bug in ctype.h

bug in ctype.h

[Levente Farkas]

hi,
I thing it's a real bug in ctype.h in cygwin:
-------------
#define _U  01
#define _L  02
#define _N  04
#define _S  010
#define _P  020
#define _C  040
#define _X  0100
#define _B  0200
--------------
and there is no undef pair of these defines. ok you can use every
name with starts with _, but it's a real nightmare (the _X is the
worst).

 -- lfarkas
 "The only thing worse than not knowing the truth is
  ruining the bliss of ignorance."

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: bug in ctype.h

[Chris Faylor]

On Mon, Mar 13, 2000 at 06:07:43PM +0100, Levente Farkas wrote:
>hi,
>I thing it's a real bug in ctype.h in cygwin:
>-------------
>#define _U  01
>#define _L  02
>#define _N  04
>#define _S  010
>#define _P  020
>#define _C  040
>#define _X  0100
>#define _B  0200
>--------------
>and there is no undef pair of these defines. ok you can use every
>name with starts with _, but it's a real nightmare (the _X is the
>worst).

Um, how would you undef these?  They're used in macros.

It was my understanding that symbols that begin with "_" were supposed
to be the province of a system library and are not supposed to be
used by user programs.  Or was it "__".  I can never remember.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: bug in ctype.h

[Levente Farkas]

Chris Faylor wrote:
> 
> On Mon, Mar 13, 2000 at 06:07:43PM +0100, Levente Farkas wrote:
> >hi,
> >I thing it's a real bug in ctype.h in cygwin:
> >-------------
> >#define _U  01
> >#define _L  02
> >#define _N  04
> >#define _S  010
> >#define _P  020
> >#define _C  040
> >#define _X  0100
> >#define _B  0200
> >--------------
> >and there is no undef pair of these defines. ok you can use every
> >name with starts with _, but it's a real nightmare (the _X is the
> >worst).
> 
> Um, how would you undef these?  They're used in macros.

that's another problem:-)
but simple after use
#undef _X
 
> It was my understanding that symbols that begin with "_" were supposed
> to be the province of a system library and are not supposed to be
> used by user programs.  Or was it "__".  I can never remember.

almost both, but such a macros are anoying anyway.

 -- lfarkas
 "The only thing worse than not knowing the truth is
  ruining the bliss of ignorance."

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: bug in ctype.h

[Chris Faylor]

On Mon, Mar 13, 2000 at 06:58:20PM +0100, Levente Farkas wrote:
>Chris Faylor wrote:
>> 
>> On Mon, Mar 13, 2000 at 06:07:43PM +0100, Levente Farkas wrote:
>> >hi,
>> >I thing it's a real bug in ctype.h in cygwin:
>> >-------------
>> >#define _U  01
>> >#define _L  02
>> >#define _N  04
>> >#define _S  010
>> >#define _P  020
>> >#define _C  040
>> >#define _X  0100
>> >#define _B  0200
>> >--------------
>> >and there is no undef pair of these defines. ok you can use every
>> >name with starts with _, but it's a real nightmare (the _X is the
>> >worst).
>> 
>> Um, how would you undef these?  They're used in macros.
>
>that's another problem:-)
>but simple after use
>#undef _X

Are you actually saying that you want to do something like this:

#include <ctype.h>

main(int argc, char **argv)
{
  printf ("%d", isalpha(*argv[0]));
#undef _U
#undef _L
#undef _N
#undef _S
#undef _P
#undef _C
#undef _X
#undef _B

Somehow, I don't see the point.

If you're advocating that these be undefined in ctype.h, then:

That won't work.

>>It was my understanding that symbols that begin with "_" were supposed
>>to be the province of a system library and are not supposed to be used
>>by user programs.  Or was it "__".  I can never remember.
>
>almost both, but such a macros are anoying anyway.

Uh huh.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: bug in ctype.h

[Levente Farkas]

Chris Faylor wrote:
> 
> On Mon, Mar 13, 2000 at 06:58:20PM +0100, Levente Farkas wrote:
> >Chris Faylor wrote:
> >>
> >> On Mon, Mar 13, 2000 at 06:07:43PM +0100, Levente Farkas wrote:
> >> >hi,
> >> >I thing it's a real bug in ctype.h in cygwin:
> >> >-------------
> >> >#define _U  01
> >> >#define _L  02
> >> >#define _N  04
> >> >#define _S  010
> >> >#define _P  020
> >> >#define _C  040
> >> >#define _X  0100
> >> >#define _B  0200
> >> >--------------
> >> >and there is no undef pair of these defines. ok you can use every
> >> >name with starts with _, but it's a real nightmare (the _X is the
> >> >worst).
> >>
> >> Um, how would you undef these?  They're used in macros.
> >
> >that's another problem:-)
> >but simple after use
> >#undef _X
> 
> Are you actually saying that you want to do something like this:
> 
> #include <ctype.h>
> 
> main(int argc, char **argv)
> {
>   printf ("%d", isalpha(*argv[0]));
> #undef _U
> #undef _L
> #undef _N
> #undef _S
> #undef _P
> #undef _C
> #undef _X
> #undef _B
> 
> Somehow, I don't see the point.
> 
> If you're advocating that these be undefined in ctype.h, then:
> 
> That won't work.
> 
> >>It was my understanding that symbols that begin with "_" were supposed
> >>to be the province of a system library and are not supposed to be used
> >>by user programs.  Or was it "__".  I can never remember.
> >
> >almost both, but such a macros are anoying anyway.
> 
> Uh huh.

simple look into glic-s header files (which don't use these macros).
why ?:-)

 -- lfarkas
 "The only thing worse than not knowing the truth is
  ruining the bliss of ignorance."

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: bug in ctype.h

[J. J. Farrell]

> From: Chris Faylor <cgf@cygnus.com>
> 
> On Mon, Mar 13, 2000 at 06:07:43PM +0100, Levente Farkas wrote:
> >
> >I thing it's a real bug in ctype.h in cygwin:
> >-------------
> >#define _U  01
> >#define _L  02
> >#define _N  04
> >#define _S  010
> >#define _P  020
> >#define _C  040
> >#define _X  0100
> >#define _B  0200
> >--------------
> >and there is no undef pair of these defines. ok you can use every
> >name with starts with _, but it's a real nightmare (the _X is the
> >worst).
> 
> Um, how would you undef these?  They're used in macros.
> 
> It was my understanding that symbols that begin with "_" were supposed
> to be the province of a system library and are not supposed to be
> used by user programs.  Or was it "__".  I can never remember.

Identifiers beginning "_" followed by another "_" or an upper-case
letter are reserved in all situations. Any "ordinary" program which
uses such identifiers is incorrect. This is not a bug.

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: bug in ctype.h

[Chris Faylor]

On Mon, Mar 13, 2000 at 05:19:28PM -0800, J. J. Farrell wrote:
>>It was my understanding that symbols that begin with "_" were supposed
>>to be the province of a system library and are not supposed to be used
>>by user programs.  Or was it "__".  I can never remember.
>
>Identifiers beginning "_" followed by another "_" or an upper-case
>letter are reserved in all situations.  Any "ordinary" program which
>uses such identifiers is incorrect.  This is not a bug.

Thanks for the clarification.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com