Have Parsoid\Config\PageConfigFactory take a rev instead of wikitext

* This let us pass mocked revisions in the parser test runner while
  running in Parsoid mode.

* This leads to improvement in wt2html tests results where a revision
  id is queried. I've verified this in the Cite extension repo as
  also the main parserTests.text file but I cannot enable Parsoid
  integrated testing on the main parser tests file without doing a
  sweep over all parser tests and adding appropriate test sections

* Currently, PageConfigFactory doesn't have unit tests. Will look
  into adding them separately in a followup.

* Moved the setupParsoidTransform function to a more suitable place
  in the ParserTestRunner.php file.

Bug: T270310
Change-Id: I94d68c8528bb2f7b367c68d80d14ebc1ab904a7f

includes/Rest/Handler/ParsoidHTMLHelper.php

@@ -34,7 +34,6 @@ use MediaWiki\Revision\RevisionRecord;
 use ParserCache;
 use ParserOptions;
 use ParserOutput;
-use TitleValue;
 use User;
 use Wikimedia\Message\MessageValue;
 use Wikimedia\ParamValidator\ParamValidator;
@@ -230,11 +229,7 @@ class ParsoidHTMLHelper {
 		// TODO: make ParsoidPageConfigFactory take PageReference as well
 		return MediaWikiServices::getInstance()
 			->get( 'ParsoidPageConfigFactory' )
-			->create(
-				TitleValue::newFromPage( $this->page ),
-				null,
-				$this->revision ? $this->revision->getId() : null
-			);
+			->create( $this->page, null, $this->revision );
 	}
 
 	/**

includes/Rest/Handler/ParsoidHandler.php

@@ -31,7 +31,9 @@ use MediaWiki\Rest\HttpException;
 use MediaWiki\Rest\LocalizedHttpException;
 use MediaWiki\Rest\Response;
 use MediaWiki\Rest\ResponseException;
+use MediaWiki\Revision\MutableRevisionRecord;
 use MediaWiki\Revision\RevisionAccessException;
+use MediaWiki\Revision\SlotRecord;
 use MobileContext;
 use ParserOutput;
 use RequestContext;
@@ -53,6 +55,7 @@ use Wikimedia\Parsoid\Utils\ContentUtils;
 use Wikimedia\Parsoid\Utils\DOMCompat;
 use Wikimedia\Parsoid\Utils\DOMUtils;
 use Wikimedia\Parsoid\Utils\Timing;
+use WikitextContent;
 
 /**
  * Base class for Parsoid handlers.
@@ -364,6 +367,24 @@ abstract class ParsoidHandler extends Handler {
 			throw new LogicException( 'Title not found!' );
 		}
 		$user = RequestContext::getMain()->getUser();
+
+		if ( $wikitextOverride === null ) {
+			$revisionRecord = null;
+		} else {
+			// Create a mutable revision record point to the same revision
+			// and set to the desired wikitext.
+			$revisionRecord = new MutableRevisionRecord( $title );
+			if ( $revision !== null ) {
+				$revisionRecord->setId( $revision );
+			}
+			$revisionRecord->setSlot(
+				SlotRecord::newUnsaved(
+					SlotRecord::MAIN,
+					new WikitextContent( $wikitextOverride )
+				)
+			);
+		}
+
 		// Note: Parsoid by design isn't supposed to use the user
 		// context right now, and all user state is expected to be
 		// introduced as a post-parse transform.  So although we pass a
@@ -371,7 +392,7 @@ abstract class ParsoidHandler extends Handler {
 		// corner cases; see PageConfigFactory::create() for more.
 		// @phan-suppress-next-line PhanUndeclaredMethod method defined in subtype
 		return $this->pageConfigFactory->create(
-			$title, $user, $revision, $wikitextOverride, $pagelanguageOverride,
+			$title, $user, $revisionRecord ?? $revision, null, $pagelanguageOverride,
 			$this->parsoidSettings
 		);
 	}

includes/parser/Parsoid/Config/PageConfigFactory.php

@@ -19,9 +19,8 @@
 
 namespace MediaWiki\Parser\Parsoid\Config;
 
-use MediaWiki\Linker\LinkTarget;
 use MediaWiki\Logger\LoggerFactory;
-use MediaWiki\Revision\MutableRevisionRecord;
+use MediaWiki\Page\PageIdentity;
 use MediaWiki\Revision\RevisionAccessException;
 use MediaWiki\Revision\RevisionRecord;
 use MediaWiki\Revision\RevisionStore;
@@ -63,41 +62,52 @@ class PageConfigFactory extends \Wikimedia\Parsoid\Config\PageConfigFactory {
 	 *
 	 * Note that Parsoid isn't supposed to use the user context by design; all
 	 * user-specific processing is expected to be introduced as a post-parse
-	 * transform.  The $user parameter is therefore usually null, especially
+	 * transform. The $user parameter is therefore usually null, especially
 	 * in background job parsing, although there are corner cases during
 	 * extension processing where a non-null $user could affect the output.
 	 *
-	 * @param LinkTarget $title The page represented by the PageConfig.
+	 * @param PageIdentity $pageId The page represented by the PageConfig.
 	 * @param ?UserIdentity $user User who is doing rendering (for parsing options).
-	 * @param ?int $revisionId The revision of the page.
-	 * @param ?string $wikitextOverride Wikitext to use instead of the
-	 *   contents of the specific $revision; used when $revision is null
-	 *   (a new page) or when we are parsing a stashed text.
+	 * @param int|RevisionRecord|null $revision Revision id or a revision record
+	 * @param ?string $unused
 	 * @param ?string $pagelanguageOverride
 	 * @param ?array $parsoidSettings Used to enable the debug API if requested
 	 * @return \Wikimedia\Parsoid\Config\PageConfig
 	 */
 	public function create(
-		LinkTarget $title,
+		PageIdentity $pageId,
 		?UserIdentity $user = null,
-		?int $revisionId = null,
-		?string $wikitextOverride = null,
+		$revision = null,
+		?string $unused = null, /* Added to mollify CI with cross-repo uses */
 		?string $pagelanguageOverride = null,
 		?array $parsoidSettings = null
 	): \Wikimedia\Parsoid\Config\PageConfig {
-		$title = Title::newFromLinkTarget( $title );
+		$title = Title::castFromPageIdentity( $pageId );
+		'@phan-var Title $title';
 
 		if ( !empty( $parsoidSettings['debugApi'] ) ) {
-			return ApiPageConfig::fromSettings( $parsoidSettings, [
-				"title" => $title->getPrefixedText(),
-				"pageContent" => $wikitextOverride,
-				"pageLanguage" => $pagelanguageOverride,
-				"revid" => $revisionId,
-				"loadData" => true,
-			] );
+			if ( $revision === null ) {
+				throw new \InvalidArgumentException(
+					"Revision not provided. Cannot lookup revision via debug API." );
+			}
+
+			$content = $revision->getContent( SlotRecord::MAIN );
+			if ( $content instanceof WikitextContent ) {
+				$wtContent = $content->getText();
+				return ApiPageConfig::fromSettings( $parsoidSettings, [
+					"title" => $title->getPrefixedText(),
+					"pageContent" => $wtContent,
+					"pageLanguage" => $pagelanguageOverride,
+					"revid" => $revision->getId(),
+					"loadData" => true,
+				] );
+			} else {
+				throw new \UnexpectedValueException(
+					"Non-wikitext content models not supported by debug API" );
+			}
 		}
 
-		if ( $revisionId === null ) {
+		if ( $revision === null ) {
 			// Fetch the 'latest' revision for the given title.
 			// Note: This initial fetch of the page context revision is
 			// *not* using Parser::fetchCurrentRevisionRecordOfTitle()
@@ -110,31 +120,31 @@ class PageConfigFactory extends \Wikimedia\Parsoid\Config\PageConfigFactory {
 			) ?: null;
 			// Note that $revisionRecord could still be null here if no
 			// page with that $title yet exists.
+		} elseif ( !is_int( $revision ) ) {
+			$revisionRecord = $revision;
 		} else {
 			// Fetch the correct revision record by the supplied id.
 			// This accesses the replica DB and may (or may not) fail over to
 			// the primary DB if the revision isn't found.
-			$revisionRecord = $this->revisionStore->getRevisionById(
-				$revisionId
-			);
+			$revisionRecord = $this->revisionStore->getRevisionById( $revision );
 			if ( $revisionRecord === null ) {
 				// This revision really ought to exist.  Check the primary DB.
 				// This *could* cause two requests to the primary DB if there
 				// were pending writes, but this codepath should be very rare.
 				// [T259855]
 				$revisionRecord = $this->revisionStore->getRevisionById(
-					$revisionId, RevisionStore::READ_LATEST
+					$revision, RevisionStore::READ_LATEST
 				);
 				$success = ( $revisionRecord !== null ) ? 'success' : 'failure';
 				LoggerFactory::getInstance( 'Parsoid' )->error(
 					"Retried revision fetch after failure: {$success}", [
-						'id' => $revisionId,
+						'id' => $revision,
 						'title' => $title->getPrefixedText(),
 					]
 				);
 			}
 			if ( $revisionRecord === null ) {
-				throw new RevisionAccessException( "Can't find revision {$revisionId}" );
+				throw new RevisionAccessException( "Can't find revision {$revision}" );
 			}
 		}
 
@@ -149,24 +159,6 @@ class PageConfigFactory extends \Wikimedia\Parsoid\Config\PageConfigFactory {
 			throw new RevisionAccessException( 'Not an available content version.' );
 		}
 
-		if ( $wikitextOverride !== null ) {
-			if ( $revisionRecord ) {
-				// PORT-FIXME this is not really the right thing to do; need
-				// a clone-like constructor for MutableRevisionRecord
-				$revisionRecord = MutableRevisionRecord::newFromParentRevision(
-					$revisionRecord
-				);
-			} else {
-				$revisionRecord = new MutableRevisionRecord( $title );
-			}
-			$revisionRecord->setSlot(
-				SlotRecord::newUnsaved(
-					SlotRecord::MAIN,
-					new WikitextContent( $wikitextOverride )
-				)
-			);
-		}
-
 		$parserOptions =
 			$user
 			? ParserOptions::newFromUser( $user )

tests/parser/ParserTestRunner.php

@@ -1572,35 +1572,6 @@ class ParserTestRunner {
 		);
 	}
 
-	private function setupParsoidTransform( ParserTest $test ): array {
-		$services = MediaWikiServices::getInstance();
-		$pageConfigFactory = $services->get( 'ParsoidPageConfigFactory' );
-		$pageConfig = null;
-		[ $title, $options, $revId ] = $this->setupParserOptions(
-			$test,
-			static function ( $context, $title, $revId, $wikitext ) use ( $pageConfigFactory, &$pageConfig ) {
-				$pageConfig = $pageConfigFactory->create(
-					$title,
-					$context->getUser(),
-					// @todo T270310: Parsoid doesn't have a mechanism
-					// to override revid with a fake revision, like the
-					// legacy parser does, so {{REVISIONID}} will be
-					// 'wrong' in parser tests.  Probably need to
-					// override
-					// ParserOptions::getCurrentRevisionRecordCallback()
-					// (like we do for the 'lastsavedrevision' option
-					// below) in order to fix this.
-					null/*$revId*/,
-					// @todo T270310: Parsoid should really accept a
-					// RevisionRecord here, instead of raw wikitext.
-					$wikitext,
-					$context->getLanguage()->getCode()
-				);
-				return $pageConfig->getParserOptions();
-			} );
-		return [ $pageConfig, $title, $options, $revId ];
-	}
-
 	/**
 	 * @param Parsoid $parsoid
 	 * @param PageConfig $pageConfig
@@ -1771,6 +1742,38 @@ class ParserTestRunner {
 		}
 	}
 
+	private function setupParsoidTransform( ParserTest $test ): array {
+		$services = MediaWikiServices::getInstance();
+		$pageConfigFactory = $services->get( 'ParsoidPageConfigFactory' );
+		$pageConfig = null;
+		$runner = $this;
+		[ $title, $options, $revId ] = $this->setupParserOptions(
+			$test,
+			static function ( $context, $title, $revId, $wikitext ) use ( $runner, $pageConfigFactory, &$pageConfig ) {
+				$user = $context->getUser();
+				$content = new WikitextContent( $wikitext );
+				$title = Title::newFromRow( (object)[
+					'page_id' => 187,
+					'page_len' => $content->getSize(),
+					'page_latest' => 1337,
+					'page_namespace' => $title->getNamespace(),
+					'page_title' => $title->getDBkey(),
+					'page_is_redirect' => 0
+				] );
+				$revRecord = new MutableRevisionRecord( $title );
+				$revRecord->setContent( SlotRecord::MAIN, $content )
+					->setUser( $user )
+					->setTimestamp( strval( $runner->getFakeTimestamp() ) )
+					->setPageId( $title->getArticleID() )
+					->setId( $title->getLatestRevID() );
+				$pageConfig = $pageConfigFactory->create(
+					$title, $user, $revRecord, $context->getLanguage()->getCode()
+				);
+				return $pageConfig->getParserOptions();
+			} );
+		return [ $pageConfig, $title, $options, $revId ];
+	}
+
 	/**
 	 * Helper function to register a Parsoid extension module in such a
 	 * way that it can be cleaned up after the test is complete.