Use Guzzle for HTTPRequest and Result

src/Network/GuzzleResponse.php

@@ -0,0 +1,152 @@
+<?php
+/**
+ * @copyright Copyright (C) 2020, Friendica
+ *
+ * @license GNU AGPL version 3 or any later version
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ *
+ */
+
+namespace Friendica\Network;
+
+use Friendica\Core\Logger;
+use Friendica\Core\System;
+use Friendica\Network\HTTPException\NotImplementedException;
+use GuzzleHttp\Psr7\Response;
+use Psr\Http\Message\ResponseInterface;
+
+/**
+ * A content wrapper class for Guzzle call results
+ */
+class GuzzleResponse extends Response implements IHTTPResult, ResponseInterface
+{
+	/** @var string The URL */
+	private $url;
+	/** @var boolean */
+	private $isTimeout;
+	/** @var boolean */
+	private $isSuccess;
+	/**
+	 * @var int the error number or 0 (zero) if no error
+	 */
+	private $errorNumber;
+
+	/**
+	 * @var string the error message or '' (the empty string) if no
+	 */
+	private $error;
+
+	public function __construct(ResponseInterface $response, string $url, $errorNumber = 0, $error = '')
+	{
+		parent::__construct($response->getStatusCode(), $response->getHeaders(), $response->getBody(), $response->getProtocolVersion(), $response->getReasonPhrase());
+		$this->url         = $url;
+		$this->error       = $error;
+		$this->errorNumber = $errorNumber;
+
+		$this->checkSuccess();
+	}
+
+	private function checkSuccess()
+	{
+		$this->isSuccess = ($this->getStatusCode() >= 200 && $this->getStatusCode() <= 299) || $this->errorNumber == 0;
+
+		// Everything higher or equal 400 is not a success
+		if ($this->getReturnCode() >= 400) {
+			$this->isSuccess = false;
+		}
+
+		if (!$this->isSuccess) {
+			Logger::notice('http error', ['url' => $this->url, 'code' => $this->getReturnCode(), 'error'  => $this->error, 'callstack' => System::callstack(20)]);
+			Logger::debug('debug', ['info' => $this->getHeaders()]);
+		}
+
+		if (!$this->isSuccess && $this->errorNumber == CURLE_OPERATION_TIMEDOUT) {
+			$this->isTimeout = true;
+		} else {
+			$this->isTimeout = false;
+		}
+	}
+
+	/** {@inheritDoc} */
+	public function getReturnCode()
+	{
+		return $this->getStatusCode();
+	}
+
+	/** {@inheritDoc} */
+	public function getContentType()
+	{
+		return $this->getHeader('Content-Type');
+	}
+
+	/** {@inheritDoc} */
+	public function inHeader(string $field)
+	{
+		return $this->hasHeader($field);
+	}
+
+	/** {@inheritDoc} */
+	public function getHeaderArray()
+	{
+		return $this->getHeaders();
+	}
+
+	/** {@inheritDoc} */
+	public function isSuccess()
+	{
+		return $this->isSuccess;
+	}
+
+	/** {@inheritDoc} */
+	public function getUrl()
+	{
+		return $this->url;
+	}
+
+	/** {@inheritDoc} */
+	public function getRedirectUrl()
+	{
+		return $this->url;
+	}
+
+	public function getInfo()
+	{
+		// TODO: Implement getInfo() method.
+	}
+
+	/** {@inheritDoc} */
+	public function isRedirectUrl()
+	{
+		throw new NotImplementedException();
+	}
+
+	/** {@inheritDoc} */
+	public function getErrorNumber()
+	{
+		return $this->errorNumber;
+	}
+
+	/** {@inheritDoc} */
+	public function getError()
+	{
+		return $this->error;
+	}
+
+	/** {@inheritDoc} */
+	public function isTimeout()
+	{
+		return $this->isTimeout;
+	}
+}

src/Network/HTTPRequest.php

@@ -28,6 +28,11 @@ use Friendica\Core\Config\IConfig;
 use Friendica\Core\System;
 use Friendica\Util\Network;
 use Friendica\Util\Profiler;
+use GuzzleHttp\Client;
+use GuzzleHttp\Exception\RequestException;
+use Psr\Http\Message\RequestInterface;
+use Psr\Http\Message\ResponseInterface;
+use Psr\Http\Message\UriInterface;
 use Psr\Log\LoggerInterface;
 
 /**
@@ -65,12 +70,8 @@ class HTTPRequest implements IHTTPRequest
 
 	/**
 	 * {@inheritDoc}
-	 *
-	 * @param int $redirects The recursion counter for internal use - default 0
-	 *
-	 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 	 */
-	public function get(string $url, array $opts = [], &$redirects = 0)
+	public function get(string $url, bool $binary = false, array $opts = [])
 	{
 		$this->profiler->startRecording('network');
 
@@ -103,120 +104,131 @@ class HTTPRequest implements IHTTPRequest
 			return CurlResult::createErrorCurl($url);
 		}
 
-		$ch = @curl_init($url);
+		$curlOptions = [];
 
-		if (($redirects > 8) || (!$ch)) {
+		$curlOptions[CURLOPT_HEADER] = true;
-			$this->profiler->stopRecording();
-			return CurlResult::createErrorCurl($url);
-		}
-
-		@curl_setopt($ch, CURLOPT_HEADER, true);
 
 		if (!empty($opts['cookiejar'])) {
-			curl_setopt($ch, CURLOPT_COOKIEJAR, $opts["cookiejar"]);
+			$curlOptions[CURLOPT_COOKIEJAR] = $opts["cookiejar"];
-			curl_setopt($ch, CURLOPT_COOKIEFILE, $opts["cookiejar"]);
+			$curlOptions[CURLOPT_COOKIEFILE] = $opts["cookiejar"];
 		}
 
 		// These settings aren't needed. We're following the location already.
-		//	@curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
+		//	$curlOptions[CURLOPT_FOLLOWLOCATION] =true;
-		//	@curl_setopt($ch, CURLOPT_MAXREDIRS, 5);
+		//	$curlOptions[CURLOPT_MAXREDIRS] = 5;
 
 		if (!empty($opts['accept_content'])) {
-			curl_setopt(
+			if (empty($curlOptions[CURLOPT_HTTPHEADER])) {
-				$ch,
+				$curlOptions[CURLOPT_HTTPHEADER] = [];
-				CURLOPT_HTTPHEADER,
+			}
-				['Accept: ' . $opts['accept_content']]
+			array_push($curlOptions[CURLOPT_HTTPHEADER], 'Accept: ' . $opts['accept_content']);
-			);
 		}
 
 		if (!empty($opts['header'])) {
-			curl_setopt($ch, CURLOPT_HTTPHEADER, $opts['header']);
+			if (empty($curlOptions[CURLOPT_HTTPHEADER])) {
+				$curlOptions[CURLOPT_HTTPHEADER] = [];
+			}
+			$curlOptions[CURLOPT_HTTPHEADER] = array_merge($opts['header'], $curlOptions[CURLOPT_HTTPHEADER]);
 		}
 
-		@curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
+		$curlOptions[CURLOPT_RETURNTRANSFER] = true;
-		@curl_setopt($ch, CURLOPT_USERAGENT, $this->getUserAgent());
+		$curlOptions[CURLOPT_USERAGENT] = $this->getUserAgent();
 
 		$range = intval($this->config->get('system', 'curl_range_bytes', 0));
 
 		if ($range > 0) {
-			@curl_setopt($ch, CURLOPT_RANGE, '0-' . $range);
+			$curlOptions[CURLOPT_RANGE] = '0-' . $range;
 		}
 
 		// Without this setting it seems as if some webservers send compressed content
 		// This seems to confuse curl so that it shows this uncompressed.
 		/// @todo  We could possibly set this value to "gzip" or something similar
-		curl_setopt($ch, CURLOPT_ENCODING, '');
+		$curlOptions[CURLOPT_ENCODING] = '';
 
 		if (!empty($opts['headers'])) {
 			$this->logger->notice('Wrong option \'headers\' used.');
-			@curl_setopt($ch, CURLOPT_HTTPHEADER, $opts['headers']);
+			if (empty($curlOptions[CURLOPT_HTTPHEADER])) {
+				$curlOptions[CURLOPT_HTTPHEADER] = [];
+			}
+			$curlOptions[CURLOPT_HTTPHEADER] = array_merge($opts['headers'], $curlOptions[CURLOPT_HTTPHEADER]);
 		}
 
 		if (!empty($opts['nobody'])) {
-			@curl_setopt($ch, CURLOPT_NOBODY, $opts['nobody']);
+			$curlOptions[CURLOPT_NOBODY] = $opts['nobody'];
 		}
 
-		@curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
+		$curlOptions[CURLOPT_CONNECTTIMEOUT] = 10;
 
 		if (!empty($opts['timeout'])) {
-			@curl_setopt($ch, CURLOPT_TIMEOUT, $opts['timeout']);
+			$curlOptions[CURLOPT_TIMEOUT] = $opts['timeout'];
 		} else {
 			$curl_time = $this->config->get('system', 'curl_timeout', 60);
-			@curl_setopt($ch, CURLOPT_TIMEOUT, intval($curl_time));
+			$curlOptions[CURLOPT_TIMEOUT] = intval($curl_time);
 		}
 
 		// by default we will allow self-signed certs
 		// but you can override this
 
 		$check_cert = $this->config->get('system', 'verifyssl');
-		@curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, (($check_cert) ? true : false));
+		$curlOptions[CURLOPT_SSL_VERIFYPEER] = ($check_cert) ? true : false;
 
 		if ($check_cert) {
-			@curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 2);
+			$curlOptions[CURLOPT_SSL_VERIFYHOST] = 2;
 		}
 
 		$proxy = $this->config->get('system', 'proxy');
 
 		if (!empty($proxy)) {
-			@curl_setopt($ch, CURLOPT_HTTPPROXYTUNNEL, 1);
+			$curlOptions[CURLOPT_HTTPPROXYTUNNEL] = 1;
-			@curl_setopt($ch, CURLOPT_PROXY, $proxy);
+			$curlOptions[CURLOPT_PROXY] = $proxy;
 			$proxyuser = $this->config->get('system', 'proxyuser');
 
 			if (!empty($proxyuser)) {
-				@curl_setopt($ch, CURLOPT_PROXYUSERPWD, $proxyuser);
+				$curlOptions[CURLOPT_PROXYUSERPWD] = $proxyuser;
 			}
 		}
 
 		if ($this->config->get('system', 'ipv4_resolve', false)) {
-			curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
+			$curlOptions[CURLOPT_IPRESOLVE] = CURL_IPRESOLVE_V4;
 		}
 
-		$s         = @curl_exec($ch);
+		if ($binary) {
-		$curl_info = @curl_getinfo($ch);
+			$curlOptions[CURLOPT_BINARYTRANSFER] = 1;
-
-		// Special treatment for HTTP Code 416
-		// See https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/416
-		if (($curl_info['http_code'] == 416) && ($range > 0)) {
-			@curl_setopt($ch, CURLOPT_RANGE, '');
-			$s         = @curl_exec($ch);
-			$curl_info = @curl_getinfo($ch);
 		}
 
-		$curlResponse = new CurlResult($url, $s, $curl_info, curl_errno($ch), curl_error($ch));
+		$onRedirect = function(
+			RequestInterface $request,
+			ResponseInterface $response,
+			UriInterface $uri
+		) {
+			$this->logger->notice('Curl redirect.', ['url' => $request->getUri(), 'to' => $uri]);
+		};
 
-		if (!Network::isRedirectBlocked($url) && $curlResponse->isRedirectUrl()) {
+		$client = new Client([
-			$redirects++;
+			'allow_redirect' => [
-			$this->logger->notice('Curl redirect.', ['url' => $url, 'to' => $curlResponse->getRedirectUrl()]);
+				'max' => 8,
-			@curl_close($ch);
+				'on_redirect' => $onRedirect,
+				'track_redirect' => true,
+				'strict' => true,
+				'referer' => true,
+			],
+			'on_headers' => $onHeaders,
+			'sink' => tempnam(get_temppath(), 'guzzle'),
+			'curl' => $curlOptions
+		]);
+
+		try {
+			$response = $client->get($url);
+			return new GuzzleResponse($response, $url);
+		} catch (RequestException $exception) {
+			if ($exception->hasResponse()) {
+				return new GuzzleResponse($exception->getResponse(), $url, $exception->getCode(), $exception->getMessage());
+			} else {
+				return new GuzzleResponse(null, $url, $exception->getCode(), $exception->getMessage());
+			}
+		} finally {
 			$this->profiler->stopRecording();
-			return $this->get($curlResponse->getRedirectUrl(), $opts, $redirects);
 		}
-
-		@curl_close($ch);
-
-		$this->profiler->stopRecording();
-
-		return $curlResponse;
 	}
 
 	/**