Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!mips!news.cs.indiana.edu!arizona.edu!arizona!dave
From: dave@cs.arizona.edu (Dave Schaumann)
Newsgroups: comp.lang.c
Subject: Re: Efficient STRing CoMPares?
Message-ID: <1193@caslon.cs.arizona.edu>
Date: 17 Mar 91 00:41:05 GMT
References: <1991Mar15.142821@mars.mpr.ca> <15486@smoke.brl.mil>
Organization: U of Arizona CS Dept, Tucson
Lines: 26

In article <15486@smoke.brl.mil> gwyn@smoke.brl.mil (Doug Gwyn) writes:
|In article <1991Mar15.142821@mars.mpr.ca> stone@mars.mpr.ca (Darren Stone) writes:
|>Does anyone know if there's a better (!?) way to compare
|>strings than with strcmp()?
|>I'm looking for any speed advantage I can get!
|
|#define  StrEq( a, b )	(*(a) == *(b) && strcmp( a, b ) == 0)	/* UNSAFE */

By "unsafe", I assume you mean it will likely crash on a NULL pointer.
I wouldn't be a bit suprised if strcmp() chokes on a NULL pointer, too.

But the main point I want to make is that this really buys you little unless
you are working under 2 conditions:

	- strcmp() is not inlined
	- function calls are expensive

I think you'll have to agree that any professional C compiler worth the name
would almost certainly inline *at least* the str???() functions; given an
environment where function calls are expensive.

Also, if you tend to compare equal successfully a lot, this will actually
be slower.

-- 
Dave Schaumann | dave@cs.arizona.edu | Short .sig's rule!