3 * @copyright Copyright (C) 2020, Friendica
5 * @license GNU APGL version 3 or any later version
7 * This program is free software: you can redistribute it and/or modify
8 * it under the terms of the GNU Affero General Public License as
9 * published by the Free Software Foundation, either version 3 of the
10 * License, or (at your option) any later version.
12 * This program is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU Affero General Public License for more details.
17 * You should have received a copy of the GNU Affero General Public License
18 * along with this program. If not, see <https://www.gnu.org/licenses/>.
22 namespace Friendica\Network;
27 use Friendica\Core\Config\IConfig;
28 use Friendica\Core\System;
29 use Friendica\Util\Network;
30 use Friendica\Util\Profiler;
31 use Psr\Log\LoggerInterface;
34 * Performs HTTP requests to a given URL
36 class HTTPRequest implements IHTTPRequest
38 /** @var LoggerInterface */
47 public function __construct(LoggerInterface $logger, Profiler $profiler, IConfig $config, App\BaseURL $baseUrl)
49 $this->logger = $logger;
50 $this->profiler = $profiler;
51 $this->config = $config;
52 $this->baseUrl = $baseUrl->get();
57 * @throws HTTPException\InternalServerErrorException
59 public function head(string $url, array $opts = [])
61 $opts['nobody'] = true;
63 return $this->get($url, $opts);
69 * @param int $redirects The recursion counter for internal use - default 0
71 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
73 public function get(string $url, array $opts = [], int &$redirects = 0)
75 $stamp1 = microtime(true);
77 if (strlen($url) > 1000) {
78 $this->logger->debug('URL is longer than 1000 characters.', ['url' => $url, 'callstack' => System::callstack(20)]);
79 return CurlResult::createErrorCurl(substr($url, 0, 200));
83 $parts = parse_url($url);
84 $path_parts = explode('/', $parts['path'] ?? '');
85 foreach ($path_parts as $part) {
86 if (strlen($part) <> mb_strlen($part)) {
87 $parts2[] = rawurlencode($part);
92 $parts['path'] = implode('/', $parts2);
93 $url = Network::unparseURL($parts);
95 if (Network::isUrlBlocked($url)) {
96 $this->logger->info('Domain is blocked.', ['url' => $url]);
97 return CurlResult::createErrorCurl($url);
100 $ch = @curl_init($url);
102 if (($redirects > 8) || (!$ch)) {
103 return CurlResult::createErrorCurl($url);
106 @curl_setopt($ch, CURLOPT_HEADER, true);
108 if (!empty($opts['cookiejar'])) {
109 curl_setopt($ch, CURLOPT_COOKIEJAR, $opts["cookiejar"]);
110 curl_setopt($ch, CURLOPT_COOKIEFILE, $opts["cookiejar"]);
113 // These settings aren't needed. We're following the location already.
114 // @curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
115 // @curl_setopt($ch, CURLOPT_MAXREDIRS, 5);
117 if (!empty($opts['accept_content'])) {
121 ['Accept: ' . $opts['accept_content']]
125 if (!empty($opts['header'])) {
126 curl_setopt($ch, CURLOPT_HTTPHEADER, $opts['header']);
129 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
130 @curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
132 $range = intval($this->config->get('system', 'curl_range_bytes', 0));
135 @curl_setopt($ch, CURLOPT_RANGE, '0-' . $range);
138 // Without this setting it seems as if some webservers send compressed content
139 // This seems to confuse curl so that it shows this uncompressed.
140 /// @todo We could possibly set this value to "gzip" or something similar
141 curl_setopt($ch, CURLOPT_ENCODING, '');
143 if (!empty($opts['headers'])) {
144 $this->logger->notice('Wrong option \'headers\' used.');
145 @curl_setopt($ch, CURLOPT_HTTPHEADER, $opts['headers']);
148 if (!empty($opts['nobody'])) {
149 @curl_setopt($ch, CURLOPT_NOBODY, $opts['nobody']);
152 @curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
154 if (!empty($opts['timeout'])) {
155 @curl_setopt($ch, CURLOPT_TIMEOUT, $opts['timeout']);
157 $curl_time = $this->config->get('system', 'curl_timeout', 60);
158 @curl_setopt($ch, CURLOPT_TIMEOUT, intval($curl_time));
161 // by default we will allow self-signed certs
162 // but you can override this
164 $check_cert = $this->config->get('system', 'verifyssl');
165 @curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, (($check_cert) ? true : false));
168 @curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 2);
171 $proxy = $this->config->get('system', 'proxy');
173 if (!empty($proxy)) {
174 @curl_setopt($ch, CURLOPT_HTTPPROXYTUNNEL, 1);
175 @curl_setopt($ch, CURLOPT_PROXY, $proxy);
176 $proxyuser = $this->config->get('system', 'proxyuser');
178 if (!empty($proxyuser)) {
179 @curl_setopt($ch, CURLOPT_PROXYUSERPWD, $proxyuser);
183 if ($this->config->get('system', 'ipv4_resolve', false)) {
184 curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
187 $s = @curl_exec($ch);
188 $curl_info = @curl_getinfo($ch);
190 // Special treatment for HTTP Code 416
191 // See https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/416
192 if (($curl_info['http_code'] == 416) && ($range > 0)) {
193 @curl_setopt($ch, CURLOPT_RANGE, '');
194 $s = @curl_exec($ch);
195 $curl_info = @curl_getinfo($ch);
198 $curlResponse = new CurlResult($url, $s, $curl_info, curl_errno($ch), curl_error($ch));
200 if (!Network::isRedirectBlocked($url) && $curlResponse->isRedirectUrl()) {
202 $this->logger->notice('Curl redirect.', ['url' => $url, 'to' => $curlResponse->getRedirectUrl()]);
204 return $this->get($curlResponse->getRedirectUrl(), $opts, $redirects);
209 $this->profiler->saveTimestamp($stamp1, 'network');
211 return $curlResponse;
217 * @param int $redirects The recursion counter for internal use - default 0
219 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
221 public function post(string $url, $params, array $headers = [], int $timeout = 0, int &$redirects = 0)
223 $stamp1 = microtime(true);
225 if (Network::isUrlBlocked($url)) {
226 $this->logger->info('Domain is blocked.' . ['url' => $url]);
227 return CurlResult::createErrorCurl($url);
230 $ch = curl_init($url);
232 if (($redirects > 8) || (!$ch)) {
233 return CurlResult::createErrorCurl($url);
236 $this->logger->debug('Post_url: start.', ['url' => $url]);
238 curl_setopt($ch, CURLOPT_HEADER, true);
239 curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
240 curl_setopt($ch, CURLOPT_POST, 1);
241 curl_setopt($ch, CURLOPT_POSTFIELDS, $params);
242 curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
244 if ($this->config->get('system', 'ipv4_resolve', false)) {
245 curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
248 @curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
250 if (intval($timeout)) {
251 curl_setopt($ch, CURLOPT_TIMEOUT, $timeout);
253 $curl_time = $this->config->get('system', 'curl_timeout', 60);
254 curl_setopt($ch, CURLOPT_TIMEOUT, intval($curl_time));
257 if (!empty($headers)) {
258 curl_setopt($ch, CURLOPT_HTTPHEADER, $headers);
261 $check_cert = $this->config->get('system', 'verifyssl');
262 curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, (($check_cert) ? true : false));
265 @curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 2);
268 $proxy = $this->config->get('system', 'proxy');
270 if (!empty($proxy)) {
271 curl_setopt($ch, CURLOPT_HTTPPROXYTUNNEL, 1);
272 curl_setopt($ch, CURLOPT_PROXY, $proxy);
273 $proxyuser = $this->config->get('system', 'proxyuser');
274 if (!empty($proxyuser)) {
275 curl_setopt($ch, CURLOPT_PROXYUSERPWD, $proxyuser);
279 // don't let curl abort the entire application
280 // if it throws any errors.
282 $s = @curl_exec($ch);
284 $curl_info = curl_getinfo($ch);
286 $curlResponse = new CurlResult($url, $s, $curl_info, curl_errno($ch), curl_error($ch));
288 if (!Network::isRedirectBlocked($url) && $curlResponse->isRedirectUrl()) {
290 $this->logger->info('Post redirect.', ['url' => $url, 'to' => $curlResponse->getRedirectUrl()]);
292 return $this->post($curlResponse->getRedirectUrl(), $params, $headers, $redirects, $timeout);
297 $this->profiler->saveTimestamp($stamp1, 'network');
299 // Very old versions of Lighttpd don't like the "Expect" header, so we remove it when needed
300 if ($curlResponse->getReturnCode() == 417) {
303 if (empty($headers)) {
304 $headers = ['Expect:'];
306 if (!in_array('Expect:', $headers)) {
307 array_push($headers, 'Expect:');
310 $this->logger->info('Server responds with 417, applying workaround', ['url' => $url]);
311 return $this->post($url, $params, $headers, $redirects, $timeout);
314 $this->logger->debug('Post_url: End.', ['url' => $url]);
316 return $curlResponse;
322 public function finalUrl(string $url, int $depth = 1, bool $fetchbody = false)
324 if (Network::isUrlBlocked($url)) {
325 $this->logger->info('Domain is blocked.', ['url' => $url]);
329 if (Network::isRedirectBlocked($url)) {
330 $this->logger->info('Domain should not be redirected.', ['url' => $url]);
334 $url = Network::stripTrackingQueryParams($url);
340 $url = trim($url, "'");
342 $stamp1 = microtime(true);
345 curl_setopt($ch, CURLOPT_URL, $url);
346 curl_setopt($ch, CURLOPT_HEADER, 1);
347 curl_setopt($ch, CURLOPT_NOBODY, 1);
348 curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
349 curl_setopt($ch, CURLOPT_TIMEOUT, 10);
350 curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
351 curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
354 $curl_info = @curl_getinfo($ch);
355 $http_code = $curl_info['http_code'];
358 $this->profiler->saveTimestamp($stamp1, "network");
360 if ($http_code == 0) {
364 if (in_array($http_code, ['301', '302'])) {
365 if (!empty($curl_info['redirect_url'])) {
366 return $this->finalUrl($curl_info['redirect_url'], ++$depth, $fetchbody);
367 } elseif (!empty($curl_info['location'])) {
368 return $this->finalUrl($curl_info['location'], ++$depth, $fetchbody);
372 // Check for redirects in the meta elements of the body if there are no redirects in the header.
374 return $this->finalUrl($url, ++$depth, true);
377 // if the file is too large then exit
378 if ($curl_info["download_content_length"] > 1000000) {
382 // if it isn't a HTML file then exit
383 if (!empty($curl_info["content_type"]) && !strstr(strtolower($curl_info["content_type"]), "html")) {
387 $stamp1 = microtime(true);
390 curl_setopt($ch, CURLOPT_URL, $url);
391 curl_setopt($ch, CURLOPT_HEADER, 0);
392 curl_setopt($ch, CURLOPT_NOBODY, 0);
393 curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
394 curl_setopt($ch, CURLOPT_TIMEOUT, 10);
395 curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
396 curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
398 $body = curl_exec($ch);
401 $this->profiler->saveTimestamp($stamp1, "network");
403 if (trim($body) == "") {
407 // Check for redirect in meta elements
408 $doc = new DOMDocument();
409 @$doc->loadHTML($body);
411 $xpath = new DomXPath($doc);
413 $list = $xpath->query("//meta[@content]");
414 foreach ($list as $node) {
416 if ($node->attributes->length) {
417 foreach ($node->attributes as $attribute) {
418 $attr[$attribute->name] = $attribute->value;
422 if (@$attr["http-equiv"] == 'refresh') {
423 $path = $attr["content"];
424 $pathinfo = explode(";", $path);
425 foreach ($pathinfo as $value) {
426 if (substr(strtolower($value), 0, 4) == "url=") {
427 return $this->finalUrl(substr($value, 4), ++$depth);
439 * @param int $redirects The recursion counter for internal use - default 0
441 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
443 public function fetch(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
445 $ret = $this->fetchFull($url, $timeout, $accept_content, $cookiejar, $redirects);
447 return $ret->getBody();
453 * @param int $redirects The recursion counter for internal use - default 0
455 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
457 public function fetchFull(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
462 'timeout' => $timeout,
463 'accept_content' => $accept_content,
464 'cookiejar' => $cookiejar
473 public function getUserAgent()
476 FRIENDICA_PLATFORM . " '" .
477 FRIENDICA_CODENAME . "' " .
478 FRIENDICA_VERSION . '-' .
479 DB_UPDATE_VERSION . '; ' .