Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
0.00% |
0 / 262 |
|
0.00% |
0 / 51 |
CRAP | |
0.00% |
0 / 1 |
| ParsoidExtensionAPI | |
0.00% |
0 / 262 |
|
0.00% |
0 / 51 |
9900 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
2 | |||
| pushError | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
2 | |||
| createInterfaceI18nFragment | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| createPageContentI18nFragment | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| createLangI18nFragment | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| addInterfaceI18nAttribute | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| addPageContentI18nAttribute | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| addLangI18nAttribute | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getErrors | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getTopLevelDoc | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| newAboutId | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getSiteConfig | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getPageConfig | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getMetadata | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getTitleUri | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getPageUri | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| makeTitle | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| inTemplate | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| isPreview | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| parentExtTag | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| parentExtTagOpts | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getContentDOM | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| clearContentDOM | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| wikitextToDOM | |
0.00% |
0 / 31 |
|
0.00% |
0 / 1 |
30 | |||
| extTagToDOM | |
0.00% |
0 / 19 |
|
0.00% |
0 / 1 |
30 | |||
| setTempNodeData | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
2 | |||
| getTempNodeData | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
6 | |||
| extArgToDOM | |
0.00% |
0 / 17 |
|
0.00% |
0 / 1 |
20 | |||
| extArgsToArray | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| findAndUpdateArg | |
0.00% |
0 / 10 |
|
0.00% |
0 / 1 |
20 | |||
| addNewArg | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| log | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| processAttributeEmbeddedHTML | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| migrateChildrenAndTransferWrapperDataAttribs | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
2 | |||
| preprocessWikitext | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| htmlToDom | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
2 | |||
| domToHtml | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
6 | |||
| setHtml2wtStateFlag | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| extStartTagToWikitext | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
2 | |||
| domToWikitext | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| htmlToWikitext | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
2 | |||
| getOrigSrc | |
0.00% |
0 / 11 |
|
0.00% |
0 / 1 |
42 | |||
| domChildrenToWikitext | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
12 | |||
| escapeWikitext | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
6 | |||
| postProcessDOM | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
2 | |||
| renderMedia | |
0.00% |
0 / 70 |
|
0.00% |
0 / 1 |
600 | |||
| serializeMedia | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
6 | |||
| addModules | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| addModuleStyles | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getExternalLinkAttribs | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| addTrackingCategory | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
2 | |||
| 1 | <?php |
| 2 | declare( strict_types = 1 ); |
| 3 | |
| 4 | namespace Wikimedia\Parsoid\Ext; |
| 5 | |
| 6 | use Closure; |
| 7 | use Wikimedia\Bcp47Code\Bcp47Code; |
| 8 | use Wikimedia\Parsoid\Config\Env; |
| 9 | use Wikimedia\Parsoid\Config\PageConfig; |
| 10 | use Wikimedia\Parsoid\Config\SiteConfig; |
| 11 | use Wikimedia\Parsoid\Core\ContentMetadataCollector; |
| 12 | use Wikimedia\Parsoid\Core\ContentMetadataCollectorStringSets as CMCSS; |
| 13 | use Wikimedia\Parsoid\Core\DomSourceRange; |
| 14 | use Wikimedia\Parsoid\Core\MediaStructure; |
| 15 | use Wikimedia\Parsoid\Core\Sanitizer; |
| 16 | use Wikimedia\Parsoid\DOM\Document; |
| 17 | use Wikimedia\Parsoid\DOM\DocumentFragment; |
| 18 | use Wikimedia\Parsoid\DOM\Element; |
| 19 | use Wikimedia\Parsoid\DOM\Node; |
| 20 | use Wikimedia\Parsoid\Html2Wt\ConstrainedText\WikiLinkText; |
| 21 | use Wikimedia\Parsoid\Html2Wt\LinkHandlerUtils; |
| 22 | use Wikimedia\Parsoid\Html2Wt\SerializerState; |
| 23 | use Wikimedia\Parsoid\NodeData\DataMwError; |
| 24 | use Wikimedia\Parsoid\Tokens\KV; |
| 25 | use Wikimedia\Parsoid\Tokens\SourceRange; |
| 26 | use Wikimedia\Parsoid\Utils\ContentUtils; |
| 27 | use Wikimedia\Parsoid\Utils\DOMCompat; |
| 28 | use Wikimedia\Parsoid\Utils\DOMDataUtils; |
| 29 | use Wikimedia\Parsoid\Utils\DOMUtils; |
| 30 | use Wikimedia\Parsoid\Utils\PipelineUtils; |
| 31 | use Wikimedia\Parsoid\Utils\Title; |
| 32 | use Wikimedia\Parsoid\Utils\TokenUtils; |
| 33 | use Wikimedia\Parsoid\Utils\Utils; |
| 34 | use Wikimedia\Parsoid\Utils\WTUtils; |
| 35 | use Wikimedia\Parsoid\Wikitext\Wikitext; |
| 36 | use Wikimedia\Parsoid\Wt2Html\DOM\Processors\AddMetaData; |
| 37 | use Wikimedia\Parsoid\Wt2Html\Frame; |
| 38 | |
| 39 | /** |
| 40 | * Extensions are expected to use only these interfaces and strongly discouraged from |
| 41 | * calling Parsoid code directly. Code review is expected to catch these discouraged |
| 42 | * code patterns. We'll have to finish grappling with the extension and hooks API |
| 43 | * to go down this path seriously. Till then, we'll have extensions leveraging existing |
| 44 | * code as in the native extension code in this repository. |
| 45 | */ |
| 46 | class ParsoidExtensionAPI { |
| 47 | /** @var Env */ |
| 48 | private $env; |
| 49 | |
| 50 | /** @var ?Frame */ |
| 51 | private $frame; |
| 52 | |
| 53 | /** @var ?ExtensionTag */ |
| 54 | public $extTag; |
| 55 | |
| 56 | /** |
| 57 | * @var ?array TokenHandler options |
| 58 | */ |
| 59 | private $wt2htmlOpts; |
| 60 | |
| 61 | /** |
| 62 | * @var ?array Serialiation options / state |
| 63 | */ |
| 64 | private $html2wtOpts; |
| 65 | |
| 66 | /** |
| 67 | * @var ?SerializerState |
| 68 | */ |
| 69 | private $serializerState; |
| 70 | |
| 71 | /** |
| 72 | * Errors collected while parsing. This is used to indicate that, although |
| 73 | * an extension is returning a dom fragment, errors were encountered while |
| 74 | * generating it and should be marked up with the mw:Error typeof. |
| 75 | * |
| 76 | * @var list<DataMwError> |
| 77 | */ |
| 78 | private $errors = []; |
| 79 | |
| 80 | /** |
| 81 | * @param Env $env |
| 82 | * @param ?array $options |
| 83 | * - wt2html: used in wt->html direction |
| 84 | * - frame: (Frame) |
| 85 | * - parseOpts: (array) |
| 86 | * - extTag: (string) |
| 87 | * - extTagOpts: (array) |
| 88 | * - inTemplate: (bool) |
| 89 | * - extTag: (ExtensionTag) |
| 90 | * - html2wt: used in html->wt direction |
| 91 | * - state: (SerializerState) |
| 92 | */ |
| 93 | public function __construct( |
| 94 | Env $env, ?array $options = null |
| 95 | ) { |
| 96 | $this->env = $env; |
| 97 | $this->wt2htmlOpts = $options['wt2html'] ?? null; |
| 98 | $this->html2wtOpts = $options['html2wt'] ?? null; |
| 99 | $this->serializerState = $this->html2wtOpts['state'] ?? null; |
| 100 | $this->frame = $this->wt2htmlOpts['frame'] ?? null; |
| 101 | $this->extTag = $this->wt2htmlOpts['extTag'] ?? null; |
| 102 | } |
| 103 | |
| 104 | /** |
| 105 | * Collect errors while parsing. If processing can't continue, an |
| 106 | * ExtensionError should be thrown instead. |
| 107 | * |
| 108 | * $key and $params are basically the arguments to wfMessage, although they |
| 109 | * will be stored in the data-mw of the encapsulation wrapper. |
| 110 | * |
| 111 | * See https://www.mediawiki.org/wiki/Specs/HTML#Error_handling |
| 112 | * |
| 113 | * The returned fragment can be inserted in the dom and will be populated |
| 114 | * with the localized message. See T266666 |
| 115 | * |
| 116 | * @unstable |
| 117 | * @param string $key |
| 118 | * @param mixed ...$params |
| 119 | * @return DocumentFragment |
| 120 | */ |
| 121 | public function pushError( string $key, ...$params ): DocumentFragment { |
| 122 | $this->errors[] = new DataMwError( $key, $params ); |
| 123 | return WTUtils::createInterfaceI18nFragment( $this->getTopLevelDoc(), $key, $params ); |
| 124 | } |
| 125 | |
| 126 | /** |
| 127 | * Creates an internationalization (i18n) message that will be localized into the user |
| 128 | * interface language. The returned DocumentFragment contains, as a single child, a span |
| 129 | * element with the appropriate information for later localization. |
| 130 | * @param string $key message key for the message to be localized |
| 131 | * @param ?array $params parameters for localization |
| 132 | * @return DocumentFragment |
| 133 | */ |
| 134 | public function createInterfaceI18nFragment( string $key, ?array $params ): DocumentFragment { |
| 135 | return WTUtils::createInterfaceI18nFragment( $this->getTopLevelDoc(), $key, $params ); |
| 136 | } |
| 137 | |
| 138 | /** |
| 139 | * Creates an internationalization (i18n) message that will be localized into the page content |
| 140 | * language. The returned DocumentFragment contains, as a single child, a span |
| 141 | * element with the appropriate information for later localization. |
| 142 | * @param string $key message key for the message to be localized |
| 143 | * @param ?array $params parameters for localization |
| 144 | * @return DocumentFragment |
| 145 | */ |
| 146 | public function createPageContentI18nFragment( string $key, ?array $params ): DocumentFragment { |
| 147 | return WTUtils::createPageContentI18nFragment( $this->getTopLevelDoc(), $key, $params ); |
| 148 | } |
| 149 | |
| 150 | /** |
| 151 | * Creates an internationalization (i18n) message that will be localized into an arbitrary |
| 152 | * language. The returned DocumentFragment contains, as a single child, a span |
| 153 | * element with the appropriate information for later localization. |
| 154 | * The use of this method is discouraged; use ::createPageContentI18nFragment(...) and |
| 155 | * ::createInterfaceI18nFragment(...) where possible rather than, respectively, |
| 156 | * ::createLangI18nFragment($wgContLang, ...) and ::createLangI18nFragment($wgLang, ...). |
| 157 | * @param Bcp47Code $lang language in which the message will be localized |
| 158 | * @param string $key message key for the message to be localized |
| 159 | * @param ?array $params parameters for localization |
| 160 | * @return DocumentFragment |
| 161 | */ |
| 162 | public function createLangI18nFragment( Bcp47Code $lang, string $key, ?array $params ): DocumentFragment { |
| 163 | return WTUtils::createLangI18nFragment( $this->getTopLevelDoc(), $lang, $key, $params ); |
| 164 | } |
| 165 | |
| 166 | /** |
| 167 | * Adds to $element the internationalization information needed for the attribute $name to be |
| 168 | * localized in a later pass into the user interface language. |
| 169 | * @param Element $element element on which to add internationalization information |
| 170 | * @param string $name name of the attribute whose value will be localized |
| 171 | * @param string $key message key used for the attribute value localization |
| 172 | * @param ?array $params parameters for localization |
| 173 | */ |
| 174 | public function addInterfaceI18nAttribute( |
| 175 | Element $element, string $name, string $key, ?array $params |
| 176 | ): void { |
| 177 | WTUtils::addInterfaceI18nAttribute( $element, $name, $key, $params ); |
| 178 | } |
| 179 | |
| 180 | /** |
| 181 | * Adds to $element the internationalization information needed for the attribute $name to be |
| 182 | * localized in a later pass into the page content language. |
| 183 | * @param Element $element element on which to add internationalization information |
| 184 | * @param string $name name of the attribute whose value will be localized |
| 185 | * @param string $key message key used for the attribute value localization |
| 186 | * @param array $params parameters for localization |
| 187 | */ |
| 188 | public function addPageContentI18nAttribute( |
| 189 | Element $element, string $name, string $key, array $params |
| 190 | ): void { |
| 191 | WTUtils::addPageContentI18nAttribute( $element, $name, $key, $params ); |
| 192 | } |
| 193 | |
| 194 | /** |
| 195 | * Adds to $element the internationalization information needed for the attribute $name to be |
| 196 | * localized in a later pass into the provided language. |
| 197 | * The use of this method is discouraged; use ::addPageContentI18nAttribute(...) and |
| 198 | * ::addInterfaceI18nAttribute(...) where possible rather than, respectively, |
| 199 | * ::addLangI18nAttribute(..., $wgContLang, ...) and ::addLangI18nAttribute(..., $wgLang, ...). |
| 200 | * @param Element $element element on which to add internationalization information |
| 201 | * @param Bcp47Code $lang language in which the attribute will be localized |
| 202 | * @param string $name name of the attribute whose value will be localized |
| 203 | * @param string $key message key used for the attribute value localization |
| 204 | * @param array $params parameters for localization |
| 205 | */ |
| 206 | public function addLangI18nAttribute( |
| 207 | Element $element, Bcp47Code $lang, string $name, string $key, array $params |
| 208 | ) { |
| 209 | WTUtils::addLangI18nAttribute( $element, $lang, $name, $key, $params ); |
| 210 | } |
| 211 | |
| 212 | /** |
| 213 | * @return list<DataMwError> |
| 214 | */ |
| 215 | public function getErrors(): array { |
| 216 | return $this->errors; |
| 217 | } |
| 218 | |
| 219 | /** |
| 220 | * Returns the main document we're parsing. Extension content is parsed |
| 221 | * to fragments of this document. |
| 222 | * |
| 223 | * @return Document |
| 224 | */ |
| 225 | public function getTopLevelDoc(): Document { |
| 226 | return $this->env->topLevelDoc; |
| 227 | } |
| 228 | |
| 229 | /** |
| 230 | * Get a new about id for marking extension output |
| 231 | * FIXME: This should never really be needed since the extension API |
| 232 | * handles this on behalf of extensions, but Cite has one use case |
| 233 | * where implicit <references /> output is added. |
| 234 | * |
| 235 | * @return string |
| 236 | */ |
| 237 | public function newAboutId(): string { |
| 238 | return $this->env->newAboutId(); |
| 239 | } |
| 240 | |
| 241 | /** |
| 242 | * Get the site configuration to let extensions customize |
| 243 | * their behavior based on how the wiki is configured. |
| 244 | * |
| 245 | * @return SiteConfig |
| 246 | */ |
| 247 | public function getSiteConfig(): SiteConfig { |
| 248 | return $this->env->getSiteConfig(); |
| 249 | } |
| 250 | |
| 251 | /** |
| 252 | * FIXME: Unsure if we need to provide this access yet |
| 253 | * Get the page configuration |
| 254 | * @return PageConfig |
| 255 | */ |
| 256 | public function getPageConfig(): PageConfig { |
| 257 | return $this->env->getPageConfig(); |
| 258 | } |
| 259 | |
| 260 | /** |
| 261 | * Get the ContentMetadataCollector corresponding to the top-level page. |
| 262 | * In Parsoid integrated mode this will typically be an instance of |
| 263 | * core's `ParserOutput` class. |
| 264 | * |
| 265 | * @return ContentMetadataCollector |
| 266 | */ |
| 267 | public function getMetadata(): ContentMetadataCollector { |
| 268 | return $this->env->getMetadata(); |
| 269 | } |
| 270 | |
| 271 | /** |
| 272 | * Get the URI to link to a title |
| 273 | * @param Title $title |
| 274 | * @return string |
| 275 | */ |
| 276 | public function getTitleUri( Title $title ): string { |
| 277 | return $this->env->makeLink( $title ); |
| 278 | } |
| 279 | |
| 280 | /** |
| 281 | * Get an URI for the current page |
| 282 | * @return string |
| 283 | */ |
| 284 | public function getPageUri(): string { |
| 285 | return $this->getTitleUri( $this->env->getContextTitle() ); |
| 286 | } |
| 287 | |
| 288 | /** |
| 289 | * Make a title from an input string |
| 290 | * @param string $str |
| 291 | * @param int $namespaceId |
| 292 | * @return ?Title |
| 293 | */ |
| 294 | public function makeTitle( string $str, int $namespaceId ): ?Title { |
| 295 | return $this->env->makeTitleFromText( $str, $namespaceId, true /* no exceptions */ ); |
| 296 | } |
| 297 | |
| 298 | /** |
| 299 | * Are we parsing in a template context? |
| 300 | * @return bool |
| 301 | */ |
| 302 | public function inTemplate(): bool { |
| 303 | return $this->wt2htmlOpts['parseOpts']['inTemplate'] ?? false; |
| 304 | } |
| 305 | |
| 306 | /** |
| 307 | * Are we parsing for a preview? |
| 308 | * FIXME: Right now, we never do; when we do, this needs to be modified to reflect reality |
| 309 | * @unstable |
| 310 | * @return bool |
| 311 | */ |
| 312 | public function isPreview(): bool { |
| 313 | return false; |
| 314 | } |
| 315 | |
| 316 | /** |
| 317 | * FIXME: Is this something that can come from the frame? |
| 318 | * If we are parsing in the context of a parent extension tag, |
| 319 | * return the name of that extension tag |
| 320 | * @return string|null |
| 321 | */ |
| 322 | public function parentExtTag(): ?string { |
| 323 | return $this->wt2htmlOpts['parseOpts']['extTag'] ?? null; |
| 324 | } |
| 325 | |
| 326 | /** |
| 327 | * FIXME: Is this something that can come from the frame? |
| 328 | * If we are parsing in the context of a parent extension tag, |
| 329 | * return the parsing options set by that tag. |
| 330 | * @return array |
| 331 | */ |
| 332 | public function parentExtTagOpts(): array { |
| 333 | return $this->wt2htmlOpts['parseOpts']['extTagOpts'] ?? []; |
| 334 | } |
| 335 | |
| 336 | /** |
| 337 | * Get the content DOM corresponding to an id |
| 338 | * @param string $contentId |
| 339 | * @return DocumentFragment |
| 340 | */ |
| 341 | public function getContentDOM( string $contentId ): DocumentFragment { |
| 342 | return $this->env->getDOMFragment( $contentId ); |
| 343 | } |
| 344 | |
| 345 | public function clearContentDOM( string $contentId ): void { |
| 346 | $this->env->removeDOMFragment( $contentId ); |
| 347 | } |
| 348 | |
| 349 | /** |
| 350 | * Parse wikitext to DOM |
| 351 | * |
| 352 | * @param string $wikitext |
| 353 | * @param array $opts |
| 354 | * - srcOffsets |
| 355 | * - processInNewFrame |
| 356 | * - clearDSROffsets |
| 357 | * - shiftDSRFn |
| 358 | * - parseOpts |
| 359 | * - extTag |
| 360 | * - extTagOpts |
| 361 | * - context "inline", "block", etc. Currently, only "inline" is supported |
| 362 | * @param bool $sol Whether tokens should be processed in start-of-line context. |
| 363 | * @return DocumentFragment |
| 364 | */ |
| 365 | public function wikitextToDOM( |
| 366 | string $wikitext, array $opts, bool $sol |
| 367 | ): DocumentFragment { |
| 368 | if ( $wikitext === '' ) { |
| 369 | $domFragment = $this->getTopLevelDoc()->createDocumentFragment(); |
| 370 | } else { |
| 371 | // Parse content to DOM and pass DOM-fragment token back to the main pipeline. |
| 372 | // The DOM will get unwrapped and integrated when processing the top level document. |
| 373 | $parseOpts = $opts['parseOpts'] ?? []; |
| 374 | $srcOffsets = $opts['srcOffsets'] ?? null; |
| 375 | $frame = $this->frame; |
| 376 | if ( !empty( $opts['processInNewFrame'] ) ) { |
| 377 | $frame = $frame->newChild( $frame->getTitle(), [], $wikitext ); |
| 378 | $srcOffsets = new SourceRange( 0, strlen( $wikitext ) ); |
| 379 | } |
| 380 | $domFragment = PipelineUtils::processContentInPipeline( |
| 381 | $this->env, $frame, $wikitext, |
| 382 | [ |
| 383 | // Full pipeline for processing content |
| 384 | 'pipelineType' => 'wikitext-to-fragment', |
| 385 | 'pipelineOpts' => [ |
| 386 | 'expandTemplates' => true, |
| 387 | 'extTag' => $parseOpts['extTag'], |
| 388 | 'extTagOpts' => $parseOpts['extTagOpts'] ?? null, |
| 389 | 'inTemplate' => $this->inTemplate(), |
| 390 | 'inlineContext' => ( $parseOpts['context'] ?? '' ) === 'inline', |
| 391 | ], |
| 392 | 'srcOffsets' => $srcOffsets, |
| 393 | 'sol' => $sol, |
| 394 | ] |
| 395 | ); |
| 396 | |
| 397 | if ( !empty( $opts['clearDSROffsets'] ) ) { |
| 398 | $dsrFn = static function ( DomSourceRange $dsr ) { |
| 399 | return null; |
| 400 | }; |
| 401 | } else { |
| 402 | $dsrFn = $opts['shiftDSRFn'] ?? null; |
| 403 | } |
| 404 | |
| 405 | if ( $dsrFn ) { |
| 406 | ContentUtils::shiftDSR( $this->env, $domFragment, $dsrFn, $this ); |
| 407 | } |
| 408 | } |
| 409 | return $domFragment; |
| 410 | } |
| 411 | |
| 412 | /** |
| 413 | * Parse extension tag to DOM. |
| 414 | * |
| 415 | * If a wrapper tag is requested, beyond parsing the contents of the extension tag, |
| 416 | * this method wraps the contents in a custom wrapper element (ex: <div>), sanitizes |
| 417 | * the arguments of the extension args and sets some content flags on the wrapper. |
| 418 | * |
| 419 | * @param array $extArgs Args sanitized and applied to wrapper |
| 420 | * @param string $wikitext Wikitext content of the tag |
| 421 | * @param array $opts |
| 422 | * - srcOffsets |
| 423 | * - wrapperTag (skip OR pass null to not add any wrapper tag) |
| 424 | * - processInNewFrame |
| 425 | * - clearDSROffsets |
| 426 | * - shiftDSRFn |
| 427 | * - parseOpts |
| 428 | * - extTag |
| 429 | * - extTagOpts |
| 430 | * - context |
| 431 | * @return DocumentFragment |
| 432 | */ |
| 433 | public function extTagToDOM( |
| 434 | array $extArgs, string $wikitext, array $opts |
| 435 | ): DocumentFragment { |
| 436 | $extTagOffsets = $this->extTag->getOffsets(); |
| 437 | if ( !isset( $opts['srcOffsets'] ) ) { |
| 438 | $opts['srcOffsets'] = new SourceRange( |
| 439 | $extTagOffsets->innerStart(), |
| 440 | $extTagOffsets->innerEnd() |
| 441 | ); |
| 442 | } |
| 443 | |
| 444 | $domFragment = $this->wikitextToDOM( $wikitext, $opts, true /* sol */ ); |
| 445 | if ( !empty( $opts['wrapperTag'] ) ) { |
| 446 | // Create a wrapper and migrate content into the wrapper |
| 447 | $wrapper = $domFragment->ownerDocument->createElement( |
| 448 | $opts['wrapperTag'] |
| 449 | ); |
| 450 | DOMUtils::migrateChildren( $domFragment, $wrapper ); |
| 451 | $domFragment->appendChild( $wrapper ); |
| 452 | |
| 453 | // Sanitize args and set on the wrapper |
| 454 | Sanitizer::applySanitizedArgs( $this->env->getSiteConfig(), $wrapper, $extArgs ); |
| 455 | |
| 456 | // Mark empty content DOMs |
| 457 | if ( $wikitext === '' ) { |
| 458 | DOMDataUtils::getDataParsoid( $wrapper )->empty = true; |
| 459 | } |
| 460 | |
| 461 | if ( !empty( $this->extTag->isSelfClosed() ) ) { |
| 462 | DOMDataUtils::getDataParsoid( $wrapper )->selfClose = true; |
| 463 | } |
| 464 | } |
| 465 | |
| 466 | return $domFragment; |
| 467 | } |
| 468 | |
| 469 | /** |
| 470 | * Set temporary data into the DOM node that will be discarded |
| 471 | * when DOM is serialized |
| 472 | * |
| 473 | * Use the tag name as the key for TempData management |
| 474 | * |
| 475 | * @param Element $node |
| 476 | * @param mixed $data |
| 477 | */ |
| 478 | public function setTempNodeData( Element $node, $data ): void { |
| 479 | $dataParsoid = DOMDataUtils::getDataParsoid( $node ); |
| 480 | $tmpData = $dataParsoid->getTemp(); |
| 481 | $key = $this->extTag->getName(); |
| 482 | $tmpData->setTagData( $key, $data ); |
| 483 | } |
| 484 | |
| 485 | /** |
| 486 | * Get temporary data into the DOM node that will be discarded |
| 487 | * when DOM is serialized. |
| 488 | * |
| 489 | * This should only be used when the ExtensionTag is not available; otherwise access the newly created data |
| 490 | * directly. |
| 491 | * |
| 492 | * @param Element $node |
| 493 | * @param string $key to access TmpData |
| 494 | * @return mixed |
| 495 | * @unstable |
| 496 | */ |
| 497 | public function getTempNodeData( Element $node, string $key ) { |
| 498 | if ( $this->extTag ) { |
| 499 | throw new \RuntimeException( |
| 500 | 'ExtensionTag is available. Data should be available directly through the DOM.' |
| 501 | ); |
| 502 | } |
| 503 | $dataParsoid = DOMDataUtils::getDataParsoid( $node ); |
| 504 | $tmpData = $dataParsoid->getTemp(); |
| 505 | return $tmpData->getTagData( $key ); |
| 506 | } |
| 507 | |
| 508 | /** |
| 509 | * Process a specific extension arg as wikitext and return its DOM equivalent. |
| 510 | * By default, this method processes the argument value in inline context and normalizes |
| 511 | * every whitespace character to a single space. |
| 512 | * @param KV[] $extArgs |
| 513 | * @param string $key should be lower-case |
| 514 | * @param string $context |
| 515 | * @return ?DocumentFragment |
| 516 | */ |
| 517 | public function extArgToDOM( |
| 518 | array $extArgs, string $key, string $context = "inline" |
| 519 | ): ?DocumentFragment { |
| 520 | $argKV = KV::lookupKV( $extArgs, strtolower( $key ) ); |
| 521 | if ( $argKV === null || !$argKV->v ) { |
| 522 | return null; |
| 523 | } |
| 524 | |
| 525 | if ( $context === "inline" ) { |
| 526 | // `normalizeExtOptions` can mess up source offsets as well as the string |
| 527 | // that ought to be processed as wikitext. So, we do our own whitespace |
| 528 | // normalization of the original source here. |
| 529 | // |
| 530 | // If 'context' is 'inline' below, it ensures indent-pre / p-wrapping is suppressed. |
| 531 | // So, the normalization is primarily for HTML string parity. |
| 532 | $argVal = preg_replace( '/[\t\r\n ]/', ' ', $argKV->vsrc ); |
| 533 | } else { |
| 534 | $argVal = $argKV->vsrc; |
| 535 | } |
| 536 | |
| 537 | return $this->wikitextToDOM( |
| 538 | $argVal, |
| 539 | [ |
| 540 | 'parseOpts' => [ |
| 541 | 'extTag' => $this->extTag->getName(), |
| 542 | 'context' => $context, |
| 543 | ], |
| 544 | 'srcOffsets' => $argKV->valueOffset(), |
| 545 | ], |
| 546 | false // inline context => no sol state |
| 547 | ); |
| 548 | } |
| 549 | |
| 550 | /** |
| 551 | * Convert the ext args representation from an array of KV objects |
| 552 | * to a plain associative array mapping arg name strings to arg value strings. |
| 553 | * @param array<KV> $extArgs |
| 554 | * @return array<string,string> |
| 555 | */ |
| 556 | public function extArgsToArray( array $extArgs ): array { |
| 557 | return TokenUtils::kvToHash( $extArgs ); |
| 558 | } |
| 559 | |
| 560 | /** |
| 561 | * This method finds a requested arg by key name and return its current value. |
| 562 | * If a closure is passed in to update the current value, it is used to update the arg. |
| 563 | * |
| 564 | * @param KV[] &$extArgs Array of extension args |
| 565 | * @param string $key Argument key whose value needs an update |
| 566 | * @param ?Closure $updater $updater will get the existing string value |
| 567 | * for the arg and is expected to return an updated value. |
| 568 | * @return ?string |
| 569 | */ |
| 570 | public function findAndUpdateArg( |
| 571 | array &$extArgs, string $key, ?Closure $updater = null |
| 572 | ): ?string { |
| 573 | // FIXME: This code will get an overhaul when T250854 is resolved. |
| 574 | foreach ( $extArgs as $i => $kv ) { |
| 575 | $k = TokenUtils::tokensToString( $kv->k ); |
| 576 | if ( strtolower( trim( $k ) ) === strtolower( $key ) ) { |
| 577 | $val = $kv->v; |
| 578 | if ( $updater ) { |
| 579 | $kv = clone $kv; |
| 580 | $kv->v = $updater( TokenUtils::tokensToString( $val ) ); |
| 581 | $extArgs[$i] = $kv; |
| 582 | } |
| 583 | return $val; |
| 584 | } |
| 585 | } |
| 586 | |
| 587 | return null; |
| 588 | } |
| 589 | |
| 590 | /** |
| 591 | * This method adds a new argument to the extension args array |
| 592 | * @param KV[] &$extArgs |
| 593 | * @param string $key |
| 594 | * @param string $value |
| 595 | */ |
| 596 | public function addNewArg( array &$extArgs, string $key, string $value ): void { |
| 597 | $extArgs[] = new KV( $key, $value ); |
| 598 | } |
| 599 | |
| 600 | // TODO: Provide support for extensions to register lints |
| 601 | // from their customized lint handlers. |
| 602 | |
| 603 | /** |
| 604 | * Forwards the logging request to the underlying logger |
| 605 | * @param string $prefix |
| 606 | * @param mixed ...$args |
| 607 | */ |
| 608 | public function log( string $prefix, ...$args ): void { |
| 609 | $this->env->log( $prefix, ...$args ); |
| 610 | } |
| 611 | |
| 612 | /** |
| 613 | * Extensions might be interested in examining (their) content embedded |
| 614 | * in data-mw attributes that don't otherwise show up in the DOM. |
| 615 | * |
| 616 | * Ex: inline media captions that aren't rendered, language variant markup, |
| 617 | * attributes that are transcluded. More scenarios might be added later. |
| 618 | * |
| 619 | * @param Element $elt The node whose data attributes need to be examined |
| 620 | * @param Closure $proc The processor that will process the embedded HTML |
| 621 | * Signature: (string) -> string |
| 622 | * This processor will be provided the HTML string as input |
| 623 | * and is expected to return a possibly modified string. |
| 624 | */ |
| 625 | public function processAttributeEmbeddedHTML( Element $elt, Closure $proc ): void { |
| 626 | ContentUtils::processAttributeEmbeddedHTML( $this, $elt, $proc ); |
| 627 | } |
| 628 | |
| 629 | /** |
| 630 | * Copy $from->childNodes to $to and clone the data attributes of $from |
| 631 | * to $to. |
| 632 | * |
| 633 | * @param Element $from |
| 634 | * @param Element $to |
| 635 | */ |
| 636 | public static function migrateChildrenAndTransferWrapperDataAttribs( |
| 637 | Element $from, Element $to |
| 638 | ): void { |
| 639 | DOMUtils::migrateChildren( $from, $to ); |
| 640 | DOMDataUtils::setDataParsoid( |
| 641 | $to, DOMDataUtils::getDataParsoid( $from )->clone() |
| 642 | ); |
| 643 | DOMDataUtils::setDataMw( |
| 644 | $to, Utils::clone( DOMDataUtils::getDataMw( $from ) ) |
| 645 | ); |
| 646 | } |
| 647 | |
| 648 | /** |
| 649 | * Equivalent of 'preprocess' from Parser.php in core. |
| 650 | * - expands templates |
| 651 | * - replaces magic variables |
| 652 | * This does not run any hooks however since that would be unexpected. |
| 653 | * This also doesn't support replacing template args from a frame. |
| 654 | * |
| 655 | * @param string $wikitext |
| 656 | * @return string preprocessed wikitext |
| 657 | */ |
| 658 | public function preprocessWikitext( string $wikitext ): string { |
| 659 | return Wikitext::preprocess( $this->env, $wikitext )['src']; |
| 660 | } |
| 661 | |
| 662 | /** |
| 663 | * Parse input string into DOM. |
| 664 | * NOTE: This leaves the DOM in Parsoid-canonical state and is the preferred method |
| 665 | * to convert HTML to DOM that will be passed into Parsoid's processing code. |
| 666 | * |
| 667 | * @param string $html |
| 668 | * @param ?Document $doc XXX You probably don't want to be doing this |
| 669 | * @param ?array $options |
| 670 | * @return DocumentFragment |
| 671 | */ |
| 672 | public function htmlToDom( |
| 673 | string $html, ?Document $doc = null, ?array $options = [] |
| 674 | ): DocumentFragment { |
| 675 | return ContentUtils::createAndLoadDocumentFragment( |
| 676 | $doc ?? $this->getTopLevelDoc(), $html, $options |
| 677 | ); |
| 678 | } |
| 679 | |
| 680 | /** |
| 681 | * Serialize DOM element to string (inner/outer HTML is controlled by flag). |
| 682 | * If $releaseDom is set to true, the DOM will be left in non-canonical form |
| 683 | * and is not safe to use after this call. This is primarily a performance optimization. |
| 684 | * |
| 685 | * @param Node $node |
| 686 | * @param bool $innerHTML if true, inner HTML of the element will be returned |
| 687 | * This flag defaults to false |
| 688 | * @param bool $releaseDom if true, the DOM will not be in canonical form after this call |
| 689 | * This flag defaults to false |
| 690 | * @return string |
| 691 | */ |
| 692 | public function domToHtml( |
| 693 | Node $node, bool $innerHTML = false, bool $releaseDom = false |
| 694 | ): string { |
| 695 | // FIXME: This is going to drop any diff markers but since |
| 696 | // the dom differ doesn't traverse into extension content (right now), |
| 697 | // none should exist anyways. |
| 698 | $html = ContentUtils::ppToXML( $node, [ 'innerXML' => $innerHTML ] ); |
| 699 | if ( !$releaseDom ) { |
| 700 | DOMDataUtils::visitAndLoadDataAttribs( $node ); |
| 701 | } |
| 702 | return $html; |
| 703 | } |
| 704 | |
| 705 | /** |
| 706 | * Bit flags describing escaping / serializing context in html -> wt mode |
| 707 | */ |
| 708 | public const IN_SOL = 1; |
| 709 | public const IN_MEDIA = 2; |
| 710 | public const IN_LINK = 4; |
| 711 | public const IN_IMG_CAPTION = 8; |
| 712 | public const IN_OPTION = 16; |
| 713 | |
| 714 | /** |
| 715 | * FIXME: This is a bit broken - shouldn't be needed ideally |
| 716 | * @param string $flag |
| 717 | */ |
| 718 | public function setHtml2wtStateFlag( string $flag ) { |
| 719 | $this->serializerState->{$flag} = true; |
| 720 | } |
| 721 | |
| 722 | /** |
| 723 | * Emit the opening tag (including attributes) for the extension |
| 724 | * represented by this node. |
| 725 | * |
| 726 | * @param Element $node |
| 727 | * @return string |
| 728 | */ |
| 729 | public function extStartTagToWikitext( Element $node ): string { |
| 730 | $state = $this->serializerState; |
| 731 | return $state->serializer->serializeExtensionStartTag( $node, $state ); |
| 732 | } |
| 733 | |
| 734 | /** |
| 735 | * Convert the input DOM to wikitext. |
| 736 | * |
| 737 | * @param array $opts |
| 738 | * - extName: (string) Name of the extension whose body we are serializing |
| 739 | * - inPHPBlock: (bool) FIXME: This needs to be removed |
| 740 | * @param Element $node DOM to serialize |
| 741 | * @param bool $releaseDom If $releaseDom is set to true, the DOM will be left in |
| 742 | * non-canonical form and is not safe to use after this call. This is primarily a |
| 743 | * performance optimization. This flag defaults to false. |
| 744 | * @return mixed |
| 745 | */ |
| 746 | public function domToWikitext( array $opts, Element $node, bool $releaseDom = false ) { |
| 747 | // FIXME: WTS expects the input DOM to be a <body> element! |
| 748 | // Till that is fixed, we have to go through this round-trip! |
| 749 | // TODO: Move $node children to a fragment and call `$serializer->domToWikitext` |
| 750 | return $this->htmlToWikitext( $opts, $this->domToHtml( $node, $releaseDom ) ); |
| 751 | } |
| 752 | |
| 753 | /** |
| 754 | * Convert the HTML body of an extension to wikitext |
| 755 | * |
| 756 | * @param array $opts |
| 757 | * - extName: (string) Name of the extension whose body we are serializing |
| 758 | * - inPHPBlock: (bool) FIXME: This needs to be removed |
| 759 | * @param string $html HTML for the extension's body |
| 760 | * @return string |
| 761 | */ |
| 762 | public function htmlToWikitext( array $opts, string $html ): string { |
| 763 | // Type cast so phan has more information to ensure type safety |
| 764 | $state = $this->serializerState; |
| 765 | $opts['env'] = $this->env; |
| 766 | return $state->serializer->htmlToWikitext( $opts, $html ); |
| 767 | } |
| 768 | |
| 769 | /** |
| 770 | * Get the original source for an element. |
| 771 | * |
| 772 | * The callable, $checkIfOrigSrcReusable, is used to determine if the $elt |
| 773 | * is unedited and therefore valid to reuse source. This is assumed to be |
| 774 | * pretty specific to the callsite so no default is provided. |
| 775 | * |
| 776 | * @param Element $elt |
| 777 | * @param bool $inner |
| 778 | * @param callable $checkIfOrigSrcReusable |
| 779 | * @return string|null |
| 780 | */ |
| 781 | public function getOrigSrc( |
| 782 | Element $elt, bool $inner, callable $checkIfOrigSrcReusable |
| 783 | ): ?string { |
| 784 | $state = $this->serializerState; |
| 785 | if ( !$state->selserMode || $state->inInsertedContent ) { |
| 786 | return null; |
| 787 | } |
| 788 | $dsr = DOMDataUtils::getDataParsoid( $elt )->dsr ?? null; |
| 789 | if ( !Utils::isValidDSR( $dsr, $inner ) ) { |
| 790 | return null; |
| 791 | } |
| 792 | if ( $checkIfOrigSrcReusable( $elt ) ) { |
| 793 | return $state->getOrigSrc( |
| 794 | $inner ? $dsr->innerRange() : $dsr |
| 795 | ); |
| 796 | } else { |
| 797 | return null; |
| 798 | } |
| 799 | } |
| 800 | |
| 801 | /** |
| 802 | * @param Element $elt |
| 803 | * @param int $context OR-ed bit flags specifying escaping / serialization context |
| 804 | * @return string |
| 805 | */ |
| 806 | public function domChildrenToWikitext( Element $elt, int $context ): string { |
| 807 | $state = $this->serializerState; |
| 808 | if ( $context & self::IN_IMG_CAPTION ) { |
| 809 | if ( $context & self::IN_OPTION ) { |
| 810 | $escapeHandler = 'mediaOptionHandler'; // Escapes "|" as well |
| 811 | } else { |
| 812 | $escapeHandler = 'wikilinkHandler'; // image captions show up in wikilink syntax |
| 813 | } |
| 814 | $out = $state->serializeCaptionChildrenToString( $elt, |
| 815 | [ $state->serializer->wteHandlers, $escapeHandler ] ); |
| 816 | } else { |
| 817 | throw new \RuntimeException( 'Not yet supported!' ); |
| 818 | } |
| 819 | return $out; |
| 820 | } |
| 821 | |
| 822 | /** |
| 823 | * Escape any wikitext like constructs in a string so that when the output |
| 824 | * is parsed, it renders as a string. The escaping is sensitive to the context |
| 825 | * in which the string is embedded. For example, a "*" is not safe at the start |
| 826 | * of a line (since it will parse as a list item), but is safe if it is not in |
| 827 | * a start of line context. Similarly the "|" character is safe outside tables, |
| 828 | * links, and transclusions. |
| 829 | * |
| 830 | * @param string $str |
| 831 | * @param Node $node |
| 832 | * @param int $context OR-ed bit flags specifying escaping / serialization context |
| 833 | * @return string |
| 834 | */ |
| 835 | public function escapeWikitext( string $str, Node $node, int $context ): string { |
| 836 | if ( $context & ( self::IN_MEDIA | self::IN_LINK ) ) { |
| 837 | $state = $this->serializerState; |
| 838 | return $state->serializer->wteHandlers->escapeLinkContent( |
| 839 | $state, $str, |
| 840 | (bool)( $context & self::IN_SOL ), |
| 841 | $node, |
| 842 | (bool)( $context & self::IN_MEDIA ) |
| 843 | ); |
| 844 | } else { |
| 845 | throw new \RuntimeException( 'Not yet supported!' ); |
| 846 | } |
| 847 | } |
| 848 | |
| 849 | /** |
| 850 | * EXTAPI-FIXME: We have to figure out what it means to run a DOM pass |
| 851 | * (and what processors and what handlers apply) on content models that are |
| 852 | * not wikitext. For now, we are only storing data attribs back to the DOM |
| 853 | * and adding metadata to the page. |
| 854 | * |
| 855 | * @param Document $doc |
| 856 | */ |
| 857 | public function postProcessDOM( Document $doc ): void { |
| 858 | $env = $this->env; |
| 859 | // From CleanUp::saveDataParsoid |
| 860 | $body = DOMCompat::getBody( $doc ); |
| 861 | DOMDataUtils::visitAndStoreDataAttribs( $body, [ |
| 862 | 'storeInPageBundle' => $env->pageBundle, |
| 863 | 'env' => $env |
| 864 | ] ); |
| 865 | // Ugh! But, this whole method needs to go away anyway |
| 866 | ( new AddMetaData( null ) )->run( $env, $body ); |
| 867 | } |
| 868 | |
| 869 | /** |
| 870 | * Produce the HTML rendering of a title string and media options as the |
| 871 | * wikitext parser would for a wikilink in the file namespace |
| 872 | * |
| 873 | * @param string $titleStr Image title string |
| 874 | * @param array $imageOpts Array of a mix of strings or arrays, |
| 875 | * the latter of which can signify that the value came from source. |
| 876 | * Where, |
| 877 | * [0] is the fully-constructed image option |
| 878 | * [1] is the full wikitext source offset for it |
| 879 | * @param ?string &$error Error string is set when the return is null. |
| 880 | * @param ?bool $forceBlock Forces the media to be rendered in a figure as |
| 881 | * opposed to a span. |
| 882 | * @param ?bool $suppressMediaFormats If any media format is present in |
| 883 | * $imageOpts, it won't be applied and will result in a linting error. |
| 884 | * @return ?Element |
| 885 | */ |
| 886 | public function renderMedia( |
| 887 | string $titleStr, array $imageOpts, ?string &$error = null, |
| 888 | ?bool $forceBlock = false, ?bool $suppressMediaFormats = false |
| 889 | ): ?Element { |
| 890 | $extTagName = $this->extTag->getName(); |
| 891 | $extTagOpts = [ 'suppressMediaFormats' => $suppressMediaFormats ]; |
| 892 | |
| 893 | $fileNs = $this->getSiteConfig()->canonicalNamespaceId( 'file' ); |
| 894 | |
| 895 | $title = $this->makeTitle( $titleStr, 0 ); |
| 896 | if ( $title === null || $title->getNamespace() !== $fileNs ) { |
| 897 | $error = "{$extTagName}_no_image"; |
| 898 | return null; |
| 899 | } |
| 900 | |
| 901 | $pieces = [ '[[' ]; |
| 902 | // Since the above two chars aren't from source, the resulting figure |
| 903 | // won't have any dsr info, so we can omit an offset for the title as |
| 904 | // well. In any case, $titleStr may not necessarily be from source, |
| 905 | // see the special case in the gallery extension. |
| 906 | $pieces[] = $titleStr; |
| 907 | $pieces = array_merge( $pieces, $imageOpts ); |
| 908 | |
| 909 | if ( $forceBlock ) { |
| 910 | // We add "none" here so that this renders in the block form |
| 911 | // (ie. figure). It's a valid media option, so shouldn't turn into |
| 912 | // a caption. And since it's first wins, it shouldn't interfere |
| 913 | // with another horizontal alignment defined in $imageOpts. |
| 914 | // We just have to remember to strip the class below. |
| 915 | // NOTE: This will have to be adjusted with T305628 |
| 916 | $pieces[] = '|none'; |
| 917 | } |
| 918 | |
| 919 | $pieces[] = ']]'; |
| 920 | |
| 921 | $shiftOffset = static function ( int $offset ) use ( $pieces ): ?int { |
| 922 | foreach ( $pieces as $p ) { |
| 923 | if ( is_string( $p ) ) { |
| 924 | $offset -= strlen( $p ); |
| 925 | if ( $offset <= 0 ) { |
| 926 | return null; |
| 927 | } |
| 928 | } else { |
| 929 | if ( $offset <= strlen( $p[0] ) && isset( $p[1] ) ) { |
| 930 | return $p[1] + $offset; |
| 931 | } |
| 932 | $offset -= strlen( $p[0] ); |
| 933 | if ( $offset <= 0 ) { |
| 934 | return null; |
| 935 | } |
| 936 | } |
| 937 | } |
| 938 | return null; |
| 939 | }; |
| 940 | |
| 941 | $imageWt = ''; |
| 942 | foreach ( $pieces as $p ) { |
| 943 | $imageWt .= ( is_string( $p ) ? $p : $p[0] ); |
| 944 | } |
| 945 | |
| 946 | $domFragment = $this->wikitextToDOM( |
| 947 | $imageWt, |
| 948 | [ |
| 949 | 'parseOpts' => [ |
| 950 | 'extTag' => $extTagName, |
| 951 | 'extTagOpts' => $extTagOpts, |
| 952 | 'context' => 'inline', |
| 953 | ], |
| 954 | // Create new frame, because $pieces doesn't literally appear |
| 955 | // on the page, it has been hand-crafted here |
| 956 | 'processInNewFrame' => true, |
| 957 | // Shift the DSRs in the DOM by startOffset, and strip DSRs |
| 958 | // for bits which aren't the caption or file, since they |
| 959 | // don't refer to actual source wikitext |
| 960 | 'shiftDSRFn' => static function ( DomSourceRange $dsr ) use ( $shiftOffset ) { |
| 961 | $start = $dsr->start === null ? null : $shiftOffset( $dsr->start ); |
| 962 | $end = $dsr->end === null ? null : $shiftOffset( $dsr->end ); |
| 963 | // If either offset is newly-invalid, remove entire DSR |
| 964 | if ( |
| 965 | ( $dsr->start !== null && $start === null ) || |
| 966 | ( $dsr->end !== null && $end === null ) |
| 967 | ) { |
| 968 | return null; |
| 969 | } |
| 970 | return new DomSourceRange( |
| 971 | $start, $end, $dsr->openWidth, $dsr->closeWidth |
| 972 | ); |
| 973 | }, |
| 974 | ], |
| 975 | true // sol |
| 976 | ); |
| 977 | |
| 978 | $thumb = $domFragment->firstChild; |
| 979 | $validWrappers = [ 'figure' ]; |
| 980 | // Downstream code expects a figcaption if we're forcing a block so we |
| 981 | // validate that we did indeed parse a figure. It might not have |
| 982 | // happened because $imageOpts has an unbalanced `]]` which closes |
| 983 | // the wikilink syntax before we get in our `|none`. |
| 984 | if ( !$forceBlock ) { |
| 985 | $validWrappers[] = 'span'; |
| 986 | } |
| 987 | if ( !in_array( DOMCompat::nodeName( $thumb ), $validWrappers, true ) ) { |
| 988 | $error = "{$extTagName}_invalid_image"; |
| 989 | return null; |
| 990 | } |
| 991 | DOMUtils::assertElt( $thumb ); |
| 992 | |
| 993 | // Detach the $thumb since the $domFragment is going out of scope |
| 994 | // See https://bugs.php.net/bug.php?id=39593 |
| 995 | DOMCompat::remove( $thumb ); |
| 996 | |
| 997 | if ( $forceBlock ) { |
| 998 | $dp = DOMDataUtils::getDataParsoid( $thumb ); |
| 999 | array_pop( $dp->optList ); |
| 1000 | $explicitNone = false; |
| 1001 | foreach ( $dp->optList as $opt ) { |
| 1002 | if ( $opt['ck'] === 'none' ) { |
| 1003 | $explicitNone = true; |
| 1004 | } |
| 1005 | } |
| 1006 | if ( !$explicitNone ) { |
| 1007 | // FIXME: Should we worry about someone adding this with the |
| 1008 | // "class=" option? |
| 1009 | DOMCompat::getClassList( $thumb )->remove( 'mw-halign-none' ); |
| 1010 | } |
| 1011 | } |
| 1012 | |
| 1013 | return $thumb; |
| 1014 | } |
| 1015 | |
| 1016 | /** |
| 1017 | * Serialize a MediaStructure to a title and media options string. |
| 1018 | * The converse to ::renderMedia. |
| 1019 | * |
| 1020 | * @param MediaStructure $ms |
| 1021 | * @return array Where, |
| 1022 | * [0] is the media title string |
| 1023 | * [1] is the string of media options |
| 1024 | */ |
| 1025 | public function serializeMedia( MediaStructure $ms ): array { |
| 1026 | $ct = LinkHandlerUtils::figureToConstrainedText( $this->serializerState, $ms ); |
| 1027 | if ( $ct instanceof WikiLinkText ) { |
| 1028 | // Remove the opening and closing square brackets |
| 1029 | $text = substr( $ct->text, 2, -2 ); |
| 1030 | return array_pad( explode( '|', $text, 2 ), 2, '' ); |
| 1031 | } else { |
| 1032 | // Note that $ct could be an AutoURLLinkText, not just null |
| 1033 | return [ '', '' ]; |
| 1034 | } |
| 1035 | } |
| 1036 | |
| 1037 | /** |
| 1038 | * @param array $modules |
| 1039 | * @deprecated Use ::getMetadata()->appendOutputStrings( MODULE, ...) instead. |
| 1040 | */ |
| 1041 | public function addModules( array $modules ) { |
| 1042 | $this->getMetadata()->appendOutputStrings( CMCSS::MODULE, $modules ); |
| 1043 | } |
| 1044 | |
| 1045 | /** |
| 1046 | * @param array $modulestyles |
| 1047 | * @deprecated Use ::getMetadata()->appendOutputStrings(MODULE_STYLE, ...) instead. |
| 1048 | */ |
| 1049 | public function addModuleStyles( array $modulestyles ) { |
| 1050 | $this->getMetadata()->appendOutputStrings( CMCSS::MODULE_STYLE, $modulestyles ); |
| 1051 | } |
| 1052 | |
| 1053 | /** |
| 1054 | * Get an array of attributes to apply to an anchor linking to $url |
| 1055 | */ |
| 1056 | public function getExternalLinkAttribs( string $url ): array { |
| 1057 | return $this->env->getExternalLinkAttribs( $url ); |
| 1058 | } |
| 1059 | |
| 1060 | /** |
| 1061 | * Add a tracking category to the current page. |
| 1062 | * @param string $key Message key (not localized) |
| 1063 | */ |
| 1064 | public function addTrackingCategory( string $key ): void { |
| 1065 | $this->env->getDataAccess()->addTrackingCategory( |
| 1066 | $this->env->getPageConfig(), |
| 1067 | $this->env->getMetadata(), |
| 1068 | $key |
| 1069 | ); |
| 1070 | } |
| 1071 | } |