Path: utzoo!utgpu!water!watmath!clyde!rutgers!umd5!trantor.umd.edu!chris
From: chris@trantor.umd.edu (Chris Torek)
Newsgroups: comp.lang.c
Subject: Re: Forward reference for structure?
Message-ID: <2295@umd5.umd.edu>
Date: 14 Feb 88 03:21:54 GMT
References: <173@heurikon.UUCP> <341@hub.ucsb.edu> <359@hub.ucsb.edu>
Sender: ris@umd5.umd.edu
Reply-To: chris@trantor.umd.edu (Chris Torek)
Organization: University of Maryland, College Park
Lines: 51

>In article <173@heurikon.UUCP> lampman@heurikon.UUCP (Ray Lampman) writes:
>>What is the best way to declare three different structures, each
>>containing a pointer to the other two?

In article <359@hub.ucsb.edu> angst%csilvax@hub.ucsb.edu writes:
>... I'm confused about this forward referencing bit.  I posted a
>response to this which included what I thought was a solution
>(because my program compiles *without* doing anything special
>w/regard to forward referencing), ....
>
>Can someone please enlighten me as to why there seems to be such a
>problem with this?

There is nothing illegal about

	% cat file.c
	struct goo { struct foo *fp; int gooval; };
	struct foo { struct goo *gp; int fooval; };
	%

The problem that occurs is when the same name is redefined, viz:

	struct global { int v; };
	f() {
		struct local { struct global *gp; } l;
		struct global /* not really */ { char *s; } oops;

The (struct global *) element of l called `gp' is a pointer to the
`really global' structure; the only member of this is `v'.  l.gp
cannot be made to point to `oops', even though it is an object of
type `struct global', because it is a different `struct global'.

The dpANS provides a way of telling the compiler `forget about any
outer definitions of this structure, because I am going to redefine
it at this nesting level'.  Changing the code to

	struct global { int v; };
	f() {
		struct global;
		struct local { struct global *gp; } l;
		struct global /* not really */ { char *s; } oops;

makes it Officially Correct.  Of course, it still does not compile
under PCC-based compilers (4.3BSD at least).

I will make no comment as to whether redefining structures locally
is a good idea.
-- 
In-Real-Life: Chris Torek, Univ of MD Computer Science, +1 301 454 7163
(hiding out on trantor.umd.edu until mimsy is reassembled in its new home)
Domain: chris@mimsy.umd.edu		Path: not easily reachable