wiki.techinc.nl/includes/content/WikitextContentHandler.php
C. Scott Ananian a0503debb0 Provide previous parse results to parser when rendering
This patch lays the groundwork for incremental/selective parsing in
Parsoid by ensuring that we can pass previous cached parses through
the parse pipeline to Parsoid.  We do this by adding a new render
hint type, `previous-output`, and ensuring it is passed along.

Because revisions can contain a ParserOutput which is the combination
of separate ParserOutput objects for each of their slots, RenderedRevision
also contains a method to unsplit the combined ParserOutput to reconstruct
an original ParserOutput for use in incremental parsing.  Currently this
is mostly a stub, but illustrates how slot combination and splitting can
work, assuming those transformations are reversible.

Extra calls to ParserCache::getDirty() are added to some code paths
in order to ensure that any previously-cached ParserOutput is available
for selective update.  In order to mitigate any performance concerns,
these are only done for the Parsoid parser at the moment.  Future
patches will add additional metrics to quantify the cost/benefit ratio
of the additional cache lookups on these paths.

Bug: T363421
Bug: T371713
Change-Id: I440884f1d7e09c1ff9806f848b7b53a636367690
2024-08-23 17:41:55 -04:00

430 lines
13 KiB
PHP

<?php
/**
* Content handler for wiki text pages.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License along
* with this program; if not, write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
* http://www.gnu.org/copyleft/gpl.html
*
* @since 1.21
*
* @file
* @ingroup Content
*/
namespace MediaWiki\Content;
use MediaWiki\Content\Renderer\ContentParseParams;
use MediaWiki\Content\Transform\PreloadTransformParams;
use MediaWiki\Content\Transform\PreSaveTransformParams;
use MediaWiki\Languages\LanguageNameUtils;
use MediaWiki\Linker\LinkRenderer;
use MediaWiki\Logger\LoggerFactory;
use MediaWiki\Parser\MagicWordFactory;
use MediaWiki\Parser\ParserOutput;
use MediaWiki\Parser\ParserOutputFlags;
use MediaWiki\Parser\Parsoid\ParsoidParserFactory;
use MediaWiki\Revision\RevisionRecord;
use MediaWiki\Title\Title;
use MediaWiki\Title\TitleFactory;
use ParserFactory;
use SearchEngine;
use SearchIndexField;
use Wikimedia\UUID\GlobalIdGenerator;
use WikiPage;
/**
* Content handler for wiki text pages.
*
* @ingroup Content
*/
class WikitextContentHandler extends TextContentHandler {
private TitleFactory $titleFactory;
private ParserFactory $parserFactory;
private GlobalIdGenerator $globalIdGenerator;
private LanguageNameUtils $languageNameUtils;
private LinkRenderer $linkRenderer;
private MagicWordFactory $magicWordFactory;
private ParsoidParserFactory $parsoidParserFactory;
public function __construct(
string $modelId,
TitleFactory $titleFactory,
ParserFactory $parserFactory,
GlobalIdGenerator $globalIdGenerator,
LanguageNameUtils $languageNameUtils,
LinkRenderer $linkRenderer,
MagicWordFactory $magicWordFactory,
ParsoidParserFactory $parsoidParserFactory
) {
// $modelId should always be CONTENT_MODEL_WIKITEXT
parent::__construct( $modelId, [ CONTENT_FORMAT_WIKITEXT ] );
$this->titleFactory = $titleFactory;
$this->parserFactory = $parserFactory;
$this->globalIdGenerator = $globalIdGenerator;
$this->languageNameUtils = $languageNameUtils;
$this->linkRenderer = $linkRenderer;
$this->magicWordFactory = $magicWordFactory;
$this->parsoidParserFactory = $parsoidParserFactory;
}
/**
* @return class-string<WikitextContent>
*/
protected function getContentClass() {
return WikitextContent::class;
}
/**
* Returns a WikitextContent object representing a redirect to the given destination page.
*
* @param Title $destination The page to redirect to.
* @param string $text Text to include in the redirect, if possible.
*
* @return Content
*
* @see ContentHandler::makeRedirectContent
*/
public function makeRedirectContent( Title $destination, $text = '' ) {
$optionalColon = '';
if ( $destination->getNamespace() === NS_CATEGORY ) {
$optionalColon = ':';
} else {
$iw = $destination->getInterwiki();
if ( $iw && $this->languageNameUtils->getLanguageName( $iw,
LanguageNameUtils::AUTONYMS,
LanguageNameUtils::DEFINED
) ) {
$optionalColon = ':';
}
}
$mwRedir = $this->magicWordFactory->get( 'redirect' );
$redirectText = $mwRedir->getSynonym( 0 ) .
' [[' . $optionalColon . $destination->getFullText() . ']]';
if ( $text != '' ) {
$redirectText .= "\n" . $text;
}
$class = $this->getContentClass();
return new $class( $redirectText );
}
/**
* Returns true because wikitext supports redirects.
*
* @return bool Always true.
*
* @see ContentHandler::supportsRedirects
*/
public function supportsRedirects() {
return true;
}
/**
* Returns true because wikitext supports sections.
*
* @return bool Always true.
*
* @see ContentHandler::supportsSections
*/
public function supportsSections() {
return true;
}
/**
* Returns true, because wikitext supports caching using the
* ParserCache mechanism.
*
* @since 1.21
*
* @return bool Always true.
*
* @see ContentHandler::isParserCacheSupported
*/
public function isParserCacheSupported() {
return true;
}
/** @inheritDoc */
public function supportsPreloadContent(): bool {
return true;
}
/**
* @return FileContentHandler
*/
protected function getFileHandler() {
return new FileContentHandler(
$this->getModelID(),
$this->titleFactory,
$this->parserFactory,
$this->globalIdGenerator,
$this->languageNameUtils,
$this->linkRenderer,
$this->magicWordFactory,
$this->parsoidParserFactory
);
}
public function getFieldsForSearchIndex( SearchEngine $engine ) {
$fields = parent::getFieldsForSearchIndex( $engine );
$fields['heading'] =
$engine->makeSearchFieldMapping( 'heading', SearchIndexField::INDEX_TYPE_TEXT );
$fields['heading']->setFlag( SearchIndexField::FLAG_SCORING );
$fields['auxiliary_text'] =
$engine->makeSearchFieldMapping( 'auxiliary_text', SearchIndexField::INDEX_TYPE_TEXT );
$fields['opening_text'] =
$engine->makeSearchFieldMapping( 'opening_text', SearchIndexField::INDEX_TYPE_TEXT );
$fields['opening_text']->setFlag(
SearchIndexField::FLAG_SCORING | SearchIndexField::FLAG_NO_HIGHLIGHT
);
// Until we have the full first-class content handler for files, we invoke it explicitly here
return array_merge( $fields, $this->getFileHandler()->getFieldsForSearchIndex( $engine ) );
}
public function getDataForSearchIndex(
WikiPage $page,
ParserOutput $parserOutput,
SearchEngine $engine,
?RevisionRecord $revision = null
) {
$fields = parent::getDataForSearchIndex( $page, $parserOutput, $engine, $revision );
$structure = new WikiTextStructure( $parserOutput );
$fields['heading'] = $structure->headings();
// text fields
$fields['opening_text'] = $structure->getOpeningText();
$fields['text'] = $structure->getMainText(); // overwrites one from ContentHandler
$fields['auxiliary_text'] = $structure->getAuxiliaryText();
$fields['defaultsort'] = $structure->getDefaultSort();
$fields['file_text'] = null;
// Until we have the full first-class content handler for files, we invoke it explicitly here
if ( $page->getTitle()->getNamespace() === NS_FILE ) {
$fields = array_merge(
$fields,
$this->getFileHandler()->getDataForSearchIndex( $page, $parserOutput, $engine, $revision )
);
}
return $fields;
}
/**
* Returns the content's text as-is.
*
* @param Content $content
* @param string|null $format The serialization format to check
*
* @return mixed
*/
public function serializeContent( Content $content, $format = null ) {
$this->checkFormat( $format );
return parent::serializeContent( $content, $format );
}
public function preSaveTransform(
Content $content,
PreSaveTransformParams $pstParams
): Content {
'@phan-var WikitextContent $content';
$text = $content->getText();
$parser = $this->parserFactory->getInstance();
$pst = $parser->preSaveTransform(
$text,
$pstParams->getPage(),
$pstParams->getUser(),
$pstParams->getParserOptions()
);
if ( $text === $pst ) {
return $content;
}
$contentClass = $this->getContentClass();
$ret = new $contentClass( $pst );
$ret->setPreSaveTransformFlags( $parser->getOutput()->getAllFlags() );
return $ret;
}
/**
* Returns a Content object with preload transformations applied (or this
* object if no transformations apply).
*
* @param Content $content
* @param PreloadTransformParams $pltParams
*
* @return Content
*/
public function preloadTransform(
Content $content,
PreloadTransformParams $pltParams
): Content {
'@phan-var WikitextContent $content';
$text = $content->getText();
$plt = $this->parserFactory->getInstance()->getPreloadText(
$text,
$pltParams->getPage(),
$pltParams->getParserOptions(),
$pltParams->getParams()
);
$contentClass = $this->getContentClass();
return new $contentClass( $plt );
}
/**
* Extract the redirect target and the remaining text on the page.
*
* @since 1.41 (used to be a method on WikitextContent since 1.23)
*
* @return array List of two elements: LinkTarget|null and WikitextContent object.
*/
public function extractRedirectTargetAndText( WikitextContent $content ): array {
$redir = $this->magicWordFactory->get( 'redirect' );
$text = ltrim( $content->getText() );
if ( !$redir->matchStartAndRemove( $text ) ) {
return [ null, $content ];
}
// Extract the first link and see if it's usable
// Ensure that it really does come directly after #REDIRECT
// Some older redirects included a colon, so don't freak about that!
$m = [];
if ( preg_match( '!^\s*:?\s*\[{2}(.*?)(?:\|.*?)?\]{2}\s*!', $text, $m ) ) {
// Strip preceding colon used to "escape" categories, etc.
// and URL-decode links
if ( strpos( $m[1], '%' ) !== false ) {
// Match behavior of inline link parsing here;
$m[1] = rawurldecode( ltrim( $m[1], ':' ) );
}
// TODO: Move isValidRedirectTarget() out Title, so we can use a TitleValue here.
$title = $this->titleFactory->newFromText( $m[1] );
// If the title is a redirect to bad special pages or is invalid, return null
if ( !$title instanceof Title || !$title->isValidRedirectTarget() ) {
return [ null, $content ];
}
$remainingContent = new WikitextContent( substr( $text, strlen( $m[0] ) ) );
return [ $title, $remainingContent ];
}
return [ null, $content ];
}
/**
* Returns a ParserOutput object resulting from parsing the content's text
* using the global Parser service.
*
* @since 1.38
*
* @param Content $content
* @param ContentParseParams $cpoParams
* @param ParserOutput &$parserOutput The output object to fill (reference).
*/
protected function fillParserOutput(
Content $content,
ContentParseParams $cpoParams,
ParserOutput &$parserOutput
) {
'@phan-var WikitextContent $content';
$title = $this->titleFactory->newFromPageReference( $cpoParams->getPage() );
$parserOptions = $cpoParams->getParserOptions();
$revId = $cpoParams->getRevId();
[ $redir, $contentWithoutRedirect ] = $this->extractRedirectTargetAndText( $content );
if ( $parserOptions->getUseParsoid() ) {
$parser = $this->parsoidParserFactory->create();
// Parsoid renders the #REDIRECT magic word as an invisible
// <link> tag and doesn't require it to be stripped.
// T349087: ...and in fact, RESTBase relies on getting
// redirect information from this <link> tag, so it needs
// to be present.
// Further, Parsoid can accept a Content in place of a string.
$text = $content;
$extraArgs = [ $cpoParams->getPreviousOutput() ];
} else {
// The legacy parser requires the #REDIRECT magic word to
// be stripped from the content before parsing.
$parser = $this->parserFactory->getInstance();
$text = $contentWithoutRedirect->getText();
$extraArgs = [];
}
$time = -microtime( true );
$parserOutput = $parser
->parse( $text, $title, $parserOptions, true, true, $revId, ...$extraArgs );
$time += microtime( true );
// Timing hack
if ( $time > 3 ) {
// TODO: Use Parser's logger (once it has one)
$channel = $parserOptions->getUseParsoid() ? 'slow-parsoid' : 'slow-parse';
$logger = LoggerFactory::getInstance( $channel );
$logger->info( 'Parsing {title} was slow, took {time} seconds', [
'time' => number_format( $time, 2 ),
'title' => (string)$title,
'trigger' => $parserOptions->getRenderReason(),
] );
}
// T330667: Record the fact that we used the value of
// 'useParsoid' to influence this parse. Note that
// ::getUseParsoid() has a side-effect on $parserOutput here
// which didn't occur when we called ::getUseParsoid() earlier
// because $parserOutput didn't exist at that time.
$parserOptions->getUseParsoid();
// Add redirect indicator at the top
if ( $redir ) {
// Make sure to include the redirect link in pagelinks
$parserOutput->addLink( $redir );
if ( $cpoParams->getGenerateHtml() ) {
$parserOutput->setRedirectHeader(
$this->linkRenderer->makeRedirectHeader(
$title->getPageLanguage(), $redir, false
)
);
$parserOutput->addModuleStyles( [ 'mediawiki.action.view.redirectPage' ] );
} else {
$parserOutput->setRawText( null );
}
}
// Pass along user-signature flag
if ( in_array( 'user-signature', $content->getPreSaveTransformFlags() ) ) {
$parserOutput->setOutputFlag( ParserOutputFlags::USER_SIGNATURE );
}
}
}
/** @deprecated class alias since 1.43 */
class_alias( WikitextContentHandler::class, 'WikitextContentHandler' );