Re: [1/2] msvcrt: Fixes multi-byte detection problem with MSVCRT_isleadbyte() (try 2)
_mbspbrk function needs to be fixed, not isleadbyte. Cheers, Piotr
_mbspbrk works fine. From what I could tell from my testing, MSVCRT_isleadbyte goes through _isctype and ignores the locale by using MSVCRT__ctype which never gets updated for locales (or if it does usually get updated, it's somehow completely failing). 837Dh for example is a マ in Japanese (Shift-JIS) but MSVCRT_isleadbyte before patch will not detect it but _ismbblead does. The only fix I can think of besides changing MSVCRT_isleadbyte directly is to change the calls to MSVCRT_isleadbyte with _ismbblead in the _mbspbrk function. That would be ignoring the real problem, though. On Wed, Nov 24, 2010 at 10:33 AM, Piotr Caban <piotr.caban(a)gmail.com> wrote:
_mbspbrk function needs to be fixed, not isleadbyte.
Cheers, Piotr
participants (2)
-
Nagato -
Piotr Caban