings. Typically, text such as source code or English is reduced by 50-60%. Compression is generally much better than that achieved by Huffman coding (as used in pack), or adaptive Huffman coding (compact), and takes less time to compute.
Under the -v option, a message is printed yielding the percentage of reduction for each file compressed.