Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site brl-tgr.ARPA
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!godot!harvard!seismo!brl-tgr!tgr!moss@Brl-Vld.ARPA
From: "Gary S. Moss (AMXBR-VLD-V)" <moss@Brl-Vld.ARPA>
Newsgroups: net.lang.c
Subject: Re:  Pattern recognition in C (wanted)
Message-ID: <7658@brl-tgr.ARPA>
Date: Wed, 23-Jan-85 13:07:11 EST
Article-I.D.: brl-tgr.7658
Posted: Wed Jan 23 13:07:11 1985
Date-Received: Fri, 25-Jan-85 07:59:06 EST
Sender: news@brl-tgr.ARPA
Organization: Ballistic Research Lab
Lines: 21

~ A recent posting to net.lang.c of a "shell style" pattern matching
~ routine by Gary Moss appeared to be incomplete so I decided to post
~ an alternate.

Now Fred, this makes me nervous, the posted routine was written for
a table-of-contents lookup routine, and complete functionality a la
UNIX shell, was not critical, however I am interested in what I left
out since I am likely to find other uses for it.

~ ... By the way, it appears that the usage message
~ for matchtest is wrong.  It apparently wants the arguments
~ [string pattern] rather than [pattern string].
Sorry Fred, it works as advertised, match [pattern string], if you leave
out the arguments, it reads pattern-string pairs on stdin until EOF.
YOUR match() function switches the arguments, i.e.,
	BOOLEAN nmatch( string, pattern ) {...}
while mine is
	int match( pattern, string ) {...}
So, if you substituted your function for mine ... :-)
Thanks for posting my sources, I should have done that originally.
__Moss__