since so many internal things hardcode 'C' locale anyway,
use own is{digit,lower,upper} macros that go via byte ranges doesn't affect utf-8 hack because these only operate on single bytes anyway saves 224t
This commit is contained in:
4
misc.c
4
misc.c
@ -3,7 +3,7 @@
|
||||
|
||||
#include "sh.h"
|
||||
|
||||
__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.35 2006/11/10 06:27:09 tg Exp $\t"
|
||||
__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.36 2006/11/10 06:53:27 tg Exp $\t"
|
||||
MKSH_SH_H_ID);
|
||||
|
||||
#undef USE_CHVT
|
||||
@ -437,7 +437,7 @@ getn(const char *s, int *ai)
|
||||
c = *s++;
|
||||
*ai = i = 0;
|
||||
do {
|
||||
if (!isdigit(c))
|
||||
if (!ksh_isdigit(c))
|
||||
goto getn_out;
|
||||
i *= 10;
|
||||
if (i < *ai)
|
||||
|
Reference in New Issue
Block a user