Xref: utzoo comp.lang.c++:13484 comp.std.c:4840
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!crdgw1!uunet!zephyr.ens.tek.com!tektronix!reed!intelhf!ichips!inews!pima!bhoughto
From: bhoughto@pima.intel.com (Blair P. Houghton)
Newsgroups: comp.lang.c++,comp.std.c
Subject: Re: Pointers to functions
Message-ID: <4253@inews.intel.com>
Date: 15 May 91 23:37:37 GMT
References: <AJ3U.91May14140642@wilbury.cs.virginia.edu> <1512@caslon.cs.arizona.edu>
Sender: news@inews.intel.com
Organization: Intel Corp, Chandler, AZ
Lines: 32

In article <1512@caslon.cs.arizona.edu> dave@cs.arizona.edu (Dave Schaumann) writes:
>You have found one of the weirdnesses of ANSI-C.  When you have a pointer
>to a function, you (normally) invoke the function by saying
>	(*funptr)()
>However, ANSI, in it's infinite wisdom, decided that you should be able to
>say
>	funptr()
>in the same context, with similar results.

It's not weird at all.  It's rather consistent.  The actual
syntax is

	<expression>(<optional list of arguments>)

where the expression must evaluate to a function pointer.
A similar thing is true of arrays, structs, and unions, where
the syntaxes are

	<expression>[<expression>]
	<expression>.<member name>
	<expression>-><member name>

The only thing remotely weird about it is that now `funptr'
and `*funptr' are the same thing, but that's why function
poiners are often singled out for different semantic
treatment in the standard (many things you can do with
a pointer to int are forbidden with a function pointer).

				--Blair
				  "Many things you can do with
				   your own nose are forbidden
				   with your neighbor's."