1.\" Copyright (c) 1986, 1990 The Regents of the University of California. 2.\" All rights reserved. 3.\" 4.\" This code is derived from software contributed to Berkeley by 5.\" James A. Woods, derived from original work by Spencer Thomas 6.\" and Joseph Orost. 7.\" 8.\" %sccs.include.redist.man% 9.\" 10.\" @(#)compress.1 6.9 (Berkeley) 03/14/91 11.\" 12.Dd 13.Dt COMPRESS 1 14.Os BSD 4.3 15.Sh NAME 16.Nm compress , 17.Nm uncompress , 18.Nm zcat 19.Nd compress and expand data 20.Sh SYNOPSIS 21.Nm compress 22.Op Fl f 23.Op Fl v 24.Op Fl c 25.Op Fl b Ar bits 26.Op Ar 27.Nm uncompress 28.Op Fl f 29.Op Fl v 30.Op Fl c 31.Ar 32.Nm zcat 33.Op Ar 34.Sh DESCRIPTION 35.Nm Compress 36reduces the size of the named files using adaptive Lempel-Ziv coding. 37Whenever possible, 38each 39.Ar file 40is replaced by one with the extension 41.Ar \&.Z , 42while keeping the same ownership modes, access and modification times. 43If no files are specified, the standard input is compressed to the 44standard output. 45Compressed files can be restored to their original form using 46.Nm uncompress 47or 48.Nm zcat 49.Bl -tag -width Ds 50.It Fl f 51Force compression of 52.Ar file , 53even if it does not actually shrink 54or the corresponding 55.Ar file.Z 56file already exists. 57Except when run in the background under 58.Pa /bin/sh , 59if 60.Fl f 61is not given the user is prompted as to whether an existing 62.Ar file.Z 63file should be overwritten. 64.It Fl c 65(``cat''). 66.Nm Compress/uncompress 67writes to the standard output; no files are changed. 68The nondestructive behavior of 69.Nm zcat 70is identical to that of 71.Nm uncompress 72.Fl c . 73.It Fl b 74Specify 75.Ar bits 76code limit (see below). 77.It Fl v 78Print the percentage reduction of each file. 79.El 80.Pp 81.Nm Compress 82uses the modified Lempel-Ziv algorithm popularized in 83"A Technique for High Performance Data Compression", 84Terry A. Welch, 85.Em IEEE Computer , 86vol. 17, 87.no 6 88(June 1984), pp. 8-19. 89Common substrings in the file are first replaced by 9-bit codes 257 and up. 90When code 512 is reached, the algorithm switches to 10-bit codes and 91continues to use more bits until the 92limit specified by the 93.Fl b 94flag is reached (default 16). 95.Ar Bits 96must be between 9 and 16. The default can be changed in the source to allow 97.Nm compress 98to be run on a smaller machine. 99.Pp 100After the 101.Ar bits 102limit is attained, 103.Nm compress 104periodically checks the compression ratio. If it is increasing, 105.Nm compress 106continues to use the existing code dictionary. However, 107if the compression ratio decreases, 108.Nm compress 109discards the table of substrings and rebuilds it from scratch. This allows 110the algorithm to adapt to the next "block" of the file. 111.Pp 112Note that the 113.Fl b 114flag is omitted for 115.Ar uncompress 116since the 117.Ar bits 118parameter specified during compression 119is encoded within the output, along with 120a magic number to ensure that neither decompression of random data nor 121recompression of compressed data is attempted. 122.Pp 123.ne 8 124The amount of compression obtained depends on the size of the 125input, the number of 126.Ar bits 127per code, and the distribution of common substrings. 128Typically, text such as source code or English 129is reduced by 50\-60%. 130Compression is generally much better than that achieved by 131Huffman coding (as used in the historical command 132pack), 133or adaptive Huffman coding (as 134used in the historical command 135compact), 136and takes less time to compute. 137.Pp 138If an error occurs, exit status is 1; 139if the last file was not compressed because it became larger, the status 140is 2; otherwise the status is 0. 141.Sh DIAGNOSTICS 142.Bl -tag -width 4n 143.It "Usage: compress [-fvc] [-b maxbits] [file ...]" 144Invalid options were specified on the command line. 145.It "Missing maxbits" 146Maxbits must follow 147.Fl b . 148.It Ar file : No "not in compressed format" 149The file specified to 150.Ar uncompress 151has not been compressed. 152.It Xo 153.Ar file : No "compressed with" 154.Ar \&xx No "bits, can only handle" Ar \&yy No bits 155.Xc 156.Ar File 157was compressed by a program that could deal with 158more 159.Ar bits 160than the compress code on this machine. 161Recompress the file with smaller 162.Ar bits . 163.It Ar file : No "already has .Z suffix -- no change" 164The file is assumed to be already compressed. 165Rename the file and try again. 166.It Ar file : No "filename too long to tack on .Z" 167The file cannot be compressed because its name is longer than 16812 characters. 169Rename and try again. 170This message does not occur on BSD systems. 171.It Ar file No "already exists; do you wish to overwrite (y or n)?" 172Respond "y" if you want the output file to be replaced; "n" if not. 173.It "uncompress: corrupt input" 174A 175.Dv SIGSEGV 176violation was detected which usually means that the input file is 177corrupted. 178.It Compression: Em "xx.xx%" 179Percentage of the input saved by compression. 180(Relevant only for 181.Fl v . ) 182.It "-- not a regular file: unchanged" 183When the input file is not a regular file, 184(e.g. a directory), it is 185left unaltered. 186.It "-- has" Ar xx No "other links: unchanged" 187The input file has links; it is left unchanged. See 188.Xr ln 1 189for more information. 190.It "-- file unchanged" 191No savings is achieved by 192compression. The input remains virgin. 193.El 194.Sh FILES 195.Bl -tag -width file.Z 196.It Pa file.Z 197compressed file is file.Z 198.El 199.Sh BUGS 200Although compressed files are compatible between machines with large memory, 201.Fl b Ns Ar 12 202should be used for file transfer to architectures with 203a small process data space (64KB or less, as exhibited by the 204.Tn DEC PDP 205series, the Intel 80286, etc.) 206.Pp 207.Nm Compress 208should be more flexible about the existence of the `.Z' suffix. 209.Sh HISTORY 210The 211.Nm 212command appeared in 213.Bx 4.3 . 214