}
/* }}} */
+/* copied from trunk's implementation of get_next_char in ext/standard/html.c */
+#define MB_FAILURE(pos, advance) do { \
+ *cursor = pos + (advance); \
+ *status = FAILURE; \
+ return 0; \
+} while (0)
+
+#define CHECK_LEN(pos, chars_need) ((str_len - (pos)) >= (chars_need))
+#define utf8_lead(c) ((c) < 0x80 || ((c) >= 0xC2 && (c) <= 0xF4))
+#define utf8_trail(c) ((c) >= 0x80 && (c) <= 0xBF)
+
+/* {{{ php_next_utf8_char
+ */
+static inline unsigned int php_next_utf8_char(
+ const unsigned char *str,
+ size_t str_len,
+ size_t *cursor,
+ int *status)
+{
+ size_t pos = *cursor;
+ unsigned int this_char = 0;
+ unsigned char c;
+
+ *status = SUCCESS;
+
+ if (!CHECK_LEN(pos, 1))
+ MB_FAILURE(pos, 1);
+
+ /* We'll follow strategy 2. from section 3.6.1 of UTR #36:
+ * "In a reported illegal byte sequence, do not include any
+ * non-initial byte that encodes a valid character or is a leading
+ * byte for a valid sequence.» */
+ c = str[pos];
+ if (c < 0x80) {
+ this_char = c;
+ pos++;
+ } else if (c < 0xc2) {
+ MB_FAILURE(pos, 1);
+ } else if (c < 0xe0) {
+ if (!CHECK_LEN(pos, 2))
+ MB_FAILURE(pos, 1);
+
+ if (!utf8_trail(str[pos + 1])) {
+ MB_FAILURE(pos, utf8_lead(str[pos + 1]) ? 1 : 2);
+ }
+ this_char = ((c & 0x1f) << 6) | (str[pos + 1] & 0x3f);
+ if (this_char < 0x80) { /* non-shortest form */
+ MB_FAILURE(pos, 2);
+ }
+ pos += 2;
+ } else if (c < 0xf0) {
+ size_t avail = str_len - pos;
+
+ if (avail < 3 ||
+ !utf8_trail(str[pos + 1]) || !utf8_trail(str[pos + 2])) {
+ if (avail < 2 || utf8_lead(str[pos + 1]))
+ MB_FAILURE(pos, 1);
+ else if (avail < 3 || utf8_lead(str[pos + 2]))
+ MB_FAILURE(pos, 2);
+ else
+ MB_FAILURE(pos, 3);
+ }
+
+ this_char = ((c & 0x0f) << 12) | ((str[pos + 1] & 0x3f) << 6) | (str[pos + 2] & 0x3f);
+ if (this_char < 0x800) { /* non-shortest form */
+ MB_FAILURE(pos, 3);
+ } else if (this_char >= 0xd800 && this_char <= 0xdfff) { /* surrogate */
+ MB_FAILURE(pos, 3);
+ }
+ pos += 3;
+ } else if (c < 0xf5) {
+ size_t avail = str_len - pos;
+
+ if (avail < 4 ||
+ !utf8_trail(str[pos + 1]) || !utf8_trail(str[pos + 2]) ||
+ !utf8_trail(str[pos + 3])) {
+ if (avail < 2 || utf8_lead(str[pos + 1]))
+ MB_FAILURE(pos, 1);
+ else if (avail < 3 || utf8_lead(str[pos + 2]))
+ MB_FAILURE(pos, 2);
+ else if (avail < 4 || utf8_lead(str[pos + 3]))
+ MB_FAILURE(pos, 3);
+ else
+ MB_FAILURE(pos, 4);
+ }
+
+ this_char = ((c & 0x07) << 18) | ((str[pos + 1] & 0x3f) << 12) | ((str[pos + 2] & 0x3f) << 6) | (str[pos + 3] & 0x3f);
+ if (this_char < 0x10000 || this_char > 0x10FFFF) { /* non-shortest form or outside range */
+ MB_FAILURE(pos, 4);
+ }
+ pos += 4;
+ } else {
+ MB_FAILURE(pos, 1);
+ }
+
+ *cursor = pos;
+ return this_char;
+}
+/* }}} */
+
+
/* {{{ xml_utf8_decode */
PHPAPI char *xml_utf8_decode(const XML_Char *s, int len, int *newlen, const XML_Char *encoding)
{
- int pos = len;
+ size_t pos = 0;
char *newbuf = emalloc(len + 1);
unsigned int c;
char (*decoder)(unsigned short) = NULL;
newbuf[*newlen] = '\0';
return newbuf;
}
- while (pos > 0) {
- c = (unsigned char)(*s);
- if (c >= 0xf0) { /* four bytes encoded, 21 bits */
- if(pos-4 >= 0) {
- c = ((s[0]&7)<<18) | ((s[1]&63)<<12) | ((s[2]&63)<<6) | (s[3]&63);
- } else {
- c = '?';
- }
- s += 4;
- pos -= 4;
- } else if (c >= 0xe0) { /* three bytes encoded, 16 bits */
- if(pos-3 >= 0) {
- c = ((s[0]&63)<<12) | ((s[1]&63)<<6) | (s[2]&63);
- } else {
- c = '?';
- }
- s += 3;
- pos -= 3;
- } else if (c >= 0xc0) { /* two bytes encoded, 11 bits */
- if(pos-2 >= 0) {
- c = ((s[0]&63)<<6) | (s[1]&63);
- } else {
- c = '?';
- }
- s += 2;
- pos -= 2;
- } else {
- s++;
- pos--;
+
+ while (pos < (size_t)len) {
+ int status = FAILURE;
+ c = php_next_utf8_char((const unsigned char*)s, (size_t) len, &pos, &status);
+
+ if (status == FAILURE || c > 0xFFU) {
+ c = '?';
}
+
newbuf[*newlen] = decoder ? decoder(c) : c;
++*newlen;
}