Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!ll-xn!husc6!seismo!mimsy!chris
From: chris@mimsy.UUCP
Newsgroups: comp.lang.c
Subject: Re: (*pointer_to_function_returning_int)(args)
Message-ID: <7088@mimsy.UUCP>
Date: Wed, 17-Jun-87 02:22:47 EDT
Article-I.D.: mimsy.7088
Posted: Wed Jun 17 02:22:47 1987
Date-Received: Fri, 19-Jun-87 01:05:41 EDT
References: <16673@cca.CCA.COM> <7048@mimsy.UUCP> <786@geac.UUCP>
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 31

>In article <7048@mimsy.UUCP> I asked
>>Why not ... accept either form?

---the forms being (*f)(x) and f(x), and hence (****f)(x) &c.

In article <786@geac.UUCP> daveb@geac.UUCP (Dave Brown) writes:
>... the more the compiler does for you, the more it has to guarantee
>that what it did will never cause future problems.  

I daresay this change, considering that it is no change at all to
some extant C compilers, is safe.

>    Friendlyness -> Complexity -> Defaults -> PL/1 

I am not sure what the arrows are supposed to mean---implication
perhaps? If so, I say only that accepting either form in fact
simplifies the compiler, for there is then only one rule for function
calls.  Precisely, in all contexts save declarations, an object of
type `function returning T' is, by finding the object's address,
converted to an object of type `pointer to function returning T'.
The only value that may appear to the left of the parentheses in
a function call is one of type `pointer to function returning T';
the value of the function call is of type T.

Do you see why this allows any number of asterisk symbols?  (If
not, consider:  What type results from applying `*' to an object
of type `pointer to function returning T'?  What must immediately
happen to this object?)
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7690)
Domain:	chris@mimsy.umd.edu	Path:	seismo!mimsy!chris