pcre32: exec: Mask bits > 21 in 32-bit UTF mode

Allow passing characters with high bits set in UTF-32 mode.

git-svn-id: svn://vcs.exim.org/pcre/code/trunk@1100 2f5784b3-3f2a-0410-8824-cb99058d5e15

1 files changed, 44 insertions, 1 deletions

diff --git a/pcre_string_utils.c b/pcre_string_utils.c
index 54a75f6..94a5126 100644
--- a/pcre_string_utils.c
+++ b/pcre_string_utils.c
@@ -81,6 +81,27 @@ while (*str1 != '\0' || *str2 != '\0')
 return 0;
 }
 
+#ifdef COMPILE_PCRE32
+
+int
+PRIV(strcmp_uc_uc_utf)(const pcre_uchar *str1, const pcre_uchar *str2)
+{
+pcre_uchar c1;
+pcre_uchar c2;
+
+while (*str1 != '\0' || *str2 != '\0')
+  {
+  c1 = RAWUCHARINC(str1);
+  c2 = RAWUCHARINC(str2);
+  if (c1 != c2)
+    return ((c1 > c2) << 1) - 1;
+  }
+/* Both length and characters must be equal. */
+return 0;
+}
+
+#endif /* COMPILE_PCRE32 */
+
 int
 PRIV(strcmp_uc_c8)(const pcre_uchar *str1, const char *str2)
 {
@@ -99,6 +120,28 @@ while (*str1 != '\0' || *ustr2 != '\0')
 return 0;
 }
 
+#ifdef COMPILE_PCRE32
+
+int
+PRIV(strcmp_uc_c8_utf)(const pcre_uchar *str1, const char *str2)
+{
+const pcre_uint8 *ustr2 = (pcre_uint8 *)str2;
+pcre_uchar c1;
+pcre_uchar c2;
+
+while (*str1 != '\0' || *ustr2 != '\0')
+  {
+  c1 = RAWUCHARINC(str1);
+  c2 = (pcre_uchar)*ustr2++;
+  if (c1 != c2)
+    return ((c1 > c2) << 1) - 1;
+  }
+/* Both length and characters must be equal. */
+return 0;
+}
+
+#endif /* COMPILE_PCRE32 */
+
 /* The following two functions compares two, fixed length
 strings. Basically an strncmp for non 8 bit characters.
 
@@ -163,6 +206,6 @@ while (*str++ != 0)
 return len;
 }
 
-#endif /* COMPILE_PCRE8 */
+#endif /* !COMPILE_PCRE8 */
 
 /* End of pcre_string_utils.c */