X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FNetwork%2FHTTPRequest.php;h=e4ff041039ed8ba225022000faa6af8f2c159af7;hb=722aada460aa023212b7d97932fe338fe54861a8;hp=131fac8a78f896e2baf82ca24d5abf896884045e;hpb=8793096c16ea97cf73fb4f04199e4aab8574dcd1;p=friendica.git diff --git a/src/Network/HTTPRequest.php b/src/Network/HTTPRequest.php index 131fac8a78..e4ff041039 100644 --- a/src/Network/HTTPRequest.php +++ b/src/Network/HTTPRequest.php @@ -21,9 +21,10 @@ namespace Friendica\Network; +use DOMDocument; +use DomXPath; use Friendica\App; use Friendica\Core\Config\IConfig; -use Friendica\Core\Logger; use Friendica\Core\System; use Friendica\Util\Network; use Friendica\Util\Profiler; @@ -32,7 +33,7 @@ use Psr\Log\LoggerInterface; /** * Performs HTTP requests to a given URL */ -class HTTPRequest +class HTTPRequest implements IHTTPRequest { /** @var LoggerInterface */ private $logger; @@ -41,36 +42,24 @@ class HTTPRequest /** @var IConfig */ private $config; /** @var string */ - private $userAgent; + private $baseUrl; - public function __construct(LoggerInterface $logger, Profiler $profiler, IConfig $config, App $a) + public function __construct(LoggerInterface $logger, Profiler $profiler, IConfig $config, App\BaseURL $baseUrl) { - $this->logger = $logger; - $this->profiler = $profiler; - $this->config = $config; - $this->userAgent = $a->getUserAgent(); + $this->logger = $logger; + $this->profiler = $profiler; + $this->config = $config; + $this->baseUrl = $baseUrl->get(); } /** - * 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, bool $binary = false, array $opts = [], int &$redirects = 0) { $stamp1 = microtime(true); @@ -127,7 +116,7 @@ class HTTPRequest } @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); - @curl_setopt($ch, CURLOPT_USERAGENT, $this->userAgent); + @curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent()); $range = intval($this->config->get('system', 'curl_range_bytes', 0)); @@ -148,6 +137,8 @@ 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 { @@ -201,30 +192,25 @@ 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 self::curl($curlResponse->getRedirectUrl(), $binary, $opts, $redirects); + return $this->get($curlResponse->getRedirectUrl(), $binary, $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) @@ -232,7 +218,7 @@ class HTTPRequest $stamp1 = microtime(true); if (Network::isUrlBlocked($url)) { - $this->logger->info('Domain is blocked.'. ['url' => $url]); + $this->logger->info('Domain is blocked.' . ['url' => $url]); return CurlResult::createErrorCurl($url); } @@ -248,12 +234,14 @@ class HTTPRequest curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); curl_setopt($ch, CURLOPT_POST, 1); curl_setopt($ch, CURLOPT_POSTFIELDS, $params); - curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent()); + curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent()); if ($this->config->get('system', 'ipv4_resolve', false)) { curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4); } + @curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10); + if (intval($timeout)) { curl_setopt($ch, CURLOPT_TIMEOUT, $timeout); } else { @@ -292,16 +280,16 @@ 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); - return self::post($curlResponse->getRedirectUrl(), $params, $headers, $redirects, $timeout); + 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 ($curlResponse->getReturnCode() == 417) { @@ -314,60 +302,156 @@ class HTTPRequest array_push($headers, 'Expect:'); } } - Logger::info('Server responds with 417, applying workaround', ['url' => $url]); - return self::post($url, $params, $headers, $redirects, $timeout); + $this->logger->info('Server responds with 417, applying workaround', ['url' => $url]); + return $this->post($url, $params, $headers, $redirects, $timeout); } - Logger::log('post_url: end ' . $url, Logger::DATA); + $this->logger->debug('Post_url: End.', ['url' => $url]); return $curlResponse; } /** - * Curl wrapper - * - * 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. + * {@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) { + return $url; + } + + $url = trim($url, "'"); + + $stamp1 = microtime(true); + + $ch = curl_init(); + curl_setopt($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, $this->getUserAgent()); + + curl_exec($ch); + $curl_info = @curl_getinfo($ch); + $http_code = $curl_info['http_code']; + curl_close($ch); + + $this->profiler->saveTimestamp($stamp1, "network"); + + if ($http_code == 0) { + return $url; + } + + if (in_array($http_code, ['301', '302'])) { + if (!empty($curl_info['redirect_url'])) { + return $this->finalUrl($curl_info['redirect_url'], ++$depth, $fetchbody); + } elseif (!empty($curl_info['location'])) { + return $this->finalUrl($curl_info['location'], ++$depth, $fetchbody); + } + } + + // Check for redirects in the meta elements of the body if there are no redirects in the header. + if (!$fetchbody) { + return $this->finalUrl($url, ++$depth, true); + } + + // if the file is too large then exit + if ($curl_info["download_content_length"] > 1000000) { + return $url; + } + + // if it isn't a HTML file then exit + if (!empty($curl_info["content_type"]) && !strstr(strtolower($curl_info["content_type"]), "html")) { + return $url; + } + + $stamp1 = microtime(true); + + $ch = curl_init(); + curl_setopt($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()); + + $body = curl_exec($ch); + curl_close($ch); + + $this->profiler->saveTimestamp($stamp1, "network"); + + if (trim($body) == "") { + return $url; + } + + // Check for redirect in meta elements + $doc = new DOMDocument(); + @$doc->loadHTML($body); + + $xpath = new DomXPath($doc); + + $list = $xpath->query("//meta[@content]"); + foreach ($list as $node) { + $attr = []; + if ($node->attributes->length) { + foreach ($node->attributes as $attribute) { + $attr[$attribute->name] = $attribute->value; + } + } + + if (@$attr["http-equiv"] == 'refresh') { + $path = $attr["content"]; + $pathinfo = explode(";", $path); + foreach ($pathinfo as $value) { + if (substr(strtolower($value), 0, 4) == "url=") { + return $this->finalUrl(substr($value, 4), ++$depth); + } + } + } + } + + return $url; + } + + /** + * {@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 string The fetched content * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ - public static function fetchUrl(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0) + public function fetch(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0) { - $ret = self::fetchUrlFull($url, $binary, $timeout, $accept_content, $cookiejar, $redirects); + $ret = $this->fetchFull($url, $binary, $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 static function fetchUrlFull(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0) + public function fetchFull(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0) { - return self::curl( + return $this->get( $url, $binary, [ @@ -378,4 +462,17 @@ class HTTPRequest $redirects ); } + + /** + * {@inheritDoc} + */ + public function getUserAgent() + { + return + FRIENDICA_PLATFORM . " '" . + FRIENDICA_CODENAME . "' " . + FRIENDICA_VERSION . '-' . + DB_UPDATE_VERSION . '; ' . + $this->baseUrl; + } }