Re: snprintf heads-up

On Fri, 8 Sep 2000, Stanislav Malyshev wrote: > ZS>> written. As it doesn't always appear to be the case, the best approach > > More precisely, it almost always is not the case, once the 'n' part comes > into play. Well, right now this is the behavior whenever there's no system-supplied snprintf() (again, assuming that's the way snprintf.c really behaves, which needs to be tested). > ZS>> IMHO would be using the bundled snprintf() always, regardless of whether > ZS>> snprintf() is available on the platform. > ZS>> As far as I can tell from glancing through snprintf.c, this version of > ZS>> snprintf() would return the number of characters that were effectively > ZS>> written under all circumstances (I could be wrong, we need to test it). > > That would be highly confusing behaviour - calling something snprintf > where it isn't snprintf. Then we better call it safe_snprintf or something > like that. Perhaps. snprintf() isn't standard yet, and as we saw, there are several different behaviors on different platforms to begin with. But I guess that seeing one thing in a manpage and experiencing a different behavior in practice could get confusing. I'd use php_snprintf(). Zeev -- Zeev Suraski <zeev@zend.com> http://www.zend.com/