Fixed: Start bytes 0xF5 and above are invalid for UTF-8

Signed-off-by: Zafer Balkan <[email protected]>
wazuh · Dec 5, 2024 · 1d50e3e · 1d50e3e
1 parent c3fbb30
commit 1d50e3e
Showing 1 changed file with 16 additions and 4 deletions.
diff --git a/src/common/utf8_op/src/utf8_op.c b/src/common/utf8_op/src/utf8_op.c
@@ -19,17 +19,29 @@
 
 /* Two bytes: 110xxxxx 10xxxxxx */
 /* Starting bytes 0xC0 and 0xC1 are forbidden (overlong) */
-#define valid_2(x) (((x)[0] & 0xE0) == 0xC0 && (x)[0] >= (char)0xC2 && ((x)[1] & 0xC0) == 0x80)
+#define valid_2(x) (((x)[0] & 0xE0) == 0xC0 && \
+                    (x)[0] >= (char)0xC2 && ((x)[1] & 0xC0) == 0x80)
 
 /* Three bytes: 1110xxxx 10xxxxxx 10xxxxxx */
 /* 0xE0 could start overlong encodings */
 /* 0xED (range U+D800–U+DFFF) is reserved for UTF-16 surrogate halves */
-#define valid_3(x) (((x)[0] & 0xF0) == 0xE0 && ((x)[1] & 0xC0) == 0x80 && ((x)[2] & 0xC0) == 0x80 && ((x)[0] != (char)0xE0 || (unsigned char)(x)[1] >= 0xA0) && ((x)[0] != (char)0xED || (unsigned char)(x)[1] < 0xA0) && ((x)[0] != (char)0xEF || (unsigned char)(x)[1] <= 0xBF))
+#define valid_3(x) (((x)[0] & 0xF0) == 0xE0 && \
+                    ((x)[1] & 0xC0) == 0x80 && \
+                    ((x)[2] & 0xC0) == 0x80 && \
+                    ((x)[0] != (char)0xE0 || (unsigned char)(x)[1] >= 0xA0) && \
+                    ((x)[0] != (char)0xED || (unsigned char)(x)[1] < 0xA0) && \
+                    ((x)[0] != (char)0xEF || (unsigned char)(x)[1] <= 0xBF))
 
 /* Four bytes: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
 /* 0xF0 could start overlong encodings */
-/* Starting bytes 111101xx are forbidden (Unicode limit) */
-#define valid_4(x) (((x)[0] & 0xF8) == 0xF0 && ((x)[1] & 0xC0) == 0x80 && ((x)[2] & 0xC0) == 0x80 && ((x)[3] & 0xC0) == 0x80 && ((x)[0] != (char)0xF0 || (unsigned char)(x)[1] >= 0x90) && ((x)[0] != (char)0xF4 || (unsigned char)(x)[1] <= 0x8F))
+/* Start bytes 0xF5 and above are invalid for UTF-8 */
+#define valid_4(x) (((x)[0] & 0xF8) == 0xF0 && \
+                    (unsigned char)(x)[0] <= 0xF4 && \
+                    ((x)[1] & 0xC0) == 0x80 && \
+                    ((x)[2] & 0xC0) == 0x80 && \
+                    ((x)[3] & 0xC0) == 0x80 && \
+                    ((x)[0] != (char)0xF0 || (unsigned char)(x)[1] >= 0x90) && \
+                    ((x)[0] != (char)0xF4 || (unsigned char)(x)[1] <= 0x8F))
 
 /* Return whether a string is UTF-8 */
 bool w_utf8_valid(const char * string) {