]> git.mxchange.org Git - friendica.git/blobdiff - src/Network/HTTPRequest.php
Merge branch '2021.03-rc' into copyright-2021
[friendica.git] / src / Network / HTTPRequest.php
index 3e5091e076f07be7a9afb82a13d5ee2fc2123968..74cf9cd0554d01a62beb0a1486b391bb734f4f69 100644 (file)
@@ -1,6 +1,6 @@
 <?php
 /**
- * @copyright Copyright (C) 2020, Friendica
+ * @copyright Copyright (C) 2010-2021, the Friendica project
  *
  * @license GNU APGL version 3 or any later version
  *
@@ -33,7 +33,7 @@ use Psr\Log\LoggerInterface;
 /**
  * Performs HTTP requests to a given URL
  */
-class HTTPRequest
+class HTTPRequest implements IHTTPRequest
 {
        /** @var LoggerInterface */
        private $logger;
@@ -52,31 +52,31 @@ class HTTPRequest
                $this->baseUrl  = $baseUrl->get();
        }
 
+       /** {@inheritDoc}
+        *
+        * @throws HTTPException\InternalServerErrorException
+        */
+       public function head(string $url, array $opts = [])
+       {
+               $opts['nobody'] = true;
+
+               return $this->get($url, $opts);
+       }
+
        /**
-        * fetches an URL.
+        * {@inheritDoc}
         *
-        * @param string $url        URL to fetch
-        * @param bool   $binary     default false
-        *                           TRUE if asked to return binary results (file download)
-        * @param array  $opts       (optional parameters) assoziative array with:
-        *                           'accept_content' => supply Accept: header with 'accept_content' as the value
-        *                           'timeout' => int Timeout in seconds, default system config value or 60 seconds
-        *                           'http_auth' => username:password
-        *                           'novalidate' => do not validate SSL certs, default is to validate using our CA list
-        *                           'nobody' => only return the header
-        *                           'cookiejar' => path to cookie jar file
-        *                           'header' => header array
-        * @param int    $redirects  The recursion counter for internal use - default 0
+        * @param int $redirects The recursion counter for internal use - default 0
         *
-        * @return CurlResult
         * @throws \Friendica\Network\HTTPException\InternalServerErrorException
         */
-       public function curl(string $url, bool $binary = false, array $opts = [], int &$redirects = 0)
+       public function get(string $url, array $opts = [], int &$redirects = 0)
        {
                $stamp1 = microtime(true);
 
                if (strlen($url) > 1000) {
                        $this->logger->debug('URL is longer than 1000 characters.', ['url' => $url, 'callstack' => System::callstack(20)]);
+                       $this->profiler->saveTimestamp($stamp1, 'network');
                        return CurlResult::createErrorCurl(substr($url, 0, 200));
                }
 
@@ -95,12 +95,14 @@ class HTTPRequest
 
                if (Network::isUrlBlocked($url)) {
                        $this->logger->info('Domain is blocked.', ['url' => $url]);
+                       $this->profiler->saveTimestamp($stamp1, 'network');
                        return CurlResult::createErrorCurl($url);
                }
 
                $ch = @curl_init($url);
 
                if (($redirects > 8) || (!$ch)) {
+                       $this->profiler->saveTimestamp($stamp1, 'network');
                        return CurlResult::createErrorCurl($url);
                }
 
@@ -142,6 +144,7 @@ class HTTPRequest
                curl_setopt($ch, CURLOPT_ENCODING, '');
 
                if (!empty($opts['headers'])) {
+                       $this->logger->notice('Wrong option \'headers\' used.');
                        @curl_setopt($ch, CURLOPT_HTTPHEADER, $opts['headers']);
                }
 
@@ -149,6 +152,8 @@ class HTTPRequest
                        @curl_setopt($ch, CURLOPT_NOBODY, $opts['nobody']);
                }
 
+               @curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
+
                if (!empty($opts['timeout'])) {
                        @curl_setopt($ch, CURLOPT_TIMEOUT, $opts['timeout']);
                } else {
@@ -182,13 +187,6 @@ class HTTPRequest
                        curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
                }
 
-               if ($binary) {
-                       @curl_setopt($ch, CURLOPT_BINARYTRANSFER, 1);
-               }
-
-               // don't let curl abort the entire application
-               // if it throws any errors.
-
                $s         = @curl_exec($ch);
                $curl_info = @curl_getinfo($ch);
 
@@ -202,30 +200,26 @@ class HTTPRequest
 
                $curlResponse = new CurlResult($url, $s, $curl_info, curl_errno($ch), curl_error($ch));
 
-               if ($curlResponse->isRedirectUrl()) {
+               if (!Network::isRedirectBlocked($url) && $curlResponse->isRedirectUrl()) {
                        $redirects++;
                        $this->logger->notice('Curl redirect.', ['url' => $url, 'to' => $curlResponse->getRedirectUrl()]);
                        @curl_close($ch);
-                       return $this->curl($curlResponse->getRedirectUrl(), $binary, $opts, $redirects);
+                       $this->profiler->saveTimestamp($stamp1, 'network');
+                       return $this->get($curlResponse->getRedirectUrl(), $opts, $redirects);
                }
 
                @curl_close($ch);
 
-               $this->profiler->saveTimestamp($stamp1, 'network', System::callstack());
+               $this->profiler->saveTimestamp($stamp1, 'network');
 
                return $curlResponse;
        }
 
        /**
-        * Send POST request to $url
+        * {@inheritDoc}
         *
-        * @param string $url       URL to post
-        * @param mixed  $params    array of POST variables
-        * @param array  $headers   HTTP headers
-        * @param int    $redirects Recursion counter for internal use - default = 0
-        * @param int    $timeout   The timeout in seconds, default system config value or 60 seconds
+        * @param int $redirects The recursion counter for internal use - default 0
         *
-        * @return CurlResult The content
         * @throws \Friendica\Network\HTTPException\InternalServerErrorException
         */
        public function post(string $url, $params, array $headers = [], int $timeout = 0, int &$redirects = 0)
@@ -234,12 +228,14 @@ class HTTPRequest
 
                if (Network::isUrlBlocked($url)) {
                        $this->logger->info('Domain is blocked.' . ['url' => $url]);
+                       $this->profiler->saveTimestamp($stamp1, 'network');
                        return CurlResult::createErrorCurl($url);
                }
 
                $ch = curl_init($url);
 
                if (($redirects > 8) || (!$ch)) {
+                       $this->profiler->saveTimestamp($stamp1, 'network');
                        return CurlResult::createErrorCurl($url);
                }
 
@@ -255,6 +251,8 @@ class HTTPRequest
                        curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
                }
 
+               @curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
+
                if (intval($timeout)) {
                        curl_setopt($ch, CURLOPT_TIMEOUT, $timeout);
                } else {
@@ -293,16 +291,17 @@ class HTTPRequest
 
                $curlResponse = new CurlResult($url, $s, $curl_info, curl_errno($ch), curl_error($ch));
 
-               if ($curlResponse->isRedirectUrl()) {
+               if (!Network::isRedirectBlocked($url) && $curlResponse->isRedirectUrl()) {
                        $redirects++;
                        $this->logger->info('Post redirect.', ['url' => $url, 'to' => $curlResponse->getRedirectUrl()]);
                        curl_close($ch);
+                       $this->profiler->saveTimestamp($stamp1, 'network');
                        return $this->post($curlResponse->getRedirectUrl(), $params, $headers, $redirects, $timeout);
                }
 
                curl_close($ch);
 
-               $this->profiler->saveTimestamp($stamp1, 'network', System::callstack());
+               $this->profiler->saveTimestamp($stamp1, 'network');
 
                // Very old versions of Lighttpd don't like the "Expect" header, so we remove it when needed
                if ($curlResponse->getReturnCode() == 417) {
@@ -325,23 +324,20 @@ class HTTPRequest
        }
 
        /**
-        * Returns the original URL of the provided URL
-        *
-        * This function strips tracking query params and follows redirections, either
-        * through HTTP code or meta refresh tags. Stops after 10 redirections.
-        *
-        * @todo  Remove the $fetchbody parameter that generates an extraneous HEAD request
-        *
-        * @see   ParseUrl::getSiteinfo
-        *
-        * @param string $url       A user-submitted URL
-        * @param int    $depth     The current redirection recursion level (internal)
-        * @param bool   $fetchbody Wether to fetch the body or not after the HEAD requests
-        * @return string A canonical URL
-        * @throws \Friendica\Network\HTTPException\InternalServerErrorException
+        * {@inheritDoc}
         */
        public function finalUrl(string $url, int $depth = 1, bool $fetchbody = false)
        {
+               if (Network::isUrlBlocked($url)) {
+                       $this->logger->info('Domain is blocked.', ['url' => $url]);
+                       return $url;
+               }
+
+               if (Network::isRedirectBlocked($url)) {
+                       $this->logger->info('Domain should not be redirected.', ['url' => $url]);
+                       return $url;
+               }
+
                $url = Network::stripTrackingQueryParams($url);
 
                if ($depth > 10) {
@@ -356,6 +352,7 @@ class HTTPRequest
                curl_setopt($ch, CURLOPT_URL, $url);
                curl_setopt($ch, CURLOPT_HEADER, 1);
                curl_setopt($ch, CURLOPT_NOBODY, 1);
+               curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
                curl_setopt($ch, CURLOPT_TIMEOUT, 10);
                curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
                curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
@@ -365,7 +362,7 @@ class HTTPRequest
                $http_code = $curl_info['http_code'];
                curl_close($ch);
 
-               $this->profiler->saveTimestamp($stamp1, "network", System::callstack());
+               $this->profiler->saveTimestamp($stamp1, "network");
 
                if ($http_code == 0) {
                        return $url;
@@ -400,6 +397,7 @@ class HTTPRequest
                curl_setopt($ch, CURLOPT_URL, $url);
                curl_setopt($ch, CURLOPT_HEADER, 0);
                curl_setopt($ch, CURLOPT_NOBODY, 0);
+               curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
                curl_setopt($ch, CURLOPT_TIMEOUT, 10);
                curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
                curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
@@ -407,7 +405,7 @@ class HTTPRequest
                $body = curl_exec($ch);
                curl_close($ch);
 
-               $this->profiler->saveTimestamp($stamp1, "network", System::callstack());
+               $this->profiler->saveTimestamp($stamp1, "network");
 
                if (trim($body) == "") {
                        return $url;
@@ -443,52 +441,30 @@ class HTTPRequest
        }
 
        /**
-        * Curl wrapper
+        * {@inheritDoc}
         *
-        * If binary flag is true, return binary results.
-        * Set the cookiejar argument to a string (e.g. "/tmp/friendica-cookies.txt")
-        * to preserve cookies from one request to the next.
+        * @param int $redirects The recursion counter for internal use - default 0
         *
-        * @param string $url             URL to fetch
-        * @param bool   $binary          default false
-        *                                TRUE if asked to return binary results (file download)
-        * @param int    $timeout         Timeout in seconds, default system config value or 60 seconds
-        * @param string $accept_content  supply Accept: header with 'accept_content' as the value
-        * @param string $cookiejar       Path to cookie jar file
-        * @param int    $redirects       The recursion counter for internal use - default 0
-        *
-        * @return string The fetched content
         * @throws \Friendica\Network\HTTPException\InternalServerErrorException
         */
-       public function fetchUrl(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
+       public function fetch(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
        {
-               $ret = $this->fetchUrlFull($url, $binary, $timeout, $accept_content, $cookiejar, $redirects);
+               $ret = $this->fetchFull($url, $timeout, $accept_content, $cookiejar, $redirects);
 
                return $ret->getBody();
        }
 
        /**
-        * Curl wrapper with array of return values.
-        *
-        * Inner workings and parameters are the same as @ref fetchUrl but returns an array with
-        * all the information collected during the fetch.
+        * {@inheritDoc}
         *
-        * @param string $url             URL to fetch
-        * @param bool   $binary          default false
-        *                                TRUE if asked to return binary results (file download)
-        * @param int    $timeout         Timeout in seconds, default system config value or 60 seconds
-        * @param string $accept_content  supply Accept: header with 'accept_content' as the value
-        * @param string $cookiejar       Path to cookie jar file
-        * @param int    $redirects       The recursion counter for internal use - default 0
+        * @param int $redirects The recursion counter for internal use - default 0
         *
-        * @return CurlResult With all relevant information, 'body' contains the actual fetched content.
         * @throws \Friendica\Network\HTTPException\InternalServerErrorException
         */
-       public function fetchUrlFull(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
+       public function fetchFull(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
        {
-               return $this->curl(
+               return $this->get(
                        $url,
-                       $binary,
                        [
                                'timeout'        => $timeout,
                                'accept_content' => $accept_content,
@@ -499,9 +475,7 @@ class HTTPRequest
        }
 
        /**
-        * Returns the current UserAgent as a String
-        *
-        * @return string the UserAgent as a String
+        * {@inheritDoc}
         */
        public function getUserAgent()
        {