Created
March 29, 2012 09:00
-
-
Save camillol/2235195 to your computer and use it in GitHub Desktop.
patch to fix glib collation on recent OS X
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
diff --git a/glib/gunicollate.c b/glib/gunicollate.c | |
index 3649786..bd66c22 100644 | |
--- a/glib/gunicollate.c | |
+++ b/glib/gunicollate.c | |
@@ -217,29 +217,25 @@ collate_key_to_string (UCCollationValue *key, | |
gsize key_len) | |
{ | |
gchar *result; | |
- gsize result_len; | |
- gsize i; | |
- | |
- /* Pretty smart algorithm here: ignore first eight bytes of the | |
- * collation key. It doesn't produce results equivalent to | |
- * UCCompareCollationKeys's, but the difference seems to be only | |
- * that UCCompareCollationKeys in some cases produces 0 where our | |
- * comparison gets -1 or 1. */ | |
+ gsize result_len = 0; | |
+ gsize i, start; | |
- if (key_len * sizeof (UCCollationValue) <= 8) | |
+ /* The first codes should be skipped: the same string on the same system can get | |
+ * different values at runtime in those positions, and they do not sort correctly. | |
+ * The exact size of the prefix depends on whether we are building 64 or 32 bit. */ | |
+ start = 2 * sizeof(void *) / sizeof(UCCollationValue); | |
+ if (key_len <= start) | |
return g_strdup (""); | |
- result_len = 0; | |
- for (i = 8; i < key_len * sizeof (UCCollationValue); i++) | |
- /* there may be nul bytes, encode byteval+1 */ | |
- result_len += utf8_encode (NULL, *((guchar*)key + i) + 1); | |
+ for (i = start; i < key_len; i++) | |
+ result_len += utf8_encode (NULL, htonl(key[i]+1)); | |
result = g_malloc (result_len + 1); | |
result_len = 0; | |
- for (i = 8; i < key_len * sizeof (UCCollationValue); i++) | |
- result_len += utf8_encode (result + result_len, *((guchar*)key + i) + 1); | |
+ for (i = start; i < key_len; i++) | |
+ result_len += utf8_encode (result + result_len, htonl(key[i]+1)); | |
- result[result_len] = 0; | |
+ result[result_len] = '\0'; | |
return result; | |
} | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment