Multiple Unicode compatibility bugs (crashes, incorrect URLs, etc. when using non-English locale different from default one. For example, Greek symbol on Cyrillic Windows.)
The bugs are related to functions like isalpha(), isdigit(), isalnum(), etc.
The following code contains bug:
char c; if( isalpha(c) ) {...}
The correct code should be:
if( isalpha( (int)(unsigned char)c ) {...}
I certainly know that we inherited a flaky unicode managemenr. However, the
truth is that I do not have the time to fix this. Therefore I would be
grateful if you suggested specific code changes or detailled specific bugs.
On Sunday, November 16, 2014 10:50:47 AM textsharik wrote:
Related
Bugs:
#248Added lots of (unsigned char) casts to protect ctype macros.
In the absence of specific tests, I cannot be certain it is complete.