+ return SUCCESS;
+}
+#endif
+
+#ifndef MAXHOSTNAMELEN
+# define MAXHOSTNAMELEN 256
+#endif
+
+#if PHP_HTTP_HAVE_IDN2
+static ZEND_RESULT_CODE parse_idn2(struct parse_state *state, size_t prev_len)
+{
+ char *idn = NULL;
+ int rv = -1;
+
+ if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
+ rv = idn2_lookup_u8((const unsigned char *) state->url.host, (unsigned char **) &idn, IDN2_NFC_INPUT);
+ }
+# ifdef PHP_HTTP_HAVE_WCHAR
+ else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
+ rv = idn2_lookup_ul(state->url.host, &idn, 0);
+ }
+# endif
+ if (rv != IDN2_OK) {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN; %s", idn2_strerror(rv));
+ return FAILURE;
+ } else {
+ size_t idnlen = strlen(idn);
+ memcpy(state->url.host, idn, idnlen + 1);
+ free(idn);
+ state->offset += idnlen - prev_len;
+ return SUCCESS;
+ }
+}
+#elif PHP_HTTP_HAVE_IDN
+static ZEND_RESULT_CODE parse_idn(struct parse_state *state, size_t prev_len)
+{
+ char *idn = NULL;
+ int rv = -1;
+
+ if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
+ rv = idna_to_ascii_8z(state->url.host, &idn, IDNA_ALLOW_UNASSIGNED|IDNA_USE_STD3_ASCII_RULES);
+ }
+# ifdef PHP_HTTP_HAVE_WCHAR
+ else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
+ rv = idna_to_ascii_lz(state->url.host, &idn, IDNA_ALLOW_UNASSIGNED|IDNA_USE_STD3_ASCII_RULES);
+ }
+# endif
+ if (rv != IDNA_SUCCESS) {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN; %s", idna_strerror(rv));
+ return FAILURE;
+ } else {
+ size_t idnlen = strlen(idn);
+ memcpy(state->url.host, idn, idnlen + 1);
+ free(idn);
+ state->offset += idnlen - prev_len;
+ return SUCCESS;
+ }
+}
+#endif
+
+#ifdef HAVE_UIDNA_IDNTOASCII
+# if HAVE_UNICODE_UIDNA_H
+# include <unicode/uidna.h>
+# else
+typedef uint16_t UChar;
+typedef enum { U_ZERO_ERROR = 0 } UErrorCode;
+int32_t uidna_IDNToASCII(const UChar *src, int32_t srcLength, UChar *dest, int32_t destCapacity, int32_t options, void *parseError, UErrorCode *status);
+# endif
+static ZEND_RESULT_CODE parse_uidn(struct parse_state *state)
+{
+ char *host_ptr;
+ uint16_t *uhost_str, ahost_str[MAXHOSTNAMELEN], *ahost_ptr;
+ size_t uhost_len, ahost_len;
+ UErrorCode error = U_ZERO_ERROR;
+
+ if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
+ if (SUCCESS != to_utf16(parse_mb_utf8, state->url.host, &uhost_str, &uhost_len)) {
+ return FAILURE;
+ }
+#ifdef PHP_HTTP_HAVE_WCHAR
+ } else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
+ if (SUCCESS != to_utf16(parse_mb_loc, state->url.host, &uhost_str, &uhost_len)) {
+ return FAILURE;
+ }
+#endif
+ } else {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN; codepage not specified");
+ return FAILURE;
+ }
+
+ ahost_len = uidna_IDNToASCII(uhost_str, uhost_len, ahost_str, MAXHOSTNAMELEN, 3, NULL, &error);
+ efree(uhost_str);
+
+ if (error != U_ZERO_ERROR) {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN; ICU error %d", error);
+ return FAILURE;
+ }
+
+ host_ptr = state->url.host;
+ ahost_ptr = ahost_str;
+ PHP_HTTP_DUFF(ahost_len, *host_ptr++ = *ahost_ptr++);
+
+ *host_ptr = '\0';
+ state->offset += host_ptr - state->url.host;
+
+ return SUCCESS;
+}
+#endif
+
+#if 0 && defined(PHP_WIN32)
+static ZEND_RESULT_CODE parse_widn(struct parse_state *state)
+{
+ char *host_ptr;
+ uint16_t *uhost_str, ahost_str[MAXHOSTNAMELEN], *ahost_ptr;
+ size_t uhost_len;
+
+ if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
+ if (SUCCESS != to_utf16(parse_mb_utf8, state->url.host, &uhost_str, &uhost_len)) {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN");
+ return FAILURE;
+ }
+#ifdef PHP_HTTP_HAVE_WCHAR
+ } else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
+ if (SUCCESS != to_utf16(parse_mb_loc, state->url.host, &uhost_str, &uhost_len)) {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN");
+ return FAILURE;
+ }
+#endif
+ } else {
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN");
+ return FAILURE;
+ }
+
+ if (!IdnToAscii(IDN_ALLOW_UNASSIGNED|IDN_USE_STD3_ASCII_RULES, uhost_str, uhost_len, ahost_str, MAXHOSTNAMELEN)) {
+ efree(uhost_str);
+ php_error_docref(NULL, E_WARNING, "Failed to parse IDN");
+ return FAILURE;
+ }
+
+ efree(uhost_str);
+ host_ptr = state->url.host;
+ ahost_ptr = ahost_str;
+ PHP_HTTP_DUFF(wcslen(ahost_str), *host_ptr++ = *ahost_ptr++);
+ efree(ahost_str);
+
+ *host_ptr = '\0';
+ state->offset += host_ptr - state->url.host;
+
+ return SUCCESS;
+}
+#endif
+
+static ZEND_RESULT_CODE parse_hostinfo(struct parse_state *state, const char *ptr)
+{
+ size_t mb, len;
+ const char *end = state->ptr, *tmp = ptr, *port = NULL;
+
+#ifdef HAVE_INET_PTON
+ if (*ptr == '[') {
+ char *error = NULL, *tmp = memchr(ptr, ']', end - ptr);
+
+ if (tmp) {
+ size_t addrlen = tmp - ptr + 1;
+ char buf[16], *addr = estrndup(ptr + 1, addrlen - 2);
+ int rv = inet_pton(AF_INET6, addr, buf);
+
+ efree(addr);
+ if (rv == 1) {
+ state->buffer[state->offset] = '[';
+ state->url.host = &state->buffer[state->offset];
+ inet_ntop(AF_INET6, buf, state->url.host + 1, state->maxlen - state->offset);
+ state->offset += strlen(state->url.host);
+ state->buffer[state->offset++] = ']';
+ state->buffer[state->offset++] = 0;
+ ptr = tmp + 1;
+ } else if (rv == -1) {
+ error = strerror(errno);
+ } else {
+ error = "unexpected '['";
+ }
+ } else {
+ error = "expected ']'";
+ }
+
+ if (error) {
+ php_error_docref(NULL, E_WARNING, "Failed to parse hostinfo; %s", error);
+ return FAILURE;
+ }
+ }
+#endif
+ if (ptr != end) do {
+ switch (*ptr) {
+ case ':':
+ if (port) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse port; unexpected ':' at pos %u in '%s'",
+ (unsigned) (ptr - tmp), tmp);
+ return FAILURE;
+ }
+ port = ptr + 1;
+ break;
+
+ case '%':
+ if (ptr[1] != '%' && (end - ptr <= 2 || !isxdigit(*(ptr+1)) || !isxdigit(*(ptr+2)))) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse hostinfo; invalid percent encoding at pos %u in '%s'",
+ (unsigned) (ptr - tmp), tmp);
+ return FAILURE;
+ }
+ state->buffer[state->offset++] = *ptr++;
+ state->buffer[state->offset++] = *ptr++;
+ state->buffer[state->offset++] = *ptr;
+ break;
+
+ case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
+ case '+': case ',': case ';': case '=': /* sub-delims */
+ case '-': case '.': case '_': case '~': /* unreserved */
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
+ case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
+ case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
+ case 'V': case 'W': case 'X': case 'Y': case 'Z':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
+ case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
+ case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
+ case 'v': case 'w': case 'x': case 'y': case 'z':
+ if (port) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse port; unexpected char '%c' at pos %u in '%s'",
+ (unsigned char) *ptr, (unsigned) (ptr - tmp), tmp);
+ return FAILURE;
+ }
+ /* no break */
+ case '0': case '1': case '2': case '3': case '4': case '5': case '6':
+ case '7': case '8': case '9':
+ /* allowed */
+ if (port) {
+ state->url.port *= 10;
+ state->url.port += *ptr - '0';
+ } else {
+ state->buffer[state->offset++] = *ptr;
+ }
+ break;
+
+ default:
+ if (ptr == end) {
+ break;
+ } else if (port) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse port; unexpected byte 0x%02x at pos %u in '%s'",
+ (unsigned char) *ptr, (unsigned) (ptr - tmp), tmp);
+ return FAILURE;
+ } else if (!(mb = parse_mb(state, PARSE_HOSTINFO, ptr, end, tmp, 0))) {
+ return FAILURE;
+ }
+ ptr += mb - 1;
+ }
+ } while (++ptr != end);
+
+ if (!state->url.host) {
+ len = (port ? port - tmp - 1 : end - tmp);
+ state->url.host = &state->buffer[state->offset - len];
+ state->buffer[state->offset++] = 0;
+ }
+
+ if (state->flags & PHP_HTTP_URL_PARSE_TOIDN) {
+#if PHP_HTTP_HAVE_IDN2
+ return parse_idn2(state, len);
+#elif PHP_HTTP_HAVE_IDN
+ return parse_idn(state, len);
+#endif
+#ifdef HAVE_UIDNA_IDNTOASCII
+ return parse_uidn(state);
+#endif
+#if 0 && defined(PHP_WIN32)
+ return parse_widn(state);
+#endif
+ }
+
+ return SUCCESS;
+}
+
+static const char *parse_authority(struct parse_state *state)
+{
+ const char *tmp = state->ptr, *host = NULL;
+
+ do {
+ switch (*state->ptr) {
+ case '@':
+ /* userinfo delimiter */
+ if (host) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse userinfo; unexpected '@'");
+ return NULL;
+ }
+ host = state->ptr + 1;
+ if (tmp != state->ptr && SUCCESS != parse_userinfo(state, tmp)) {
+ return NULL;
+ }
+ tmp = state->ptr + 1;
+ break;
+
+ case '/':
+ case '?':
+ case '#':
+ case '\0':
+ EOD:
+ /* host delimiter */
+ if (tmp != state->ptr && SUCCESS != parse_hostinfo(state, tmp)) {
+ return NULL;
+ }
+ return state->ptr;
+ }
+ } while (++state->ptr <= state->end);
+
+ --state->ptr;
+ goto EOD;
+}
+
+static const char *parse_path(struct parse_state *state)
+{
+ size_t mb;
+ const char *tmp;
+
+ /* is there actually a path to parse? */
+ if (!*state->ptr) {
+ return state->ptr;
+ }
+ tmp = state->ptr;
+ state->url.path = &state->buffer[state->offset];
+
+ do {
+ switch (*state->ptr) {
+ case '#':
+ case '?':
+ goto done;
+
+ case '%':
+ if (state->ptr[1] != '%' && (state->end - state->ptr <= 2 || !isxdigit(*(state->ptr+1)) || !isxdigit(*(state->ptr+2)))) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse path; invalid percent encoding at pos %u in '%s'",
+ (unsigned) (state->ptr - tmp), tmp);
+ return NULL;
+ }
+ state->buffer[state->offset++] = *state->ptr++;
+ state->buffer[state->offset++] = *state->ptr++;
+ state->buffer[state->offset++] = *state->ptr;
+ break;
+
+ case '/': /* yeah, well */
+ case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
+ case '+': case ',': case ';': case '=': /* sub-delims */
+ case '-': case '.': case '_': case '~': /* unreserved */
+ case ':': case '@': /* pchar */
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
+ case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
+ case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
+ case 'V': case 'W': case 'X': case 'Y': case 'Z':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
+ case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
+ case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
+ case 'v': case 'w': case 'x': case 'y': case 'z':
+ case '0': case '1': case '2': case '3': case '4': case '5': case '6':
+ case '7': case '8': case '9':
+ /* allowed */
+ state->buffer[state->offset++] = *state->ptr;
+ break;
+
+ default:
+ if (!(mb = parse_mb(state, PARSE_PATH, state->ptr, state->end, tmp, 0))) {
+ return NULL;
+ }
+ state->ptr += mb - 1;
+ }
+ } while (++state->ptr < state->end);
+
+ done:
+ /* did we have any path component ? */
+ if (tmp != state->ptr) {
+ state->buffer[state->offset++] = 0;
+ } else {
+ state->url.path = NULL;
+ }
+ return state->ptr;
+}
+
+static const char *parse_query(struct parse_state *state)
+{
+ size_t mb;
+ const char *tmp = state->ptr + !!*state->ptr;
+
+ /* is there actually a query to parse? */
+ if (*state->ptr != '?') {
+ return state->ptr;
+ }
+
+ /* skip initial '?' */
+ tmp = ++state->ptr;
+ state->url.query = &state->buffer[state->offset];
+
+ while (state->ptr < state->end) {
+ switch (*state->ptr) {
+ case '#':
+ goto done;
+
+ case '%':
+ if (state->ptr[1] != '%' && (state->end - state->ptr <= 2 || !isxdigit(*(state->ptr+1)) || !isxdigit(*(state->ptr+2)))) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse query; invalid percent encoding at pos %u in '%s'",
+ (unsigned) (state->ptr - tmp), tmp);
+ return NULL;
+ }
+ state->buffer[state->offset++] = *state->ptr++;
+ state->buffer[state->offset++] = *state->ptr++;
+ state->buffer[state->offset++] = *state->ptr;
+ break;
+
+ /* RFC1738 unsafe */
+ case '{': case '}':
+ case '<': case '>':
+ case '[': case ']':
+ case '|': case '\\': case '^': case '`': case '"': case ' ':
+ if (state->flags & PHP_HTTP_URL_PARSE_TOPCT) {
+ state->buffer[state->offset++] = '%';
+ state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) >> 4];
+ state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) & 0xf];
+ break;
+ }
+ /* no break */
+
+ case '?': case '/': /* yeah, well */
+ case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
+ case '+': case ',': case ';': case '=': /* sub-delims */
+ case '-': case '.': case '_': case '~': /* unreserved */
+ case ':': case '@': /* pchar */
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
+ case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
+ case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
+ case 'V': case 'W': case 'X': case 'Y': case 'Z':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
+ case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
+ case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
+ case 'v': case 'w': case 'x': case 'y': case 'z':
+ case '0': case '1': case '2': case '3': case '4': case '5': case '6':
+ case '7': case '8': case '9':
+ /* allowed */
+ state->buffer[state->offset++] = *state->ptr;
+ break;
+
+ default:
+ if (!(mb = parse_mb(state, PARSE_QUERY, state->ptr, state->end, tmp, 0))) {
+ return NULL;
+ }
+ state->ptr += mb - 1;
+ }
+
+ ++state->ptr;
+ }
+
+ done:
+ state->buffer[state->offset++] = 0;
+ return state->ptr;
+}
+
+static const char *parse_fragment(struct parse_state *state)
+{
+ size_t mb;
+ const char *tmp;
+
+ /* is there actually a fragment to parse? */
+ if (*state->ptr != '#') {
+ return state->ptr;
+ }
+
+ /* skip initial '#' */
+ tmp = ++state->ptr;
+ state->url.fragment = &state->buffer[state->offset];
+
+ do {
+ switch (*state->ptr) {
+ case '%':
+ if (state->ptr[1] != '%' && (state->end - state->ptr <= 2 || !isxdigit(*(state->ptr+1)) || !isxdigit(*(state->ptr+2)))) {
+ php_error_docref(NULL, E_WARNING,
+ "Failed to parse fragment; invalid percent encoding at pos %u in '%s'",
+ (unsigned) (state->ptr - tmp), tmp);
+ return NULL;
+ }
+ state->buffer[state->offset++] = *state->ptr++;
+ state->buffer[state->offset++] = *state->ptr++;
+ state->buffer[state->offset++] = *state->ptr;
+ break;
+
+ /* RFC1738 unsafe */
+ case '{': case '}':
+ case '<': case '>':
+ case '[': case ']':
+ case '|': case '\\': case '^': case '`': case '"': case ' ':
+ if (state->flags & PHP_HTTP_URL_PARSE_TOPCT) {
+ state->buffer[state->offset++] = '%';
+ state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) >> 4];
+ state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) & 0xf];
+ break;
+ }
+ /* no break */
+
+ case '?': case '/':
+ case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
+ case '+': case ',': case ';': case '=': /* sub-delims */
+ case '-': case '.': case '_': case '~': /* unreserved */
+ case ':': case '@': /* pchar */
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
+ case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
+ case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
+ case 'V': case 'W': case 'X': case 'Y': case 'Z':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
+ case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
+ case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
+ case 'v': case 'w': case 'x': case 'y': case 'z':
+ case '0': case '1': case '2': case '3': case '4': case '5': case '6':
+ case '7': case '8': case '9':
+ /* allowed */
+ state->buffer[state->offset++] = *state->ptr;
+ break;
+
+ default:
+ if (!(mb = parse_mb(state, PARSE_FRAGMENT, state->ptr, state->end, tmp, 0))) {
+ return NULL;
+ }
+ state->ptr += mb - 1;
+ }
+ } while (++state->ptr < state->end);
+
+ state->buffer[state->offset++] = 0;
+ return state->ptr;
+}
+
+static const char *parse_hier(struct parse_state *state)
+{
+ if (*state->ptr == '/') {
+ if (state->end - state->ptr > 1) {
+ if (*(state->ptr + 1) == '/') {
+ state->ptr += 2;
+ if (!(state->ptr = parse_authority(state))) {
+ return NULL;