feat: Migrate LinkReferenceProvider to OCP

This would be useful to eleminate the need for using this OC class
when an app dev wants to implement a custom reference provider
for the web client but wants to fall back on opengraph for mobile
clients.

Signed-off-by: Anupam Kumar <kyteinsky@gmail.com>

4 files changed, 221 insertions, 155 deletions

diff --git a/lib/composer/composer/autoload_classmap.php b/lib/composer/composer/autoload_classmap.php
index fa765537953..1ed555d7a72 100644
--- a/lib/composer/composer/autoload_classmap.php
+++ b/lib/composer/composer/autoload_classmap.php
@@ -184,6 +184,7 @@ return array(
     'OCP\\Collaboration\\Reference\\IReferenceManager' => $baseDir . '/lib/public/Collaboration/Reference/IReferenceManager.php',
     'OCP\\Collaboration\\Reference\\IReferenceProvider' => $baseDir . '/lib/public/Collaboration/Reference/IReferenceProvider.php',
     'OCP\\Collaboration\\Reference\\ISearchableReferenceProvider' => $baseDir . '/lib/public/Collaboration/Reference/ISearchableReferenceProvider.php',
+    'OCP\\Collaboration\\Reference\\LinkReferenceProvider' => $baseDir . '/lib/public/Collaboration/Reference/LinkReferenceProvider.php',
     'OCP\\Collaboration\\Reference\\Reference' => $baseDir . '/lib/public/Collaboration/Reference/Reference.php',
     'OCP\\Collaboration\\Reference\\RenderReferenceEvent' => $baseDir . '/lib/public/Collaboration/Reference/RenderReferenceEvent.php',
     'OCP\\Collaboration\\Resources\\CollectionException' => $baseDir . '/lib/public/Collaboration/Resources/CollectionException.php',
diff --git a/lib/composer/composer/autoload_static.php b/lib/composer/composer/autoload_static.php
index 2899125595e..aa5951dc44f 100644
--- a/lib/composer/composer/autoload_static.php
+++ b/lib/composer/composer/autoload_static.php
@@ -217,6 +217,7 @@ class ComposerStaticInit749170dad3f5e7f9ca158f5a9f04f6a2
         'OCP\\Collaboration\\Reference\\IReferenceManager' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Reference/IReferenceManager.php',
         'OCP\\Collaboration\\Reference\\IReferenceProvider' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Reference/IReferenceProvider.php',
         'OCP\\Collaboration\\Reference\\ISearchableReferenceProvider' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Reference/ISearchableReferenceProvider.php',
+        'OCP\\Collaboration\\Reference\\LinkReferenceProvider' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Reference/LinkReferenceProvider.php',
         'OCP\\Collaboration\\Reference\\Reference' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Reference/Reference.php',
         'OCP\\Collaboration\\Reference\\RenderReferenceEvent' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Reference/RenderReferenceEvent.php',
         'OCP\\Collaboration\\Resources\\CollectionException' => __DIR__ . '/../../..' . '/lib/public/Collaboration/Resources/CollectionException.php',
diff --git a/lib/private/Collaboration/Reference/LinkReferenceProvider.php b/lib/private/Collaboration/Reference/LinkReferenceProvider.php
index df6c6cc9da9..08b388b47a4 100644
--- a/lib/private/Collaboration/Reference/LinkReferenceProvider.php
+++ b/lib/private/Collaboration/Reference/LinkReferenceProvider.php
@@ -5,6 +5,7 @@ declare(strict_types=1);
  * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
  *
  * @author Julius Härtl <jus@bitgrid.net>
+ * @author Anupam Kumar <kyteinsky@gmail.com>
  *
  * @license GNU AGPL version 3 or any later version
  *
@@ -24,160 +25,8 @@ declare(strict_types=1);
 
 namespace OC\Collaboration\Reference;
 
-use Fusonic\OpenGraph\Consumer;
-use GuzzleHttp\Exception\GuzzleException;
-use GuzzleHttp\Psr7\LimitStream;
-use GuzzleHttp\Psr7\Utils;
-use OC\Security\RateLimiting\Exception\RateLimitExceededException;
-use OC\Security\RateLimiting\Limiter;
-use OC\SystemConfig;
-use OCP\Collaboration\Reference\IReference;
-use OCP\Collaboration\Reference\IReferenceProvider;
-use OCP\Collaboration\Reference\Reference;
-use OCP\Files\AppData\IAppDataFactory;
-use OCP\Files\NotFoundException;
-use OCP\Http\Client\IClientService;
-use OCP\IRequest;
-use OCP\IURLGenerator;
-use OCP\IUserSession;
-use Psr\Log\LoggerInterface;
+use OCP\Collaboration\Reference\LinkReferenceProvider as OCPLinkReferenceProvider;
 
-class LinkReferenceProvider implements IReferenceProvider {
-	public const MAX_PREVIEW_SIZE = 1024 * 1024;
-
-	public const ALLOWED_CONTENT_TYPES = [
-		'image/png',
-		'image/jpg',
-		'image/jpeg',
-		'image/gif',
-		'image/svg+xml',
-		'image/webp'
-	];
-
-	public function __construct(
-		private IClientService $clientService,
-		private LoggerInterface $logger,
-		private SystemConfig $systemConfig,
-		private IAppDataFactory $appDataFactory,
-		private IURLGenerator $urlGenerator,
-		private Limiter $limiter,
-		private IUserSession $userSession,
-		private IRequest $request,
-	) {
-	}
-
-	public function matchReference(string $referenceText): bool {
-		if ($this->systemConfig->getValue('reference_opengraph', true) !== true) {
-			return false;
-		}
-
-		return (bool)preg_match(IURLGenerator::URL_REGEX, $referenceText);
-	}
-
-	public function resolveReference(string $referenceText): ?IReference {
-		if ($this->matchReference($referenceText)) {
-			$reference = new Reference($referenceText);
-			$this->fetchReference($reference);
-			return $reference;
-		}
-
-		return null;
-	}
-
-	private function fetchReference(Reference $reference): void {
-		try {
-			$user = $this->userSession->getUser();
-			if ($user) {
-				$this->limiter->registerUserRequest('opengraph', 10, 120, $user);
-			} else {
-				$this->limiter->registerAnonRequest('opengraph', 10, 120, $this->request->getRemoteAddress());
-			}
-		} catch (RateLimitExceededException $e) {
-			return;
-		}
-
-		$client = $this->clientService->newClient();
-		try {
-			$headResponse = $client->head($reference->getId(), [ 'timeout' => 10 ]);
-		} catch (\Exception $e) {
-			$this->logger->debug('Failed to perform HEAD request to get target metadata', ['exception' => $e]);
-			return;
-		}
-		$linkContentLength = $headResponse->getHeader('Content-Length');
-		if (is_numeric($linkContentLength) && (int) $linkContentLength > 5 * 1024 * 1024) {
-			$this->logger->debug('Skip resolving links pointing to content length > 5 MB');
-			return;
-		}
-		$linkContentType = $headResponse->getHeader('Content-Type');
-		$expectedContentType = 'text/html';
-		$suffixedExpectedContentType = $expectedContentType . ';';
-		$startsWithSuffixed = str_starts_with($linkContentType, $suffixedExpectedContentType);
-		// check the header begins with the expected content type
-		if ($linkContentType !== $expectedContentType && !$startsWithSuffixed) {
-			$this->logger->debug('Skip resolving links pointing to content type that is not "text/html"');
-			return;
-		}
-		try {
-			$response = $client->get($reference->getId(), [ 'timeout' => 10 ]);
-		} catch (\Exception $e) {
-			$this->logger->debug('Failed to fetch link for obtaining open graph data', ['exception' => $e]);
-			return;
-		}
-
-		$responseBody = (string)$response->getBody();
-
-		// OpenGraph handling
-		$consumer = new Consumer();
-		$consumer->useFallbackMode = true;
-		$object = $consumer->loadHtml($responseBody);
-
-		$reference->setUrl($reference->getId());
-
-		if ($object->title) {
-			$reference->setTitle($object->title);
-		}
-
-		if ($object->description) {
-			$reference->setDescription($object->description);
-		}
-
-		if ($object->images) {
-			try {
-				$host = parse_url($object->images[0]->url, PHP_URL_HOST);
-				if ($host === false || $host === null) {
-					$this->logger->warning('Could not detect host of open graph image URI for ' . $reference->getId());
-				} else {
-					$appData = $this->appDataFactory->get('core');
-					try {
-						$folder = $appData->getFolder('opengraph');
-					} catch (NotFoundException $e) {
-						$folder = $appData->newFolder('opengraph');
-					}
-					$response = $client->get($object->images[0]->url, ['timeout' => 10]);
-					$contentType = $response->getHeader('Content-Type');
-					$contentLength = $response->getHeader('Content-Length');
-
-					if (in_array($contentType, self::ALLOWED_CONTENT_TYPES, true) && $contentLength < self::MAX_PREVIEW_SIZE) {
-						$stream = Utils::streamFor($response->getBody());
-						$bodyStream = new LimitStream($stream, self::MAX_PREVIEW_SIZE, 0);
-						$reference->setImageContentType($contentType);
-						$folder->newFile(md5($reference->getId()), $bodyStream->getContents());
-						$reference->setImageUrl($this->urlGenerator->linkToRouteAbsolute('core.Reference.preview', ['referenceId' => md5($reference->getId())]));
-					}
-				}
-			} catch (GuzzleException $e) {
-				$this->logger->info('Failed to fetch and store the open graph image for ' . $reference->getId(), ['exception' => $e]);
-			} catch (\Throwable $e) {
-				$this->logger->error('Failed to fetch and store the open graph image for ' . $reference->getId(), ['exception' => $e]);
-			}
-		}
-	}
-
-	public function getCachePrefix(string $referenceId): string {
-		return $referenceId;
-	}
-
-	public function getCacheKey(string $referenceId): ?string {
-		return null;
-	}
+/** @deprecated 29.0.0 Use OCP\Collaboration\Reference\LinkReferenceProvider instead */
+class LinkReferenceProvider extends OCPLinkReferenceProvider {
 }
diff --git a/lib/public/Collaboration/Reference/LinkReferenceProvider.php b/lib/public/Collaboration/Reference/LinkReferenceProvider.php
new file mode 100644
index 00000000000..76f977f2eb0
--- /dev/null
+++ b/lib/public/Collaboration/Reference/LinkReferenceProvider.php
@@ -0,0 +1,215 @@
+<?php
+
+declare(strict_types=1);
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ * @author Anupam Kumar <kyteinsky@gmail.com>
+ *
+ * @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 <http://www.gnu.org/licenses/>.
+ */
+
+namespace OCP\Collaboration\Reference;
+
+use Fusonic\OpenGraph\Consumer;
+use GuzzleHttp\Exception\GuzzleException;
+use GuzzleHttp\Psr7\LimitStream;
+use GuzzleHttp\Psr7\Utils;
+use OC\Security\RateLimiting\Exception\RateLimitExceededException;
+use OC\Security\RateLimiting\Limiter;
+use OC\SystemConfig;
+use OCP\Files\AppData\IAppDataFactory;
+use OCP\Files\NotFoundException;
+use OCP\Http\Client\IClientService;
+use OCP\IRequest;
+use OCP\IURLGenerator;
+use OCP\IUserSession;
+use Psr\Log\LoggerInterface;
+
+/**
+ * @since 29.0.0
+ */
+class LinkReferenceProvider implements IReferenceProvider {
+
+	/* for image size and webpage header */
+	private const MAX_CONTENT_LENGTH = 5 * 1024 * 1024;
+
+	private const ALLOWED_CONTENT_TYPES = [
+		'image/png',
+		'image/jpg',
+		'image/jpeg',
+		'image/gif',
+		'image/svg+xml',
+		'image/webp'
+	];
+
+	/**
+	 * @since 29.0.0
+	 */
+	public function __construct(
+		private IClientService $clientService,
+		private LoggerInterface $logger,
+		private SystemConfig $systemConfig,
+		private IAppDataFactory $appDataFactory,
+		private IURLGenerator $urlGenerator,
+		private Limiter $limiter,
+		private IUserSession $userSession,
+		private IRequest $request,
+	) {
+	}
+
+	/**
+	 * @inheritDoc
+	 * @since 29.0.0
+	 */
+	public function matchReference(string $referenceText): bool {
+		if ($this->systemConfig->getValue('reference_opengraph', true) !== true) {
+			return false;
+		}
+
+		return (bool)preg_match(IURLGenerator::URL_REGEX, $referenceText);
+	}
+
+	/**
+	 * @inheritDoc
+	 * @since 29.0.0
+	 */
+	public function resolveReference(string $referenceText): ?IReference {
+		if ($this->matchReference($referenceText)) {
+			$reference = new Reference($referenceText);
+			$this->fetchReference($reference);
+			return $reference;
+		}
+
+		return null;
+	}
+
+	/**
+	 * Populates the reference with OpenGraph data
+	 *
+	 * @param Reference $reference
+	 * @since 29.0.0
+	 */
+	private function fetchReference(Reference $reference): void {
+		try {
+			$user = $this->userSession->getUser();
+			if ($user) {
+				$this->limiter->registerUserRequest('opengraph', 10, 120, $user);
+			} else {
+				$this->limiter->registerAnonRequest('opengraph', 10, 120, $this->request->getRemoteAddress());
+			}
+		} catch (RateLimitExceededException $e) {
+			return;
+		}
+
+		$client = $this->clientService->newClient();
+		try {
+			$headResponse = $client->head($reference->getId(), [ 'timeout' => 10 ]);
+		} catch (\Exception $e) {
+			$this->logger->debug('Failed to perform HEAD request to get target metadata', ['exception' => $e]);
+			return;
+		}
+
+		$linkContentLength = $headResponse->getHeader('Content-Length');
+		if (is_numeric($linkContentLength) && (int) $linkContentLength > self::MAX_CONTENT_LENGTH) {
+			$this->logger->debug('Skip resolving links pointing to content length > 5 MiB');
+			return;
+		}
+
+		$linkContentType = $headResponse->getHeader('Content-Type');
+		$expectedContentTypeRegex = '/^text\/html;?/i';
+
+		// check the header begins with the expected content type
+		if (!preg_match($expectedContentTypeRegex, $linkContentType)) {
+			$this->logger->debug('Skip resolving links pointing to content type that is not "text/html"');
+			return;
+		}
+
+		try {
+			$response = $client->get($reference->getId(), [ 'timeout' => 10 ]);
+		} catch (\Exception $e) {
+			$this->logger->debug('Failed to fetch link for obtaining open graph data', ['exception' => $e]);
+			return;
+		}
+
+		$responseBody = (string)$response->getBody();
+
+		// OpenGraph handling
+		$consumer = new Consumer();
+		$consumer->useFallbackMode = true;
+		$object = $consumer->loadHtml($responseBody);
+
+		$reference->setUrl($reference->getId());
+
+		if ($object->title) {
+			$reference->setTitle($object->title);
+		}
+
+		if ($object->description) {
+			$reference->setDescription($object->description);
+		}
+
+		if ($object->images) {
+			try {
+				$host = parse_url($object->images[0]->url, PHP_URL_HOST);
+				if ($host === false || $host === null) {
+					$this->logger->warning('Could not detect host of open graph image URI for ' . $reference->getId());
+					return;
+				}
+
+				$appData = $this->appDataFactory->get('core');
+				try {
+					$folder = $appData->getFolder('opengraph');
+				} catch (NotFoundException $e) {
+					$folder = $appData->newFolder('opengraph');
+				}
+
+				$response = $client->get($object->images[0]->url, ['timeout' => 10]);
+				$contentType = $response->getHeader('Content-Type');
+				$contentLength = $response->getHeader('Content-Length');
+
+				if (in_array($contentType, self::ALLOWED_CONTENT_TYPES, true) && $contentLength < self::MAX_CONTENT_LENGTH) {
+					$stream = Utils::streamFor($response->getBody());
+					$bodyStream = new LimitStream($stream, self::MAX_CONTENT_LENGTH, 0);
+					$reference->setImageContentType($contentType);
+					$folder->newFile(md5($reference->getId()), $bodyStream->getContents());
+					$reference->setImageUrl($this->urlGenerator->linkToRouteAbsolute('core.Reference.preview', ['referenceId' => md5($reference->getId())]));
+				}
+			} catch (GuzzleException $e) {
+				$this->logger->info('Failed to fetch and store the open graph image for ' . $reference->getId(), ['exception' => $e]);
+			} catch (\Throwable $e) {
+				$this->logger->error('Failed to fetch and store the open graph image for ' . $reference->getId(), ['exception' => $e]);
+			}
+		}
+	}
+
+	/**
+	 * @inheritDoc
+	 * @since 29.0.0
+	 */
+	public function getCachePrefix(string $referenceId): string {
+		return $referenceId;
+	}
+
+	/**
+	 * @inheritDoc
+	 * @since 29.0.0
+	 */
+	public function getCacheKey(string $referenceId): ?string {
+		return null;
+	}
+}