aboutsummaryrefslogtreecommitdiffstats
path: root/src/util.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/util.c')
-rw-r--r--src/util.c75
1 files changed, 75 insertions, 0 deletions
diff --git a/src/util.c b/src/util.c
index 81926c6..f73634a 100644
--- a/src/util.c
+++ b/src/util.c
@@ -152,3 +152,78 @@ void msleep(int ms)
nanosleep(&ts, NULL);
}
+/**
+ * Check whether a NUL-terminated string is encoded in UTF-8
+ *
+ * @param string The string
+ * @param allow_modified_nul Whether Modified UTF-8 is allowed, which allows a two-byte encoding for NUL
+ * @return Zero if good, -1 on encoding error
+ */
+int verify_utf8(const char* string, int allow_modified_nul)
+{
+ static long BYTES_TO_MIN_BITS[] = {0, 0, 8, 12, 17, 22, 37};
+ static long BYTES_TO_MAX_BITS[] = {0, 7, 11, 16, 21, 26, 31};
+ long bytes = 0, read_bytes = 0, bits = 0, c, character;
+
+ /* min bits max bits
+ 0....... 0 7
+ 110..... 10...... 8 11
+ 1110.... 10...... 10...... 12 16
+ 11110... 10...... 10...... 10...... 17 21
+ 111110.. 10...... 10...... 10...... 10...... 22 26
+ 1111110. 10...... 10...... 10...... 10...... 10...... 27 31
+ */
+
+ while ((c = (long)(*string++)))
+ if (read_bytes == 0)
+ {
+ /* First byte of the character. */
+
+ if ((c & 0x80) == 0x00)
+ /* Single-byte character. */
+ continue;
+
+ if ((c & 0xC0) == 0x80)
+ /* Single-byte character marked as multibyte, or
+ a non-first byte in a multibyte character. */
+ return -1;
+
+ /* Multibyte character. */
+ while ((c & 0x80))
+ bytes++, c <<= 1;
+ read_bytes = 1;
+ character = c & 0x7F;
+ if (bytes > 6)
+ /* 31-bit characters can be encoded with 6-bytes,
+ and UTF-8 does not cover higher code points. */
+ return -1;
+ }
+ else
+ {
+ /* Not first byte of the character. */
+
+ if ((c & 0xC0) != 0x80)
+ /* Beginning of new character before a
+ multibyte character has ended. */
+ return -1;
+
+ character = (character << 6) | (c & 0x7F);
+
+ if (++read_bytes < bytes)
+ /* Not at last byte yet. */
+ continue;
+
+ /* Check that the character is not unnecessarily long. */
+ while (character)
+ character >>= 1, bits++;
+ bits = ((bits == 0) && (bytes == 2) && allow_modified_nul) ? 8 : bits;
+ if ((bits < BYTES_TO_MIN_BITS[bytes]) || (BYTES_TO_MAX_BITS[bytes] < bits))
+ return -1;
+
+ read_bytes = bytes = bits = 0;
+ }
+
+ /* Make sure we did not stop at the middle of a multibyte character. */
+ return read_bytes == 0 ? 0 : -1;
+}
+