50 parent::__construct();
51 $this->
addDescription(
'Compares variant conversion output between Parser and HtmlOutputRendererHelper' );
54 'Name of the page to be parsed and compared',
59 'Target variant language code to transform the content to',
68 $pageName = $this->
getArg(
'page-title' );
69 $pageTitle = Title::newFromText( $pageName );
71 if ( !$pageTitle || !$pageTitle->
exists() ) {
72 $this->
fatalError(
"Title with name $pageName not found" );
75 $targetVariantCode = $this->
getArg(
'target-variant' );
76 $languageNameUtils = $mwInstance->getLanguageNameUtils();
77 if ( !$languageNameUtils->isValidBuiltInCode( $targetVariantCode ) ) {
78 $this->
fatalError(
"$targetVariantCode is not a supported variant" );
80 $targetVariant = $mwInstance->getLanguageFactory()->getLanguage(
84 $user = User::newSystemUser( User::MAINTENANCE_SCRIPT_USER, [
'steal' =>
true ] );
87 $parserOutput = $this->getParserOutput( $pageTitle, $baseLanguage, $targetVariant );
88 $parsoidOutput = $this->getParsoidOutput( $pageTitle, $targetVariant, $user );
89 $converterUsed = $this->getConverterUsed( $parsoidOutput );
91 $this->compareOutput( $parserOutput->getContentHolderText(), $parsoidOutput->
getContentHolderText(),
100 new ServiceOptions( PageRestHelperFactory::CONSTRUCTOR_OPTIONS, $services->getMainConfig() ),
101 $services->getRevisionLookup(),
102 $services->getRevisionRenderer(),
103 $services->getTitleFormatter(),
104 $services->getPageStore(),
105 $services->getParsoidOutputStash(),
106 $services->getParserOutputAccess(),
107 $services->getParsoidSiteConfig(),
108 $services->getHtmlTransformFactory(),
109 $services->getContentHandlerFactory(),
110 $services->getLanguageFactory(),
111 $services->getRedirectStore(),
112 $services->getLanguageConverterFactory(),
113 $services->getTitleFactory(),
114 $services->getConnectionProvider(),
115 $services->getChangeTagsStore(),
116 StatsFactory::newNull()
123 $parserOpts->disableContentConversion(
false );
124 $parserOpts->disableTitleConversion(
false );
129 private function getParserOutput(
139 $mwInstance = $this->getServiceContainer();
141 $languageFactory = $mwInstance->getLanguageFactory();
142 $parser = $mwInstance->getParser();
143 $parserOptions = $this->getParserOptions(
144 $languageFactory->getParentLanguage( $baseLanguage )
147 $content = $mwInstance->getRevisionLookup()
148 ->getRevisionByTitle( $pageTitle )
149 ->getContent( SlotRecord::MAIN );
150 $wikiContent = ( $content instanceof
TextContent ) ? $content->getText() :
'';
152 $po = $parser->parse( $wikiContent, $pageTitle, $parserOptions );
154 $pipeline = $mwInstance->getDefaultOutputPipeline();
155 $options = [
'deduplicateStyles' => false ];
156 return $pipeline->run( $po, $parserOptions, $options );
159 private function getParsoidOutput(
161 Bcp47Code $targetVariant,
165 $htmlOutputRendererHelper = $this->newPageRestHelperFactory()->newHtmlOutputRendererHelper( $pageTitle, [
168 ], $user,
null,
false, $parserOptions );
169 $htmlOutputRendererHelper->setVariantConversionLanguage( $targetVariant );
171 $po = $htmlOutputRendererHelper->getHtml();
172 $pipeline = $this->getServiceContainer()->getDefaultOutputPipeline();
173 $options = [
'deduplicateStyles' => false ];
174 return $pipeline->run( $po, $parserOptions, $options );
177 private function getWords(
string $output ): array {
178 $tagsRemoved = strip_tags( $output );
179 $words = preg_split(
'/\s+/', trim( $tagsRemoved ), -1, PREG_SPLIT_NO_EMPTY );
183 private function getBody(
string $output ): string {
184 $dom = new DOMDocument();
186 @$dom->loadHTML( $output );
187 $body = $dom->getElementsByTagName(
'body' )->item( 0 );
188 if ( $body ===
null ) {
193 return $body->textContent;
196 private function compareOutput(
199 string $converterUsed
201 $parsoidWords = $this->getWords( $this->getBody( $parsoidText ) );
202 $parserWords = $this->getWords( $parserText );
204 $parserWordCount = count( $parserWords );
205 $parsoidWordCount = count( $parsoidWords );
206 $this->output(
"Word count: Parsoid: $parsoidWordCount; Parser: $parserWordCount\n" );
208 $this->outputSimilarity( $parsoidWords, $parserWords );
209 $this->output(
"\n" );
210 $this->outputDiff( $parsoidWords, $parserWords, $converterUsed );
213 private function getConverterUsed(
ParserOutput $parsoidOutput ): string {
214 $isCoreConverterUsed = strpos(
215 $parsoidOutput->getRawText(),
216 'Variant conversion performed using the core LanguageConverter'
219 if ( $isCoreConverterUsed ) {
220 return 'Core LanguageConverter';
222 return 'Parsoid LanguageConverter';
227 private function mb_sprintf(
string $format, ...$args ): string {
231 preg_replace_callback(
232 '/(?<=%|%-)\d+(?=s)/',
233 static function ( array
$matches ) use ( &$params ) {
234 $value = array_shift( $params );
236 return (
string)( strlen( $value ) - mb_strlen( $value ) +
$matches[0] );
244 private function outputSimilarity( array $parsoidWords, array $parserWords ): void {
245 $parsoidOutput = implode(
' ', $parsoidWords );
246 $parserOutput = implode(
' ', $parserWords );
248 'Total characters: Parsoid: ' . strlen( $parsoidOutput ) .
249 '; Parser: ' . strlen( $parserOutput ) .
"\n"
252 $similarityPercent = 0;
253 $similarCharacters = similar_text( $parsoidOutput, $parserOutput, $similarityPercent );
254 $similarityPercent = round( $similarityPercent, 2 );
257 "Similarity via similar_text(): $similarityPercent%; Similar characters: $similarCharacters"
261 private function outputDiff( array $parsoidWords, array $parserWords,
string $converterUsed ): void {
262 $out = str_repeat(
'-', 96 ) .
"\n";
263 $out .= sprintf(
"| %5s | %-35s | %-35s | %-8s |\n",
'Line',
'Parsoid',
'Parser',
'Diff' );
264 $out .= sprintf(
"| %5s | %-35s | %-35s | %-8s |\n",
'',
"($converterUsed)",
'',
'' );
265 $out .= str_repeat(
'-', 96 ) .
"\n";
268 $diff =
new Diff( $parsoidWords, $parserWords );
270 $this->output( $e->getMessage() );
271 $this->error(
'Encountered ComplexityException while computing diff' );
276 foreach ( $wordDiffFormat as $index => $wordDiff ) {
277 $action = $wordDiff[
'action'];
278 $old = $wordDiff[
'old'] ??
null;
279 $new = $wordDiff[
'new'] ??
null;
281 $out .= $this->mb_sprintf(
282 "| %5s | %-35s | %-35s | %-8s |\n",
283 str_pad( (
string)( $index + 1 ), 5,
' ', STR_PAD_LEFT ),
284 mb_strimwidth( $old ??
'- N/A -', 0, 35,
'…' ),
285 mb_strimwidth( $new ??
'- N/A -', 0, 35,
'…' ),
291 $out .= str_repeat(
'-', 96 ) .
"\n";
292 $this->output(
"\n" . $out );