Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
36.27% |
103 / 284 |
|
7.04% |
5 / 71 |
CRAP | |
0.00% |
0 / 1 |
| Env | |
36.27% |
103 / 284 |
|
7.04% |
5 / 71 |
4169.83 | |
0.00% |
0 / 1 |
| __construct | |
88.00% |
44 / 50 |
|
0.00% |
0 / 1 |
6.06 | |||
| checkPlatform | |
12.50% |
2 / 16 |
|
0.00% |
0 / 1 |
9.03 | |||
| profiling | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getCurrentProfile | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| pushNewProfile | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
12 | |||
| popProfile | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| hasTraceFlags | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| hasTraceFlag | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| hasDumpFlags | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| hasDumpFlag | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| writeDump | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getSiteConfig | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| getPageConfig | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| getDataAccess | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getMetadata | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getTOCData | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| nativeTemplateExpansionEnabled | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getUID | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getFID | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getWrapSections | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getPipelineFactory | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getRequestOffsetType | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getCurrentOffsetType | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| setCurrentOffsetType | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getContextTitle | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
2 | |||
| resolveTitle | |
96.67% |
29 / 30 |
|
0.00% |
0 / 1 |
15 | |||
| normalizedTitleKey | |
66.67% |
4 / 6 |
|
0.00% |
0 / 1 |
3.33 | |||
| makeTitle | |
37.50% |
3 / 8 |
|
0.00% |
0 / 1 |
7.91 | |||
| makeTitleFromText | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| makeTitleFromURLDecodedStr | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| makeLink | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
2 | |||
| isValidLinkTarget | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
2 | |||
| generateUID | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| generateAnnotationUID | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| newAnnotationId | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| newAboutId | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| setDOMDiff | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getDOMDiff | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| newFragmentId | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| setupTopLevelDoc | |
66.67% |
12 / 18 |
|
0.00% |
0 / 1 |
2.15 | |||
| getTopLevelDoc | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| fetchRemexPipeline | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
6 | |||
| setBehaviorSwitch | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getBehaviorSwitch | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getDOMFragmentMap | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getDOMFragment | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| setDOMFragment | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
2 | |||
| removeDOMFragment | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
2 | |||
| getPFragment | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| addToPFragmentMap | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| pFragmentMapToString | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
6 | |||
| recordLint | |
0.00% |
0 / 8 |
|
0.00% |
0 / 1 |
12 | |||
| getLints | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| setLints | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| log | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| trace | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
12 | |||
| bumpWt2HtmlResourceUse | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
6 | |||
| compareWt2HtmlLimit | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
6 | |||
| bumpHtml2WtResourceUse | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
12 | |||
| getContentHandler | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
12 | |||
| langConverterEnabled | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
2 | |||
| getInputContentVersion | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getOutputContentVersion | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getHtmlVariantLanguageBcp47 | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getWtVariantLanguageBcp47 | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getSkipLanguageConversionPass | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| htmlVary | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
6 | |||
| htmlContentLanguageBcp47 | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
2 | |||
| getExternalLinkAttribs | |
0.00% |
0 / 19 |
|
0.00% |
0 / 1 |
56 | |||
| getLinterConfig | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| linting | |
0.00% |
0 / 14 |
|
0.00% |
0 / 1 |
42 | |||
| 1 | <?php |
| 2 | declare( strict_types = 1 ); |
| 3 | |
| 4 | namespace Wikimedia\Parsoid\Config; |
| 5 | |
| 6 | use Wikimedia\Assert\Assert; |
| 7 | use Wikimedia\Bcp47Code\Bcp47Code; |
| 8 | use Wikimedia\Parsoid\Core\ContentMetadataCollector; |
| 9 | use Wikimedia\Parsoid\Core\ContentModelHandler; |
| 10 | use Wikimedia\Parsoid\Core\DomPageBundle; |
| 11 | use Wikimedia\Parsoid\Core\ResourceLimitExceededException; |
| 12 | use Wikimedia\Parsoid\Core\Sanitizer; |
| 13 | use Wikimedia\Parsoid\Core\TOCData; |
| 14 | use Wikimedia\Parsoid\DOM\Document; |
| 15 | use Wikimedia\Parsoid\DOM\DocumentFragment; |
| 16 | use Wikimedia\Parsoid\Fragments\PFragment; |
| 17 | use Wikimedia\Parsoid\Logger\ParsoidLogger; |
| 18 | use Wikimedia\Parsoid\Parsoid; |
| 19 | use Wikimedia\Parsoid\Tokens\Token; |
| 20 | use Wikimedia\Parsoid\Utils\DOMCompat; |
| 21 | use Wikimedia\Parsoid\Utils\DOMDataUtils; |
| 22 | use Wikimedia\Parsoid\Utils\PHPUtils; |
| 23 | use Wikimedia\Parsoid\Utils\Title; |
| 24 | use Wikimedia\Parsoid\Utils\TitleException; |
| 25 | use Wikimedia\Parsoid\Utils\TokenUtils; |
| 26 | use Wikimedia\Parsoid\Utils\UrlUtils; |
| 27 | use Wikimedia\Parsoid\Utils\Utils; |
| 28 | use Wikimedia\Parsoid\Wikitext\ContentModelHandler as WikitextContentModelHandler; |
| 29 | use Wikimedia\Parsoid\Wt2Html\Frame; |
| 30 | use Wikimedia\Parsoid\Wt2Html\PageConfigFrame; |
| 31 | use Wikimedia\Parsoid\Wt2Html\ParserPipelineFactory; |
| 32 | use Wikimedia\Parsoid\Wt2Html\TreeBuilder\RemexPipeline; |
| 33 | |
| 34 | /** |
| 35 | * Environment/Envelope class for Parsoid |
| 36 | * |
| 37 | * Carries around the SiteConfig and PageConfig during an operation |
| 38 | * and provides certain other services. |
| 39 | */ |
| 40 | class Env { |
| 41 | private SiteConfig $siteConfig; |
| 42 | private PageConfig $pageConfig; |
| 43 | private DataAccess $dataAccess; |
| 44 | private ContentMetadataCollector $metadata; |
| 45 | |
| 46 | /** Table of Contents metadata for the article */ |
| 47 | private TOCData $tocData; |
| 48 | |
| 49 | /** |
| 50 | * The top-level frame for this conversion. |
| 51 | * This largely wraps the PageConfig. |
| 52 | * In the future we may replace PageConfig with the Frame |
| 53 | */ |
| 54 | public Frame $topFrame; |
| 55 | // XXX In the future, perhaps replace PageConfig with the Frame, and |
| 56 | // add $this->currentFrame (relocated from TokenHandlerPipeline) if/when |
| 57 | // we've removed async parsing. |
| 58 | |
| 59 | /** |
| 60 | * Are we using native template expansion? |
| 61 | * |
| 62 | * Parsoid implements native template expansion, which is currently |
| 63 | * only used during parser tests; in production, template expansion |
| 64 | * is done via MediaWiki's legacy preprocessor. |
| 65 | * |
| 66 | * FIXME: Hopefully this distinction can be removed when we're entirely |
| 67 | * in PHP land. |
| 68 | */ |
| 69 | private bool $nativeTemplateExpansion; |
| 70 | |
| 71 | /** @var array<string,int> */ |
| 72 | private array $wt2htmlUsage = []; |
| 73 | |
| 74 | /** @var array<string,int> */ |
| 75 | private array $html2wtUsage = []; |
| 76 | private bool $profiling = false; |
| 77 | |
| 78 | /** @var array<Profile> */ |
| 79 | private array $profileStack = []; |
| 80 | private bool $wrapSections; |
| 81 | |
| 82 | /** @var ('byte'|'ucs2'|'char') */ |
| 83 | private string $requestOffsetType = 'byte'; |
| 84 | |
| 85 | /** @var ('byte'|'ucs2'|'char') */ |
| 86 | private string $currentOffsetType = 'byte'; |
| 87 | private bool $skipLanguageConversionPass = false; |
| 88 | |
| 89 | /** @var array<string,mixed> */ |
| 90 | private array $behaviorSwitches = []; |
| 91 | |
| 92 | /** |
| 93 | * Maps fragment id to the fragment forest (array of Nodes). |
| 94 | * @var array<string,DocumentFragment> |
| 95 | */ |
| 96 | private array $fragmentMap = []; |
| 97 | |
| 98 | /** |
| 99 | * Maps pfragment id to a PFragment. |
| 100 | * @var array<string,PFragment> |
| 101 | */ |
| 102 | private array $pFragmentMap = []; |
| 103 | |
| 104 | /** |
| 105 | * Used to generate fragment ids as needed during parse |
| 106 | */ |
| 107 | private int $fid = 1; |
| 108 | |
| 109 | /** Used to generate uids as needed during this parse */ |
| 110 | private int $uid = 1; |
| 111 | |
| 112 | /** Used to generate annotation uids as needed during this parse */ |
| 113 | private int $annUid = 0; |
| 114 | |
| 115 | /** Lints recorded */ |
| 116 | private array $lints = []; |
| 117 | public bool $logLinterData = false; |
| 118 | private array $linterOverrides = []; |
| 119 | |
| 120 | /** @var bool[] */ |
| 121 | private array $traceFlags; |
| 122 | |
| 123 | /** @var bool[] */ |
| 124 | private array $dumpFlags; |
| 125 | |
| 126 | /** @var bool[] */ |
| 127 | private array $debugFlags; |
| 128 | private ParsoidLogger $parsoidLogger; |
| 129 | |
| 130 | /** |
| 131 | * The default content version that Parsoid assumes it's serializing or |
| 132 | * updating in the pb2pb endpoints |
| 133 | */ |
| 134 | private string $inputContentVersion; |
| 135 | |
| 136 | /** |
| 137 | * The default content version that Parsoid will generate. |
| 138 | */ |
| 139 | private string $outputContentVersion; |
| 140 | |
| 141 | /** |
| 142 | * If non-null, the language variant used for Parsoid HTML; |
| 143 | * we convert to this if wt2html, or from this if html2wt. |
| 144 | */ |
| 145 | private ?Bcp47Code $htmlVariantLanguage; |
| 146 | |
| 147 | /** |
| 148 | * If non-null, the language variant to be used for wikitext. |
| 149 | * If null, heuristics will be used to identify the original wikitext variant |
| 150 | * in wt2html mode, and in html2wt mode new or edited HTML will be left unconverted. |
| 151 | */ |
| 152 | private ?Bcp47Code $wtVariantLanguage; |
| 153 | private ParserPipelineFactory $pipelineFactory; |
| 154 | |
| 155 | /** |
| 156 | * FIXME Used in DedupeStyles::dedupe() |
| 157 | */ |
| 158 | public array $styleTagKeys = []; |
| 159 | |
| 160 | /** |
| 161 | * The DomPageBundle holding the JSON data for data-parsoid and data-mw |
| 162 | * attributes, or `null` if these are to be encoded as inline HTML |
| 163 | * attributes. |
| 164 | */ |
| 165 | public ?DomPageBundle $pageBundle = null; |
| 166 | private ?Document $domDiff = null; |
| 167 | public bool $hasAnnotations = false; |
| 168 | |
| 169 | /** |
| 170 | * Cache of wikitext source for a title; only used for ParserTests. |
| 171 | */ |
| 172 | public array $pageCache = []; |
| 173 | |
| 174 | /** |
| 175 | * The current top-level document. During wt2html, this will be the document |
| 176 | * associated with the RemexPipeline. During html2wt, this will be the |
| 177 | * input document, typically passed as a constructor option. |
| 178 | * |
| 179 | * This document should be prepared and loaded; see |
| 180 | * ContentUtils::createAndLoadDocument(). |
| 181 | */ |
| 182 | private Document $topLevelDoc; |
| 183 | |
| 184 | /** |
| 185 | * The RemexPipeline used during a wt2html operation. |
| 186 | */ |
| 187 | private ?RemexPipeline $remexPipeline; |
| 188 | private WikitextContentModelHandler $wikitextContentModelHandler; |
| 189 | private ?Title $cachedContextTitle = null; |
| 190 | |
| 191 | /** |
| 192 | * @param SiteConfig $siteConfig |
| 193 | * @param PageConfig $pageConfig |
| 194 | * @param DataAccess $dataAccess |
| 195 | * @param ContentMetadataCollector $metadata |
| 196 | * @param ?array $options |
| 197 | * - wrapSections: (bool) Whether `<section>` wrappers should be added. |
| 198 | * - pageBundle: (bool) When true, sets ids on nodes and stores |
| 199 | * data-* attributes in a JSON blob in Env::$pageBundle |
| 200 | * - traceFlags: (array) Flags indicating which components need to be traced |
| 201 | * - dumpFlags: (bool[]) Dump flags |
| 202 | * - debugFlags: (bool[]) Debug flags |
| 203 | * - nativeTemplateExpansion: boolean |
| 204 | * - offsetType: 'byte' (default), 'ucs2', 'char' |
| 205 | * See `Parsoid\Wt2Html\DOM\Processors\ConvertOffsets`. |
| 206 | * - logLinterData: (bool) Should we log linter data if linting is enabled? |
| 207 | * - linterOverrides: (array) Override the site linting configs. |
| 208 | * - skipLanguageConversionPass: (bool) Should we skip the language |
| 209 | * conversion pass? (defaults to false) |
| 210 | * - htmlVariantLanguage: Bcp47Code|null |
| 211 | * If non-null, the language variant used for Parsoid HTML |
| 212 | * as a BCP 47 object. |
| 213 | * We convert to this if wt2html, or from this if html2wt. |
| 214 | * - wtVariantLanguage: Bcp47Code|null |
| 215 | * If non-null, the language variant to be used for wikitext |
| 216 | * as a BCP 47 object. |
| 217 | * If null, heuristics will be used to identify the original |
| 218 | * wikitext variant in wt2html mode, and in html2wt mode new |
| 219 | * or edited HTML will be left unconverted. |
| 220 | * - logLevels: (string[]) Levels to log |
| 221 | * - topLevelDoc: Document Set explicitly |
| 222 | * when serializing otherwise it gets initialized for parsing. |
| 223 | * This should be a "prepared & loaded" document. |
| 224 | */ |
| 225 | public function __construct( |
| 226 | SiteConfig $siteConfig, |
| 227 | PageConfig $pageConfig, |
| 228 | DataAccess $dataAccess, |
| 229 | ContentMetadataCollector $metadata, |
| 230 | ?array $options = null |
| 231 | ) { |
| 232 | self::checkPlatform(); |
| 233 | $options ??= []; |
| 234 | $this->siteConfig = $siteConfig; |
| 235 | $this->pageConfig = $pageConfig; |
| 236 | $this->dataAccess = $dataAccess; |
| 237 | $this->metadata = $metadata; |
| 238 | $this->tocData = new TOCData(); |
| 239 | $this->topFrame = new PageConfigFrame( $this, $pageConfig, $siteConfig ); |
| 240 | $this->wrapSections = (bool)( $options['wrapSections'] ?? true ); |
| 241 | $this->pipelineFactory = new ParserPipelineFactory( $this ); |
| 242 | $defaultContentVersion = Parsoid::defaultHTMLVersion(); |
| 243 | $this->inputContentVersion = $options['inputContentVersion'] ?? $defaultContentVersion; |
| 244 | // FIXME: We should have a check for the supported input content versions as well. |
| 245 | // That will require a separate constant. |
| 246 | $this->outputContentVersion = $options['outputContentVersion'] ?? $defaultContentVersion; |
| 247 | if ( !in_array( $this->outputContentVersion, Parsoid::AVAILABLE_VERSIONS, true ) ) { |
| 248 | throw new \UnexpectedValueException( |
| 249 | $this->outputContentVersion . ' is not an available content version.' ); |
| 250 | } |
| 251 | $this->skipLanguageConversionPass = |
| 252 | $options['skipLanguageConversionPass'] ?? false; |
| 253 | $this->htmlVariantLanguage = !empty( $options['htmlVariantLanguage'] ) ? |
| 254 | Utils::mwCodeToBcp47( |
| 255 | $options['htmlVariantLanguage'], |
| 256 | // Be strict in what we accept here. |
| 257 | true, $this->siteConfig->getLogger() |
| 258 | ) : null; |
| 259 | $this->wtVariantLanguage = !empty( $options['wtVariantLanguage'] ) ? |
| 260 | Utils::mwCodeToBcp47( |
| 261 | $options['wtVariantLanguage'], |
| 262 | // Be strict in what we accept here. |
| 263 | true, $this->siteConfig->getLogger() |
| 264 | ) : null; |
| 265 | $this->nativeTemplateExpansion = !empty( $options['nativeTemplateExpansion'] ); |
| 266 | $this->requestOffsetType = $options['offsetType'] ?? 'byte'; |
| 267 | $this->logLinterData = !empty( $options['logLinterData'] ); |
| 268 | $this->linterOverrides = $options['linterOverrides'] ?? []; |
| 269 | $this->traceFlags = $options['traceFlags'] ?? []; |
| 270 | $this->dumpFlags = $options['dumpFlags'] ?? []; |
| 271 | $this->debugFlags = $options['debugFlags'] ?? []; |
| 272 | $this->parsoidLogger = new ParsoidLogger( $this->siteConfig->getLogger(), [ |
| 273 | 'logLevels' => $options['logLevels'] ?? [ 'fatal', 'error', 'warn', 'info' ], |
| 274 | 'debugFlags' => $this->debugFlags, |
| 275 | 'dumpFlags' => $this->dumpFlags, |
| 276 | 'traceFlags' => $this->traceFlags |
| 277 | ] ); |
| 278 | if ( $this->hasTraceFlag( 'time' ) ) { |
| 279 | $this->profiling = true; |
| 280 | } |
| 281 | $this->setupTopLevelDoc( $options['topLevelDoc'] ?? null ); |
| 282 | if ( $options['pageBundle'] ?? false ) { |
| 283 | $this->pageBundle = DomPageBundle::newEmpty( |
| 284 | $this->topLevelDoc |
| 285 | ); |
| 286 | } |
| 287 | // NOTE: |
| 288 | // Don't try to do this in setupTopLevelDoc since it is called on existing Env objects |
| 289 | // in a couple of places. That then leads to a multiple-write to tocdata property on |
| 290 | // the metadata object. |
| 291 | // |
| 292 | // setupTopLevelDoc is called outside Env in these couple cases: |
| 293 | // 1. html2wt in ContentModelHandler for dealing with |
| 294 | // missing original HTML. |
| 295 | // 2. ParserTestRunner's html2html tests |
| 296 | // |
| 297 | // That is done to either reuse an existing Env object (as in 1.) |
| 298 | // OR to refresh the attached DOC (html2html as in 2.). |
| 299 | // Constructing a new Env in both cases could eliminate this issue. |
| 300 | $this->metadata->setTOCData( $this->tocData ); |
| 301 | |
| 302 | $this->wikitextContentModelHandler = new WikitextContentModelHandler( $this ); |
| 303 | } |
| 304 | |
| 305 | /** |
| 306 | * Check to see if the PHP platform is sensible |
| 307 | */ |
| 308 | private static function checkPlatform() { |
| 309 | static $checked; |
| 310 | if ( !$checked ) { |
| 311 | $highBytes = |
| 312 | "\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8a\x8b\x8c\x8d\x8e\x8f" . |
| 313 | "\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9a\x9b\x9c\x9d\x9e\x9f" . |
| 314 | "\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9\xaa\xab\xac\xad\xae\xaf" . |
| 315 | "\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf" . |
| 316 | "\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf" . |
| 317 | "\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xd7\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf" . |
| 318 | "\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef" . |
| 319 | "\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff"; |
| 320 | if ( strtolower( 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' . $highBytes ) |
| 321 | !== 'abcdefghijklmnopqrstuvwxyz' . $highBytes |
| 322 | ) { |
| 323 | throw new \RuntimeException( 'strtolower() doesn\'t work -- ' . |
| 324 | 'please set the locale to C or a UTF-8 variant such as C.UTF-8' ); |
| 325 | } |
| 326 | $checked = true; |
| 327 | } |
| 328 | } |
| 329 | |
| 330 | /** |
| 331 | * Is profiling enabled? |
| 332 | * @return bool |
| 333 | */ |
| 334 | public function profiling(): bool { |
| 335 | return $this->profiling; |
| 336 | } |
| 337 | |
| 338 | /** |
| 339 | * Get the profile at the top of the stack |
| 340 | * |
| 341 | * FIXME: This implicitly assumes sequential in-order processing |
| 342 | * This wouldn't have worked in Parsoid/JS and may not work in the future |
| 343 | * depending on how / if we restructure the pipeline for concurrency, etc. |
| 344 | * |
| 345 | * @return Profile |
| 346 | */ |
| 347 | public function getCurrentProfile(): Profile { |
| 348 | return PHPUtils::lastItem( $this->profileStack ); |
| 349 | } |
| 350 | |
| 351 | /** |
| 352 | * New pipeline started. Push profile. |
| 353 | * @return Profile |
| 354 | */ |
| 355 | public function pushNewProfile(): Profile { |
| 356 | $currProfile = count( $this->profileStack ) > 0 ? $this->getCurrentProfile() : null; |
| 357 | $profile = new Profile(); |
| 358 | $this->profileStack[] = $profile; |
| 359 | if ( $currProfile !== null ) { |
| 360 | $currProfile->pushNestedProfile( $profile ); |
| 361 | } |
| 362 | return $profile; |
| 363 | } |
| 364 | |
| 365 | /** |
| 366 | * Pipeline ended. Pop profile. |
| 367 | * @return Profile |
| 368 | */ |
| 369 | public function popProfile(): Profile { |
| 370 | return array_pop( $this->profileStack ); |
| 371 | } |
| 372 | |
| 373 | public function hasTraceFlags(): bool { |
| 374 | return !empty( $this->traceFlags ); |
| 375 | } |
| 376 | |
| 377 | /** |
| 378 | * Test which trace information to log |
| 379 | * |
| 380 | * @param string $flag Flag name. |
| 381 | * @return bool |
| 382 | */ |
| 383 | public function hasTraceFlag( string $flag ): bool { |
| 384 | return isset( $this->traceFlags[$flag] ); |
| 385 | } |
| 386 | |
| 387 | public function hasDumpFlags(): bool { |
| 388 | return !empty( $this->dumpFlags ); |
| 389 | } |
| 390 | |
| 391 | /** |
| 392 | * Test which state to dump |
| 393 | * |
| 394 | * @param string $flag Flag name. |
| 395 | * @return bool |
| 396 | */ |
| 397 | public function hasDumpFlag( string $flag ): bool { |
| 398 | return isset( $this->dumpFlags[$flag] ); |
| 399 | } |
| 400 | |
| 401 | /** |
| 402 | * Write out a string (because it was requested by dumpFlags) |
| 403 | * @param string $str |
| 404 | */ |
| 405 | public function writeDump( string $str ) { |
| 406 | $this->log( 'dump', $str ); |
| 407 | } |
| 408 | |
| 409 | /** |
| 410 | * Get the site config |
| 411 | * @return SiteConfig |
| 412 | */ |
| 413 | public function getSiteConfig(): SiteConfig { |
| 414 | return $this->siteConfig; |
| 415 | } |
| 416 | |
| 417 | /** |
| 418 | * Get the page config |
| 419 | * @return PageConfig |
| 420 | */ |
| 421 | public function getPageConfig(): PageConfig { |
| 422 | return $this->pageConfig; |
| 423 | } |
| 424 | |
| 425 | /** |
| 426 | * Get the data access object |
| 427 | * @return DataAccess |
| 428 | */ |
| 429 | public function getDataAccess(): DataAccess { |
| 430 | return $this->dataAccess; |
| 431 | } |
| 432 | |
| 433 | /** |
| 434 | * Return the ContentMetadataCollector. |
| 435 | * @return ContentMetadataCollector |
| 436 | */ |
| 437 | public function getMetadata(): ContentMetadataCollector { |
| 438 | return $this->metadata; |
| 439 | } |
| 440 | |
| 441 | /** |
| 442 | * Return the Table of Contents information for the article. |
| 443 | * @return TOCData |
| 444 | */ |
| 445 | public function getTOCData(): TOCData { |
| 446 | return $this->tocData; |
| 447 | } |
| 448 | |
| 449 | public function nativeTemplateExpansionEnabled(): bool { |
| 450 | return $this->nativeTemplateExpansion; |
| 451 | } |
| 452 | |
| 453 | /** |
| 454 | * Get the current uid counter value |
| 455 | * @return int |
| 456 | */ |
| 457 | public function getUID(): int { |
| 458 | return $this->uid; |
| 459 | } |
| 460 | |
| 461 | /** |
| 462 | * Get the current fragment id counter value |
| 463 | * @return int |
| 464 | */ |
| 465 | public function getFID(): int { |
| 466 | return $this->fid; |
| 467 | } |
| 468 | |
| 469 | /** |
| 470 | * Whether `<section>` wrappers should be added. |
| 471 | * @todo Does this actually belong here? Should it be a behavior switch? |
| 472 | * @return bool |
| 473 | */ |
| 474 | public function getWrapSections(): bool { |
| 475 | return $this->wrapSections; |
| 476 | } |
| 477 | |
| 478 | /** |
| 479 | * Get the pipeline factory. |
| 480 | * @return ParserPipelineFactory |
| 481 | */ |
| 482 | public function getPipelineFactory(): ParserPipelineFactory { |
| 483 | return $this->pipelineFactory; |
| 484 | } |
| 485 | |
| 486 | /** |
| 487 | * Return the external format of character offsets in source ranges. |
| 488 | * Internally we always keep DomSourceRange and SourceRange information |
| 489 | * as UTF-8 byte offsets for efficiency (matches the native string |
| 490 | * representation), but for external use we can convert these to |
| 491 | * other formats when we output wt2html or input for html2wt. |
| 492 | * |
| 493 | * @see Parsoid\Wt2Html\DOM\Processors\ConvertOffsets |
| 494 | * @return ('byte'|'ucs2'|'char') |
| 495 | */ |
| 496 | public function getRequestOffsetType(): string { |
| 497 | return $this->requestOffsetType; |
| 498 | } |
| 499 | |
| 500 | /** |
| 501 | * Return the current format of character offsets in source ranges. |
| 502 | * This allows us to track whether the internal byte offsets have |
| 503 | * been converted to the external format (as returned by |
| 504 | * `getRequestOffsetType`) yet. |
| 505 | * |
| 506 | * @see Parsoid\Wt2Html\DOM\Processors\ConvertOffsets |
| 507 | * @return ('byte'|'ucs2'|'char') |
| 508 | */ |
| 509 | public function getCurrentOffsetType(): string { |
| 510 | return $this->currentOffsetType; |
| 511 | } |
| 512 | |
| 513 | /** |
| 514 | * Update the current offset type. Only |
| 515 | * Parsoid\Wt2Html\DOM\Processors\ConvertOffsets should be doing this. |
| 516 | * @param ('byte'|'ucs2'|'char') $offsetType 'byte', 'ucs2', or 'char' |
| 517 | */ |
| 518 | public function setCurrentOffsetType( string $offsetType ) { |
| 519 | $this->currentOffsetType = $offsetType; |
| 520 | } |
| 521 | |
| 522 | /** |
| 523 | * Return the title from the PageConfig, as a Parsoid title. |
| 524 | * @return Title |
| 525 | */ |
| 526 | public function getContextTitle(): Title { |
| 527 | if ( $this->cachedContextTitle === null ) { |
| 528 | $this->cachedContextTitle = Title::newFromLinkTarget( |
| 529 | $this->pageConfig->getLinkTarget(), $this->siteConfig |
| 530 | ); |
| 531 | } |
| 532 | return $this->cachedContextTitle; |
| 533 | } |
| 534 | |
| 535 | /** |
| 536 | * Resolve strings that are page-fragments or subpage references with |
| 537 | * respect to the current page name. |
| 538 | * |
| 539 | * @param string $str Page fragment or subpage reference. Not URL encoded. |
| 540 | * @param bool $resolveOnly If true, only trim and add the current title to |
| 541 | * lone fragments. TODO: This parameter seems poorly named. |
| 542 | * @return string Resolved title |
| 543 | */ |
| 544 | public function resolveTitle( string $str, bool $resolveOnly = false ): string { |
| 545 | $origName = $str; |
| 546 | $str = trim( $str ); |
| 547 | |
| 548 | $pageConfig = $this->getPageConfig(); |
| 549 | $title = $this->getContextTitle(); |
| 550 | |
| 551 | // Resolve lonely fragments (important if the current page is a subpage, |
| 552 | // otherwise the relative link will be wrong) |
| 553 | if ( $str !== '' && $str[0] === '#' ) { |
| 554 | return $title->getPrefixedText() . $str; |
| 555 | } |
| 556 | |
| 557 | // Default return value |
| 558 | $titleKey = $str; |
| 559 | if ( $this->getSiteConfig()->namespaceHasSubpages( $title->getNamespace() ) ) { |
| 560 | // Resolve subpages |
| 561 | $reNormalize = false; |
| 562 | if ( preg_match( '!^(?:\.\./)+!', $str, $relUp ) ) { |
| 563 | $levels = strlen( $relUp[0] ) / 3; // Levels are indicated by '../'. |
| 564 | $titleBits = explode( '/', $title->getPrefixedText() ); |
| 565 | if ( $titleBits[0] === '' ) { |
| 566 | // FIXME: Punt on subpages of titles starting with "/" for now |
| 567 | return $origName; |
| 568 | } |
| 569 | if ( count( $titleBits ) <= $levels ) { |
| 570 | // Too many levels -- invalid relative link |
| 571 | return $origName; |
| 572 | } |
| 573 | $newBits = array_slice( $titleBits, 0, -$levels ); |
| 574 | if ( $str !== $relUp[0] ) { |
| 575 | $newBits[] = substr( $str, $levels * 3 ); |
| 576 | } |
| 577 | $str = implode( '/', $newBits ); |
| 578 | $reNormalize = true; |
| 579 | } elseif ( $str !== '' && $str[0] === '/' ) { |
| 580 | // Resolve absolute subpage links |
| 581 | $str = $title->getPrefixedText() . $str; |
| 582 | $reNormalize = true; |
| 583 | } |
| 584 | |
| 585 | if ( $reNormalize && !$resolveOnly ) { |
| 586 | // Remove final slashes if present. |
| 587 | // See https://gerrit.wikimedia.org/r/173431 |
| 588 | $str = rtrim( $str, '/' ); |
| 589 | $titleKey = (string)$this->normalizedTitleKey( $str ); |
| 590 | } |
| 591 | } |
| 592 | |
| 593 | // Strip leading ':' |
| 594 | if ( $titleKey !== '' && $titleKey[0] === ':' && !$resolveOnly ) { |
| 595 | $titleKey = substr( $titleKey, 1 ); |
| 596 | } |
| 597 | return $titleKey; |
| 598 | } |
| 599 | |
| 600 | /** |
| 601 | * Get normalized title key for a title string. |
| 602 | * |
| 603 | * @param string $str Should be in url-decoded format. |
| 604 | * @param bool $noExceptions Return null instead of throwing exceptions. |
| 605 | * @param bool $ignoreFragment Ignore the fragment, if any. |
| 606 | * @return string|null Normalized title key for a title string (or null for invalid titles). |
| 607 | */ |
| 608 | public function normalizedTitleKey( |
| 609 | string $str, bool $noExceptions = false, bool $ignoreFragment = false |
| 610 | ): ?string { |
| 611 | $title = $this->makeTitleFromURLDecodedStr( $str, 0, $noExceptions ); |
| 612 | if ( !$title ) { |
| 613 | return null; |
| 614 | } |
| 615 | return $ignoreFragment ? |
| 616 | $title->getPrefixedDBKey() : |
| 617 | $title->getFullDBKey(); |
| 618 | } |
| 619 | |
| 620 | /** |
| 621 | * Create a Title object |
| 622 | * @param string $text URL-decoded text |
| 623 | * @param ?int $defaultNs |
| 624 | * @param bool $noExceptions |
| 625 | * @return Title|null |
| 626 | */ |
| 627 | private function makeTitle( string $text, ?int $defaultNs = null, bool $noExceptions = false ): ?Title { |
| 628 | try { |
| 629 | if ( preg_match( '!^(?:[#/]|\.\./)!', $text ) ) { |
| 630 | $defaultNs = $this->getContextTitle()->getNamespace(); |
| 631 | } |
| 632 | $text = $this->resolveTitle( $text ); |
| 633 | return Title::newFromText( $text, $this->getSiteConfig(), $defaultNs ); |
| 634 | } catch ( TitleException $e ) { |
| 635 | if ( $noExceptions ) { |
| 636 | return null; |
| 637 | } |
| 638 | throw $e; |
| 639 | } |
| 640 | } |
| 641 | |
| 642 | /** |
| 643 | * Create a Title object |
| 644 | * @see Title::newFromURL in MediaWiki |
| 645 | * @param string $str URL-encoded text |
| 646 | * @param ?int $defaultNs |
| 647 | * @param bool $noExceptions |
| 648 | * @return Title|null |
| 649 | */ |
| 650 | public function makeTitleFromText( |
| 651 | string $str, ?int $defaultNs = null, bool $noExceptions = false |
| 652 | ): ?Title { |
| 653 | return $this->makeTitle( Utils::decodeURIComponent( $str ), $defaultNs, $noExceptions ); |
| 654 | } |
| 655 | |
| 656 | /** |
| 657 | * Create a Title object |
| 658 | * @see Title::newFromText in MediaWiki |
| 659 | * @param string $str URL-decoded text |
| 660 | * @param ?int $defaultNs |
| 661 | * @param bool $noExceptions |
| 662 | * @return Title|null |
| 663 | */ |
| 664 | public function makeTitleFromURLDecodedStr( |
| 665 | string $str, ?int $defaultNs = null, bool $noExceptions = false |
| 666 | ): ?Title { |
| 667 | return $this->makeTitle( $str, $defaultNs, $noExceptions ); |
| 668 | } |
| 669 | |
| 670 | /** |
| 671 | * Make a link to a local Title |
| 672 | * @param Title $title |
| 673 | * @return string |
| 674 | */ |
| 675 | public function makeLink( Title $title ): string { |
| 676 | // T380676: This method *should* be used only for local titles, |
| 677 | // (ie $title->getInterwiki() should be '') but apparently we |
| 678 | // are using it for interwiki/interlanguage links as well. |
| 679 | return $this->getSiteConfig()->relativeLinkPrefix() . Sanitizer::sanitizeTitleURI( |
| 680 | $title->getFullDBKey(), |
| 681 | false |
| 682 | ); |
| 683 | } |
| 684 | |
| 685 | /** |
| 686 | * Test if an href attribute value could be a valid link target |
| 687 | * @param string|(Token|string)[] $href |
| 688 | * @return bool |
| 689 | */ |
| 690 | public function isValidLinkTarget( $href ): bool { |
| 691 | $href = TokenUtils::tokensToString( $href ); |
| 692 | |
| 693 | // decode percent-encoding so that we can reliably detect |
| 694 | // bad page title characters |
| 695 | $hrefToken = Utils::decodeURIComponent( $href ); |
| 696 | return $this->normalizedTitleKey( $this->resolveTitle( $hrefToken, true ), true ) !== null; |
| 697 | } |
| 698 | |
| 699 | /** |
| 700 | * Generate a new uid |
| 701 | * @return int |
| 702 | */ |
| 703 | public function generateUID(): int { |
| 704 | return $this->uid++; |
| 705 | } |
| 706 | |
| 707 | /** |
| 708 | * Generate a new annotation uid |
| 709 | * @return int |
| 710 | */ |
| 711 | public function generateAnnotationUID(): int { |
| 712 | return $this->annUid++; |
| 713 | } |
| 714 | |
| 715 | /** |
| 716 | * Generate a new annotation id |
| 717 | * @return string |
| 718 | */ |
| 719 | public function newAnnotationId(): string { |
| 720 | return "mwa" . $this->generateAnnotationUID(); |
| 721 | } |
| 722 | |
| 723 | /** |
| 724 | * Generate a new about id |
| 725 | * @return string |
| 726 | */ |
| 727 | public function newAboutId(): string { |
| 728 | return '#mwt' . $this->generateUID(); |
| 729 | } |
| 730 | |
| 731 | /** |
| 732 | * Store reference to DOM diff document |
| 733 | * @param Document $doc |
| 734 | */ |
| 735 | public function setDOMDiff( $doc ): void { |
| 736 | $this->domDiff = $doc; |
| 737 | } |
| 738 | |
| 739 | /** |
| 740 | * Return reference to DOM diff document |
| 741 | * @return Document|null |
| 742 | */ |
| 743 | public function getDOMDiff(): ?Document { |
| 744 | return $this->domDiff; |
| 745 | } |
| 746 | |
| 747 | /** |
| 748 | * Generate a new fragment id |
| 749 | * @return string |
| 750 | */ |
| 751 | public function newFragmentId(): string { |
| 752 | return "mwf" . (string)$this->fid++; |
| 753 | } |
| 754 | |
| 755 | /** |
| 756 | * When an environment is constructed, we initialize a document (and |
| 757 | * RemexPipeline) to be used throughout the parse. |
| 758 | * |
| 759 | * @param ?Document $topLevelDoc if non-null, |
| 760 | * the document should be prepared and loaded. |
| 761 | */ |
| 762 | public function setupTopLevelDoc( ?Document $topLevelDoc = null ): void { |
| 763 | if ( $topLevelDoc ) { |
| 764 | $this->remexPipeline = null; |
| 765 | // This is a prepared & loaded Document. |
| 766 | Assert::invariant( |
| 767 | DOMDataUtils::isPreparedAndLoaded( $topLevelDoc ), |
| 768 | "toplevelDoc should be prepared and loaded already" |
| 769 | ); |
| 770 | $this->topLevelDoc = $topLevelDoc; |
| 771 | } else { |
| 772 | $this->remexPipeline = new RemexPipeline( $this ); |
| 773 | $this->topLevelDoc = $this->remexPipeline->doc; |
| 774 | // Prepare and load. |
| 775 | // (Loading should be easy since the doc is expected to be empty.) |
| 776 | $options = [ |
| 777 | 'validateXMLNames' => true, |
| 778 | // Don't mark the <body> tag as new! |
| 779 | 'markNew' => false, |
| 780 | ]; |
| 781 | DOMDataUtils::prepareDoc( $this->topLevelDoc ); |
| 782 | DOMDataUtils::visitAndLoadDataAttribs( |
| 783 | DOMCompat::getBody( $this->topLevelDoc ), $options |
| 784 | ); |
| 785 | // Mark the document as loaded so we can try to catch errors which |
| 786 | // might try to reload this again later. |
| 787 | DOMDataUtils::getBag( $this->topLevelDoc )->loaded = true; |
| 788 | } |
| 789 | } |
| 790 | |
| 791 | /** |
| 792 | * Return the current top-level document. During wt2html, this |
| 793 | * will be the document associated with the RemexPipeline. During |
| 794 | * html2wt, this will be the input document, typically passed as a |
| 795 | * constructor option. |
| 796 | * |
| 797 | * This document will be prepared and loaded; see |
| 798 | * ContentUtils::createAndLoadDocument(). |
| 799 | */ |
| 800 | public function getTopLevelDoc(): Document { |
| 801 | return $this->topLevelDoc; |
| 802 | } |
| 803 | |
| 804 | public function fetchRemexPipeline( bool $toFragment ): RemexPipeline { |
| 805 | if ( !$toFragment ) { |
| 806 | return $this->remexPipeline; |
| 807 | } else { |
| 808 | $pipeline = new RemexPipeline( $this ); |
| 809 | // Attach the top-level bag to the document, for the convenience |
| 810 | // of code that modifies the data within the RemexHtml TreeBuilder |
| 811 | // pipeline, prior to the migration of nodes to the top-level |
| 812 | // document. |
| 813 | DOMDataUtils::prepareChildDoc( $this->topLevelDoc, $pipeline->doc ); |
| 814 | return $pipeline; |
| 815 | } |
| 816 | } |
| 817 | |
| 818 | /** |
| 819 | * Record a behavior switch. |
| 820 | * |
| 821 | * @param string $switch Switch name |
| 822 | * @param mixed $state Relevant state data to record |
| 823 | */ |
| 824 | public function setBehaviorSwitch( string $switch, $state ): void { |
| 825 | $this->behaviorSwitches[$switch] = $state; |
| 826 | } |
| 827 | |
| 828 | /** |
| 829 | * Fetch the state of a previously-recorded behavior switch. |
| 830 | * |
| 831 | * @param string $switch Switch name |
| 832 | * @param mixed $default Default value if the switch was never set |
| 833 | * @return mixed State data that was previously passed to setBehaviorSwitch(), or $default |
| 834 | */ |
| 835 | public function getBehaviorSwitch( string $switch, $default = null ) { |
| 836 | return $this->behaviorSwitches[$switch] ?? $default; |
| 837 | } |
| 838 | |
| 839 | /** |
| 840 | * @return array<string,DocumentFragment> |
| 841 | */ |
| 842 | public function getDOMFragmentMap(): array { |
| 843 | return $this->fragmentMap; |
| 844 | } |
| 845 | |
| 846 | /** |
| 847 | * @param string $id Fragment id |
| 848 | * @return DocumentFragment |
| 849 | */ |
| 850 | public function getDOMFragment( string $id ): DocumentFragment { |
| 851 | return $this->fragmentMap[$id]; |
| 852 | } |
| 853 | |
| 854 | /** |
| 855 | * @param string $id Fragment id |
| 856 | * @param DocumentFragment $forest DOM forest |
| 857 | * to store against the fragment id |
| 858 | */ |
| 859 | public function setDOMFragment( |
| 860 | string $id, DocumentFragment $forest |
| 861 | ): void { |
| 862 | Assert::invariant( |
| 863 | $forest->ownerDocument === $this->topLevelDoc, |
| 864 | "fragment should belong to the top level document" |
| 865 | ); |
| 866 | $this->fragmentMap[$id] = $forest; |
| 867 | } |
| 868 | |
| 869 | public function removeDOMFragment( string $id ): void { |
| 870 | $domFragment = $this->fragmentMap[$id]; |
| 871 | Assert::invariant( |
| 872 | !$domFragment->hasChildNodes(), 'Fragment should be empty.' |
| 873 | ); |
| 874 | unset( $this->fragmentMap[$id] ); |
| 875 | } |
| 876 | |
| 877 | public function getPFragment( string $id ): PFragment { |
| 878 | return $this->pFragmentMap[$id]; |
| 879 | } |
| 880 | |
| 881 | /** @param array<string,PFragment> $mapping */ |
| 882 | public function addToPFragmentMap( array $mapping ): void { |
| 883 | $this->pFragmentMap += $mapping; |
| 884 | } |
| 885 | |
| 886 | /** |
| 887 | * @internal |
| 888 | * Serialize pfragment map to string for debugging dumps |
| 889 | */ |
| 890 | public function pFragmentMapToString(): string { |
| 891 | $codec = DOMDataUtils::getCodec( $this->getTopLevelDoc() ); |
| 892 | $buf = ''; |
| 893 | foreach ( $this->pFragmentMap as $k => $v ) { |
| 894 | $buf .= "$k = " . $codec->toJsonString( $v, PFragment::hint() ); |
| 895 | } |
| 896 | return $buf; |
| 897 | } |
| 898 | |
| 899 | /** |
| 900 | * Record a lint |
| 901 | * @param string $type Lint type key |
| 902 | * @param array $lintData Data for the lint. |
| 903 | * - dsr: (SourceRange) |
| 904 | * - params: (array) |
| 905 | * - templateInfo: (array|null) |
| 906 | */ |
| 907 | public function recordLint( string $type, array $lintData ): void { |
| 908 | if ( !$this->linting( $type ) ) { |
| 909 | return; |
| 910 | } |
| 911 | |
| 912 | if ( empty( $lintData['dsr'] ) ) { |
| 913 | $this->log( 'error/lint', "Missing DSR; msg=", $lintData ); |
| 914 | return; |
| 915 | } |
| 916 | |
| 917 | // This will always be recorded as a native 'byte' offset |
| 918 | $lintData['dsr'] = $lintData['dsr']->toJsonArray(); |
| 919 | $lintData['params'] ??= []; |
| 920 | |
| 921 | $this->lints[] = [ 'type' => $type ] + $lintData; |
| 922 | } |
| 923 | |
| 924 | /** |
| 925 | * Retrieve recorded lints |
| 926 | * @return array[] |
| 927 | */ |
| 928 | public function getLints(): array { |
| 929 | return $this->lints; |
| 930 | } |
| 931 | |
| 932 | /** |
| 933 | * Init lints to the passed array. |
| 934 | * |
| 935 | * FIXME: This is currently needed to reset lints after converting |
| 936 | * DSR offsets because of ordering of DOM passes. So, in reality, |
| 937 | * there should be no real use case for setting this anywhere else |
| 938 | * but from that single callsite. |
| 939 | * |
| 940 | * @param array $lints |
| 941 | */ |
| 942 | public function setLints( array $lints ): void { |
| 943 | $this->lints = $lints; |
| 944 | } |
| 945 | |
| 946 | /** |
| 947 | * @param string $prefix |
| 948 | * @param mixed ...$args |
| 949 | */ |
| 950 | public function log( string $prefix, ...$args ): void { |
| 951 | $this->parsoidLogger->log( $prefix, ...$args ); |
| 952 | } |
| 953 | |
| 954 | /** |
| 955 | * Shortcut helper that also allows early exit if tracing is not enabled. |
| 956 | * @param string $prefix |
| 957 | * @param mixed ...$args |
| 958 | */ |
| 959 | public function trace( string $prefix, ...$args ): void { |
| 960 | if ( $this->traceFlags ) { |
| 961 | $this->parsoidLogger->log( $prefix ? "trace/$prefix" : "trace", ...$args ); |
| 962 | } |
| 963 | } |
| 964 | |
| 965 | /** |
| 966 | * Bump usage of some limited parser resource |
| 967 | * (ex: tokens, # transclusions, # list items, etc.) |
| 968 | * |
| 969 | * @param string $resource |
| 970 | * @param int $count How much of the resource is used? |
| 971 | * @return ?bool Returns `null` if the limit was already reached, `false` when exceeded |
| 972 | */ |
| 973 | public function bumpWt2HtmlResourceUse( string $resource, int $count = 1 ): ?bool { |
| 974 | $n = $this->wt2htmlUsage[$resource] ?? 0; |
| 975 | if ( !$this->compareWt2HtmlLimit( $resource, $n ) ) { |
| 976 | return null; |
| 977 | } |
| 978 | $n += $count; |
| 979 | $this->wt2htmlUsage[$resource] = $n; |
| 980 | return $this->compareWt2HtmlLimit( $resource, $n ); |
| 981 | } |
| 982 | |
| 983 | /** |
| 984 | * @param string $resource |
| 985 | * @param int $n |
| 986 | * @return bool Return `false` when exceeded |
| 987 | */ |
| 988 | public function compareWt2HtmlLimit( string $resource, int $n ): bool { |
| 989 | $wt2htmlLimits = $this->siteConfig->getWt2HtmlLimits(); |
| 990 | return !( isset( $wt2htmlLimits[$resource] ) && $n > $wt2htmlLimits[$resource] ); |
| 991 | } |
| 992 | |
| 993 | /** |
| 994 | * Bump usage of some limited serializer resource |
| 995 | * (ex: html size) |
| 996 | * |
| 997 | * @param string $resource |
| 998 | * @param int $count How much of the resource is used? (defaults to 1) |
| 999 | * @throws ResourceLimitExceededException |
| 1000 | */ |
| 1001 | public function bumpHtml2WtResourceUse( string $resource, int $count = 1 ): void { |
| 1002 | $n = $this->html2wtUsage[$resource] ?? 0; |
| 1003 | $n += $count; |
| 1004 | $this->html2wtUsage[$resource] = $n; |
| 1005 | $html2wtLimits = $this->siteConfig->getHtml2WtLimits(); |
| 1006 | if ( |
| 1007 | isset( $html2wtLimits[$resource] ) && |
| 1008 | $n > $html2wtLimits[$resource] |
| 1009 | ) { |
| 1010 | throw new ResourceLimitExceededException( "html2wt: $resource limit exceeded: $n" ); |
| 1011 | } |
| 1012 | } |
| 1013 | |
| 1014 | /** |
| 1015 | * Get an appropriate content handler, given a contentmodel. |
| 1016 | * |
| 1017 | * @param ?string &$contentmodel An optional content model which |
| 1018 | * will override whatever the source specifies. It gets set to the |
| 1019 | * handler which is used. |
| 1020 | * @return ContentModelHandler An appropriate content handler |
| 1021 | */ |
| 1022 | public function getContentHandler( |
| 1023 | ?string &$contentmodel = null |
| 1024 | ): ContentModelHandler { |
| 1025 | $contentmodel ??= $this->pageConfig->getContentModel(); |
| 1026 | $handler = $this->siteConfig->getContentModelHandler( $contentmodel ); |
| 1027 | if ( !$handler && $contentmodel !== 'wikitext' ) { |
| 1028 | // For now, fallback to 'wikitext' as the default handler |
| 1029 | // FIXME: This is bogus, but this is just so suppress noise in our |
| 1030 | // logs till we get around to handling all these other content models. |
| 1031 | // $this->log( 'warn', "Unknown contentmodel $contentmodel" ); |
| 1032 | } |
| 1033 | return $handler ?? $this->wikitextContentModelHandler; |
| 1034 | } |
| 1035 | |
| 1036 | /** |
| 1037 | * Is the language converter enabled on this page? |
| 1038 | * |
| 1039 | * @return bool |
| 1040 | */ |
| 1041 | public function langConverterEnabled(): bool { |
| 1042 | return $this->siteConfig->langConverterEnabledBcp47( |
| 1043 | $this->pageConfig->getPageLanguageBcp47() |
| 1044 | ); |
| 1045 | } |
| 1046 | |
| 1047 | /** |
| 1048 | * The HTML content version of the input document (for html2wt and html2html conversions). |
| 1049 | * @see https://www.mediawiki.org/wiki/Parsoid/API#Content_Negotiation |
| 1050 | * @see https://www.mediawiki.org/wiki/Specs/HTML#Versioning |
| 1051 | * @return string A semver version number |
| 1052 | */ |
| 1053 | public function getInputContentVersion(): string { |
| 1054 | return $this->inputContentVersion; |
| 1055 | } |
| 1056 | |
| 1057 | /** |
| 1058 | * The HTML content version of the input document (for html2wt and html2html conversions). |
| 1059 | * @see https://www.mediawiki.org/wiki/Parsoid/API#Content_Negotiation |
| 1060 | * @see https://www.mediawiki.org/wiki/Specs/HTML#Versioning |
| 1061 | * @return string A semver version number |
| 1062 | */ |
| 1063 | public function getOutputContentVersion(): string { |
| 1064 | return $this->outputContentVersion; |
| 1065 | } |
| 1066 | |
| 1067 | /** |
| 1068 | * If non-null, the language variant used for Parsoid HTML; we convert |
| 1069 | * to this if wt2html, or from this (if html2wt). |
| 1070 | * |
| 1071 | * @return ?Bcp47Code a BCP-47 language code |
| 1072 | */ |
| 1073 | public function getHtmlVariantLanguageBcp47(): ?Bcp47Code { |
| 1074 | return $this->htmlVariantLanguage; // Stored as BCP-47 |
| 1075 | } |
| 1076 | |
| 1077 | /** |
| 1078 | * If non-null, the language variant to be used for wikitext. If null, |
| 1079 | * heuristics will be used to identify the original wikitext variant |
| 1080 | * in wt2html mode, and in html2wt mode new or edited HTML will be left |
| 1081 | * unconverted. |
| 1082 | * |
| 1083 | * @return ?Bcp47Code a BCP-47 language code |
| 1084 | */ |
| 1085 | public function getWtVariantLanguageBcp47(): ?Bcp47Code { |
| 1086 | return $this->wtVariantLanguage; |
| 1087 | } |
| 1088 | |
| 1089 | public function getSkipLanguageConversionPass(): bool { |
| 1090 | return $this->skipLanguageConversionPass; |
| 1091 | } |
| 1092 | |
| 1093 | /** |
| 1094 | * Determine appropriate vary headers for the HTML form of this page. |
| 1095 | * @return string |
| 1096 | */ |
| 1097 | public function htmlVary(): string { |
| 1098 | $varies = [ 'Accept' ]; // varies on Content-Type |
| 1099 | if ( $this->langConverterEnabled() ) { |
| 1100 | $varies[] = 'Accept-Language'; |
| 1101 | } |
| 1102 | |
| 1103 | sort( $varies ); |
| 1104 | return implode( ', ', $varies ); |
| 1105 | } |
| 1106 | |
| 1107 | /** |
| 1108 | * Determine an appropriate content-language for the HTML form of this page. |
| 1109 | * @return Bcp47Code a BCP-47 language code. |
| 1110 | */ |
| 1111 | public function htmlContentLanguageBcp47(): Bcp47Code { |
| 1112 | // PageConfig::htmlVariant is set iff we do variant conversion on the |
| 1113 | // HTML |
| 1114 | return $this->pageConfig->getVariantBcp47() ?? |
| 1115 | $this->pageConfig->getPageLanguageBcp47(); |
| 1116 | } |
| 1117 | |
| 1118 | /** |
| 1119 | * Get an array of attributes to apply to an anchor linking to $url |
| 1120 | */ |
| 1121 | public function getExternalLinkAttribs( string $url ): array { |
| 1122 | $siteConfig = $this->getSiteConfig(); |
| 1123 | $noFollowConfig = $siteConfig->getNoFollowConfig(); |
| 1124 | $attribs = []; |
| 1125 | $ns = $this->getContextTitle()->getNamespace(); |
| 1126 | if ( |
| 1127 | $noFollowConfig['nofollow'] && |
| 1128 | !in_array( $ns, $noFollowConfig['nsexceptions'], true ) && |
| 1129 | !UrlUtils::matchesDomainList( |
| 1130 | $url, |
| 1131 | // Cast to an array because parserTests sets it as a string |
| 1132 | (array)$noFollowConfig['domainexceptions'] |
| 1133 | ) |
| 1134 | ) { |
| 1135 | $attribs['rel'] = [ 'nofollow' ]; |
| 1136 | } |
| 1137 | $target = $siteConfig->getExternalLinkTarget(); |
| 1138 | if ( $target ) { |
| 1139 | $attribs['target'] = $target; |
| 1140 | if ( !in_array( $target, [ '_self', '_parent', '_top' ], true ) ) { |
| 1141 | // T133507. New windows can navigate parent cross-origin. |
| 1142 | // Including noreferrer due to lacking browser |
| 1143 | // support of noopener. Eventually noreferrer should be removed. |
| 1144 | if ( !isset( $attribs['rel'] ) ) { |
| 1145 | $attribs['rel'] = []; |
| 1146 | } |
| 1147 | array_push( $attribs['rel'], 'noreferrer', 'noopener' ); |
| 1148 | } |
| 1149 | } |
| 1150 | return $attribs; |
| 1151 | } |
| 1152 | |
| 1153 | /** |
| 1154 | * @return array |
| 1155 | */ |
| 1156 | public function getLinterConfig(): array { |
| 1157 | return $this->linterOverrides + $this->getSiteConfig()->getLinterSiteConfig(); |
| 1158 | } |
| 1159 | |
| 1160 | /** |
| 1161 | * Whether to enable linter Backend. |
| 1162 | * Consults the allow list and block list from ::getLinterConfig(). |
| 1163 | * |
| 1164 | * @param string|null $type If $type is null or omitted, returns true if *any* linting |
| 1165 | * type is enabled; otherwise returns true only if the specified |
| 1166 | * linting type is enabled. |
| 1167 | * @return bool If $type is null or omitted, returns true if *any* linting |
| 1168 | * type is enabled; otherwise returns true only if the specified |
| 1169 | * linting type is enabled. |
| 1170 | */ |
| 1171 | public function linting( ?string $type = null ) { |
| 1172 | if ( !$this->getSiteConfig()->linterEnabled() ) { |
| 1173 | return false; |
| 1174 | } |
| 1175 | $lintConfig = $this->getLinterConfig(); |
| 1176 | // Allow list |
| 1177 | $allowList = $lintConfig['enabled'] ?? null; |
| 1178 | if ( is_array( $allowList ) ) { |
| 1179 | if ( $type === null ) { |
| 1180 | return count( $allowList ) > 0; |
| 1181 | } |
| 1182 | return in_array( $type, $allowList, true ); |
| 1183 | } |
| 1184 | // Block list |
| 1185 | if ( $type === null ) { |
| 1186 | return true; |
| 1187 | } |
| 1188 | $blockList = $lintConfig['disabled'] ?? null; |
| 1189 | if ( is_array( $blockList ) ) { |
| 1190 | return !in_array( $type, $blockList, true ); |
| 1191 | } |
| 1192 | // No specific configuration |
| 1193 | return true; |
| 1194 | } |
| 1195 | } |