NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/std/c/330

Internet Message Format | 1996-08-06 | 2.0 KB

Path: mail2news.demon.co.uk!genesis.demon.co.uk
From: Lawrence Kirby <[e-mail]>
Newsgroups: comp.lang.c,comp.std.c
Subject: Re: Integral conversion e.t.c. (was: Re: Hungarian notation)
Date: Thu, 08 Feb 96 23:20:26 GMT
Organization: none
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <4epvo3$[e-mail]>
Reply-To: [e-mail]
X-NNTP-Posting-Host: genesis.demon.co.uk
X-Newsreader: Demon Internet Simple News v1.27
X-Mail2News-Path: genesis.demon.co.uk

In article <4epvo3$[e-mail]> [e-mail] "Jerry Coffin" writes:

>Lawrence Kirby <[e-mail]> wrote:
>
>>Unfortunately there are some. For instance POSIX defines fdopen() and
>>fileno() in stdio.h. It would have made more sense to put these in unistd.h
>>alongsize read, write and close, even though that means defining FILE
>>in unistd.h
>
>However, this only happens if _POSIX_SOURCE is defined.  _POSIX_SOURCE
>is in the implementor's name space, so no strictly conforming program
>defines it.  As such, stdio.h can (and certainly should) be defined to
>allow POSIX conformance without bothering C programs that want to use
>the same names for other purposes.

It remains extremely unfortunate that I have to define _POSIX_SOURCE (with
all of its other consequences) just to get fdopen() (plus various other
things) declared. It just makes much more sense to keep POSIX headers
entirely separate from ANSI headers.

I'll try to get back on topic now! :-)

Also it depends on what exactly defines _POSIX_SOURCE. If it is in the
program source then all bets are off. If on the other hand the implemenation
defined it (say in response to a compiler command line argument) then it
ceases to be ANSI conforming by declaring fdopen().

--
-----------------------------------------
Lawrence Kirby | [e-mail]
Wilts, England | [e-mail]
-----------------------------------------