Make iThenticate reports work with new TCA integration

.copypatrol.ini.dist

@@ -0,0 +1,30 @@
+; Used by CopyPatrolBot, which populates the database that gets feed into the frontend.
+; For more information, see https://github.com/JJMC89/copypatrol-backend#configuration
+[copypatrol]
+version = 2023.08.10
+url-ignore-list-title = example-url-title
+user-ignore-list-title = example-user-title
+
+[copypatrol:en.wikipedia.org]
+enabled = true
+namespaces = 0,2,118
+pagetriage-namespaces = 0,118
+
+[copypatrol:es.wikipedia.org]
+enabled = true
+namespaces = 0,2
+
+[copypatrol:fr.wikipedia.org]
+enabled = false
+
+[client]
+drivername = mysql+pymysql
+username = example-db-user
+password = example-db-password
+database = example-db-name
+host = localhost
+port = 3306
+
+[tca]
+domain = example-tca-domain.com
+key = example-tca-key

.gitignore

@@ -1,4 +1,5 @@
 .idea/
+.copypatrol.ini
 
 ###> symfony/framework-bundle ###
 /.env.local

README.md

@@ -6,7 +6,8 @@ A tool that allows you to see recent Wikipedia edits that are flagged as possibl
 
 * User documentation: https://meta.wikimedia.org/wiki/Special:MyLanguage/CopyPatrol
 * Issue tracker: https://phabricator.wikimedia.org/tag/copypatrol/
-* Source code: https://gitlab.wikimedia.org/repos/commtech/copypatrol
+* Frontend source code (this repo): https://github.com/wikimedia/CopyPatrol
+* Bot source code: https://github.com/JJMC89/copypatrol-backend
 
 ## Installing manually
 
@@ -18,7 +19,11 @@ A tool that allows you to see recent Wikipedia edits that are flagged as possibl
 * [Toolforge access](https://wikitech.wikimedia.org/wiki/Help:Toolforge/Quickstart)
 
 This application makes use of the [Symfony framework](https://symfony.com/) and
-the [ToolforgeBundle](https://github.com/wikimedia/ToolforgeBundle).
+the [ToolforgeBundle](https://github.com/wikimedia/ToolforgeBundle). A [bot](https://meta.wikimedia.org/wiki/User:CopyPatrolBot) is used
+to continually query recent changes against the Turnitin Core API (TCA), record possible copyright
+violations in a user database, and CopyPatrol then reads from that database. Unless you need to work
+on the [bot code](https://github.com/JJMC89/copypatrol-backend), there's no reason to bother with bot
+integration, and instead connect to the existing user database on Toolforge (more on this below).
 
 ### Instructions
 
@@ -123,10 +128,11 @@ docker run -ti -p 80:80 -v $(pwd)/.env.local:/app/.env.local wikimedia/copypatro
    Any languages that are not regularly being used should be removed.
 3. Make sure the corresponding `-wikipedia` message in [i18n/en.json](i18n/en.json) (and [qqq.json](i18n/qqq.json)
    exists and is translated in the desired language.
-4. Add the language code to `APP_ENABLED_LANGS` in the [.env](.env) files.
-5. _We now use https://github.com/JJMC89/copypatrol-backend as the bot backend; technical instructions TBD_
+4. Update the `.copypatrol.ini` file accordingly (which is used by the bot),
+   and add the new languages to the `APP_ENABLED_LANGS` variable in [.env](.env).
 
 ## Removing a language
 
-1. Remove the language from `APP_ENABLED_LANGS`.
-2. _TBD_: Remove the job from CopyPatrolBot.
+1. In `.copypatrol.ini`, set the `enabled` key for the desired language to `false`,
+   or remove the definition entirely.
+2. Remove the language from the `APP_ENABLED_LANGS` variable in [.env](.env).

composer.json

@@ -7,6 +7,7 @@
         "php": ">=7.4",
         "ext-ctype": "*",
         "ext-iconv": "*",
+        "ext-json": "*",
         "doctrine/annotations": "^2.0",
         "doctrine/doctrine-bundle": "^2.2",
         "phpxmlrpc/phpxmlrpc": "^4.10",

src/Controller/AppController.php

@@ -8,6 +8,7 @@
 use DateTime;
 use Doctrine\DBAL\Exception\DriverException;
 use Exception;
+use Krinkle\Intuition\Intuition;
 use PhpXmlRpc\Client;
 use PhpXmlRpc\Value;
 use stdClass;
@@ -18,8 +19,10 @@
 use Symfony\Component\HttpFoundation\RequestStack;
 use Symfony\Component\HttpFoundation\Response;
 use Symfony\Component\HttpFoundation\Session\SessionInterface;
+use Symfony\Component\HttpKernel\Exception\HttpException;
 // phpcs:ignore MediaWiki.Classes.UnusedUseStatement.UnusedUse
 use Symfony\Component\Routing\Annotation\Route;
+use Symfony\Contracts\HttpClient\HttpClientInterface;
 
 class AppController extends AbstractController {
 
@@ -278,26 +281,39 @@ public function undoReviewAction(
 	}
 
 	/**
-	 * @Route("/ithenticate/{id}", name="ithenticate", requirements={"id"="\d+|[a-z\d+\-]"})
+	 * @Route("/ithenticate/{id}", name="ithenticate", requirements={"id"="\d+|[a-z\d+\-]+"})
+	 * @param HttpClientInterface $httpClient
+	 * @param RequestStack $requestStack
 	 * @param CopyPatrolRepository $copyPatrolRepo
+	 * @param Intuition $intuition
 	 * @param string $iThenticateUser
 	 * @param string $iThenticatePassword
-	 * @param int $id
+	 * @param string $id
 	 * @return RedirectResponse
 	 * @throws Exception
 	 */
 	public function iThenticateAction(
+		HttpClientInterface $httpClient,
+		RequestStack $requestStack,
 		CopyPatrolRepository $copyPatrolRepo,
+		Intuition $intuition,
 		string $iThenticateUser,
 		string $iThenticatePassword,
-		int $id
+		string $id
 	): RedirectResponse {
 		$record = $copyPatrolRepo->getRecordById( $id );
 		$v2DateTime = new DateTime( self::ITHENTICATE_V2_TIMESTAMP );
 		if ( new DateTime( $record['rev_timestamp'] ) > $v2DateTime ) {
-			// New system
-			// TODO: make this work
-			return new RedirectResponse( '' );
+			// New system.
+			$loggedInUser = $requestStack->getSession()->get( 'logged_in_user' )->username;
+			if ( $loggedInUser === null ) {
+				return $this->redirectToRoute( 'toolforge_login', [
+					'callback' => $this->generateUrl( 'toolforge_oauth_callback', [
+						'redirect' => $requestStack->getCurrentRequest()->getUri(),
+					] )
+				] );
+			}
+			return $this->redirectToTcaViewer( $httpClient, $loggedInUser, $intuition->getLang(), $id );
 		}
 
 		$client = new Client( 'https://api.ithenticate.com/rpc' );
@@ -315,6 +331,63 @@ public function iThenticateAction(
 		return $this->redirect( $response['view_only_url']->scalarval() );
 	}
 
+	/**
+	 * @param HttpClientInterface $client
+	 * @param string $loggedInUser
+	 * @param string $locale
+	 * @param string $id
+	 * @param int $retries
+	 * @return RedirectResponse
+	 */
+	private function redirectToTcaViewer(
+		HttpClientInterface $client,
+		string $loggedInUser,
+		string $locale,
+		string $id,
+		int $retries = 0
+	): RedirectResponse {
+		// Load config settings. This lives in .copypatrol.ini and is shared with the bot.
+		$projectDir = $this->getParameter( 'kernel.project_dir' );
+		$config = parse_ini_file( "$projectDir/.copypatrol.ini" );
+		// Request the viewer URL from Turnitin.
+		$response = $client->request(
+			'POST',
+			"https://{$config['domain']}/api/v1/submissions/$id/viewer-url",
+			[
+				'json' => [
+					'viewer_user_id' => $loggedInUser,
+					'locale' => $locale,
+					'viewer_permissions' => [
+						'may_view_submission_full_source' => true,
+						'may_view_match_submission_info' => true,
+						'may_view_document_details_panel' => true,
+						'may_view_sections_exclusion_panel' => true,
+					],
+				],
+				'headers' => [
+					'Authorization' => "Bearer {$config['key']}",
+					'From' => '[email protected]',
+					'User-Agent' => "copypatrol/{$config['version']}",
+					'X-Turnitin-Integration-Name' => 'CopyPatrol',
+					'X-Turnitin-Integration-Version' => $config['version'],
+				],
+			]
+		);
+
+		if ( $response->getStatusCode() !== Response::HTTP_OK ) {
+			if ( $retries > 5 ) {
+				throw new HttpException(
+					Response::HTTP_BAD_GATEWAY,
+					'Failed to fetch URL from the Turnitin Core API'
+				);
+			}
+			sleep( $retries + 1 );
+			return $this->redirectToTcaViewer( $client, $loggedInUser, $locale, $id, $retries + 1 );
+		}
+
+		return $this->redirect( json_decode( $response->getContent() )->viewer_url );
+	}
+
 	/**
 	 * ToolforgeBundle's logout apparently doesn't work :(
 	 *

src/Repository/CopyPatrolRepository.php

@@ -197,10 +197,10 @@ public function updateCopyvioAssessment(
 	/**
 	 * Get a particular record by submission ID.
 	 *
-	 * @param int $submissionId ID of record.
+	 * @param string $submissionId ID of record.
 	 * @return array Query result.
 	 */
-	public function getRecordById( int $submissionId ) {
+	public function getRecordById( string $submissionId ) {
 		$sql = "SELECT * FROM diffs WHERE submission_id = :id";
 		return $this->client->fetchAssociative( $sql, [
 			'id' => $submissionId