Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!seismo!mimsy!chris
From: chris@mimsy.UUCP
Newsgroups: comp.sources.d
Subject: Re: "look" utility?
Message-ID: <5912@mimsy.UUCP>
Date: Sat, 21-Mar-87 13:09:38 EST
Article-I.D.: mimsy.5912
Posted: Sat Mar 21 13:09:38 1987
Date-Received: Sun, 22-Mar-87 17:46:44 EST
References: <837@maynard.BSW.COM> <4670001@hpihoah.HP.COM>
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 19

In article <4670001@hpihoah.HP.COM> cire@hpihoah.HP.COM (Eric B.
Decker) writes:
>the key assumption is the file is already sorted so a binary search
>works.  this is generally not true.

`Look' operates on sorted dictionaries.  It is not `generally not
true' *in the cases for which the program was designed*.

>to be true one then needs to include the time needed to sort files
>of the size we are talking about.  fgrep wins.

Only if the file changes often:  If not, it is better first to sort
it.  In any case, fgrep does not win; see James A. Woods's article
<706@ames.UUCP>, `Get Hep to Kanji-Ready Five-Algorithm [ef]?grep'
for details.  If you have only the standard family of greps, use
egrep.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7690)
UUCP:	seismo!mimsy!chris	ARPA/CSNet:	chris@mimsy.umd.edu