Improved user agent string

src/Network/HTTPClient/Capability/ICanSendHttpRequests.php

@@ -95,10 +95,11 @@ interface ICanSendHttpRequests
 	 * @param mixed  $params         POST variables (if an array is passed, it will automatically set as formular parameters)
 	 * @param array  $headers        HTTP headers
 	 * @param int    $timeout        The timeout in seconds, default system config value or 60 seconds
+	 * @param string $request        The type of the request. This is set in the user agent string
 	 *
 	 * @return ICanHandleHttpResponses The content
 	 */
-	public function post(string $url, $params, array $headers = [], int $timeout = 0): ICanHandleHttpResponses;
+	public function post(string $url, $params, array $headers = [], int $timeout = 0, string $request = ''): ICanHandleHttpResponses;
 
 	/**
 	 * Sends an HTTP request to a given url

src/Network/HTTPClient/Client/HttpClient.php

@@ -21,6 +21,7 @@
 
 namespace Friendica\Network\HTTPClient\Client;
 
+use Friendica\App;
 use Friendica\Core\System;
 use Friendica\Network\HTTPClient\Response\CurlResult;
 use Friendica\Network\HTTPClient\Response\GuzzleResponse;
@@ -51,13 +52,16 @@ class HttpClient implements ICanSendHttpRequests
 	private $client;
 	/** @var URLResolver */
 	private $resolver;
+	/** @var App\BaseURL */
+	private $baseUrl;
 
-	public function __construct(LoggerInterface $logger, Profiler $profiler, Client $client, URLResolver $resolver)
+	public function __construct(LoggerInterface $logger, Profiler $profiler, Client $client, URLResolver $resolver, App\BaseURL $baseUrl)
 	{
 		$this->logger   = $logger;
 		$this->profiler = $profiler;
 		$this->client   = $client;
 		$this->resolver = $resolver;
+		$this->baseUrl  = $baseUrl;
 	}
 
 	/**
@@ -73,7 +77,7 @@ class HttpClient implements ICanSendHttpRequests
 			throw new \InvalidArgumentException('Unable to retrieve the host in URL: ' . $url);
 		}
 
-		if(!filter_var($host, FILTER_VALIDATE_IP) && !@dns_get_record($host . '.', DNS_A) && !@dns_get_record($host . '.', DNS_AAAA)) {
+		if (!filter_var($host, FILTER_VALIDATE_IP) && !@dns_get_record($host . '.', DNS_A) && !@dns_get_record($host . '.', DNS_AAAA)) {
 			$this->logger->debug('URL cannot be resolved.', ['url' => $url]);
 			$this->profiler->stopRecording();
 			return CurlResult::createErrorCurl($this->logger, $url);
@@ -115,7 +119,7 @@ class HttpClient implements ICanSendHttpRequests
 			$conf[RequestOptions::COOKIES] = $jar;
 		}
 
-		$headers = [];
+		$headers = ['User-Agent' => $this->getUserAgent($opts[HttpClientOptions::REQUEST] ?? '')];
 
 		if (!empty($opts[HttpClientOptions::ACCEPT_CONTENT])) {
 			$headers['Accept'] = $opts[HttpClientOptions::ACCEPT_CONTENT];
@@ -153,8 +157,10 @@ class HttpClient implements ICanSendHttpRequests
 		}
 
 		$conf[RequestOptions::ON_HEADERS] = function (ResponseInterface $response) use ($opts) {
-			if (!empty($opts[HttpClientOptions::CONTENT_LENGTH]) &&
-				(int)$response->getHeaderLine('Content-Length') > $opts[HttpClientOptions::CONTENT_LENGTH]) {
+			if (
+				!empty($opts[HttpClientOptions::CONTENT_LENGTH]) &&
+				(int)$response->getHeaderLine('Content-Length') > $opts[HttpClientOptions::CONTENT_LENGTH]
+			) {
 				throw new TransferException('The file is too big!');
 			}
 		};
@@ -172,8 +178,10 @@ class HttpClient implements ICanSendHttpRequests
 			$response = $this->client->request($method, $url, $conf);
 			return new GuzzleResponse($response, $url);
 		} catch (TransferException $exception) {
-			if ($exception instanceof RequestException &&
-				$exception->hasResponse()) {
+			if (
+				$exception instanceof RequestException &&
+				$exception->hasResponse()
+			) {
 				return new GuzzleResponse($exception->getResponse(), $url, $exception->getCode(), '');
 			} else {
 				return new CurlResult($this->logger, $url, '', ['http_code' => 500], $exception->getCode(), '');
@@ -209,7 +217,7 @@ class HttpClient implements ICanSendHttpRequests
 	/**
 	 * {@inheritDoc}
 	 */
-	public function post(string $url, $params, array $headers = [], int $timeout = 0): ICanHandleHttpResponses
+	public function post(string $url, $params, array $headers = [], int $timeout = 0, string $request = ''): ICanHandleHttpResponses
 	{
 		$opts = [];
 
@@ -227,6 +235,10 @@ class HttpClient implements ICanSendHttpRequests
 			$opts[HttpClientOptions::TIMEOUT] = $timeout;
 		}
 
+		if (!empty($request)) {
+			$opts[HttpClientOptions::REQUEST] = $request;
+		}
+
 		return $this->request('post', $url, $opts);
 	}
 
@@ -255,6 +267,7 @@ class HttpClient implements ICanSendHttpRequests
 
 		$url = trim($url, "'");
 
+		$this->resolver->setUserAgent($this->getUserAgent(HttpClientRequest::RESOLVER));
 		$urlResult = $this->resolver->resolveURL($url);
 
 		if ($urlResult->didErrorOccur()) {
@@ -288,4 +301,15 @@ class HttpClient implements ICanSendHttpRequests
 			]
 		);
 	}
+
+	private function getUserAgent(string $type = ''): string
+	{
+		// @see https://developers.whatismybrowser.com/learn/browser-detection/user-agents/user-agent-best-practices
+		$userAgent = App::PLATFORM . '/' . App::VERSION  . ' DatabaseVersion/' . DB_UPDATE_VERSION;
+		if ($type != '') {
+			$userAgent .= ' Request/' . $type;
+		}
+		$userAgent .= ' +' . $this->baseUrl;
+		return $userAgent;
+	}
 }

src/Network/HTTPClient/Client/HttpClientOptions.php

@@ -52,7 +52,10 @@ class HttpClientOptions
 	 * content_length: (int) maximum File content length
 	 */
 	const CONTENT_LENGTH = 'content_length';
-
+	/**
+	 * Request: (string) Type of request (ActivityPub, Diaspora, server discovery, ...)
+	 */
+	const REQUEST = 'request';
 	/**
 	 * verify: (bool|string, default=true) Describes the SSL certificate
 	 */

src/Network/HTTPClient/Client/HttpClientRequest.php

@@ -0,0 +1,39 @@
+<?php
+/**
+ * @copyright Copyright (C) 2010-2024, the Friendica project
+ *
+ * @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\HTTPClient\Client;
+
+/**
+ * This class contains a list of request types that are set in the user agent string
+ */
+class HttpClientRequest
+{
+	public const ACTIVITYPUB = 'ActivityPub/1';
+	public const CONTENTTYPE = 'ContentTypeChecker/1';
+	public const DFRN        = 'DFRN/1';
+	public const DIASPORA    = 'Diaspora/1';
+	public const MAGICAUTH   = 'MagicAuth/1';
+	public const MEDIAPROXY  = 'MediaProxy/1';
+	public const SALMON      = 'Salmon/1';
+	public const PUBSUB      = 'PubSub/1';
+	public const RESOLVER    = 'URLResolver/1';
+	public const VERIFIER    = 'URLVerifier/1';
+}

src/Network/HTTPClient/Factory/HttpClient.php

@@ -86,12 +86,6 @@ class HttpClient extends BaseFactory
 			$logger->info('Curl redirect.', ['url' => $request->getUri(), 'to' => $uri, 'method' => $request->getMethod()]);
 		};
 
-		$userAgent = App::PLATFORM . " '" .
-					 App::CODENAME . "' " .
-					 App::VERSION . '-' .
-					 DB_UPDATE_VERSION . '; ' .
-					 $this->baseUrl;
-
 		$guzzle = new GuzzleHttp\Client([
 			RequestOptions::ALLOW_REDIRECTS => [
 				'max'             => 8,
@@ -112,22 +106,19 @@ class HttpClient extends BaseFactory
 			// but it can be overridden
 			RequestOptions::VERIFY  => (bool)$this->config->get('system', 'verifyssl'),
 			RequestOptions::PROXY   => $proxy,
-			RequestOptions::HEADERS => [
-				'User-Agent' => $userAgent,
-			],
+			RequestOptions::HEADERS => [],
 			'handler' => $handlerStack ?? HandlerStack::create(),
 		]);
 
 		$resolver = new URLResolver();
-		$resolver->setUserAgent($userAgent);
 		$resolver->setMaxRedirects(10);
 		$resolver->setRequestTimeout(10);
 		// if the file is too large then exit
 		$resolver->setMaxResponseDataSize($this->config->get('performance', 'max_response_data_size', 1000000));
 		// Designate a temporary file that will store cookies during the session.
 		// Some websites test the browser for cookie support, so this enhances results.
-		$resolver->setCookieJar(System::getTempPath() .'/resolver-cookie-' . Strings::getRandomName(10));
+		$resolver->setCookieJar(System::getTempPath() . '/resolver-cookie-' . Strings::getRandomName(10));
 
-		return new Client\HttpClient($logger, $this->profiler, $guzzle, $resolver);
+		return new Client\HttpClient($logger, $this->profiler, $guzzle, $resolver, $this->baseUrl);
 	}
 }