.\"
.\" $Id: sq.1,v 1.4 89/04/28 01:12:28 geoff Exp $
.\"
.\" $Log:	sq.1,v $
.\" Revision 1.4  89/04/28  01:12:28  geoff
.\" Change Header to Id;  nobody cares about my pathnames.
.\" 
.\" Revision 1.3  88/03/27  01:04:20  geoff
.\" Fix a couple of small typos.  Add a paragraph on how the squeezing is
.\" done.  Credit Mike Wexler as author.
.\" 
.\" Revision 1.2  88/03/13  00:58:29  geoff
.\" Add ID keywords
.\" 
.\"
.TH SQ 1 LOCAL
.SH NAME
sq \- squeeze a sorted word list
unsq \- unsqueeze a sorted word list
.SH SYNOPSIS
.B sq
< infile > outfile
.PP
.B unsq
< infile > outfile
.SH DESCRIPTION
.I sq
compresses a sorted list of words (a dictionary). 
For example:
.RS
sort /usr/dict/words | sq | compress > words.sq.Z
.RE
will compress dict by about a factor of 4.
.PP
.I unsq
uncompress the output of
.I sq.
For example:
.RS
compress -d < words.sq.Z | unsq | sort -f -o words
.RE
will uncompress a dictionary compressed with
.I sq.
.P
The squeezing is achieved by eliminating common prefixes, and replacing
them with a single character which encodes the number of characters
shared with the preceding word.
The prefix size is encoded as a single printable character:
0-9 represent 0-9, A-Z represent 10-35, and a-z represent 36-61.
.SH AUTHOR
Mike Wexler
.SH SEE ALSO
compress(1), sort(1).