Path: utzoo!mnetor!uunet!husc6!bbn!rochester!PT.CS.CMU.EDU!IUS1.CS.CMU.EDU!edw
From: edw@IUS1.CS.CMU.EDU (Eddie Wyatt)
Newsgroups: comp.lang.c
Subject: Re: Priorities of = and == again (c'mon you ANSI freaks)
Message-ID: <665@PT.CS.CMU.EDU>
Date: 18 Jan 88 15:05:43 GMT
References: <1171@ark.cs.vu.nl>
Sender: netnews@PT.CS.CMU.EDU
Organization: Carnegie-Mellon University, CS/RI
Lines: 41

In article <1171@ark.cs.vu.nl>, maart@cs.vu.nl (Maarten Litmaath) writes:
> Lately I asked why the priorities of = and == were not reversed:
> 
> 	while ((c = getchar()) == ' ')
> 
> is used much more than
> 
> 	while (c = getchar() == ' ')
> .
> I argued that ONLY programs containing code of the second form,
> need to be rewritten using parentheses; occurrences of those
> expressions are not hard to find using shelltools (SUN, yeah).
> Concluding, I'm not very impressed by the 'backward compatibility'
> argument used by ANSI and by Doug Gwyn, who responded to my previous
> article (thanks, Doug).
> Any further thoughts on this ?

   Changing the precedence rules will will break more than just
those while loops that you are complaining about.  Consider

	found = key == entry.key;

Perfectly valid and it is pretty obviously that the programmer
didn't intend the expression to be parsed as:

	(found = key) == entry.key;

which is what would happen if you change the precedence of "=" to be
higher than "==".  Then again, we could change it such that in the
context of either "while" or odd number Tuesdays of an even number
month "=" has a  higher precedent than "==". :-)

BTW I usually parenthesises expressions involving boolean assignment
to make it clearly as to what is going on:

	found = (key == entry.key);


-- 

Eddie Wyatt 				e-mail: edw@ius1.cs.cmu.edu