tommy/libpostal
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[numex] regen data file. utf8_is_hyphen requires a character, all other methods use category

Browse Source
This commit is contained in:
Al
2015-06-08 21:32:01 -04:00
parent c1d0afa52c
commit 81be8e771e
3 changed files with 19 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/string_utils.c
View File

@@ -142,8 +142,7 @@ error_free_output:
return NULL;
}
inline bool utf8_is_letter(int32_t ch) {
int cat = utf8proc_category(ch);
inline bool utf8_is_letter(int cat) {
return cat == UTF8PROC_CATEGORY_LL || cat == UTF8PROC_CATEGORY_LU \
|| cat == UTF8PROC_CATEGORY_LT || cat == UTF8PROC_CATEGORY_LO \
|| cat == UTF8PROC_CATEGORY_LM;
@@ -160,7 +159,8 @@ inline bool utf8_is_letter_or_number(int cat) {
|| cat == UTF8PROC_CATEGORY_NL || cat == UTF8PROC_CATEGORY_NO;
}
inline bool utf8_is_hyphen(int cat) {
inline bool utf8_is_hyphen(int32_t ch) {
int cat = utf8proc_category(ch);
return cat == UTF8PROC_CATEGORY_PD || ch == 0x2212;
}