r13335: Fix the build and add an utf8 safe ldb_hadler_fold function

based on ldb_casefold
(This used to be commit 6104f900863c688707809d42c5429a42d654d5fb)

3 files changed, 44 insertions, 15 deletions

diff --git a/source4/lib/ldb/common/attrib_handlers.c b/source4/lib/ldb/common/attrib_handlers.c
index 2109e9b2da..4b6a7af1ee 100644
--- a/source4/lib/ldb/common/attrib_handlers.c
+++ b/source4/lib/ldb/common/attrib_handlers.c
@@ -46,32 +46,60 @@ int ldb_handler_copy(struct ldb_context *ldb, void *mem_ctx,
 /*
   a case folding copy handler, removing leading and trailing spaces and
   multiple internal spaces
+
+  We exploit the fact that utf8 never uses the space octet except for
+  the space itself
 */
 static int ldb_handler_fold(struct ldb_context *ldb, void *mem_ctx,
 			    const struct ldb_val *in, struct ldb_val *out)
 {
-	uint8_t *s1, *s2;
-	out->data = talloc_size(mem_ctx, strlen((char *)in->data)+1);
+	char *s, *t;
+	int l;
+	if (!in || !out || !(in->data)) {
+		return -1;
+	}
+
+	out->data = (uint8_t *)ldb_casefold(ldb, mem_ctx, (const char *)(in->data));
 	if (out->data == NULL) {
-		ldb_oom(ldb);
+		ldb_debug(ldb, LDB_DEBUG_ERROR, "ldb_handler_fold: unable to casefold string [%s]", in->data);
 		return -1;
 	}
-	s1 = in->data;
-	s2 = out->data;
-	while (*s1 == ' ') s1++;
-	while (*s1) {
-		*s2 = toupper(*s1);
-		if (s1[0] == ' ') {
-			while (s1[0] == s1[1]) s1++;
+
+	s = (char *)(out->data);
+	
+	/* remove trailing spaces if any */
+	l = strlen(s);
+	while (s[l - 1] == ' ') l--;
+	s[l] = '\0';
+	
+	/* remove leading spaces if any */
+	if (*s == ' ') {
+		for (t = s; *s == ' '; s++) ;
+
+		/* remove leading spaces by moving down the string */
+		memmove(t, s, l);
+
+		s = t;
+	}
+
+	/* check middle spaces */
+	while ((t = strchr(s, ' ')) != NULL) {
+		for (s = t; *s == ' '; s++) ;
+
+		if ((s - t) > 1) {
+			l = strlen(s);
+
+			/* remove all spaces but one by moving down the string */
+			memmove(t + 1, s, l);
 		}
-		s2++; s1++;
 	}
-	*s2 = 0;
+
 	out->length = strlen((char *)out->data);
 	return 0;
 }
 
 
+
 /*
   canonicalise a ldap Integer
   rfc2252 specifies it should be in decimal form
@@ -114,8 +142,8 @@ int ldb_comparison_binary(struct ldb_context *ldb, void *mem_ctx,
 }
 
 /*
-  compare two case insensitive strings, ignoring multiple whitespace
-  and leading and trailing whitespace
+  compare two case insensitive strings, ignoring multiple whitespaces
+  and leading and trailing whitespaces
   see rfc2252 section 8.1
 */
 static int ldb_comparison_fold(struct ldb_context *ldb, void *mem_ctx,
diff --git a/source4/lib/ldb/include/ldb.h b/source4/lib/ldb/include/ldb.h
index 7c39aeeeb9..4e457b028f 100644
--- a/source4/lib/ldb/include/ldb.h
+++ b/source4/lib/ldb/include/ldb.h
@@ -970,6 +970,7 @@ int ldb_dn_cmp(struct ldb_context *ldb, const char *dn1, const char *dn2);
    case; non-zero if there are any differences
 */
 int ldb_attr_cmp(const char *attr1, const char *attr2);
+char *ldb_attr_casefold(void *mem_ctx, const char *s);
 int ldb_attr_dn(const char *attr);
 char *ldb_dn_escape_value(void *mem_ctx, struct ldb_val value);
 
diff --git a/source4/lib/ldb/ldb_tdb/ldb_index.c b/source4/lib/ldb/ldb_tdb/ldb_index.c
index ac3063ef28..5c601f9ea8 100644
--- a/source4/lib/ldb/ldb_tdb/ldb_index.c
+++ b/source4/lib/ldb/ldb_tdb/ldb_index.c
@@ -106,7 +106,7 @@ static struct ldb_dn *ldb_dn_key(struct ldb_context *ldb,
 	const struct ldb_attrib_handler *h;
 	char *attr_folded;
 
-	attr_folded = ldb_attr_casefold(ldb, ldb, attr);
+	attr_folded = ldb_attr_casefold(ldb, attr);
 	if (!attr_folded) {
 		return NULL;
 	}