Walter

I know I was the one who asked for it, but it's _too_ effective. Or, at least, is there a suppression pragma, and, if so, can you arrange to have that specified in the DMC++ header files?

<G>


-- 
Matthew Wilson

Author: "Imperfect C++", Addison-Wesley, 2004
    (http://www.imperfectcplusplus.com)
Contributing editor, C/C++ Users Journal
    (http://www.synesis.com.au/articles.html#columns)
STLSoft moderator
    (http://www.stlsoft.org)

"I can't sleep nights till I found out who hurled what ball through what apparatus" -- Dr Niles Crane

-------------------------------------------------------------------------------

"Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cuppnc$84h$1@digitaldaemon.com...
> Walter
>
> I know I was the one who asked for it, but it's _too_ effective. Or, at least, is there a suppression pragma, and, if so, can you arrange to have that specified in the DMC++ header files?
>
> <G>

Suppression pragmas? Uggh. <g>

"Walter" <newshound@digitalmars.com> wrote in message news:cvmcdp$2qt3$1@digitaldaemon.com...
>
> "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cuppnc$84h$1@digitaldaemon.com...
>> Walter
>>
>> I know I was the one who asked for it, but it's _too_ effective. Or,
>> at
>> least, is there a suppression pragma, and, if so, can you arrange to
>> have that specified in the DMC++ header files?
>>
>> <G>
>
> Suppression pragmas? Uggh. <g>

So what's the answer?

"Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cvmgdk$2ut2$1@digitaldaemon.com...
>
> "Walter" <newshound@digitalmars.com> wrote in message news:cvmcdp$2qt3$1@digitaldaemon.com...
> >
> > "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cuppnc$84h$1@digitaldaemon.com...
> >> Walter
> >>
> >> I know I was the one who asked for it, but it's _too_ effective. Or,
> >> at
> >> least, is there a suppression pragma, and, if so, can you arrange to
> >> have that specified in the DMC++ header files?
> >>
> >> <G>
> >
> > Suppression pragmas? Uggh. <g>
>
> So what's the answer?

I'd rather not. Switches to turn warnings on, pragmas to turn them off - I don't think there's much to be gained here that's worth the complexity.

"Walter" <newshound@digitalmars.com> wrote in message news:cvt8q3$mra$1@digitaldaemon.com...
>
> "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cvmgdk$2ut2$1@digitaldaemon.com...
>>
>> "Walter" <newshound@digitalmars.com> wrote in message news:cvmcdp$2qt3$1@digitaldaemon.com...
>> >
>> > "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cuppnc$84h$1@digitaldaemon.com...
>> >> Walter
>> >>
>> >> I know I was the one who asked for it, but it's _too_ effective.
>> >> Or,
>> >> at
>> >> least, is there a suppression pragma, and, if so, can you arrange
>> >> to
>> >> have that specified in the DMC++ header files?
>> >>
>> >> <G>
>> >
>> > Suppression pragmas? Uggh. <g>
>>
>> So what's the answer?
>
> I'd rather not. Switches to turn warnings on, pragmas to turn them
> off - I
> don't think there's much to be gained here that's worth the
> complexity.

Ok. The alternative is to fix the C-style casts in the DMC++ headers. Are you amenable to adopting fixes, should I be motivated, from time to time, to detect and effect them?

"Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cvtkn4$12j7$1@digitaldaemon.com...
> > I'd rather not. Switches to turn warnings on, pragmas to turn them
> > off - I
> > don't think there's much to be gained here that's worth the
> > complexity.
>
> Ok. The alternative is to fix the C-style casts in the DMC++ headers. Are you amenable to adopting fixes, should I be motivated, from time to time, to detect and effect them?

Sure.

Cool. I'll stick it on the to-do. ;)

"Walter" <newshound@digitalmars.com> wrote in message news:cvvt4t$fot$2@digitaldaemon.com...
>
> "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cvtkn4$12j7$1@digitaldaemon.com...
>> > I'd rather not. Switches to turn warnings on, pragmas to turn them
>> > off - I
>> > don't think there's much to be gained here that's worth the
>> > complexity.
>>
>> Ok. The alternative is to fix the C-style casts in the DMC++ headers.
>> Are you amenable to adopting fixes, should I be motivated, from time
>> to
>> time, to detect and effect them?
>
> Sure.
>
> 

string.h, lines 312-324

inline const char *strchr(const char *s, int n) { return
strchr(const_cast<char*>(s), n); }
inline const char *strrchr(const char *s, int n) { return
strrchr(const_cast<char*>(s), n); }
inline const char *strpbrk(const char *s1,const char *s2) { return
strpbrk(const_cast<char*>(s1), s2); }
inline const char *strstr(const char *s1,const char *s2) { return
strstr(const_cast<char*>(s1), s2); }
inline const void *memchr(const void *s, int c, size_t n) { return
memchr(const_cast<void*>(s), c, n); }

#ifndef __WCS_INLINE
#define __WCS_INLINE
inline const wchar_t *wcschr(const wchar_t *s, wchar_t c) { return
wcschr(const_cast<wchar_t*>(s), c); }
inline const wchar_t *wcsrchr(const wchar_t *s, wchar_t c) { return
wcsrchr(const_cast<wchar_t*>(s), c); }
inline const wchar_t *wcspbrk(const wchar_t *s1,const wchar_t *s2) { return
wcspbrk(const_cast<wchar_t*>(s1), s2); }
inline const wchar_t *wcswcs(const wchar_t *s1,const wchar_t *s2) { return
wcswcs(const_cast<wchar_t*>(s1), s2); }
inline const void *wmemchr(const void *s, wchar_t c, size_t n) { return
wmemchr(const_cast<void*>(s), c, n); }


Note: the strpbrk/wcspbrk were actually wrong. They were:

inline const char *strpbrk(const char *s1,const char *s2) { return (char *)
strpbrk((const char *)s1,s2); }
inline const wchar_t *wcspbrk(const wchar_t *s1,const wchar_t *s2) { return
(wchar_t *) wcspbrk((const wchar_t *)s1,s2); }


The cast on the return type is not necessary, but benign.
The incorrect cast on the s1 argument looks like it'd result in a nasty stack
exhaustion. ;)

The following test program produces correct behaviour when used with the modified string.h, and just dies in the strpbrk() call in the original (i.e. we never see "Exiting" or "Something bad happened" or "s: uff")


    //#include <string.h.old>
    #include <string.h>
    #include <stdio.h>

    int main()
    {
        try
        {
            const char  stuff[] =   "stuff";

            printf("Calling strpbrk()\n");

            char const  *s  =   strpbrk(stuff, "fu");

            printf("s: %s\n", s);

            return 0;
        }
        catch(...)
        {
            printf("Something bad happened\n");

            return 1;
        }

        printf("Exiting\n");

        return 0;
    }



"Walter" <newshound@digitalmars.com> wrote in message news:cvvt4t$fot$2@digitaldaemon.com...
>
> "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cvtkn4$12j7$1@digitaldaemon.com...
>> > I'd rather not. Switches to turn warnings on, pragmas to turn them
>> > off - I
>> > don't think there's much to be gained here that's worth the
>> > complexity.
>>
>> Ok. The alternative is to fix the C-style casts in the DMC++ headers. Are you amenable to adopting fixes, should I be motivated, from time to time, to detect and effect them?
>
> Sure.
>
> 

stlport/stl/_algobase.h

Lines 159-150

    (static_cast<char*>(memmove(__result, __first, (static_cast<char const*>(__last) - static_cast<char
const*>(__first))))) +
    (static_cast<char const*>(__last) - static_cast<char const*>(__first));

Lines 183-184

  const ptrdiff_t _Num = static_cast<char const*>(__last) - static_cast<char const*>(__first);
  return (_Num > 0) ? memmove(static_cast<char*>(__result) - _Num, __first, _Num) : __result ;

stlport/stl/_uninitialized.h

Line 84:

  return  static_cast<char*>(__copy_trivial (__first, __last, __result));


Line 90:

  return  static_cast<wchar_t*>(__copy_trivial (__first, __last, __result));


"Walter" <newshound@digitalmars.com> wrote in message news:cvvt4t$fot$2@digitaldaemon.com...
>
> "Matthew" <admin@stlsoft.dot.dot.dot.dot.org> wrote in message news:cvtkn4$12j7$1@digitaldaemon.com...
>> > I'd rather not. Switches to turn warnings on, pragmas to turn them
>> > off - I
>> > don't think there's much to be gained here that's worth the
>> > complexity.
>>
>> Ok. The alternative is to fix the C-style casts in the DMC++ headers. Are you amenable to adopting fixes, should I be motivated, from time to time, to detect and effect them?
>
> Sure.
>
>