API for fetching reference metadata

Signed-off-by: Julius Härtl <jus@bitgrid.net>

core/Controller/ReferenceController.php

@@ -0,0 +1,82 @@
+<?php
+
+declare(strict_types=1);
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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 OC\Core\Controller;
+
+use OCP\AppFramework\Http\DataResponse;
+use OC\Collaboration\Reference\ReferenceManager;
+use OCP\IRequest;
+
+class ReferenceController extends \OCP\AppFramework\OCSController {
+	private ReferenceManager $referenceManager;
+
+	public function __construct($appName, IRequest $request, ReferenceManager $referenceManager) {
+		parent::__construct($appName, $request);
+		$this->referenceManager = $referenceManager;
+	}
+
+	/**
+	 * @NoAdminRequired
+	 *
+	 * @param string $text
+	 * @param bool $resolve
+	 * @return DataResponse
+	 */
+	public function extract(string $text, bool $resolve = false, int $limit = 1): DataResponse {
+		$references = $this->referenceManager->extractReferences($text);
+
+		$result = [];
+		$index = 0;
+		foreach ($references as $reference) {
+			if ($index++ < $limit) {
+				$result[$reference] = $resolve ? $this->referenceManager->resolveReference($reference) : null;
+			}
+		}
+
+		return new DataResponse([
+			'references' => $result
+		]);
+	}
+
+
+	/**
+	 * @NoAdminRequired
+	 *
+	 * @param array $references
+	 * @return DataResponse
+	 */
+	public function resolve(array $references, int $limit = 1): DataResponse {
+		$result = [];
+		$index = 0;
+		foreach ($references as $reference) {
+			if ($index++ < $limit) {
+				$result[$reference] = $this->referenceManager->resolveReference($reference);
+			}
+		}
+
+		return new DataResponse([
+			'references' => array_filter($result)
+		]);
+	}
+}

core/routes.php

@@ -120,6 +120,9 @@ $application->registerRoutes($this, [
 		['root' => '/collaboration', 'name' => 'CollaborationResources#getCollectionsByResource', 'url' => '/resources/{resourceType}/{resourceId}', 'verb' => 'GET'],
 		['root' => '/collaboration', 'name' => 'CollaborationResources#createCollectionOnResource', 'url' => '/resources/{baseResourceType}/{baseResourceId}', 'verb' => 'POST'],
 
+		['root' => '/references', 'name' => 'Reference#extract', 'url' => '/extract', 'verb' => 'POST'],
+		['root' => '/references', 'name' => 'Reference#resolve', 'url' => '/resolve', 'verb' => 'POST'],
+
 		['root' => '/profile', 'name' => 'ProfileApi#setVisibility', 'url' => '/{targetUserId}', 'verb' => 'PUT'],
 
 		// Unified search

lib/private/Collaboration/Reference/FileReferenceProvider.php

@@ -0,0 +1,93 @@
+<?php
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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 OC\Collaboration\Reference;
+
+use OCP\Collaboration\Reference\IReference;
+use OCP\Collaboration\Reference\IReferenceProvider;
+use OCP\Files\IRootFolder;
+use OCP\Files\Node;
+use OCP\Files\NotFoundException;
+use OCP\IURLGenerator;
+use OCP\IUserSession;
+
+class FileReferenceProvider implements IReferenceProvider {
+	private IURLGenerator $urlGenerator;
+	private IRootFolder $rootFolder;
+	private ?string $userId;
+
+	public function __construct(IURLGenerator $urlGenerator, IRootFolder $rootFolder, IUserSession $userSession) {
+		$this->urlGenerator = $urlGenerator;
+		$this->rootFolder = $rootFolder;
+		$this->userId = $userSession->getUser() ? $userSession->getUser()->getUID() : null;
+	}
+
+	public function resolveReference(string $referenceText): ?IReference {
+		if (str_starts_with($referenceText, $this->urlGenerator->getAbsoluteURL('/index.php/f/')) || str_starts_with($referenceText, $this->urlGenerator->getAbsoluteURL('/f/'))) {
+			$reference = new Reference($referenceText);
+			try {
+				$this->fetchReference($reference);
+			} catch (NotFoundException $e) {
+				$reference->setAccessible(false);
+			}
+			return $reference;
+		}
+
+		return null;
+	}
+
+	/**
+	 * @throws \OCP\Files\NotPermittedException
+	 * @throws \OCP\Files\InvalidPathException
+	 * @throws NotFoundException
+	 * @throws \OC\User\NoUserException
+	 */
+	private function fetchReference(Reference $reference) {
+		$fileId = str_replace($this->urlGenerator->getAbsoluteURL('/index.php/f/'), '', $reference->getId());
+		$fileId = str_replace($this->urlGenerator->getAbsoluteURL('/f/'), '', $fileId);
+
+		$userFolder = $this->rootFolder->getUserFolder($this->userId);
+		$files = $userFolder->getById((int)$fileId);
+
+		if (empty($files)) {
+			throw new NotFoundException();
+		}
+
+		/** @var Node $file */
+		$file = array_shift($files);
+
+		$reference->setTitle($file->getName());
+		$reference->setDescription($file->getMimetype());
+		$reference->setUrl($this->urlGenerator->getAbsoluteURL('/index.php/f/' . $fileId));
+		$reference->setImageUrl($this->urlGenerator->linkToRouteAbsolute('core.Preview.getPreviewByFileId', ['x' => 1600, 'y' => 630, 'fileId' => $fileId]));
+
+		$reference->setRichObject('file', [
+			'id' => $file->getId(),
+			'name' => $file->getName(),
+			'size' => $file->getSize(),
+			'path' => $file->getPath(),
+			'link' => $reference->getUrl(),
+			'mimetype' => $file->getMimetype(),
+			'preview-available' => false
+		]);
+	}
+}

lib/private/Collaboration/Reference/LinkReferenceProvider.php

@@ -0,0 +1,83 @@
+<?php
+
+declare(strict_types=1);
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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 OC\Collaboration\Reference;
+
+use Fusonic\OpenGraph\Consumer;
+use OCP\Collaboration\Reference\IReference;
+use OCP\Collaboration\Reference\IReferenceProvider;
+use OCP\Http\Client\IClientService;
+use Psr\Log\LoggerInterface;
+
+class LinkReferenceProvider implements IReferenceProvider {
+	public const URL_PATTERN = '/(\s|^)(https?:\/\/)?((?:[-A-Z0-9+_]+\.)+[-A-Z]+(?:\/[-A-Z0-9+&@#%?=~_|!:,.;()]*)*)(\s|$)/i';
+
+	private IClientService $clientService;
+	private LoggerInterface $logger;
+
+	public function __construct(IClientService $clientService, LoggerInterface $logger) {
+		$this->clientService = $clientService;
+		$this->logger = $logger;
+	}
+
+	public function resolveReference(string $referenceText): ?IReference {
+		if (preg_match(self::URL_PATTERN, $referenceText)) {
+			$reference = new Reference($referenceText);
+			$this->fetchReference($reference);
+			return $reference;
+		}
+
+		return null;
+	}
+
+	public function fetchReference(Reference $reference) {
+		$client = $this->clientService->newClient();
+		try {
+			$response = $client->get($reference->getId());
+		} catch (\Exception $e) {
+			$this->logger->debug('Failed to fetch link for obtaining open graph data', ['exception' => $e]);
+			return;
+		}
+
+		$responseBody = (string)$response->getBody();
+
+		$reference->setUrl($reference->getId());
+
+		// OpenGraph handling
+		$consumer = new Consumer();
+		$object = $consumer->loadHtml($responseBody);
+
+		if ($object->title) {
+			$reference->setTitle($object->title);
+		}
+
+		if ($object->description) {
+			$reference->setDescription($object->description);
+		}
+
+		if ($object->images) {
+			$reference->setImageUrl($object->images[0]->url);
+		}
+	}
+}

lib/private/Collaboration/Reference/Reference.php

@@ -0,0 +1,146 @@
+<?php
+
+declare(strict_types=1);
+
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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 OC\Collaboration\Reference;
+
+class Reference implements \OCP\Collaboration\Reference\IReference, \JsonSerializable {
+	private string $reference;
+
+	private bool $accessible = true;
+
+	private ?string $title = null;
+	private ?string $description = null;
+	private ?string $imageUrl = null;
+	private ?string $url = null;
+
+	private ?string $richObjectType = null;
+	private ?array $richObject = null;
+
+	public function __construct(string $reference) {
+		$this->reference = $reference;
+	}
+
+	public function getId(): string {
+		return $this->reference;
+	}
+
+	public function setAccessible(bool $accessible): void {
+		$this->accessible = $accessible;
+	}
+
+	public function setTitle(string $title): void {
+		$this->title = $title;
+	}
+
+	public function getTitle(): string {
+		return $this->title ?? $this->reference;
+	}
+
+	public function setDescription(?string $description): void {
+		$this->description = $description;
+	}
+
+	public function getDescription(): ?string {
+		return $this->description;
+	}
+
+	public function setImageUrl(?string $imageUrl): void {
+		$this->imageUrl = $imageUrl;
+	}
+
+	public function getImageUrl(): ?string {
+		return $this->imageUrl;
+	}
+
+	public function setUrl(?string $url): void {
+		$this->url = $url;
+	}
+
+	public function getUrl(): ?string {
+		return $this->url;
+	}
+
+	public function setRichObject(string $type, array $richObject): void {
+		$this->richObjectType = $type;
+		$this->richObject = $richObject;
+	}
+
+	public function getRichObjectType(): string {
+		if (!$this->richObjectType) {
+			return 'open-graph';
+		}
+		return $this->richObjectType;
+	}
+
+	public function getRichObject(): array {
+		if (!$this->richObject) {
+			return $this->getOpenGraphObject();
+		}
+		return $this->richObject;
+	}
+
+	public function getOpenGraphObject(): array {
+		return [
+			'id' => $this->getId(),
+			'name' => $this->getTitle(),
+			'description' => $this->getDescription(),
+			'thumb' => $this->getImageUrl(),
+			'link' => $this->getUrl()
+		];
+	}
+
+	public static function toCache(Reference $reference): array {
+		return [
+			'id' => $reference->getId(),
+			'title' => $reference->getTitle(),
+			'imageUrl' => $reference->getImageUrl(),
+			'description' => $reference->getDescription(),
+			'link' => $reference->getUrl(),
+			'accessible' => $reference->accessible,
+			'richObjectType' => $reference->getRichObjectType(),
+			'richObject' => $reference->getRichObject(),
+		];
+	}
+
+	public static function fromCache(array $cache): Reference {
+		$reference = new Reference($cache['id']);
+		$reference->setTitle($cache['title']);
+		$reference->setDescription($cache['description']);
+		$reference->setImageUrl($cache['imageUrl']);
+		$reference->setUrl($cache['link']);
+		$reference->setRichObject($cache['richObjectType'], $cache['richObject']);
+		$reference->setAccessible($cache['accessible']);
+		return $reference;
+	}
+
+	public function jsonSerialize() {
+		return [
+			'richObjectType' => $this->getRichObjectType(),
+			'richObject' => $this->getRichObject(),
+			'openGraphObject' => $this->getOpenGraphObject(),
+			'accessible' => $this->accessible
+		];
+	}
+}

lib/private/Collaboration/Reference/ReferenceManager.php

@@ -0,0 +1,79 @@
+<?php
+
+declare(strict_types=1);
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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 OC\Collaboration\Reference;
+
+use OCP\Collaboration\Reference\IReference;
+use OCP\Collaboration\Reference\IReferenceManager;
+use OCP\Collaboration\Reference\IReferenceProvider;
+use OCP\ICache;
+use OCP\ICacheFactory;
+
+class ReferenceManager implements IReferenceManager {
+	public const URL_PATTERN = '/(\s|\n|^)(https?:\/\/)?((?:[-A-Z0-9+_]+\.)+[-A-Z]+(?:\/[-A-Z0-9+&@#%?=~_|!:,.;()]*)*)(\s|\n|$)/mi';
+	public const REF_PATTERN = '/#[A-z0-9_]+\/[A-z0-9_]+/i';
+
+	/** @var IReferenceProvider[] */
+	private array $providers = [];
+	private ICache $cache;
+
+	public function __construct(LinkReferenceProvider $linkReferenceProvider, FileReferenceProvider $fileReferenceProvider, ICacheFactory $cacheFactory) {
+		$this->registerReferenceProvider($fileReferenceProvider);
+		$this->registerReferenceProvider($linkReferenceProvider);
+		$this->cache = $cacheFactory->createDistributed('reference');
+	}
+
+	public function extractReferences(string $text): array {
+		$matches = [];
+		preg_match_all(self::REF_PATTERN, $text, $matches);
+		$references = $matches[0] ?? [];
+
+		preg_match_all(self::URL_PATTERN, $text, $matches);
+		$references = array_merge($references, $matches[0] ?? []);
+		return array_map(function ($reference) {
+			return trim($reference);
+		}, $references);
+	}
+
+	public function resolveReference(string $referenceId): ?IReference {
+		$cached = $this->cache->get($referenceId);
+		if ($cached) {
+			// TODO: Figure out caching for references that depend on the viewer user
+			return Reference::fromCache($cached);
+		}
+		foreach ($this->providers as $provider) {
+			$reference = $provider->resolveReference($referenceId);
+			if ($reference) {
+				$this->cache->set($referenceId, Reference::toCache($reference), 60);
+				return $reference;
+			}
+		}
+
+		return null;
+	}
+
+	public function registerReferenceProvider(IReferenceProvider $provider): void {
+		$this->providers[] = $provider;
+	}
+}

lib/public/Collaboration/Reference/IReference.php

@@ -0,0 +1,31 @@
+<?php
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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 OC\Collaboration\Reference\Reference;
+
+interface IReference {
+	public function getRichObject(): array;
+
+	public static function toCache(Reference $reference): array;
+}

lib/public/Collaboration/Reference/IReferenceManager.php

@@ -0,0 +1,31 @@
+<?php
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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;
+
+interface IReferenceManager {
+	public function extractReferences(string $text): array;
+
+	public function resolveReference(string $reference): ?IReference;
+
+	public function registerReferenceProvider(IReferenceProvider $provider): void;
+}

lib/public/Collaboration/Reference/IReferenceProvider.php

@@ -0,0 +1,27 @@
+<?php
+/**
+ * @copyright Copyright (c) 2022 Julius Härtl <jus@bitgrid.net>
+ *
+ * @author Julius Härtl <jus@bitgrid.net>
+ *
+ * @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;
+
+interface IReferenceProvider {
+	public function resolveReference(string $referenceText): ?IReference;
+}