E*

[Richard Tobin]

7.26.3 Errors <errno.h>
Macros that begin with E and a digit or E and an uppercase letter
may be added to the declarations in the <errno.hheade r.

May an implementation define an error macro EOF with a value different
from the EOF in <stdio.h>?

-- Richard
--
:wq

[Jack Klein]

Re: E*

On 10 May 2008 21:12:28 GMT, richard@cogsci. ed.ac.uk (Richard Tobin)
wrote in comp.lang.c:

7.26.3 Errors <errno.h>
Macros that begin with E and a digit or E and an uppercase letter
may be added to the declarations in the <errno.hheade r.
>
May an implementation define an error macro EOF with a value different
from the EOF in <stdio.h>?

Perhaps in a non-standard extra include file, or guarded by some
conditional compilation guard in <errno.h>.

It could not unconditionally include such a definition in <errno.h>
when the compiler is invoked in conforming mode because it would break
conforming programs that included both <stdio.hand <errno.h>.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://c-faq.com/
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.l earn.c-c++

http://www.club.cc.cmu.edu/~ajo/docs/FAQ-acllc.html

[lawrence.jones@siemens.com]

Re: E*

Richard Tobin <richard@cogsci .ed.ac.ukwrote:

7.26.3 Errors <errno.h>
Macros that begin with E and a digit or E and an uppercase letter
may be added to the declarations in the <errno.hheade r.
>
May an implementation define an error macro EOF with a value different
from the EOF in <stdio.h>?

No. A strictly conforming program is allowed to include both headers
and the conflicting definitions of EOF would violate the constraint in
6.10.3p2. Since a conforming implementation is not allowed to have
entensions that invalidate strictly conforming programs, such a
definition is not allowed.

Conformance questions like this really belong in comp.std.c where
there's less traffic, much less heat, and far more light than there is
here.

-- Larry Jones

I think we need to change the rules. -- Calvin

[Eric Sosman]

Re: E*

Richard Tobin wrote:

7.26.3 Errors <errno.h>
Macros that begin with E and a digit or E and an uppercase letter
may be added to the declarations in the <errno.hheade r.
>
May an implementation define an error macro EOF with a value different
from the EOF in <stdio.h>?

No, because a conforming program can include both <stdio.h>
and <errno.h>. Note that the freedom to add more EXXX macros
in <errno.hdoes not imply that every EXXX macro everywhere
must be an error number.

The paragraph can be read as a warning and as an implied
promise. First, it warns you not to use EEYORE for your own
purposes, because a future <errno.hmight define it. And
second, it implies that future <errno.hversion s will *not*
define E12MANY or EiEiO or E_IS_FOR_ELEPHA NT, so you can use
those identifiers as you wish. However, implications are not
actual promises: Witness C99's encroachment on names reserved
to the programmer in C90.

--
Eric Sosman
esosman@ieee-dot-org.invalid

[Harald van =?UTF-8?b?RMSzaw==?=]

Re: E*

On Sun, 11 May 2008 09:01:36 -0400, Eric Sosman wrote:

Richard Tobin wrote:

> 7.26.3 Errors <errno.h>
> Macros that begin with E and a digit or E and an uppercase letter
> may be added to the declarations in the <errno.hheade r.

>
The paragraph can be read as a warning and as an implied
promise. First, it warns you not to use EEYORE for your own purposes,
because a future <errno.hmight define it. And second, it implies that
future <errno.hversion s will *not* define E12MANY or EiEiO or
E_IS_FOR_ELEPHA NT, so you can use those identifiers as you wish.

Why can't E12MANY be defined? It's a macro that begins with E and a digit,
which is allowed, right? I'm asking because E2BIG is a fairly common real-
world macro provided by <errno.h>.

[Eric Sosman]

Re: E*

Harald van Dijk wrote:

On Sun, 11 May 2008 09:01:36 -0400, Eric Sosman wrote:

>Richard Tobin wrote:

>> 7.26.3 Errors <errno.h>
>> Macros that begin with E and a digit or E and an uppercase letter
>> may be added to the declarations in the <errno.hheade r.

> The paragraph can be read as a warning and as an implied
>promise. First, it warns you not to use EEYORE for your own purposes,
>because a future <errno.hmight define it. And second, it implies that
>future <errno.hversion s will *not* define E12MANY or EiEiO or
>E_IS_FOR_ELEPH ANT, so you can use those identifiers as you wish.

>
Why can't E12MANY be defined? It's a macro that begins with E and a digit,
which is allowed, right? I'm asking because E2BIG is a fairly common real-
world macro provided by <errno.h>.

Oops! My blunder; sorry.

--
Eric Sosman
esosman@ieee-dot-org.invalid