compress [-fv][-b bits][file
...]
compress [-cfv][-b bits][file]
The compress utility shall attempt to reduce the size of the named files by using adaptive Lempel-Ziv coding algorithm.
Lempel-Ziv-Welch compression is covered by US Patent 4558302, issued to Terry A. Welch on December 10th, 1985, and assigned to Sperry Corporation.
On systems not supporting adaptive Lempel-Ziv coding algorithm, the input files shall not be changed and an error value greater than two shall be returned. Except when the output is to the standard output, each file shall be replaced by one with the extension .Z. If the invoking process has appropriate privileges, the ownership, modes, access time, and modification time of the original file are preserved. If appending the .Z to the filename would make the name exceed {NAME_MAX} bytes, the command shall fail. If no files are specified, the standard input shall be compressed to the standard output.
The compress utility shall conform to the Base Definitions volume of IEEE Std 1003.1-2001, Section 12.2, Utility Syntax Guidelines.
The following options shall be supported:
9 <= bits <= 14
The implementation may allow bits values of greater than 14. The default is 14, 15, or 16.
The following operand shall be supported:
The standard input shall be used only if no file operands are specified, or if a file operand is '-' .
If file operands are specified, the input files contain the data to be compressed.
The following environment variables shall affect the execution of compress:
If no file operands are specified, or if a file operand is '-' , or if the -c option is specified, the standard output contains the compressed output.
The standard error shall be used only for diagnostic and prompt messages and the output from -v.
The output files shall contain the compressed output. The format of compressed files is unspecified and interchange of such files between implementations (including access via unspecified file sharing mechanisms) is not required by IEEE Std 1003.1-2001.
The following exit values shall be returned:
The input file shall remain unmodified.
The following sections are informative.
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 or adaptive Huffman coding ( compact), and takes less time to compute.
Although compress strictly follows the default actions upon receipt of a signal or when an error occurs, some unexpected results may occur. In some implementations it is likely that a partially compressed file is left in place, alongside its uncompressed input file. Since the general operation of compress is to delete the uncompressed file only after the .Z file has been successfully filled, an application should always carefully check the exit status of compress before arbitrarily deleting files that have like-named neighbors with .Z suffixes.
The limit of 14 on the bits option-argument is to achieve portability to all systems (within the restrictions imposed by the lack of an explicit published file format). Some implementations based on 16-bit architectures cannot support 15 or 16-bit uncompression.