Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
20.81% |
31 / 149 |
|
0.00% |
0 / 5 |
CRAP | |
0.00% |
0 / 1 |
| ParsoidParser | |
20.81% |
31 / 149 |
|
0.00% |
0 / 5 |
335.43 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
2 | |||
| genParserOutput | |
0.00% |
0 / 80 |
|
0.00% |
0 / 1 |
110 | |||
| parse | |
88.57% |
31 / 35 |
|
0.00% |
0 / 1 |
9.12 | |||
| parseFakeRevision | |
0.00% |
0 / 13 |
|
0.00% |
0 / 1 |
12 | |||
| makeLimitReport | |
0.00% |
0 / 17 |
|
0.00% |
0 / 1 |
6 | |||
| 1 | <?php |
| 2 | |
| 3 | namespace MediaWiki\Parser\Parsoid; |
| 4 | |
| 5 | use MediaWiki\Content\TextContent; |
| 6 | use MediaWiki\Content\WikitextContent; |
| 7 | use MediaWiki\Context\RequestContext; |
| 8 | use MediaWiki\Languages\LanguageConverterFactory; |
| 9 | use MediaWiki\MainConfigNames; |
| 10 | use MediaWiki\MediaWikiServices; |
| 11 | use MediaWiki\Page\PageReference; |
| 12 | use MediaWiki\Parser\ParserFactory; |
| 13 | use MediaWiki\Parser\ParserOptions; |
| 14 | use MediaWiki\Parser\ParserOutput; |
| 15 | use MediaWiki\Parser\Parsoid\Config\PageConfigFactory; |
| 16 | use MediaWiki\Revision\MutableRevisionRecord; |
| 17 | use MediaWiki\Revision\RevisionRecord; |
| 18 | use MediaWiki\Revision\SlotRecord; |
| 19 | use MediaWiki\Title\Title; |
| 20 | use MediaWiki\WikiMap\WikiMap; |
| 21 | use Wikimedia\Assert\Assert; |
| 22 | use Wikimedia\Parsoid\Config\PageConfig; |
| 23 | use Wikimedia\Parsoid\Parsoid; |
| 24 | |
| 25 | /** |
| 26 | * Parser implementation which uses Parsoid. |
| 27 | * |
| 28 | * Currently incomplete; see T236809 for the long-term plan. |
| 29 | * |
| 30 | * @since 1.41 |
| 31 | * @unstable since 1.41; see T236809 for plan. |
| 32 | */ |
| 33 | class ParsoidParser /* eventually this will extend \Parser */ { |
| 34 | /** |
| 35 | * @unstable |
| 36 | * This should not be used widely right now since this may go away. |
| 37 | * This is being added to support DiscussionTools with Parsoid HTML |
| 38 | * and after initial exploration, this may be implemented differently. |
| 39 | */ |
| 40 | public const PARSOID_TITLE_KEY = "parsoid:title-dbkey"; |
| 41 | private Parsoid $parsoid; |
| 42 | private PageConfigFactory $pageConfigFactory; |
| 43 | private LanguageConverterFactory $languageConverterFactory; |
| 44 | private ParserFactory $legacyParserFactory; |
| 45 | |
| 46 | /** |
| 47 | * @param Parsoid $parsoid |
| 48 | * @param PageConfigFactory $pageConfigFactory |
| 49 | * @param LanguageConverterFactory $languageConverterFactory |
| 50 | * @param ParserFactory $legacyParserFactory |
| 51 | */ |
| 52 | public function __construct( |
| 53 | Parsoid $parsoid, |
| 54 | PageConfigFactory $pageConfigFactory, |
| 55 | LanguageConverterFactory $languageConverterFactory, |
| 56 | ParserFactory $legacyParserFactory |
| 57 | ) { |
| 58 | $this->parsoid = $parsoid; |
| 59 | $this->pageConfigFactory = $pageConfigFactory; |
| 60 | $this->languageConverterFactory = $languageConverterFactory; |
| 61 | $this->legacyParserFactory = $legacyParserFactory; |
| 62 | } |
| 63 | |
| 64 | /** |
| 65 | * Internal helper to avoid code deuplication across two methods |
| 66 | * |
| 67 | * @param PageConfig $pageConfig |
| 68 | * @param ParserOptions $options |
| 69 | * @return ParserOutput |
| 70 | */ |
| 71 | private function genParserOutput( |
| 72 | PageConfig $pageConfig, ParserOptions $options, ?ParserOutput $previousOutput |
| 73 | ): ParserOutput { |
| 74 | $parserOutput = new ParserOutput(); |
| 75 | |
| 76 | // Parsoid itself does not vary output by parser options right now. |
| 77 | // But, ensure that any option use by extensions, parser functions, |
| 78 | // recursive parses, or (in the unlikely future scenario) Parsoid itself |
| 79 | // are recorded as used. |
| 80 | $options->registerWatcher( [ $parserOutput, 'recordOption' ] ); |
| 81 | |
| 82 | // The enable/disable logic here matches that in Parser::internalParseHalfParsed(), |
| 83 | // although __NOCONTENTCONVERT__ is handled internal to Parsoid. |
| 84 | // |
| 85 | // T349137: It might be preferable to handle __NOCONTENTCONVERT__ here rather than |
| 86 | // by inspecting the DOM inside Parsoid. That will come in a separate patch. |
| 87 | $htmlVariantLanguage = null; |
| 88 | if ( !( $options->getDisableContentConversion() || $options->getInterfaceMessage() ) ) { |
| 89 | // NOTES (some of these are TODOs for read views integration) |
| 90 | // 1. This html variant conversion is a pre-cache transform. HtmlOutputRendererHelper |
| 91 | // has another variant conversion that is a post-cache transform based on the |
| 92 | // 'Accept-Language' header. If that header is set, there is really no reason to |
| 93 | // do this conversion here. So, eventually, we are likely to either not pass in |
| 94 | // the htmlVariantLanguage option below OR disable language conversion from the |
| 95 | // wt2html path in Parsoid and this and the Accept-Language variant conversion |
| 96 | // both would have to be handled as post-cache transforms. |
| 97 | // |
| 98 | // 2. Parser.php calls convert() which computes a preferred variant from the |
| 99 | // target language. But, we cannot do that unconditionally here because REST API |
| 100 | // requests specify the exact variant via the 'Content-Language' header. |
| 101 | // |
| 102 | // For Parsoid page views, either the callers will have to compute the |
| 103 | // preferred variant and set it in ParserOptions OR the REST API will have |
| 104 | // to set some other flag indicating that the preferred variant should not |
| 105 | // be computed. For now, I am adding a temporary hack, but this should be |
| 106 | // replaced with something more sensible (T267067). |
| 107 | // |
| 108 | // 3. Additionally, Parsoid's callers will have to set targetLanguage in ParserOptions |
| 109 | // to mimic the logic in Parser.php (missing right now). |
| 110 | $langCode = $pageConfig->getPageLanguageBcp47(); |
| 111 | if ( $options->getRenderReason() === 'page-view' ) { // TEMPORARY HACK |
| 112 | $langFactory = MediaWikiServices::getInstance()->getLanguageFactory(); |
| 113 | $lang = $langFactory->getLanguage( $langCode ); |
| 114 | $langConv = $this->languageConverterFactory->getLanguageConverter( $lang ); |
| 115 | $htmlVariantLanguage = $langFactory->getLanguage( $langConv->getPreferredVariant() ); |
| 116 | } else { |
| 117 | $htmlVariantLanguage = $langCode; |
| 118 | } |
| 119 | } |
| 120 | $oldPageConfig = null; |
| 121 | $oldPageBundle = null; |
| 122 | |
| 123 | // T371713: Temporary statistics collection code to determine |
| 124 | // feasibility of Parsoid selective update |
| 125 | $sampleRate = MediaWikiServices::getInstance()->getMainConfig()->get( |
| 126 | MainConfigNames::ParsoidSelectiveUpdateSampleRate |
| 127 | ); |
| 128 | $doSample = ( $sampleRate && mt_rand( 1, $sampleRate ) === 1 ); |
| 129 | if ( $doSample && $previousOutput !== null && $previousOutput->getCacheRevisionId() ) { |
| 130 | // Allow fetching the old wikitext corresponding to the |
| 131 | // $previousOutput |
| 132 | $oldPageConfig = $this->pageConfigFactory->create( |
| 133 | Title::newFromLinkTarget( $pageConfig->getLinkTarget() ), |
| 134 | $options->getUserIdentity(), |
| 135 | $previousOutput->getCacheRevisionId(), |
| 136 | null, |
| 137 | $previousOutput->getLanguage(), |
| 138 | ); |
| 139 | $oldPageBundle = |
| 140 | PageBundleParserOutputConverter::pageBundleFromParserOutput( |
| 141 | $previousOutput |
| 142 | ); |
| 143 | } |
| 144 | |
| 145 | $defaultOptions = [ |
| 146 | 'pageBundle' => true, |
| 147 | 'wrapSections' => true, |
| 148 | 'logLinterData' => true, |
| 149 | 'body_only' => false, |
| 150 | 'htmlVariantLanguage' => $htmlVariantLanguage, |
| 151 | 'offsetType' => 'byte', |
| 152 | 'outputContentVersion' => Parsoid::defaultHTMLVersion(), |
| 153 | 'previousOutput' => $oldPageBundle, |
| 154 | 'previousInput' => $oldPageConfig, |
| 155 | // The following are passed for metrics & labelling |
| 156 | 'sampleStats' => $doSample, |
| 157 | 'renderReason' => $options->getRenderReason(), |
| 158 | 'userAgent' => RequestContext::getMain()->getRequest()->getHeader( 'User-Agent' ), |
| 159 | ]; |
| 160 | |
| 161 | $parserOutput->resetParseStartTime(); |
| 162 | |
| 163 | // This can throw ClientError or ResourceLimitExceededException. |
| 164 | // Callers are responsible for figuring out how to handle them. |
| 165 | $pageBundle = $this->parsoid->wikitext2html( |
| 166 | $pageConfig, |
| 167 | $defaultOptions, |
| 168 | $headers, |
| 169 | $parserOutput ); |
| 170 | |
| 171 | $parserOutput = PageBundleParserOutputConverter::parserOutputFromPageBundle( $pageBundle, $parserOutput ); |
| 172 | |
| 173 | // Record the page title in dbkey form so that post-cache transforms |
| 174 | // have access to the title. |
| 175 | $parserOutput->setExtensionData( |
| 176 | self::PARSOID_TITLE_KEY, |
| 177 | Title::newFromLinkTarget( $pageConfig->getLinkTarget() )->getPrefixedDBkey() |
| 178 | ); |
| 179 | |
| 180 | // Register a watcher again because the $parserOutput arg |
| 181 | // and $parserOutput return value above are different objects! |
| 182 | $options->registerWatcher( [ $parserOutput, 'recordOption' ] ); |
| 183 | |
| 184 | $parserOutput->setFromParserOptions( $options ); |
| 185 | |
| 186 | $parserOutput->recordTimeProfile(); |
| 187 | $this->makeLimitReport( $options, $parserOutput ); |
| 188 | |
| 189 | // T371713: Collect statistics on parsing time -vs- presence of |
| 190 | // $previousOutput |
| 191 | $stats = MediaWikiServices::getInstance()->getStatsFactory(); |
| 192 | $labels = [ |
| 193 | 'type' => $previousOutput === null ? 'full' : 'selective', |
| 194 | 'wiki' => WikiMap::getCurrentWikiId(), |
| 195 | 'reason' => $options->getRenderReason() ?: 'unknown', |
| 196 | ]; |
| 197 | $stats |
| 198 | ->getCounter( 'Parsoid_parse_cpu_seconds' ) |
| 199 | ->setLabels( $labels ) |
| 200 | ->incrementBy( $parserOutput->getTimeProfile( 'cpu' ) ); |
| 201 | $stats |
| 202 | ->getCounter( 'Parsoid_parse_total' ) |
| 203 | ->setLabels( $labels ) |
| 204 | ->increment(); |
| 205 | |
| 206 | // Add Parsoid skinning module |
| 207 | $parserOutput->addModuleStyles( [ 'mediawiki.skinning.content.parsoid' ] ); |
| 208 | |
| 209 | // Record Parsoid version in extension data; this allows |
| 210 | // us to use the onRejectParserCacheValue hook to selectively |
| 211 | // expire "bad" generated content in the event of a rollback. |
| 212 | $parserOutput->setExtensionData( |
| 213 | 'core:parsoid-version', Parsoid::version() |
| 214 | ); |
| 215 | $parserOutput->setExtensionData( |
| 216 | 'core:html-version', Parsoid::defaultHTMLVersion() |
| 217 | ); |
| 218 | |
| 219 | return $parserOutput; |
| 220 | } |
| 221 | |
| 222 | /** |
| 223 | * Convert wikitext to HTML |
| 224 | * Do not call this function recursively. |
| 225 | * |
| 226 | * @param string|TextContent $text Text we want to parse |
| 227 | * @param-taint $text escapes_htmlnoent |
| 228 | * @param PageReference $page |
| 229 | * @param ParserOptions $options |
| 230 | * @param bool $linestart |
| 231 | * @param bool $clearState |
| 232 | * @param int|null $revId ID of the revision being rendered. This is used to render |
| 233 | * REVISION* magic words. 0 means that any current revision will be used. Null means |
| 234 | * that {{REVISIONID}}/{{REVISIONUSER}} will be empty and {{REVISIONTIMESTAMP}} will |
| 235 | * use the current timestamp. |
| 236 | * @param ?ParserOutput $previousOutput The (optional) result of a |
| 237 | * previous parse of this page, which can be used for selective update. |
| 238 | * @return ParserOutput |
| 239 | * @return-taint escaped |
| 240 | * @unstable since 1.41 |
| 241 | */ |
| 242 | public function parse( |
| 243 | $text, PageReference $page, ParserOptions $options, |
| 244 | bool $linestart = true, bool $clearState = true, ?int $revId = null, |
| 245 | ?ParserOutput $previousOutput = null |
| 246 | ): ParserOutput { |
| 247 | Assert::invariant( $linestart, '$linestart=false is not yet supported' ); |
| 248 | Assert::invariant( $clearState, '$clearState=false is not yet supported' ); |
| 249 | $title = Title::newFromPageReference( $page ); |
| 250 | $lang = $options->getTargetLanguage(); |
| 251 | if ( $lang === null && $options->getInterfaceMessage() ) { |
| 252 | $lang = $options->getUserLangObj(); |
| 253 | } |
| 254 | $pageConfig = $revId === null || $revId === 0 ? null : $this->pageConfigFactory->create( |
| 255 | $title, |
| 256 | $options->getUserIdentity(), |
| 257 | $revId, |
| 258 | null, // unused |
| 259 | $lang // defaults to title page language if null |
| 260 | ); |
| 261 | $content = null; |
| 262 | if ( $text instanceof TextContent ) { |
| 263 | $content = $text; |
| 264 | $text = $content->getText(); |
| 265 | } |
| 266 | if ( !( $pageConfig && $pageConfig->getPageMainContent() === $text ) ) { |
| 267 | // This is a bit awkward! But we really need to parse $text, which |
| 268 | // may or may not correspond to the $revId provided! |
| 269 | // T332928 suggests one solution: splitting the "have revid" |
| 270 | // callers from the "bare text, no associated revision" callers. |
| 271 | $revisionRecord = new MutableRevisionRecord( $title ); |
| 272 | if ( $revId !== null ) { |
| 273 | $revisionRecord->setId( $revId ); |
| 274 | } |
| 275 | $revisionRecord->setSlot( |
| 276 | SlotRecord::newUnsaved( |
| 277 | SlotRecord::MAIN, |
| 278 | $content ?? new WikitextContent( $text ) |
| 279 | ) |
| 280 | ); |
| 281 | $pageConfig = $this->pageConfigFactory->create( |
| 282 | $title, |
| 283 | $options->getUserIdentity(), |
| 284 | $revisionRecord, |
| 285 | null, // unused |
| 286 | $lang // defaults to title page language if null |
| 287 | ); |
| 288 | } |
| 289 | |
| 290 | return $this->genParserOutput( $pageConfig, $options, $previousOutput ); |
| 291 | } |
| 292 | |
| 293 | /** |
| 294 | * @internal |
| 295 | * |
| 296 | * Convert custom wikitext (stored in main slot of the $fakeRev arg) to HTML. |
| 297 | * Callers are expected NOT to stuff the result into ParserCache. |
| 298 | * |
| 299 | * @param RevisionRecord $fakeRev Revision to parse |
| 300 | * @param PageReference $page |
| 301 | * @param ParserOptions $options |
| 302 | * @return ParserOutput |
| 303 | * @unstable since 1.41 |
| 304 | */ |
| 305 | public function parseFakeRevision( |
| 306 | RevisionRecord $fakeRev, PageReference $page, ParserOptions $options |
| 307 | ): ParserOutput { |
| 308 | wfDeprecated( __METHOD__, '1.43' ); |
| 309 | $title = Title::newFromPageReference( $page ); |
| 310 | $lang = $options->getTargetLanguage(); |
| 311 | if ( $lang === null && $options->getInterfaceMessage() ) { |
| 312 | $lang = $options->getUserLangObj(); |
| 313 | } |
| 314 | $pageConfig = $this->pageConfigFactory->create( |
| 315 | $title, |
| 316 | $options->getUserIdentity(), |
| 317 | $fakeRev, |
| 318 | null, // unused |
| 319 | $lang // defaults to title page language if null |
| 320 | ); |
| 321 | |
| 322 | return $this->genParserOutput( $pageConfig, $options, null ); |
| 323 | } |
| 324 | |
| 325 | /** |
| 326 | * Set the limit report data in the current ParserOutput. |
| 327 | * This is ported from Parser::makeLimitReport() and should eventually |
| 328 | * use the method from the superclass directly. |
| 329 | */ |
| 330 | protected function makeLimitReport( |
| 331 | ParserOptions $parserOptions, ParserOutput $parserOutput |
| 332 | ) { |
| 333 | $maxIncludeSize = $parserOptions->getMaxIncludeSize(); |
| 334 | |
| 335 | $cpuTime = $parserOutput->getTimeProfile( 'cpu' ); |
| 336 | if ( $cpuTime !== null ) { |
| 337 | $parserOutput->setLimitReportData( 'limitreport-cputime', |
| 338 | sprintf( "%.3f", $cpuTime ) |
| 339 | ); |
| 340 | } |
| 341 | |
| 342 | $wallTime = $parserOutput->getTimeProfile( 'wall' ); |
| 343 | $parserOutput->setLimitReportData( 'limitreport-walltime', |
| 344 | sprintf( "%.3f", $wallTime ) |
| 345 | ); |
| 346 | |
| 347 | $parserOutput->setLimitReportData( 'limitreport-timingprofile', [ 'not yet supported' ] ); |
| 348 | |
| 349 | // Add other cache related metadata |
| 350 | $parserOutput->setLimitReportData( 'cachereport-timestamp', |
| 351 | $parserOutput->getCacheTime() ); |
| 352 | $parserOutput->setLimitReportData( 'cachereport-ttl', |
| 353 | $parserOutput->getCacheExpiry() ); |
| 354 | $parserOutput->setLimitReportData( 'cachereport-transientcontent', |
| 355 | $parserOutput->hasReducedExpiry() ); |
| 356 | } |
| 357 | |
| 358 | } |