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;
/** @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();
}
/**
* @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)
{
$stamp1 = microtime(true);
}
@curl_setopt($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));
$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);
$stamp1 = microtime(true);
if (Network::isUrlBlocked($url)) {
- $this->logger->info('Domain is blocked.'. ['url' => $url]);
+ $this->logger->info('Domain is blocked.' . ['url' => $url]);
return CurlResult::createErrorCurl($url);
}
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);
$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);
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;
}
+ /**
+ * 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
+ */
+ public function finalUrl(string $url, int $depth = 1, bool $fetchbody = false)
+ {
+ $url = Network::stripTrackingQueryParams($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_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", System::callstack());
+
+ 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_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", System::callstack());
+
+ 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;
+ }
+
/**
* Curl wrapper
*
*/
public function fetchUrlFull(string $url, bool $binary = false, int $timeout = 0, string $accept_content = '', string $cookiejar = '', int &$redirects = 0)
{
- return $this->curl(
+ return $this->get(
$url,
$binary,
[
$redirects
);
}
+
+ /**
+ * Returns the current UserAgent as a String
+ *
+ * @return string the UserAgent as a String
+ */
+ public function getUserAgent()
+ {
+ return
+ FRIENDICA_PLATFORM . " '" .
+ FRIENDICA_CODENAME . "' " .
+ FRIENDICA_VERSION . '-' .
+ DB_UPDATE_VERSION . '; ' .
+ $this->baseUrl;
+ }
}