diff options
| -rw-r--r-- | ChangeLog | 3 | ||||
| -rw-r--r-- | locale/programs/ld-collate.c | 25 |
2 files changed, 24 insertions, 4 deletions
diff --git a/ChangeLog b/ChangeLog index ef1d856bce..9965df01af 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,8 @@ 2000-06-28 Ulrich Drepper <drepper@redhat.com> + * locale/programs/ld-collate.c (insert_weights): Handle <Uxxxx> + symbols in multi-character weights correctly. + * wcsmbs/mbrtowc.c: Set flush to 1 for conversion function calls depending on whether the input string is empty or not. diff --git a/locale/programs/ld-collate.c b/locale/programs/ld-collate.c index 73a186df26..97059c2593 100644 --- a/locale/programs/ld-collate.c +++ b/locale/programs/ld-collate.c @@ -724,8 +724,10 @@ insert_weights (struct linereader *ldfile, struct element_t *elem, { if (*cp == '<') { - /* Ahh, it's a bsymbol. That's what we want. */ + /* Ahh, it's a bsymbol or an UCS4 value. If it's + the latter we have to unify the name. */ const char *startp = ++cp; + size_t len; while (*cp != '>') { @@ -738,9 +740,24 @@ insert_weights (struct linereader *ldfile, struct element_t *elem, ++cp; } - charelem = find_element (ldfile, collate, startp, - cp - startp); - ++cp; + if (cp - startp == 5 && startp[0] == 'U' + && isxdigit (startp[1]) && isxdigit (startp[2]) + && isxdigit (startp[3]) && isxdigit (startp[4])) + { + unsigned int ucs4 = strtoul (startp + 1, NULL, 16); + char *newstr; + + newstr = (char *) xmalloc (10); + snprintf (newstr, 10, "U%08X", ucs4); + startp = newstr; + + len = 9; + } + else + len = cp - startp; + + charelem = find_element (ldfile, collate, startp, len); + ++cp; } else { |