Path: utzoo!news-server.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!magnus.ircc.ohio-state.edu!zaphod.mps.ohio-state.edu!think.com!hsdndev!spdcc!dirtydog!karl
From: karl@ima.isc.com (Karl Heuer)
Newsgroups: comp.lang.c
Subject: Re: Tough stupid array question
Message-ID: <1991Mar04.020619.22787@ima.isc.com>
Date: 4 Mar 91 02:06:19 GMT
References: <1991Feb27.054609.10792@nlm.nih.gov> <10362@dog.ee.lbl.gov>
Sender: news@ima.isc.com (NEWS ADMIN)
Reply-To: karl@ima.isc.com (Karl Heuer)
Distribution: usa
Organization: Interactive Systems, Cambridge, MA 02138-5302
Lines: 20

In article <10362@dog.ee.lbl.gov> torek@elf.ee.lbl.gov (Chris Torek) writes:
>[You need to use "foo(&iray)" to make the program type-correct]
>(actually, given your defintion for DIM1 &c, it would probably be
>clearer to write
>	for (i = 0; i < DIM1(*ray); ++i)

Note that, if all mention of "ray" is via the construct "*ray" (I believe this
is true given the above modification), it may be reasonable to use a reference
rather than a pointer.  C++ has builtin syntax for this; in C it can be
emulated with a macro (using #undef to emulate scope):
	f(int (*ptr_ray)[5][1]) {
	#define jray (*ptr_ray)
	    for (i = 0; i < DIM1(jray); ++i) ...
	    jray[0][0] = ...
	#undef jray
	}
This may clean up the perceived "ugliness" of the pointer-to-array code.
Whether it's worth the added ugliness of the macro is a judgement call.

Karl W. Z. Heuer (karl@ima.isc.com or uunet!ima!karl), The Walking Lint