projects / m6w6 / ext-http / blob
? search:
summary | shortlog | log | commit | commitdiff | tree | github
history | raw | HEAD
fix PHP-73185
[m6w6/ext-http] / src / php_http_url.c
1 /*
2 +--------------------------------------------------------------------+
3 | PECL :: http |
4 +--------------------------------------------------------------------+
5 | Redistribution and use in source and binary forms, with or without |
6 | modification, are permitted provided that the conditions mentioned |
7 | in the accompanying LICENSE file are met. |
8 +--------------------------------------------------------------------+
9 | Copyright (c) 2004-2014, Michael Wallner <mike@php.net> |
10 +--------------------------------------------------------------------+
11 */
12
13 #include "php_http_api.h"
14
15 #if PHP_HTTP_HAVE_IDN2
16 # include <idn2.h>
17 #elif PHP_HTTP_HAVE_IDN
18 # include <idna.h>
19 #endif
20
21 #ifdef PHP_HTTP_HAVE_WCHAR
22 # include <wchar.h>
23 # include <wctype.h>
24 #endif
25
26 #ifdef HAVE_ARPA_INET_H
27 # include <arpa/inet.h>
28 #endif
29
30 #include "php_http_utf8.h"
31
32 static inline char *localhostname(void)
33 {
34 char hostname[1024] = {0};
35
36 #ifdef PHP_WIN32
37 if (SUCCESS == gethostname(hostname, lenof(hostname))) {
38 return estrdup(hostname);
39 }
40 #elif defined(HAVE_GETHOSTNAME)
41 if (SUCCESS == gethostname(hostname, lenof(hostname))) {
42 # if defined(HAVE_GETDOMAINNAME)
43 size_t hlen = strlen(hostname);
44 if (hlen <= lenof(hostname) - lenof("(none)")) {
45 hostname[hlen++] = '.';
46 if (SUCCESS == getdomainname(&hostname[hlen], lenof(hostname) - hlen)) {
47 if (!strcmp(&hostname[hlen], "(none)")) {
48 hostname[hlen - 1] = '\0';
49 }
50 return estrdup(hostname);
51 }
52 }
53 # endif
54 if (strcmp(hostname, "(none)")) {
55 return estrdup(hostname);
56 }
57 }
58 #endif
59 return estrndup("localhost", lenof("localhost"));
60 }
61
62 #define url(buf) ((php_http_url_t *) (buf).data)
63
64 static php_http_url_t *php_http_url_from_env(TSRMLS_D)
65 {
66 zval *https, *zhost, *zport;
67 long port;
68 php_http_buffer_t buf;
69
70 php_http_buffer_init_ex(&buf, MAX(PHP_HTTP_BUFFER_DEFAULT_SIZE, sizeof(php_http_url_t)<<2), PHP_HTTP_BUFFER_INIT_PREALLOC);
71 php_http_buffer_account(&buf, sizeof(php_http_url_t));
72 memset(buf.data, 0, buf.used);
73
74 /* scheme */
75 url(buf)->scheme = &buf.data[buf.used];
76 https = php_http_env_get_server_var(ZEND_STRL("HTTPS"), 1 TSRMLS_CC);
77 if (https && !strcasecmp(Z_STRVAL_P(https), "ON")) {
78 php_http_buffer_append(&buf, "https", sizeof("https"));
79 } else {
80 php_http_buffer_append(&buf, "http", sizeof("http"));
81 }
82
83 /* host */
84 url(buf)->host = &buf.data[buf.used];
85 if ((((zhost = php_http_env_get_server_var(ZEND_STRL("HTTP_HOST"), 1 TSRMLS_CC)) ||
86 (zhost = php_http_env_get_server_var(ZEND_STRL("SERVER_NAME"), 1 TSRMLS_CC)) ||
87 (zhost = php_http_env_get_server_var(ZEND_STRL("SERVER_ADDR"), 1 TSRMLS_CC)))) && Z_STRLEN_P(zhost)) {
88 size_t stop_at = strspn(Z_STRVAL_P(zhost), "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-.");
89
90 php_http_buffer_append(&buf, Z_STRVAL_P(zhost), stop_at);
91 php_http_buffer_append(&buf, "", 1);
92 } else {
93 char *host_str = localhostname();
94
95 php_http_buffer_append(&buf, host_str, strlen(host_str) + 1);
96 efree(host_str);
97 }
98
99 /* port */
100 zport = php_http_env_get_server_var(ZEND_STRL("SERVER_PORT"), 1 TSRMLS_CC);
101 if (zport && IS_LONG == is_numeric_string(Z_STRVAL_P(zport), Z_STRLEN_P(zport), &port, NULL, 0)) {
102 url(buf)->port = port;
103 }
104
105 /* path */
106 if (SG(request_info).request_uri && SG(request_info).request_uri[0]) {
107 const char *q = strchr(SG(request_info).request_uri, '?');
108
109 url(buf)->path = &buf.data[buf.used];
110
111 if (q) {
112 php_http_buffer_append(&buf, SG(request_info).request_uri, q - SG(request_info).request_uri);
113 php_http_buffer_append(&buf, "", 1);
114 } else {
115 php_http_buffer_append(&buf, SG(request_info).request_uri, strlen(SG(request_info).request_uri) + 1);
116 }
117 }
118
119 /* query */
120 if (SG(request_info).query_string && SG(request_info).query_string[0]) {
121 url(buf)->query = &buf.data[buf.used];
122 php_http_buffer_append(&buf, SG(request_info).query_string, strlen(SG(request_info).query_string) + 1);
123 }
124
125 return url(buf);
126 }
127
128 #define url_isset(u,n) \
129 ((u)&&(u)->n)
130 #define url_append(buf, append) do { \
131 char *_ptr = (buf)->data; \
132 php_http_url_t *_url = (php_http_url_t *) _ptr, _mem = *_url; \
133 append; \
134 /* relocate */ \
135 if (_ptr != (buf)->data) { \
136 ptrdiff_t diff = (buf)->data - _ptr; \
137 _url = (php_http_url_t *) (buf)->data; \
138 if (_mem.scheme) _url->scheme += diff; \
139 if (_mem.user) _url->user += diff; \
140 if (_mem.pass) _url->pass += diff; \
141 if (_mem.host) _url->host += diff; \
142 if (_mem.path) _url->path += diff; \
143 if (_mem.query) _url->query += diff; \
144 if (_mem.fragment) _url->fragment += diff; \
145 } \
146 } while (0)
147 #define url_copy(n) do { \
148 if (url_isset(new_url, n)) { \
149 url(buf)->n = &buf.data[buf.used]; \
150 url_append(&buf, php_http_buffer_append(&buf, new_url->n, strlen(new_url->n) + 1)); \
151 } else if (url_isset(old_url, n)) { \
152 url(buf)->n = &buf.data[buf.used]; \
153 url_append(&buf, php_http_buffer_append(&buf, old_url->n, strlen(old_url->n) + 1)); \
154 } \
155 } while (0)
156
157 php_http_url_t *php_http_url_mod(const php_http_url_t *old_url, const php_http_url_t *new_url, unsigned flags TSRMLS_DC)
158 {
159 php_http_url_t *tmp_url = NULL;
160 php_http_buffer_t buf;
161
162 php_http_buffer_init_ex(&buf, MAX(PHP_HTTP_BUFFER_DEFAULT_SIZE, sizeof(php_http_url_t)<<2), PHP_HTTP_BUFFER_INIT_PREALLOC);
163 php_http_buffer_account(&buf, sizeof(php_http_url_t));
164 memset(buf.data, 0, buf.used);
165
166 /* set from env if requested */
167 if (flags & PHP_HTTP_URL_FROM_ENV) {
168 php_http_url_t *env_url = php_http_url_from_env(TSRMLS_C);
169
170 old_url = tmp_url = php_http_url_mod(env_url, old_url, flags ^ PHP_HTTP_URL_FROM_ENV TSRMLS_CC);
171 php_http_url_free(&env_url);
172 }
173
174 url_copy(scheme);
175
176 if (!(flags & PHP_HTTP_URL_STRIP_USER)) {
177 url_copy(user);
178 }
179
180 if (!(flags & PHP_HTTP_URL_STRIP_PASS)) {
181 url_copy(pass);
182 }
183
184 url_copy(host);
185
186 if (!(flags & PHP_HTTP_URL_STRIP_PORT)) {
187 url(buf)->port = url_isset(new_url, port) ? new_url->port : ((old_url) ? old_url->port : 0);
188 }
189
190 if (!(flags & PHP_HTTP_URL_STRIP_PATH)) {
191 if ((flags & PHP_HTTP_URL_JOIN_PATH) && url_isset(old_url, path) && url_isset(new_url, path) && *new_url->path != '/') {
192 size_t old_path_len = strlen(old_url->path), new_path_len = strlen(new_url->path);
193 char *path = ecalloc(1, old_path_len + new_path_len + 1 + 1);
194
195 strcat(path, old_url->path);
196 if (path[old_path_len - 1] != '/') {
197 php_dirname(path, old_path_len);
198 strcat(path, "/");
199 }
200 strcat(path, new_url->path);
201
202 url(buf)->path = &buf.data[buf.used];
203 if (path[0] != '/') {
204 url_append(&buf, php_http_buffer_append(&buf, "/", 1));
205 }
206 url_append(&buf, php_http_buffer_append(&buf, path, strlen(path) + 1));
207 efree(path);
208 } else {
209 const char *path = NULL;
210
211 if (url_isset(new_url, path)) {
212 path = new_url->path;
213 } else if (url_isset(old_url, path)) {
214 path = old_url->path;
215 }
216
217 if (path) {
218 url(buf)->path = &buf.data[buf.used];
219
220 url_append(&buf, php_http_buffer_append(&buf, path, strlen(path) + 1));
221 }
222
223
224 }
225 }
226
227 if (!(flags & PHP_HTTP_URL_STRIP_QUERY)) {
228 if ((flags & PHP_HTTP_URL_JOIN_QUERY) && url_isset(new_url, query) && url_isset(old_url, query)) {
229 zval qarr, qstr;
230
231 INIT_PZVAL(&qstr);
232 INIT_PZVAL(&qarr);
233 array_init(&qarr);
234
235 ZVAL_STRING(&qstr, old_url->query, 0);
236 php_http_querystring_update(&qarr, &qstr, NULL TSRMLS_CC);
237 ZVAL_STRING(&qstr, new_url->query, 0);
238 php_http_querystring_update(&qarr, &qstr, NULL TSRMLS_CC);
239
240 ZVAL_NULL(&qstr);
241 php_http_querystring_update(&qarr, NULL, &qstr TSRMLS_CC);
242
243 url(buf)->query = &buf.data[buf.used];
244 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL(qstr), Z_STRLEN(qstr) + 1));
245
246 zval_dtor(&qstr);
247 zval_dtor(&qarr);
248 } else {
249 url_copy(query);
250 }
251 }
252
253 if (!(flags & PHP_HTTP_URL_STRIP_FRAGMENT)) {
254 url_copy(fragment);
255 }
256
257 /* done with copy & combine & strip */
258
259 if (flags & PHP_HTTP_URL_FROM_ENV) {
260 /* free old_url we tainted above */
261 php_http_url_free(&tmp_url);
262 }
263
264 /* replace directory references if path is not a single slash */
265 if ((flags & PHP_HTTP_URL_SANITIZE_PATH)
266 && url(buf)->path
267 && url(buf)->path[0] && url(buf)->path[1]) {
268 char *ptr, *end = url(buf)->path + strlen(url(buf)->path) + 1;
269
270 for (ptr = strchr(url(buf)->path, '/'); ptr; ptr = strchr(ptr, '/')) {
271 switch (ptr[1]) {
272 case '/':
273 memmove(&ptr[1], &ptr[2], end - &ptr[2]);
274 break;
275
276 case '.':
277 switch (ptr[2]) {
278 case '\0':
279 ptr[1] = '\0';
280 break;
281
282 case '/':
283 memmove(&ptr[1], &ptr[3], end - &ptr[3]);
284 break;
285
286 case '.':
287 if (ptr[3] == '/') {
288 char *pos = &ptr[4];
289 while (ptr != url(buf)->path) {
290 if (*--ptr == '/') {
291 break;
292 }
293 }
294 memmove(&ptr[1], pos, end - pos);
295 break;
296 } else if (!ptr[3]) {
297 /* .. at the end */
298 ptr[1] = '\0';
299 }
300 /* no break */
301
302 default:
303 /* something else */
304 ++ptr;
305 break;
306 }
307 break;
308
309 default:
310 ++ptr;
311 break;
312 }
313 }
314 }
315 /* unset default ports */
316 if (url(buf)->port) {
317 if ( ((url(buf)->port == 80) && url(buf)->scheme && !strcmp(url(buf)->scheme, "http"))
318 || ((url(buf)->port ==443) && url(buf)->scheme && !strcmp(url(buf)->scheme, "https"))
319 ) {
320 url(buf)->port = 0;
321 }
322 }
323
324 return url(buf);
325 }
326
327 char *php_http_url_to_string(const php_http_url_t *url, char **url_str, size_t *url_len, zend_bool persistent)
328 {
329 php_http_buffer_t buf;
330
331 php_http_buffer_init_ex(&buf, PHP_HTTP_BUFFER_DEFAULT_SIZE, persistent ?
332 PHP_HTTP_BUFFER_INIT_PERSISTENT : 0);
333
334 if (url->scheme && *url->scheme) {
335 php_http_buffer_appendl(&buf, url->scheme);
336 php_http_buffer_appends(&buf, "://");
337 } else if ((url->user && *url->user) || (url->host && *url->host)) {
338 php_http_buffer_appends(&buf, "//");
339 }
340
341 if (url->user && *url->user) {
342 php_http_buffer_appendl(&buf, url->user);
343 if (url->pass && *url->pass) {
344 php_http_buffer_appends(&buf, ":");
345 php_http_buffer_appendl(&buf, url->pass);
346 }
347 php_http_buffer_appends(&buf, "@");
348 }
349
350 if (url->host && *url->host) {
351 php_http_buffer_appendl(&buf, url->host);
352 if (url->port) {
353 php_http_buffer_appendf(&buf, ":%hu", url->port);
354 }
355 }
356
357 if (url->path && *url->path) {
358 if (*url->path != '/') {
359 php_http_buffer_appends(&buf, "/");
360 }
361 php_http_buffer_appendl(&buf, url->path);
362 } else if (buf.used) {
363 php_http_buffer_appends(&buf, "/");
364 }
365
366 if (url->query && *url->query) {
367 php_http_buffer_appends(&buf, "?");
368 php_http_buffer_appendl(&buf, url->query);
369 }
370
371 if (url->fragment && *url->fragment) {
372 php_http_buffer_appends(&buf, "#");
373 php_http_buffer_appendl(&buf, url->fragment);
374 }
375
376 php_http_buffer_shrink(&buf);
377 php_http_buffer_fix(&buf);
378
379 if (url_len) {
380 *url_len = buf.used;
381 }
382
383 if (url_str) {
384 *url_str = buf.data;
385 }
386
387 return buf.data;
388 }
389
390 char *php_http_url_authority_to_string(const php_http_url_t *url, char **url_str, size_t *url_len)
391 {
392 php_http_buffer_t buf;
393
394 php_http_buffer_init(&buf);
395
396 if (url->user && *url->user) {
397 php_http_buffer_appendl(&buf, url->user);
398 if (url->pass && *url->pass) {
399 php_http_buffer_appends(&buf, ":");
400 php_http_buffer_appendl(&buf, url->pass);
401 }
402 php_http_buffer_appends(&buf, "@");
403 }
404
405 if (url->host && *url->host) {
406 php_http_buffer_appendl(&buf, url->host);
407 if (url->port) {
408 php_http_buffer_appendf(&buf, ":%hu", url->port);
409 }
410 }
411
412 php_http_buffer_shrink(&buf);
413 php_http_buffer_fix(&buf);
414
415 if (url_len) {
416 *url_len = buf.used;
417 }
418
419 if (url_str) {
420 *url_str = buf.data;
421 }
422
423 return buf.data;
424 }
425
426 php_http_url_t *php_http_url_from_zval(zval *value, unsigned flags TSRMLS_DC)
427 {
428 zval *zcpy;
429 php_http_url_t *purl;
430
431 switch (Z_TYPE_P(value)) {
432 case IS_ARRAY:
433 case IS_OBJECT:
434 purl = php_http_url_from_struct(HASH_OF(value));
435 break;
436
437 default:
438 zcpy = php_http_ztyp(IS_STRING, value);
439 purl = php_http_url_parse(Z_STRVAL_P(zcpy), Z_STRLEN_P(zcpy), flags TSRMLS_CC);
440 zval_ptr_dtor(&zcpy);
441 }
442
443 return purl;
444 }
445
446 php_http_url_t *php_http_url_from_struct(HashTable *ht)
447 {
448 zval **e;
449 php_http_buffer_t buf;
450
451 php_http_buffer_init_ex(&buf, MAX(PHP_HTTP_BUFFER_DEFAULT_SIZE, sizeof(php_http_url_t)<<2), PHP_HTTP_BUFFER_INIT_PREALLOC);
452 php_http_buffer_account(&buf, sizeof(php_http_url_t));
453 memset(buf.data, 0, buf.used);
454
455 if (SUCCESS == zend_hash_find(ht, "scheme", sizeof("scheme"), (void *) &e)) {
456 zval *cpy = php_http_ztyp(IS_STRING, *e);
457 url(buf)->scheme = &buf.data[buf.used];
458 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
459 zval_ptr_dtor(&cpy);
460 }
461 if (SUCCESS == zend_hash_find(ht, "user", sizeof("user"), (void *) &e)) {
462 zval *cpy = php_http_ztyp(IS_STRING, *e);
463 url(buf)->user = &buf.data[buf.used];
464 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
465 zval_ptr_dtor(&cpy);
466 }
467 if (SUCCESS == zend_hash_find(ht, "pass", sizeof("pass"), (void *) &e)) {
468 zval *cpy = php_http_ztyp(IS_STRING, *e);
469 url(buf)->pass = &buf.data[buf.used];
470 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
471 zval_ptr_dtor(&cpy);
472 }
473 if (SUCCESS == zend_hash_find(ht, "host", sizeof("host"), (void *) &e)) {
474 zval *cpy = php_http_ztyp(IS_STRING, *e);
475 url(buf)->host = &buf.data[buf.used];
476 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
477 zval_ptr_dtor(&cpy);
478 }
479 if (SUCCESS == zend_hash_find(ht, "port", sizeof("port"), (void *) &e)) {
480 zval *cpy = php_http_ztyp(IS_LONG, *e);
481 url(buf)->port = (unsigned short) Z_LVAL_P(cpy);
482 zval_ptr_dtor(&cpy);
483 }
484 if (SUCCESS == zend_hash_find(ht, "path", sizeof("path"), (void *) &e)) {
485 zval *cpy = php_http_ztyp(IS_STRING, *e);
486 url(buf)->path = &buf.data[buf.used];
487 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
488 zval_ptr_dtor(&cpy);
489 }
490 if (SUCCESS == zend_hash_find(ht, "query", sizeof("query"), (void *) &e)) {
491 zval *cpy = php_http_ztyp(IS_STRING, *e);
492 url(buf)->query = &buf.data[buf.used];
493 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
494 zval_ptr_dtor(&cpy);
495 }
496 if (SUCCESS == zend_hash_find(ht, "fragment", sizeof("fragment"), (void *) &e)) {
497 zval *cpy = php_http_ztyp(IS_STRING, *e);
498 url(buf)->fragment = &buf.data[buf.used];
499 url_append(&buf, php_http_buffer_append(&buf, Z_STRVAL_P(cpy), Z_STRLEN_P(cpy) + 1));
500 zval_ptr_dtor(&cpy);
501 }
502
503 return url(buf);
504 }
505
506 HashTable *php_http_url_to_struct(const php_http_url_t *url, zval *strct TSRMLS_DC)
507 {
508 zval arr;
509
510 if (strct) {
511 switch (Z_TYPE_P(strct)) {
512 default:
513 zval_dtor(strct);
514 array_init(strct);
515 /* no break */
516 case IS_ARRAY:
517 case IS_OBJECT:
518 INIT_PZVAL_ARRAY((&arr), HASH_OF(strct));
519 break;
520 }
521 } else {
522 INIT_PZVAL(&arr);
523 array_init(&arr);
524 }
525
526 if (url) {
527 if (url->scheme) {
528 add_assoc_string(&arr, "scheme", url->scheme, 1);
529 }
530 if (url->user) {
531 add_assoc_string(&arr, "user", url->user, 1);
532 }
533 if (url->pass) {
534 add_assoc_string(&arr, "pass", url->pass, 1);
535 }
536 if (url->host) {
537 add_assoc_string(&arr, "host", url->host, 1);
538 }
539 if (url->port) {
540 add_assoc_long(&arr, "port", (long) url->port);
541 }
542 if (url->path) {
543 add_assoc_string(&arr, "path", url->path, 1);
544 }
545 if (url->query) {
546 add_assoc_string(&arr, "query", url->query, 1);
547 }
548 if (url->fragment) {
549 add_assoc_string(&arr, "fragment", url->fragment, 1);
550 }
551 }
552
553 return Z_ARRVAL(arr);
554 }
555
556 ZEND_RESULT_CODE php_http_url_encode_hash(HashTable *hash, const char *pre_encoded_str, size_t pre_encoded_len, char **encoded_str, size_t *encoded_len TSRMLS_DC)
557 {
558 const char *arg_sep_str = "&";
559 size_t arg_sep_len = 1;
560 php_http_buffer_t *qstr = php_http_buffer_new();
561
562 php_http_url_argsep(&arg_sep_str, &arg_sep_len TSRMLS_CC);
563
564 if (SUCCESS != php_http_url_encode_hash_ex(hash, qstr, arg_sep_str, arg_sep_len, "=", 1, pre_encoded_str, pre_encoded_len TSRMLS_CC)) {
565 php_http_buffer_free(&qstr);
566 return FAILURE;
567 }
568
569 php_http_buffer_data(qstr, encoded_str, encoded_len);
570 php_http_buffer_free(&qstr);
571
572 return SUCCESS;
573 }
574
575 ZEND_RESULT_CODE php_http_url_encode_hash_ex(HashTable *hash, php_http_buffer_t *qstr, const char *arg_sep_str, size_t arg_sep_len, const char *val_sep_str, size_t val_sep_len, const char *pre_encoded_str, size_t pre_encoded_len TSRMLS_DC)
576 {
577 if (pre_encoded_len && pre_encoded_str) {
578 php_http_buffer_append(qstr, pre_encoded_str, pre_encoded_len);
579 }
580
581 if (!php_http_params_to_string(qstr, hash, arg_sep_str, arg_sep_len, "", 0, val_sep_str, val_sep_len, PHP_HTTP_PARAMS_QUERY TSRMLS_CC)) {
582 return FAILURE;
583 }
584
585 return SUCCESS;
586 }
587
588 struct parse_state {
589 php_http_url_t url;
590 #ifdef ZTS
591 void ***ts;
592 #endif
593 const char *ptr;
594 const char *end;
595 size_t maxlen;
596 off_t offset;
597 unsigned flags;
598 char buffer[1]; /* last member */
599 };
600
601 void php_http_url_free(php_http_url_t **url)
602 {
603 if (*url) {
604 efree(*url);
605 *url = NULL;
606 }
607 }
608
609 php_http_url_t *php_http_url_copy(const php_http_url_t *url, zend_bool persistent)
610 {
611 php_http_url_t *cpy;
612 const char *end = NULL, *url_ptr = (const char *) url;
613 char *cpy_ptr;
614
615 end = MAX(url->scheme, end);
616 end = MAX(url->pass, end);
617 end = MAX(url->user, end);
618 end = MAX(url->host, end);
619 end = MAX(url->path, end);
620 end = MAX(url->query, end);
621 end = MAX(url->fragment, end);
622
623 if (end) {
624 end += strlen(end) + 1;
625 cpy_ptr = pecalloc(1, end - url_ptr, persistent);
626 cpy = (php_http_url_t *) cpy_ptr;
627
628 memcpy(cpy_ptr + sizeof(*cpy), url_ptr + sizeof(*url), end - url_ptr - sizeof(*url));
629
630 cpy->scheme = url->scheme ? cpy_ptr + (url->scheme - url_ptr) : NULL;
631 cpy->pass = url->pass ? cpy_ptr + (url->pass - url_ptr) : NULL;
632 cpy->user = url->user ? cpy_ptr + (url->user - url_ptr) : NULL;
633 cpy->host = url->host ? cpy_ptr + (url->host - url_ptr) : NULL;
634 cpy->path = url->path ? cpy_ptr + (url->path - url_ptr) : NULL;
635 cpy->query = url->query ? cpy_ptr + (url->query - url_ptr) : NULL;
636 cpy->fragment = url->fragment ? cpy_ptr + (url->fragment - url_ptr) : NULL;
637 } else {
638 cpy = ecalloc(1, sizeof(*url));
639 }
640
641 cpy->port = url->port;
642
643 return cpy;
644 }
645
646 static size_t parse_mb_utf8(unsigned *wc, const char *ptr, const char *end)
647 {
648 unsigned wchar;
649 size_t consumed = utf8towc(&wchar, (const unsigned char *) ptr, end - ptr);
650
651 if (!consumed || consumed == (size_t) -1) {
652 return 0;
653 }
654
655 if (wc) {
656 *wc = wchar;
657 }
658 return consumed;
659 }
660
661 #ifdef PHP_HTTP_HAVE_WCHAR
662 static size_t parse_mb_loc(unsigned *wc, const char *ptr, const char *end)
663 {
664 wchar_t wchar;
665 size_t consumed = 0;
666 #if defined(HAVE_MBRTOWC)
667 mbstate_t ps;
668
669 memset(&ps, 0, sizeof(ps));
670 consumed = mbrtowc(&wchar, ptr, end - ptr, &ps);
671 #elif defined(HAVE_MBTOWC)
672 consumed = mbtowc(&wchar, ptr, end - ptr);
673 #endif
674
675 if (!consumed || consumed == (size_t) -1) {
676 return 0;
677 }
678
679 if (wc) {
680 *wc = wchar;
681 }
682 return consumed;
683 }
684 #endif
685
686 typedef enum parse_mb_what {
687 PARSE_SCHEME,
688 PARSE_USERINFO,
689 PARSE_HOSTINFO,
690 PARSE_PATH,
691 PARSE_QUERY,
692 PARSE_FRAGMENT
693 } parse_mb_what_t;
694
695 static const char * const parse_what[] = {
696 "scheme",
697 "userinfo",
698 "hostinfo",
699 "path",
700 "query",
701 "fragment"
702 };
703
704 static const char parse_xdigits[] = "0123456789ABCDEF";
705
706 static size_t parse_mb(struct parse_state *state, parse_mb_what_t what, const char *ptr, const char *end, const char *begin, zend_bool silent)
707 {
708 unsigned wchar;
709 size_t consumed = 0;
710
711 if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
712 consumed = parse_mb_utf8(&wchar, ptr, end);
713 }
714 #ifdef PHP_HTTP_HAVE_WCHAR
715 else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
716 consumed = parse_mb_loc(&wchar, ptr, end);
717 }
718 #endif
719
720 while (consumed) {
721 if (!(state->flags & PHP_HTTP_URL_PARSE_TOPCT) || what == PARSE_HOSTINFO || what == PARSE_SCHEME) {
722 if (what == PARSE_HOSTINFO && (state->flags & PHP_HTTP_URL_PARSE_TOIDN)) {
723 /* idna */
724 } else if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
725 if (!isualnum(wchar)) {
726 break;
727 }
728 #ifdef PHP_HTTP_HAVE_WCHAR
729 } else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
730 if (!iswalnum(wchar)) {
731 break;
732 }
733 #endif
734 }
735 PHP_HTTP_DUFF(consumed, state->buffer[state->offset++] = *ptr++);
736 } else {
737 int i = 0;
738
739 PHP_HTTP_DUFF(consumed,
740 state->buffer[state->offset++] = '%';
741 state->buffer[state->offset++] = parse_xdigits[((unsigned char) ptr[i]) >> 4];
742 state->buffer[state->offset++] = parse_xdigits[((unsigned char) ptr[i]) & 0xf];
743 ++i;
744 );
745 }
746
747 return consumed;
748 }
749
750 if (!silent) {
751 TSRMLS_FETCH_FROM_CTX(state->ts);
752 if (consumed) {
753 php_error_docref(NULL TSRMLS_CC, E_WARNING,
754 "Failed to parse %s; unexpected multibyte sequence 0x%x at pos %u in '%s'",
755 parse_what[what], wchar, (unsigned) (ptr - begin), begin);
756 } else {
757 php_error_docref(NULL TSRMLS_CC, E_WARNING,
758 "Failed to parse %s; unexpected byte 0x%02x at pos %u in '%s'",
759 parse_what[what], (unsigned char) *ptr, (unsigned) (ptr - begin), begin);
760 }
761 }
762
763 return 0;
764 }
765
766 static ZEND_RESULT_CODE parse_userinfo(struct parse_state *state, const char *ptr)
767 {
768 size_t mb;
769 const char *password = NULL, *end = state->ptr, *tmp = ptr;
770 TSRMLS_FETCH_FROM_CTX(state->ts);
771
772 state->url.user = &state->buffer[state->offset];
773
774 do {
775 switch (*ptr) {
776 case ':':
777 if (password) {
778 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
779 php_error_docref(NULL TSRMLS_CC, E_WARNING,
780 "Failed to parse password; duplicate ':' at pos %u in '%s'",
781 (unsigned) (ptr - tmp), tmp);
782 }
783 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
784 return FAILURE;
785 }
786 state->buffer[state->offset++] = *ptr;
787 break;
788 }
789 password = ptr + 1;
790 state->buffer[state->offset++] = 0;
791 state->url.pass = &state->buffer[state->offset];
792 break;
793
794 case '%':
795 if (ptr[1] != '%' && (end - ptr <= 2 || !isxdigit(*(ptr+1)) || !isxdigit(*(ptr+2)))) {
796 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
797 php_error_docref(NULL TSRMLS_CC, E_WARNING,
798 "Failed to parse userinfo; invalid percent encoding at pos %u in '%s'",
799 (unsigned) (ptr - tmp), tmp);
800 }
801 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
802 return FAILURE;
803 }
804 state->buffer[state->offset++] = *ptr++;
805 break;
806 }
807 state->buffer[state->offset++] = *ptr++;
808 state->buffer[state->offset++] = *ptr++;
809 state->buffer[state->offset++] = *ptr;
810 break;
811
812 default:
813 if ((mb = parse_mb(state, PARSE_USERINFO, ptr, end, tmp, state->flags & PHP_HTTP_URL_SILENT_ERRORS))) {
814 ptr += mb - 1;
815 break;
816 }
817 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
818 return FAILURE;
819 }
820 /* no break */
821 case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
822 case '+': case ',': case ';': case '=': /* sub-delims */
823 case '-': case '.': case '_': case '~': /* unreserved */
824 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
825 case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
826 case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
827 case 'V': case 'W': case 'X': case 'Y': case 'Z':
828 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
829 case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
830 case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
831 case 'v': case 'w': case 'x': case 'y': case 'z':
832 case '0': case '1': case '2': case '3': case '4': case '5': case '6':
833 case '7': case '8': case '9':
834 /* allowed */
835 state->buffer[state->offset++] = *ptr;
836 break;
837
838 }
839 } while(++ptr < end);
840
841
842 state->buffer[state->offset++] = 0;
843
844 return SUCCESS;
845 }
846
847 #if defined(PHP_WIN32) || defined(HAVE_UIDNA_IDNTOASCII)
848 typedef size_t (*parse_mb_func)(unsigned *wc, const char *ptr, const char *end);
849 static ZEND_RESULT_CODE to_utf16(parse_mb_func fn, const char *u8, uint16_t **u16, size_t *len TSRMLS_DC)
850 {
851 size_t offset = 0, u8_len = strlen(u8);
852
853 *u16 = ecalloc(4 * sizeof(uint16_t), u8_len + 1);
854 *len = 0;
855
856 while (offset < u8_len) {
857 unsigned wc;
858 uint16_t buf[2], *ptr = buf;
859 size_t consumed = fn(&wc, &u8[offset], &u8[u8_len]);
860
861 if (!consumed) {
862 efree(*u16);
863 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse UTF-8 at pos %zu of '%s'", offset, u8);
864 return FAILURE;
865 } else {
866 offset += consumed;
867 }
868
869 switch (wctoutf16(buf, wc)) {
870 case 2:
871 (*u16)[(*len)++] = *ptr++;
872 /* no break */
873 case 1:
874 (*u16)[(*len)++] = *ptr++;
875 break;
876 case 0:
877 default:
878 efree(*u16);
879 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to convert UTF-32 'U+%X' to UTF-16", wc);
880 return FAILURE;
881 }
882 }
883
884 return SUCCESS;
885 }
886 #endif
887
888 #ifndef MAXHOSTNAMELEN
889 # define MAXHOSTNAMELEN 256
890 #endif
891
892 #if PHP_HTTP_HAVE_IDN2
893 static ZEND_RESULT_CODE parse_idn2(struct parse_state *state, size_t prev_len)
894 {
895 char *idn = NULL;
896 int rv = -1;
897 TSRMLS_FETCH_FROM_CTX(state->ts);
898
899 if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
900 rv = idn2_lookup_u8((const unsigned char *) state->url.host, (unsigned char **) &idn, IDN2_NFC_INPUT);
901 }
902 # ifdef PHP_HTTP_HAVE_WCHAR
903 else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
904 rv = idn2_lookup_ul(state->url.host, &idn, 0);
905 }
906 # endif
907 if (rv != IDN2_OK) {
908 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
909 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN; %s", idn2_strerror(rv));
910 }
911 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
912 return FAILURE;
913 }
914 } else {
915 size_t idnlen = strlen(idn);
916 memcpy(state->url.host, idn, idnlen + 1);
917 free(idn);
918 state->offset += idnlen - prev_len;
919 }
920 return SUCCESS;
921 }
922 #elif PHP_HTTP_HAVE_IDN
923 static ZEND_RESULT_CODE parse_idn(struct parse_state *state, size_t prev_len)
924 {
925 char *idn = NULL;
926 int rv = -1;
927 TSRMLS_FETCH_FROM_CTX(state->ts);
928
929 if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
930 rv = idna_to_ascii_8z(state->url.host, &idn, IDNA_ALLOW_UNASSIGNED|IDNA_USE_STD3_ASCII_RULES);
931 }
932 # ifdef PHP_HTTP_HAVE_WCHAR
933 else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
934 rv = idna_to_ascii_lz(state->url.host, &idn, IDNA_ALLOW_UNASSIGNED|IDNA_USE_STD3_ASCII_RULES);
935 }
936 # endif
937 if (rv != IDNA_SUCCESS) {
938 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
939 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN; %s", idna_strerror(rv));
940 }
941 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
942 return FAILURE;
943 }
944 } else {
945 size_t idnlen = strlen(idn);
946 memcpy(state->url.host, idn, idnlen + 1);
947 free(idn);
948 state->offset += idnlen - prev_len;
949 }
950 return SUCCESS;
951 }
952 #endif
953
954 #ifdef HAVE_UIDNA_IDNTOASCII
955 # if HAVE_UNICODE_UIDNA_H
956 # include <unicode/uidna.h>
957 # else
958 typedef uint16_t UChar;
959 typedef enum { U_ZERO_ERROR = 0 } UErrorCode;
960 int32_t uidna_IDNToASCII(const UChar *src, int32_t srcLength, UChar *dest, int32_t destCapacity, int32_t options, void *parseError, UErrorCode *status);
961 # endif
962 static ZEND_RESULT_CODE parse_uidn(struct parse_state *state)
963 {
964 char *host_ptr;
965 uint16_t *uhost_str, ahost_str[MAXHOSTNAMELEN], *ahost_ptr;
966 size_t uhost_len, ahost_len;
967 UErrorCode error = U_ZERO_ERROR;
968 TSRMLS_FETCH_FROM_CTX(state->ts);
969
970 if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
971 if (SUCCESS != to_utf16(parse_mb_utf8, state->url.host, &uhost_str, &uhost_len TSRMLS_CC)) {
972 return FAILURE;
973 }
974 #ifdef PHP_HTTP_HAVE_WCHAR
975 } else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
976 if (SUCCESS != to_utf16(parse_mb_loc, state->url.host, &uhost_str, &uhost_len TSRMLS_CC)) {
977 return FAILURE;
978 }
979 #endif
980 } else {
981 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN; codepage not specified");
982 return FAILURE;
983 }
984
985 ahost_len = uidna_IDNToASCII(uhost_str, uhost_len, ahost_str, MAXHOSTNAMELEN, 3, NULL, &error);
986 efree(uhost_str);
987
988 if (error != U_ZERO_ERROR) {
989 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN; ICU error %d", error);
990 return FAILURE;
991 }
992
993 host_ptr = state->url.host;
994 ahost_ptr = ahost_str;
995 PHP_HTTP_DUFF(ahost_len, *host_ptr++ = *ahost_ptr++);
996
997 *host_ptr = '\0';
998 state->offset += host_ptr - state->url.host;
999
1000 return SUCCESS;
1001 }
1002 #endif
1003
1004 #if 0 && defined(PHP_WIN32)
1005 static ZEND_RESULT_CODE parse_widn(struct parse_state *state)
1006 {
1007 char *host_ptr;
1008 uint16_t *uhost_str, ahost_str[MAXHOSTNAMELEN], *ahost_ptr;
1009 size_t uhost_len;
1010 TSRMLS_FETCH_FROM_CTX(state->ts);
1011
1012 if (state->flags & PHP_HTTP_URL_PARSE_MBUTF8) {
1013 if (SUCCESS != to_utf16(parse_mb_utf8, state->url.host, &uhost_str, &uhost_len)) {
1014 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN");
1015 return FAILURE;
1016 }
1017 #ifdef PHP_HTTP_HAVE_WCHAR
1018 } else if (state->flags & PHP_HTTP_URL_PARSE_MBLOC) {
1019 if (SUCCESS != to_utf16(parse_mb_loc, state->url.host, &uhost_str, &uhost_len)) {
1020 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN");
1021 return FAILURE;
1022 }
1023 #endif
1024 } else {
1025 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN");
1026 return FAILURE;
1027 }
1028
1029 if (!IdnToAscii(IDN_ALLOW_UNASSIGNED|IDN_USE_STD3_ASCII_RULES, uhost_str, uhost_len, ahost_str, MAXHOSTNAMELEN)) {
1030 efree(uhost_str);
1031 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse IDN");
1032 return FAILURE;
1033 }
1034
1035 efree(uhost_str);
1036 host_ptr = state->url.host;
1037 ahost_ptr = ahost_str;
1038 PHP_HTTP_DUFF(wcslen(ahost_str), *host_ptr++ = *ahost_ptr++);
1039 efree(ahost_str);
1040
1041 *host_ptr = '\0';
1042 state->offset += host_ptr - state->url.host;
1043
1044 return SUCCESS;
1045 }
1046 #endif
1047
1048 #ifdef HAVE_INET_PTON
1049 static const char *parse_ip6(struct parse_state *state, const char *ptr)
1050 {
1051 unsigned pos = 0;
1052 const char *error = NULL, *end = state->ptr, *tmp = memchr(ptr, ']', end - ptr);
1053 TSRMLS_FETCH_FROM_CTX(state->ts);
1054
1055 if (tmp) {
1056 size_t addrlen = tmp - ptr + 1;
1057 char buf[16], *addr = estrndup(ptr + 1, addrlen - 2);
1058 int rv = inet_pton(AF_INET6, addr, buf);
1059
1060 if (rv == 1) {
1061 state->buffer[state->offset] = '[';
1062 state->url.host = &state->buffer[state->offset];
1063 inet_ntop(AF_INET6, buf, state->url.host + 1, state->maxlen - state->offset);
1064 state->offset += strlen(state->url.host);
1065 state->buffer[state->offset++] = ']';
1066 state->buffer[state->offset++] = 0;
1067 ptr = tmp + 1;
1068 } else if (rv == -1) {
1069 pos = 1;
1070 error = strerror(errno);
1071 } else {
1072 error = "unexpected '['";
1073 }
1074 efree(addr);
1075 } else {
1076 pos = tmp ? tmp - ptr : end - ptr;
1077 error = "expected ']'";
1078 }
1079
1080 if (error) {
1081 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1082 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse hostinfo; %s at pos %u in '%s'", error, pos, ptr);
1083 }
1084 return NULL;
1085 }
1086
1087 return ptr;
1088 }
1089 #endif
1090
1091 static ZEND_RESULT_CODE parse_hostinfo(struct parse_state *state, const char *ptr)
1092 {
1093 size_t mb, len = state->offset;
1094 const char *end = state->ptr, *tmp = ptr, *port = NULL, *label = NULL;
1095 TSRMLS_FETCH_FROM_CTX(state->ts);
1096
1097 #ifdef HAVE_INET_PTON
1098 if (*ptr == '[' && !(ptr = parse_ip6(state, ptr))) {
1099 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1100 return FAILURE;
1101 }
1102 ptr = tmp;
1103 }
1104 #endif
1105
1106 if (ptr != end) do {
1107 switch (*ptr) {
1108 case ':':
1109 if (port) {
1110 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1111 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1112 "Failed to parse port; unexpected ':' at pos %u in '%s'",
1113 (unsigned) (ptr - tmp), tmp);
1114 }
1115 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1116 return FAILURE;
1117 }
1118 }
1119 port = ptr + 1;
1120 break;
1121
1122 case '%':
1123 if (ptr[1] != '%' && (end - ptr <= 2 || !isxdigit(*(ptr+1)) || !isxdigit(*(ptr+2)))) {
1124 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1125 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1126 "Failed to parse hostinfo; invalid percent encoding at pos %u in '%s'",
1127 (unsigned) (ptr - tmp), tmp);
1128 }
1129 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1130 return FAILURE;
1131 }
1132 state->buffer[state->offset++] = *ptr++;
1133 break;
1134 }
1135 state->buffer[state->offset++] = *ptr++;
1136 state->buffer[state->offset++] = *ptr++;
1137 state->buffer[state->offset++] = *ptr;
1138 break;
1139
1140 case '.':
1141 if (port || !label) {
1142 /* sort of a compromise, just ensure we don't end up
1143 * with a dot at the beginning or two consecutive dots
1144 */
1145 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1146 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1147 "Failed to parse %s; unexpected '%c' at pos %u in '%s'",
1148 port ? "port" : "host",
1149 (unsigned char) *ptr, (unsigned) (ptr - tmp), tmp);
1150 }
1151 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1152 return FAILURE;
1153 }
1154 break;
1155 }
1156 state->buffer[state->offset++] = *ptr;
1157 label = NULL;
1158 break;
1159
1160 case '-':
1161 if (!label) {
1162 /* sort of a compromise, just ensure we don't end up
1163 * with a hyphen at the beginning
1164 */
1165 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1166 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1167 "Failed to parse %s; unexpected '%c' at pos %u in '%s'",
1168 port ? "port" : "host",
1169 (unsigned char) *ptr, (unsigned) (ptr - tmp), tmp);
1170 }
1171 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1172 return FAILURE;
1173 }
1174 break;
1175 }
1176 /* no break */
1177 case '_': case '~': /* unreserved */
1178 case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
1179 case '+': case ',': case ';': case '=': /* sub-delims */
1180 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
1181 case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
1182 case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
1183 case 'V': case 'W': case 'X': case 'Y': case 'Z':
1184 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
1185 case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
1186 case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
1187 case 'v': case 'w': case 'x': case 'y': case 'z':
1188 if (port) {
1189 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1190 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1191 "Failed to parse port; unexpected char '%c' at pos %u in '%s'",
1192 (unsigned char) *ptr, (unsigned) (ptr - tmp), tmp);
1193 }
1194 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1195 return FAILURE;
1196 }
1197 break;
1198 }
1199 /* no break */
1200 case '0': case '1': case '2': case '3': case '4': case '5': case '6':
1201 case '7': case '8': case '9':
1202 /* allowed */
1203 if (port) {
1204 state->url.port *= 10;
1205 state->url.port += *ptr - '0';
1206 } else {
1207 label = ptr;
1208 state->buffer[state->offset++] = *ptr;
1209 }
1210 break;
1211
1212 default:
1213 if (ptr == end) {
1214 break;
1215 } else if (port) {
1216 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1217 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1218 "Failed to parse port; unexpected byte 0x%02x at pos %u in '%s'",
1219 (unsigned char) *ptr, (unsigned) (ptr - tmp), tmp);
1220 }
1221 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1222 return FAILURE;
1223 }
1224 break;
1225 } else if (!(mb = parse_mb(state, PARSE_HOSTINFO, ptr, end, tmp, state->flags & PHP_HTTP_URL_SILENT_ERRORS))) {
1226 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1227 return FAILURE;
1228 }
1229 break;
1230 }
1231 label = ptr;
1232 ptr += mb - 1;
1233 }
1234 } while (++ptr < end);
1235
1236 if (!state->url.host) {
1237 len = state->offset - len;
1238 state->url.host = &state->buffer[state->offset - len];
1239 state->buffer[state->offset++] = 0;
1240 }
1241
1242 if (state->flags & PHP_HTTP_URL_PARSE_TOIDN) {
1243 #if PHP_HTTP_HAVE_IDN2
1244 return parse_idn2(state, len);
1245 #elif PHP_HTTP_HAVE_IDN
1246 return parse_idn(state, len);
1247 #endif
1248 #ifdef HAVE_UIDNA_IDNTOASCII
1249 return parse_uidn(state);
1250 #endif
1251 #if 0 && defined(PHP_WIN32)
1252 return parse_widn(state);
1253 #endif
1254 }
1255
1256 return SUCCESS;
1257 }
1258
1259 static const char *parse_authority(struct parse_state *state)
1260 {
1261 const char *tmp = state->ptr, *host = NULL;
1262
1263 do {
1264 switch (*state->ptr) {
1265 case '@':
1266 /* userinfo delimiter */
1267 if (host) {
1268 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1269 TSRMLS_FETCH_FROM_CTX(state->ts);
1270 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1271 "Failed to parse userinfo; unexpected '@'");
1272 }
1273 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1274 return NULL;
1275 }
1276 break;
1277 }
1278 host = state->ptr + 1;
1279 if (tmp != state->ptr && SUCCESS != parse_userinfo(state, tmp)) {
1280 return NULL;
1281 }
1282 tmp = state->ptr + 1;
1283 break;
1284
1285 case '/':
1286 case '?':
1287 case '#':
1288 case '\0':
1289 EOD:
1290 /* host delimiter */
1291 if (tmp != state->ptr && SUCCESS != parse_hostinfo(state, tmp)) {
1292 return NULL;
1293 }
1294 return state->ptr;
1295 }
1296 } while (++state->ptr <= state->end);
1297
1298 --state->ptr;
1299 goto EOD;
1300 }
1301
1302 static const char *parse_path(struct parse_state *state)
1303 {
1304 size_t mb;
1305 const char *tmp;
1306 TSRMLS_FETCH_FROM_CTX(state->ts);
1307
1308 /* is there actually a path to parse? */
1309 if (!*state->ptr) {
1310 return state->ptr;
1311 }
1312 tmp = state->ptr;
1313 state->url.path = &state->buffer[state->offset];
1314
1315 do {
1316 switch (*state->ptr) {
1317 case '#':
1318 case '?':
1319 goto done;
1320
1321 case '%':
1322 if (state->ptr[1] != '%' && (state->end - state->ptr <= 2 || !isxdigit(*(state->ptr+1)) || !isxdigit(*(state->ptr+2)))) {
1323 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1324 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1325 "Failed to parse path; invalid percent encoding at pos %u in '%s'",
1326 (unsigned) (state->ptr - tmp), tmp);
1327 }
1328 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1329 return NULL;
1330 }
1331 state->buffer[state->offset++] = *state->ptr;
1332 break;
1333 }
1334 state->buffer[state->offset++] = *state->ptr++;
1335 state->buffer[state->offset++] = *state->ptr++;
1336 state->buffer[state->offset++] = *state->ptr;
1337 break;
1338
1339 case '/': /* yeah, well */
1340 case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
1341 case '+': case ',': case ';': case '=': /* sub-delims */
1342 case '-': case '.': case '_': case '~': /* unreserved */
1343 case ':': case '@': /* pchar */
1344 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
1345 case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
1346 case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
1347 case 'V': case 'W': case 'X': case 'Y': case 'Z':
1348 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
1349 case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
1350 case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
1351 case 'v': case 'w': case 'x': case 'y': case 'z':
1352 case '0': case '1': case '2': case '3': case '4': case '5': case '6':
1353 case '7': case '8': case '9':
1354 /* allowed */
1355 state->buffer[state->offset++] = *state->ptr;
1356 break;
1357
1358 default:
1359 if (!(mb = parse_mb(state, PARSE_PATH, state->ptr, state->end, tmp, state->flags & PHP_HTTP_URL_SILENT_ERRORS))) {
1360 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1361 return NULL;
1362 }
1363 break;
1364 }
1365 state->ptr += mb - 1;
1366 }
1367 } while (++state->ptr < state->end);
1368
1369 done:
1370 /* did we have any path component ? */
1371 if (tmp != state->ptr) {
1372 state->buffer[state->offset++] = 0;
1373 } else {
1374 state->url.path = NULL;
1375 }
1376 return state->ptr;
1377 }
1378
1379 static const char *parse_query(struct parse_state *state)
1380 {
1381 size_t mb;
1382 const char *tmp = state->ptr + !!*state->ptr;
1383 TSRMLS_FETCH_FROM_CTX(state->ts);
1384
1385 /* is there actually a query to parse? */
1386 if (*state->ptr != '?') {
1387 return state->ptr;
1388 }
1389
1390 /* skip initial '?' */
1391 tmp = ++state->ptr;
1392 state->url.query = &state->buffer[state->offset];
1393
1394 while (state->ptr < state->end) {
1395 switch (*state->ptr) {
1396 case '#':
1397 goto done;
1398
1399 case '%':
1400 if (state->ptr[1] != '%' && (state->end - state->ptr <= 2 || !isxdigit(*(state->ptr+1)) || !isxdigit(*(state->ptr+2)))) {
1401 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1402 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1403 "Failed to parse query; invalid percent encoding at pos %u in '%s'",
1404 (unsigned) (state->ptr - tmp), tmp);
1405 }
1406 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1407 return NULL;
1408 }
1409 /* fallthrough, pct-encode the percent sign */
1410 } else {
1411 state->buffer[state->offset++] = *state->ptr++;
1412 state->buffer[state->offset++] = *state->ptr++;
1413 state->buffer[state->offset++] = *state->ptr;
1414 break;
1415 }
1416 /* no break */
1417 case '{': case '}':
1418 case '<': case '>':
1419 case '[': case ']':
1420 case '|': case '\\': case '^': case '`': case '"': case ' ':
1421 /* RFC1738 unsafe */
1422 if (state->flags & PHP_HTTP_URL_PARSE_TOPCT) {
1423 state->buffer[state->offset++] = '%';
1424 state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) >> 4];
1425 state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) & 0xf];
1426 break;
1427 }
1428 /* no break */
1429
1430 case '?': case '/': /* yeah, well */
1431 case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
1432 case '+': case ',': case ';': case '=': /* sub-delims */
1433 case '-': case '.': case '_': case '~': /* unreserved */
1434 case ':': case '@': /* pchar */
1435 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
1436 case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
1437 case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
1438 case 'V': case 'W': case 'X': case 'Y': case 'Z':
1439 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
1440 case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
1441 case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
1442 case 'v': case 'w': case 'x': case 'y': case 'z':
1443 case '0': case '1': case '2': case '3': case '4': case '5': case '6':
1444 case '7': case '8': case '9':
1445 /* allowed */
1446 state->buffer[state->offset++] = *state->ptr;
1447 break;
1448
1449 default:
1450 if (!(mb = parse_mb(state, PARSE_QUERY, state->ptr, state->end, tmp, state->flags & PHP_HTTP_URL_SILENT_ERRORS))) {
1451 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1452 return NULL;
1453 }
1454 break;
1455 }
1456 state->ptr += mb - 1;
1457 }
1458
1459 ++state->ptr;
1460 }
1461
1462 done:
1463 state->buffer[state->offset++] = 0;
1464 return state->ptr;
1465 }
1466
1467 static const char *parse_fragment(struct parse_state *state)
1468 {
1469 size_t mb;
1470 const char *tmp;
1471 TSRMLS_FETCH_FROM_CTX(state->ts);
1472
1473 /* is there actually a fragment to parse? */
1474 if (*state->ptr != '#') {
1475 return state->ptr;
1476 }
1477
1478 /* skip initial '#' */
1479 tmp = ++state->ptr;
1480 state->url.fragment = &state->buffer[state->offset];
1481
1482 do {
1483 switch (*state->ptr) {
1484 case '#':
1485 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1486 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1487 "Failed to parse fragment; invalid fragment identifier at pos %u in '%s'",
1488 (unsigned) (state->ptr - tmp), tmp);
1489 }
1490 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1491 return NULL;
1492 }
1493 state->buffer[state->offset++] = *state->ptr;
1494 break;
1495
1496 case '%':
1497 if (state->ptr[1] != '%' && (state->end - state->ptr <= 2 || !isxdigit(*(state->ptr+1)) || !isxdigit(*(state->ptr+2)))) {
1498 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1499 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1500 "Failed to parse fragment; invalid percent encoding at pos %u in '%s'",
1501 (unsigned) (state->ptr - tmp), tmp);
1502 }
1503 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1504 return NULL;
1505 }
1506 /* fallthrough */
1507 } else {
1508 state->buffer[state->offset++] = *state->ptr++;
1509 state->buffer[state->offset++] = *state->ptr++;
1510 state->buffer[state->offset++] = *state->ptr;
1511 break;
1512 }
1513 /* no break */
1514
1515 case '{': case '}':
1516 case '<': case '>':
1517 case '[': case ']':
1518 case '|': case '\\': case '^': case '`': case '"': case ' ':
1519 /* RFC1738 unsafe */
1520 if (state->flags & PHP_HTTP_URL_PARSE_TOPCT) {
1521 state->buffer[state->offset++] = '%';
1522 state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) >> 4];
1523 state->buffer[state->offset++] = parse_xdigits[((unsigned char) *state->ptr) & 0xf];
1524 break;
1525 }
1526 /* no break */
1527
1528 case '?': case '/':
1529 case '!': case '$': case '&': case '\'': case '(': case ')': case '*':
1530 case '+': case ',': case ';': case '=': /* sub-delims */
1531 case '-': case '.': case '_': case '~': /* unreserved */
1532 case ':': case '@': /* pchar */
1533 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
1534 case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
1535 case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
1536 case 'V': case 'W': case 'X': case 'Y': case 'Z':
1537 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
1538 case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
1539 case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
1540 case 'v': case 'w': case 'x': case 'y': case 'z':
1541 case '0': case '1': case '2': case '3': case '4': case '5': case '6':
1542 case '7': case '8': case '9':
1543 /* allowed */
1544 state->buffer[state->offset++] = *state->ptr;
1545 break;
1546
1547 default:
1548 if (!(mb = parse_mb(state, PARSE_FRAGMENT, state->ptr, state->end, tmp, state->flags & PHP_HTTP_URL_SILENT_ERRORS))) {
1549 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1550 return NULL;
1551 }
1552 break;
1553 }
1554 state->ptr += mb - 1;
1555 }
1556 } while (++state->ptr < state->end);
1557
1558 state->buffer[state->offset++] = 0;
1559 return state->ptr;
1560 }
1561
1562 static const char *parse_hier(struct parse_state *state)
1563 {
1564 if (*state->ptr == '/') {
1565 if (state->end - state->ptr > 1) {
1566 if (*(state->ptr + 1) == '/') {
1567 state->ptr += 2;
1568 if (!(state->ptr = parse_authority(state))) {
1569 return NULL;
1570 }
1571 }
1572 }
1573 }
1574 return parse_path(state);
1575 }
1576
1577 static const char *parse_scheme(struct parse_state *state)
1578 {
1579 size_t mb;
1580 const char *tmp = state->ptr;
1581
1582 do {
1583 switch (*state->ptr) {
1584 case ':':
1585 /* scheme delimiter */
1586 state->url.scheme = &state->buffer[0];
1587 state->buffer[state->offset++] = 0;
1588 return ++state->ptr;
1589
1590 case '0': case '1': case '2': case '3': case '4': case '5': case '6':
1591 case '7': case '8': case '9':
1592 case '+': case '-': case '.':
1593 if (state->ptr == tmp) {
1594 goto softfail;
1595 }
1596 /* no break */
1597 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
1598 case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
1599 case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
1600 case 'V': case 'W': case 'X': case 'Y': case 'Z':
1601 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
1602 case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
1603 case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
1604 case 'v': case 'w': case 'x': case 'y': case 'z':
1605 /* scheme part */
1606 state->buffer[state->offset++] = *state->ptr;
1607 break;
1608
1609 default:
1610 if (!(mb = parse_mb(state, PARSE_SCHEME, state->ptr, state->end, tmp, 1))) {
1611 goto softfail;
1612 }
1613 state->ptr += mb - 1;
1614 }
1615 } while (++state->ptr < state->end);
1616
1617 softfail:
1618 state->offset = 0;
1619 return state->ptr = tmp;
1620 }
1621
1622 php_http_url_t *php_http_url_parse(const char *str, size_t len, unsigned flags TSRMLS_DC)
1623 {
1624 size_t maxlen = 3 * len + 8 /* null bytes for all components */;
1625 struct parse_state *state = ecalloc(1, sizeof(*state) + maxlen);
1626
1627 state->end = str + len;
1628 state->ptr = str;
1629 state->flags = flags;
1630 state->maxlen = maxlen;
1631 TSRMLS_SET_CTX(state->ts);
1632
1633 if (!parse_scheme(state)) {
1634 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse URL scheme: '%s'", state->ptr);
1635 efree(state);
1636 return NULL;
1637 }
1638
1639 if (!parse_hier(state)) {
1640 efree(state);
1641 return NULL;
1642 }
1643
1644 if (!parse_query(state)) {
1645 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse URL query: '%s'", state->ptr);
1646 efree(state);
1647 return NULL;
1648 }
1649
1650 if (!parse_fragment(state)) {
1651 php_error_docref(NULL TSRMLS_CC, E_WARNING, "Failed to parse URL fragment: '%s'", state->ptr);
1652 efree(state);
1653 return NULL;
1654 }
1655
1656 return (php_http_url_t *) state;
1657 }
1658
1659 php_http_url_t *php_http_url_parse_authority(const char *str, size_t len, unsigned flags TSRMLS_DC)
1660 {
1661 size_t maxlen = 3 * len;
1662 struct parse_state *state = ecalloc(1, sizeof(*state) + maxlen);
1663
1664 state->end = str + len;
1665 state->ptr = str;
1666 state->flags = flags;
1667 state->maxlen = maxlen;
1668 TSRMLS_SET_CTX(state->ts);
1669
1670 if (!(state->ptr = parse_authority(state))) {
1671 efree(state);
1672 return NULL;
1673 }
1674
1675 if (state->ptr != state->end) {
1676 if (!(state->flags & PHP_HTTP_URL_SILENT_ERRORS)) {
1677 php_error_docref(NULL TSRMLS_CC, E_WARNING,
1678 "Failed to parse URL authority, unexpected character at pos %u in '%s'",
1679 (unsigned) (state->ptr - str), str);
1680 }
1681 if (!(state->flags & PHP_HTTP_URL_IGNORE_ERRORS)) {
1682 efree(state);
1683 return NULL;
1684 }
1685 }
1686
1687 return (php_http_url_t *) state;
1688 }
1689
1690 ZEND_BEGIN_ARG_INFO_EX(ai_HttpUrl___construct, 0, 0, 0)
1691 ZEND_ARG_INFO(0, old_url)
1692 ZEND_ARG_INFO(0, new_url)
1693 ZEND_ARG_INFO(0, flags)
1694 ZEND_END_ARG_INFO();
1695 PHP_METHOD(HttpUrl, __construct)
1696 {
1697 zval *new_url = NULL, *old_url = NULL;
1698 long flags = PHP_HTTP_URL_FROM_ENV;
1699 zend_error_handling zeh;
1700
1701 php_http_expect(SUCCESS == zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "|z!z!l", &old_url, &new_url, &flags), invalid_arg, return);
1702
1703 if (flags & PHP_HTTP_URL_SILENT_ERRORS) {
1704 zend_replace_error_handling(EH_SUPPRESS, NULL, &zeh TSRMLS_CC);
1705 } else if (flags & PHP_HTTP_URL_IGNORE_ERRORS) {
1706 zend_replace_error_handling(EH_NORMAL, NULL, &zeh TSRMLS_CC);
1707 } else {
1708 zend_replace_error_handling(EH_THROW, php_http_exception_bad_url_class_entry, &zeh TSRMLS_CC);
1709 }
1710 {
1711 php_http_url_t *res_purl, *new_purl = NULL, *old_purl = NULL;
1712
1713 if (new_url) {
1714 new_purl = php_http_url_from_zval(new_url, flags TSRMLS_CC);
1715 if (!new_purl) {
1716 zend_restore_error_handling(&zeh TSRMLS_CC);
1717 return;
1718 }
1719 }
1720 if (old_url) {
1721 old_purl = php_http_url_from_zval(old_url, flags TSRMLS_CC);
1722 if (!old_purl) {
1723 if (new_purl) {
1724 php_http_url_free(&new_purl);
1725 }
1726 zend_restore_error_handling(&zeh TSRMLS_CC);
1727 return;
1728 }
1729 }
1730
1731 res_purl = php_http_url_mod(old_purl, new_purl, flags TSRMLS_CC);
1732 php_http_url_to_struct(res_purl, getThis() TSRMLS_CC);
1733
1734 php_http_url_free(&res_purl);
1735 if (old_purl) {
1736 php_http_url_free(&old_purl);
1737 }
1738 if (new_purl) {
1739 php_http_url_free(&new_purl);
1740 }
1741 }
1742 zend_restore_error_handling(&zeh TSRMLS_CC);
1743 }
1744
1745 ZEND_BEGIN_ARG_INFO_EX(ai_HttpUrl_mod, 0, 0, 1)
1746 ZEND_ARG_INFO(0, more_url_parts)
1747 ZEND_ARG_INFO(0, flags)
1748 ZEND_END_ARG_INFO();
1749 PHP_METHOD(HttpUrl, mod)
1750 {
1751 zval *new_url = NULL;
1752 long flags = PHP_HTTP_URL_JOIN_PATH | PHP_HTTP_URL_JOIN_QUERY | PHP_HTTP_URL_SANITIZE_PATH;
1753 zend_error_handling zeh;
1754
1755 php_http_expect(SUCCESS == zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "z!|l", &new_url, &flags), invalid_arg, return);
1756
1757 if (flags & PHP_HTTP_URL_SILENT_ERRORS) {
1758 zend_replace_error_handling(EH_SUPPRESS, NULL, &zeh TSRMLS_CC);
1759 } else if (flags & PHP_HTTP_URL_IGNORE_ERRORS) {
1760 zend_replace_error_handling(EH_NORMAL, NULL, &zeh TSRMLS_CC);
1761 } else {
1762 zend_replace_error_handling(EH_THROW, php_http_exception_bad_url_class_entry, &zeh TSRMLS_CC);
1763 }
1764 {
1765 php_http_url_t *new_purl = NULL, *old_purl = NULL;
1766
1767 if (new_url) {
1768 new_purl = php_http_url_from_zval(new_url, flags TSRMLS_CC);
1769 if (!new_purl) {
1770 zend_restore_error_handling(&zeh TSRMLS_CC);
1771 return;
1772 }
1773 }
1774
1775 if ((old_purl = php_http_url_from_struct(HASH_OF(getThis())))) {
1776 php_http_url_t *res_purl;
1777
1778 ZVAL_OBJVAL(return_value, zend_objects_clone_obj(getThis() TSRMLS_CC), 0);
1779
1780 res_purl = php_http_url_mod(old_purl, new_purl, flags TSRMLS_CC);
1781 php_http_url_to_struct(res_purl, return_value TSRMLS_CC);
1782
1783 php_http_url_free(&res_purl);
1784 php_http_url_free(&old_purl);
1785 }
1786 if (new_purl) {
1787 php_http_url_free(&new_purl);
1788 }
1789 }
1790 zend_restore_error_handling(&zeh TSRMLS_CC);
1791 }
1792
1793 ZEND_BEGIN_ARG_INFO_EX(ai_HttpUrl_toString, 0, 0, 0)
1794 ZEND_END_ARG_INFO();
1795 PHP_METHOD(HttpUrl, toString)
1796 {
1797 if (SUCCESS == zend_parse_parameters_none()) {
1798 php_http_url_t *purl;
1799
1800 if ((purl = php_http_url_from_struct(HASH_OF(getThis())))) {
1801 char *str;
1802 size_t len;
1803
1804 php_http_url_to_string(purl, &str, &len, 0);
1805 php_http_url_free(&purl);
1806 RETURN_STRINGL(str, len, 0);
1807 }
1808 }
1809 RETURN_EMPTY_STRING();
1810 }
1811
1812 ZEND_BEGIN_ARG_INFO_EX(ai_HttpUrl_toArray, 0, 0, 0)
1813 ZEND_END_ARG_INFO();
1814 PHP_METHOD(HttpUrl, toArray)
1815 {
1816 php_http_url_t *purl;
1817
1818 if (SUCCESS != zend_parse_parameters_none()) {
1819 return;
1820 }
1821
1822 /* strip any non-URL properties */
1823 purl = php_http_url_from_struct(HASH_OF(getThis()));
1824 php_http_url_to_struct(purl, return_value TSRMLS_CC);
1825 php_http_url_free(&purl);
1826 }
1827
1828 static zend_function_entry php_http_url_methods[] = {
1829 PHP_ME(HttpUrl, __construct, ai_HttpUrl___construct, ZEND_ACC_PUBLIC|ZEND_ACC_CTOR)
1830 PHP_ME(HttpUrl, mod, ai_HttpUrl_mod, ZEND_ACC_PUBLIC)
1831 PHP_ME(HttpUrl, toString, ai_HttpUrl_toString, ZEND_ACC_PUBLIC)
1832 ZEND_MALIAS(HttpUrl, __toString, toString, ai_HttpUrl_toString, ZEND_ACC_PUBLIC)
1833 PHP_ME(HttpUrl, toArray, ai_HttpUrl_toArray, ZEND_ACC_PUBLIC)
1834 EMPTY_FUNCTION_ENTRY
1835 };
1836
1837 zend_class_entry *php_http_url_class_entry;
1838
1839 PHP_MINIT_FUNCTION(http_url)
1840 {
1841 zend_class_entry ce = {0};
1842
1843 INIT_NS_CLASS_ENTRY(ce, "http", "Url", php_http_url_methods);
1844 php_http_url_class_entry = zend_register_internal_class(&ce TSRMLS_CC);
1845
1846 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("scheme"), ZEND_ACC_PUBLIC TSRMLS_CC);
1847 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("user"), ZEND_ACC_PUBLIC TSRMLS_CC);
1848 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("pass"), ZEND_ACC_PUBLIC TSRMLS_CC);
1849 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("host"), ZEND_ACC_PUBLIC TSRMLS_CC);
1850 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("port"), ZEND_ACC_PUBLIC TSRMLS_CC);
1851 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("path"), ZEND_ACC_PUBLIC TSRMLS_CC);
1852 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("query"), ZEND_ACC_PUBLIC TSRMLS_CC);
1853 zend_declare_property_null(php_http_url_class_entry, ZEND_STRL("fragment"), ZEND_ACC_PUBLIC TSRMLS_CC);
1854
1855 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("REPLACE"), PHP_HTTP_URL_REPLACE TSRMLS_CC);
1856 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("JOIN_PATH"), PHP_HTTP_URL_JOIN_PATH TSRMLS_CC);
1857 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("JOIN_QUERY"), PHP_HTTP_URL_JOIN_QUERY TSRMLS_CC);
1858 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_USER"), PHP_HTTP_URL_STRIP_USER TSRMLS_CC);
1859 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_PASS"), PHP_HTTP_URL_STRIP_PASS TSRMLS_CC);
1860 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_AUTH"), PHP_HTTP_URL_STRIP_AUTH TSRMLS_CC);
1861 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_PORT"), PHP_HTTP_URL_STRIP_PORT TSRMLS_CC);
1862 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_PATH"), PHP_HTTP_URL_STRIP_PATH TSRMLS_CC);
1863 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_QUERY"), PHP_HTTP_URL_STRIP_QUERY TSRMLS_CC);
1864 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_FRAGMENT"), PHP_HTTP_URL_STRIP_FRAGMENT TSRMLS_CC);
1865 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STRIP_ALL"), PHP_HTTP_URL_STRIP_ALL TSRMLS_CC);
1866 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("FROM_ENV"), PHP_HTTP_URL_FROM_ENV TSRMLS_CC);
1867 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("SANITIZE_PATH"), PHP_HTTP_URL_SANITIZE_PATH TSRMLS_CC);
1868
1869 #ifdef PHP_HTTP_HAVE_WCHAR
1870 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("PARSE_MBLOC"), PHP_HTTP_URL_PARSE_MBLOC TSRMLS_CC);
1871 #endif
1872 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("PARSE_MBUTF8"), PHP_HTTP_URL_PARSE_MBUTF8 TSRMLS_CC);
1873 #if defined(PHP_HTTP_HAVE_IDN2) || defined(PHP_HTTP_HAVE_IDN) || defined(HAVE_UIDNA_IDNTOASCII)
1874 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("PARSE_TOIDN"), PHP_HTTP_URL_PARSE_TOIDN TSRMLS_CC);
1875 #endif
1876 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("PARSE_TOPCT"), PHP_HTTP_URL_PARSE_TOPCT TSRMLS_CC);
1877
1878 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("IGNORE_ERRORS"), PHP_HTTP_URL_IGNORE_ERRORS TSRMLS_CC);
1879 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("SILENT_ERRORS"), PHP_HTTP_URL_SILENT_ERRORS TSRMLS_CC);
1880
1881 zend_declare_class_constant_long(php_http_url_class_entry, ZEND_STRL("STDFLAGS"), PHP_HTTP_URL_STDFLAGS TSRMLS_CC);
1882
1883 return SUCCESS;
1884 }
1885
1886
1887 /*
1888 * Local variables:
1889 * tab-width: 4
1890 * c-basic-offset: 4
1891 * End:
1892 * vim600: noet sw=4 ts=4 fdm=marker
1893 * vim<600: noet sw=4 ts=4
1894 */
1895
Extended HTTP Support