[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
> No, I meant that the routine, if outputting a terminating NULL, should
> ensure that it does so in the bounds of the given size (effectively making
> sure it substracts itself the needed space for the NULL from the given
> size argument).
But the ANSI strncat() doesn't do it. Should Allegro do it then ?
> Inconsistency between Allegro routines ? If these are the same as the
> inconsistencies between libc functions, then I think this is OK.
I disagree here. Unless someone can give me a good reason for the behaviour
of strncpy() compared with that of strncat(), I'll keep viewing it as an
inconsistency.
> If you prefer another (more self consistent) behavior, it is still
> possible to add either other functions or macros which alter the
> size parameter, as:
>
> #define ustrnconcat(s,size) ustrncat((s),(size)-ucwidth(0))
I'm ok with ustrncat(): it follows ANSI standard's behaviour.
ustrncpy() is the problem.
--
Eric Botcazou
ebotcazou@xxxxxxxxxx