Path: utzoo!attcan!uunet!husc6!cs.utexas.edu!pp!riunite!rfg
From: rfg@riunite.ACA.MCC.COM (Ron Guilmette)
Newsgroups: comp.lang.c++
Subject: Uses of incomplete types... which ones are legal?
Message-ID: <90@riunite.ACA.MCC.COM>
Date: 22 Jan 89 18:32:27 GMT
Reply-To: rfg@riunite.UUCP (Ron Guilmette)
Organization: MCC  Austin, Texas
Lines: 61


I am reposting the following response from ekrell to my question about
an apparent bug in GNU G++ with incomplete types.  Some people may be
interested in seeing how AT&T cfront 2.0 responds.

------------------------------------------------------------------------
>From ekrell%ulysses@research.att.com Sat Jan 21 13:06:09 1989

This is AT&T cfront 2.0:

x.C:
  1	class first_class;
  2
  3	class second_class {
  4	public:
  5		first_class some_function ();
  6	};
  7
  8	another_function ()
  9	{
 10		second_class second_class_object;
 11	
 12		second_class_object.some_function ();
 13	}

nCC -c x.C:

"x.C", line 12: error: first_class  undefined, size not known
"x.C", line 10: warning:  second_class_object used but not set
------------------------------------------------------------------------

Since G++ issues *no* errors for the code above, it seems apparent
that there is at least one bug in G++ (1.32.0) regarding incomplete
types.  I have also noted that G++ issues no errors even if the
incomplete type name (first_class) is used as a formal parameter
type in the declaration of "some_function".

Note that the above example is similar to one on page 290 of "The
Book".  The example there shows why one would like to be able to
use the names of "incomplete" types in the specification of the
parameter and return types for a function.  Unfortunately, the
usage shown seems to conflict with my understanding of the rules
regarding allowed uses of incomplete type in ANSI C, i.e. that
these type names can only be used to declare pointer types & objects.
I don't see anywhere in The Book where the allowed deviations from
this simple ANSI rule are explicitly stated.  Did I miss something
in The Book, or did I miss something in the ANSI C draft?

Please... I don't mind reading TFM, but just point me to where to
look.  (The index in The Book doesn't help much.  I looked for
"incomplete types" and it isn't even in there).

// Ron Guilmette  -  MCC  -  Experimental (parallel) Systems Kit Project
// 3500 West Balcones Center Drive,  Austin, TX  78759  -  (512)338-3740
// ARPA: rfg@mcc.com
// UUCP: {ihnp4,seismo,ucb-vax,gatech}!cs.utexas.edu!pp!rfg
-- 
// Ron Guilmette  -  MCC  -  Experimental (parallel) Systems Kit Project
// 3500 West Balcones Center Drive,  Austin, TX  78759  -  (512)338-3740
// ARPA: rfg@mcc.com
// UUCP: {ihnp4,seismo,ucb-vax,gatech}!cs.utexas.edu!pp!rfg