Fix g_strtrim() regression

Commit 36ea4a3 failed to take UTF-8 characters into account on
platforms with signed char.
This commit is contained in:
matt335672 2023-12-24 20:14:14 +00:00
parent 4d03d69125
commit 62a51dead9
2 changed files with 37 additions and 3 deletions

View File

@ -715,6 +715,7 @@ g_strstr(const char *haystack, const char *needle)
int int
g_strtrim(char *str, int trim_flags) g_strtrim(char *str, int trim_flags)
{ {
#define TRIMMABLE_CHAR(c) ((unsigned char)(c) <= ' ')
int rv = 0; int rv = 0;
int index; int index;
int j; int j;
@ -726,7 +727,7 @@ g_strtrim(char *str, int trim_flags)
j = 0; j = 0;
for (index = 0; str[index] != '\0'; index++) for (index = 0; str[index] != '\0'; index++)
{ {
if (str[index] > ' ') if (!TRIMMABLE_CHAR(str[index]))
{ {
str[j++] = str[index]; str[j++] = str[index];
} }
@ -741,7 +742,7 @@ g_strtrim(char *str, int trim_flags)
case 2: /* trim right */ case 2: /* trim right */
index = strlen(str); index = strlen(str);
while (index > 0 && str[index - 1] <= ' ') while (index > 0 && TRIMMABLE_CHAR(str[index - 1]))
{ {
--index; --index;
} }
@ -750,7 +751,7 @@ g_strtrim(char *str, int trim_flags)
case 1: /* trim left */ case 1: /* trim left */
index = 0; index = 0;
while (str[index] != '\0' && str[index] <= ' ') while (str[index] != '\0' && TRIMMABLE_CHAR(str[index]))
{ {
++index; ++index;
} }
@ -765,6 +766,7 @@ g_strtrim(char *str, int trim_flags)
} }
return rv; return rv;
#undef TRIMMABLE_CHAR
} }
/*****************************************************************************/ /*****************************************************************************/

View File

@ -14,6 +14,24 @@
#define RESULT_LEN 1024 #define RESULT_LEN 1024
/* Universal character names need a C99 compiler */
#if __STDC_VERSION__ >= 199901L
# define CJK_UNIFIED_IDEOGRAPH_5E78 "\u5e78"
# define CJK_UNIFIED_IDEOGRAPH_798F "\u798f"
# define CJK_UNIFIED_IDEOGRAPH_5B89 "\u5b89"
# define CJK_UNIFIED_IDEOGRAPH_5EB7 "\u5eb7"
#else
// Assume we're using UTF-8
# define CJK_UNIFIED_IDEOGRAPH_5E78 "\xe5\xb9\xb8"
# define CJK_UNIFIED_IDEOGRAPH_798F "\xe7\xa6\x8f"
# define CJK_UNIFIED_IDEOGRAPH_5B89 "\xe5\xae\x89"
# define CJK_UNIFIED_IDEOGRAPH_5EB7 "\xe5\xba\xb7"
#endif
#define HAPPINESS_AND_WELL_BEING \
CJK_UNIFIED_IDEOGRAPH_5E78 CJK_UNIFIED_IDEOGRAPH_798F \
CJK_UNIFIED_IDEOGRAPH_5B89 CJK_UNIFIED_IDEOGRAPH_5EB7
START_TEST(test_strnjoin__when_src_is_null__returns_empty_string) START_TEST(test_strnjoin__when_src_is_null__returns_empty_string)
{ {
/* setup */ /* setup */
@ -1038,6 +1056,19 @@ START_TEST(test_strtrim__trim_through)
} }
END_TEST END_TEST
START_TEST(test_strtrim__chinese_chars)
{
/* setup */
char output[] = "\t\t \t" HAPPINESS_AND_WELL_BEING "\t\t \n\n";
/* test */
g_strtrim(output, 4);
/* verify */
ck_assert_str_eq(output, HAPPINESS_AND_WELL_BEING);
}
END_TEST
/******************************************************************************/ /******************************************************************************/
START_TEST(test_sigs__common) START_TEST(test_sigs__common)
@ -1192,6 +1223,7 @@ make_suite_test_string(void)
tcase_add_test(tc_strtrim, test_strtrim__trim_right); tcase_add_test(tc_strtrim, test_strtrim__trim_right);
tcase_add_test(tc_strtrim, test_strtrim__trim_both); tcase_add_test(tc_strtrim, test_strtrim__trim_both);
tcase_add_test(tc_strtrim, test_strtrim__trim_through); tcase_add_test(tc_strtrim, test_strtrim__trim_through);
tcase_add_test(tc_strtrim, test_strtrim__chinese_chars);
tc_sigs = tcase_create("signals"); tc_sigs = tcase_create("signals");
suite_add_tcase(s, tc_sigs); suite_add_tcase(s, tc_sigs);