Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
7.09% |
21 / 296 |
|
1.85% |
1 / 54 |
CRAP | |
0.00% |
0 / 1 |
| ParsoidExtensionAPI | |
7.09% |
21 / 296 |
|
1.85% |
1 / 54 |
10171.10 | |
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 / 41 |
|
0.00% |
0 / 1 |
56 | |||
| 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 | |||
| updateAllArgs | |
0.00% |
0 / 8 |
|
0.00% |
0 / 1 |
12 | |||
| normalizeWhiteSpaceInArgs | |
100.00% |
21 / 21 |
|
100.00% |
1 / 1 |
7 | |||
| 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 | |||
| processAttributeEmbeddedDom | |
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 / 1 |
|
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\Fragments\PFragment; |
| 21 | use Wikimedia\Parsoid\Fragments\WikitextPFragment; |
| 22 | use Wikimedia\Parsoid\Html2Wt\ConstrainedText\WikiLinkText; |
| 23 | use Wikimedia\Parsoid\Html2Wt\LinkHandlerUtils; |
| 24 | use Wikimedia\Parsoid\Html2Wt\SerializerState; |
| 25 | use Wikimedia\Parsoid\NodeData\DataMwError; |
| 26 | use Wikimedia\Parsoid\Tokens\KV; |
| 27 | use Wikimedia\Parsoid\Tokens\SourceRange; |
| 28 | use Wikimedia\Parsoid\Utils\ContentUtils; |
| 29 | use Wikimedia\Parsoid\Utils\DOMCompat; |
| 30 | use Wikimedia\Parsoid\Utils\DOMDataUtils; |
| 31 | use Wikimedia\Parsoid\Utils\DOMUtils; |
| 32 | use Wikimedia\Parsoid\Utils\PipelineUtils; |
| 33 | use Wikimedia\Parsoid\Utils\Title; |
| 34 | use Wikimedia\Parsoid\Utils\TokenUtils; |
| 35 | use Wikimedia\Parsoid\Utils\Utils; |
| 36 | use Wikimedia\Parsoid\Utils\WTUtils; |
| 37 | use Wikimedia\Parsoid\Wikitext\Wikitext; |
| 38 | use Wikimedia\Parsoid\Wt2Html\DOM\Processors\AddMetaData; |
| 39 | use Wikimedia\Parsoid\Wt2Html\Frame; |
| 40 | |
| 41 | /** |
| 42 | * Extensions are expected to use only these interfaces and strongly discouraged from |
| 43 | * calling Parsoid code directly. Code review is expected to catch these discouraged |
| 44 | * code patterns. We'll have to finish grappling with the extension and hooks API |
| 45 | * to go down this path seriously. Till then, we'll have extensions leveraging existing |
| 46 | * code as in the native extension code in this repository. |
| 47 | */ |
| 48 | class ParsoidExtensionAPI { |
| 49 | /** @var Env */ |
| 50 | private $env; |
| 51 | |
| 52 | /** @var ?Frame */ |
| 53 | private $frame; |
| 54 | |
| 55 | /** @var ?ExtensionTag */ |
| 56 | public $extTag; |
| 57 | |
| 58 | /** |
| 59 | * @var ?array TokenHandler options |
| 60 | */ |
| 61 | private $wt2htmlOpts; |
| 62 | |
| 63 | /** |
| 64 | * @var ?array Serialiation options / state |
| 65 | */ |
| 66 | private $html2wtOpts; |
| 67 | |
| 68 | /** |
| 69 | * @var ?SerializerState |
| 70 | */ |
| 71 | private $serializerState; |
| 72 | |
| 73 | /** |
| 74 | * Errors collected while parsing. This is used to indicate that, although |
| 75 | * an extension is returning a dom fragment, errors were encountered while |
| 76 | * generating it and should be marked up with the mw:Error typeof. |
| 77 | * |
| 78 | * @var list<DataMwError> |
| 79 | */ |
| 80 | private $errors = []; |
| 81 | |
| 82 | /** |
| 83 | * @param Env $env |
| 84 | * @param ?array $options |
| 85 | * - wt2html: used in wt->html direction |
| 86 | * - frame: (Frame) |
| 87 | * - parseOpts: (array) |
| 88 | * - extTag: (string) |
| 89 | * - extTagOpts: (array) |
| 90 | * - inTemplate: (bool) |
| 91 | * - extTag: (ExtensionTag) |
| 92 | * - html2wt: used in html->wt direction |
| 93 | * - state: (SerializerState) |
| 94 | */ |
| 95 | public function __construct( |
| 96 | Env $env, ?array $options = null |
| 97 | ) { |
| 98 | $this->env = $env; |
| 99 | $this->wt2htmlOpts = $options['wt2html'] ?? null; |
| 100 | $this->html2wtOpts = $options['html2wt'] ?? null; |
| 101 | $this->serializerState = $this->html2wtOpts['state'] ?? null; |
| 102 | $this->frame = $this->wt2htmlOpts['frame'] ?? $env->topFrame ?? null; |
| 103 | $this->extTag = $this->wt2htmlOpts['extTag'] ?? null; |
| 104 | } |
| 105 | |
| 106 | /** |
| 107 | * Collect errors while parsing. If processing can't continue, an |
| 108 | * ExtensionError should be thrown instead. |
| 109 | * |
| 110 | * $key and $params are basically the arguments to wfMessage, although they |
| 111 | * will be stored in the data-mw of the encapsulation wrapper. |
| 112 | * |
| 113 | * See https://www.mediawiki.org/wiki/Specs/HTML#Error_handling |
| 114 | * |
| 115 | * The returned fragment can be inserted in the dom and will be populated |
| 116 | * with the localized message. See T266666 |
| 117 | * |
| 118 | * @unstable |
| 119 | * @param string $key |
| 120 | * @param mixed ...$params |
| 121 | * @return DocumentFragment |
| 122 | */ |
| 123 | public function pushError( string $key, ...$params ): DocumentFragment { |
| 124 | $this->errors[] = new DataMwError( $key, $params ); |
| 125 | return WTUtils::createInterfaceI18nFragment( $this->getTopLevelDoc(), $key, $params ); |
| 126 | } |
| 127 | |
| 128 | /** |
| 129 | * Creates an internationalization (i18n) message that will be localized into the user |
| 130 | * interface language. The returned DocumentFragment contains, as a single child, a span |
| 131 | * element with the appropriate information for later localization. |
| 132 | * @param string $key message key for the message to be localized |
| 133 | * @param ?array $params parameters for localization |
| 134 | * @return DocumentFragment |
| 135 | */ |
| 136 | public function createInterfaceI18nFragment( string $key, ?array $params ): DocumentFragment { |
| 137 | return WTUtils::createInterfaceI18nFragment( $this->getTopLevelDoc(), $key, $params ); |
| 138 | } |
| 139 | |
| 140 | /** |
| 141 | * Creates an internationalization (i18n) message that will be localized into the page content |
| 142 | * language. The returned DocumentFragment contains, as a single child, a span |
| 143 | * element with the appropriate information for later localization. |
| 144 | * @param string $key message key for the message to be localized |
| 145 | * @param ?array $params parameters for localization |
| 146 | * @return DocumentFragment |
| 147 | */ |
| 148 | public function createPageContentI18nFragment( string $key, ?array $params ): DocumentFragment { |
| 149 | return WTUtils::createPageContentI18nFragment( $this->getTopLevelDoc(), $key, $params ); |
| 150 | } |
| 151 | |
| 152 | /** |
| 153 | * Creates an internationalization (i18n) message that will be localized into an arbitrary |
| 154 | * language. The returned DocumentFragment contains, as a single child, a span |
| 155 | * element with the appropriate information for later localization. |
| 156 | * The use of this method is discouraged; use ::createPageContentI18nFragment(...) and |
| 157 | * ::createInterfaceI18nFragment(...) where possible rather than, respectively, |
| 158 | * ::createLangI18nFragment($wgContLang, ...) and ::createLangI18nFragment($wgLang, ...). |
| 159 | * @param Bcp47Code $lang language in which the message will be localized |
| 160 | * @param string $key message key for the message to be localized |
| 161 | * @param ?array $params parameters for localization |
| 162 | * @return DocumentFragment |
| 163 | */ |
| 164 | public function createLangI18nFragment( Bcp47Code $lang, string $key, ?array $params ): DocumentFragment { |
| 165 | return WTUtils::createLangI18nFragment( $this->getTopLevelDoc(), $lang, $key, $params ); |
| 166 | } |
| 167 | |
| 168 | /** |
| 169 | * Adds to $element the internationalization information needed for the attribute $name to be |
| 170 | * localized in a later pass into the user interface language. |
| 171 | * @param Element $element element on which to add internationalization information |
| 172 | * @param string $name name of the attribute whose value will be localized |
| 173 | * @param string $key message key used for the attribute value localization |
| 174 | * @param ?array $params parameters for localization |
| 175 | */ |
| 176 | public function addInterfaceI18nAttribute( |
| 177 | Element $element, string $name, string $key, ?array $params |
| 178 | ): void { |
| 179 | WTUtils::addInterfaceI18nAttribute( $element, $name, $key, $params ); |
| 180 | } |
| 181 | |
| 182 | /** |
| 183 | * Adds to $element the internationalization information needed for the attribute $name to be |
| 184 | * localized in a later pass into the page content language. |
| 185 | * @param Element $element element on which to add internationalization information |
| 186 | * @param string $name name of the attribute whose value will be localized |
| 187 | * @param string $key message key used for the attribute value localization |
| 188 | * @param array $params parameters for localization |
| 189 | */ |
| 190 | public function addPageContentI18nAttribute( |
| 191 | Element $element, string $name, string $key, array $params |
| 192 | ): void { |
| 193 | WTUtils::addPageContentI18nAttribute( $element, $name, $key, $params ); |
| 194 | } |
| 195 | |
| 196 | /** |
| 197 | * Adds to $element the internationalization information needed for the attribute $name to be |
| 198 | * localized in a later pass into the provided language. |
| 199 | * The use of this method is discouraged; use ::addPageContentI18nAttribute(...) and |
| 200 | * ::addInterfaceI18nAttribute(...) where possible rather than, respectively, |
| 201 | * ::addLangI18nAttribute(..., $wgContLang, ...) and ::addLangI18nAttribute(..., $wgLang, ...). |
| 202 | * @param Element $element element on which to add internationalization information |
| 203 | * @param Bcp47Code $lang language in which the attribute will be localized |
| 204 | * @param string $name name of the attribute whose value will be localized |
| 205 | * @param string $key message key used for the attribute value localization |
| 206 | * @param array $params parameters for localization |
| 207 | */ |
| 208 | public function addLangI18nAttribute( |
| 209 | Element $element, Bcp47Code $lang, string $name, string $key, array $params |
| 210 | ) { |
| 211 | WTUtils::addLangI18nAttribute( $element, $lang, $name, $key, $params ); |
| 212 | } |
| 213 | |
| 214 | /** |
| 215 | * @return list<DataMwError> |
| 216 | */ |
| 217 | public function getErrors(): array { |
| 218 | return $this->errors; |
| 219 | } |
| 220 | |
| 221 | /** |
| 222 | * Returns the main document we're parsing. Extension content is parsed |
| 223 | * to fragments of this document. |
| 224 | * |
| 225 | * @return Document |
| 226 | */ |
| 227 | public function getTopLevelDoc(): Document { |
| 228 | return $this->env->getTopLevelDoc(); |
| 229 | } |
| 230 | |
| 231 | /** |
| 232 | * Get a new about id for marking extension output |
| 233 | * FIXME: This should never really be needed since the extension API |
| 234 | * handles this on behalf of extensions, but Cite has one use case |
| 235 | * where implicit <references /> output is added. |
| 236 | * |
| 237 | * @return string |
| 238 | */ |
| 239 | public function newAboutId(): string { |
| 240 | return $this->env->newAboutId(); |
| 241 | } |
| 242 | |
| 243 | /** |
| 244 | * Get the site configuration to let extensions customize |
| 245 | * their behavior based on how the wiki is configured. |
| 246 | * |
| 247 | * @return SiteConfig |
| 248 | */ |
| 249 | public function getSiteConfig(): SiteConfig { |
| 250 | return $this->env->getSiteConfig(); |
| 251 | } |
| 252 | |
| 253 | /** |
| 254 | * FIXME: Unsure if we need to provide this access yet |
| 255 | * Get the page configuration |
| 256 | * @return PageConfig |
| 257 | */ |
| 258 | public function getPageConfig(): PageConfig { |
| 259 | return $this->env->getPageConfig(); |
| 260 | } |
| 261 | |
| 262 | /** |
| 263 | * Get the ContentMetadataCollector corresponding to the top-level page. |
| 264 | * In Parsoid integrated mode this will typically be an instance of |
| 265 | * core's `ParserOutput` class. |
| 266 | * |
| 267 | * @return ContentMetadataCollector |
| 268 | */ |
| 269 | public function getMetadata(): ContentMetadataCollector { |
| 270 | return $this->env->getMetadata(); |
| 271 | } |
| 272 | |
| 273 | /** |
| 274 | * Get the URI to link to a title |
| 275 | * @param Title $title |
| 276 | * @return string |
| 277 | */ |
| 278 | public function getTitleUri( Title $title ): string { |
| 279 | return $this->env->makeLink( $title ); |
| 280 | } |
| 281 | |
| 282 | /** |
| 283 | * Get an URI for the current page |
| 284 | * @return string |
| 285 | */ |
| 286 | public function getPageUri(): string { |
| 287 | return $this->getTitleUri( $this->env->getContextTitle() ); |
| 288 | } |
| 289 | |
| 290 | /** |
| 291 | * Make a title from an input string |
| 292 | * @param string $str |
| 293 | * @param int $namespaceId |
| 294 | * @return ?Title |
| 295 | */ |
| 296 | public function makeTitle( string $str, int $namespaceId ): ?Title { |
| 297 | return $this->env->makeTitleFromText( $str, $namespaceId, true /* no exceptions */ ); |
| 298 | } |
| 299 | |
| 300 | /** |
| 301 | * Are we parsing in a template context? |
| 302 | * @return bool |
| 303 | */ |
| 304 | public function inTemplate(): bool { |
| 305 | return $this->wt2htmlOpts['parseOpts']['inTemplate'] ?? false; |
| 306 | } |
| 307 | |
| 308 | /** |
| 309 | * Are we parsing for a preview? |
| 310 | * FIXME: Right now, we never do; when we do, this needs to be modified to reflect reality |
| 311 | * @unstable |
| 312 | * @return bool |
| 313 | */ |
| 314 | public function isPreview(): bool { |
| 315 | return false; |
| 316 | } |
| 317 | |
| 318 | /** |
| 319 | * FIXME: Is this something that can come from the frame? |
| 320 | * If we are parsing in the context of a parent extension tag, |
| 321 | * return the name of that extension tag |
| 322 | * @return string|null |
| 323 | */ |
| 324 | public function parentExtTag(): ?string { |
| 325 | return $this->wt2htmlOpts['parseOpts']['extTag'] ?? null; |
| 326 | } |
| 327 | |
| 328 | /** |
| 329 | * FIXME: Is this something that can come from the frame? |
| 330 | * If we are parsing in the context of a parent extension tag, |
| 331 | * return the parsing options set by that tag. |
| 332 | * @return array |
| 333 | */ |
| 334 | public function parentExtTagOpts(): array { |
| 335 | return $this->wt2htmlOpts['parseOpts']['extTagOpts'] ?? []; |
| 336 | } |
| 337 | |
| 338 | /** |
| 339 | * Get the content DOM corresponding to an id |
| 340 | * @param string $contentId |
| 341 | * @return DocumentFragment |
| 342 | */ |
| 343 | public function getContentDOM( string $contentId ): DocumentFragment { |
| 344 | return $this->env->getDOMFragment( $contentId ); |
| 345 | } |
| 346 | |
| 347 | public function clearContentDOM( string $contentId ): void { |
| 348 | $this->env->removeDOMFragment( $contentId ); |
| 349 | } |
| 350 | |
| 351 | /** |
| 352 | * Parse wikitext to DOM |
| 353 | * |
| 354 | * @param string|PFragment $wikitextOrPFragment |
| 355 | * @param array $opts |
| 356 | * - srcOffsets |
| 357 | * - processInNewFrame |
| 358 | * - clearDSROffsets |
| 359 | * - shiftDSRFn |
| 360 | * - parseOpts |
| 361 | * - extTag |
| 362 | * - extTagOpts |
| 363 | * - context "inline", "block", etc. Currently, only "inline" is supported |
| 364 | * @param bool $sol Whether tokens should be processed in start-of-line context. |
| 365 | * @return DocumentFragment |
| 366 | */ |
| 367 | public function wikitextToDOM( |
| 368 | $wikitextOrPFragment, array $opts, bool $sol |
| 369 | ): DocumentFragment { |
| 370 | if ( is_string( $wikitextOrPFragment ) ) { |
| 371 | $srcOffsets = $opts['srcOffsets'] ?? null; |
| 372 | if ( $srcOffsets !== null && !$srcOffsets instanceof DomSourceRange ) { |
| 373 | $srcOffsets = DomSourceRange::fromTsr( $srcOffsets ); |
| 374 | } |
| 375 | $pFragment = WikitextPFragment::newFromWt( |
| 376 | $wikitextOrPFragment, $srcOffsets |
| 377 | ); |
| 378 | } else { |
| 379 | $pFragment = $wikitextOrPFragment; |
| 380 | } |
| 381 | if ( $pFragment->isEmpty() ) { |
| 382 | $domFragment = $this->getTopLevelDoc()->createDocumentFragment(); |
| 383 | } else { |
| 384 | // Parse content to DOM and pass DOM-fragment token back to the main pipeline. |
| 385 | // The DOM will get unwrapped and integrated when processing the top level document. |
| 386 | [ |
| 387 | 'frame' => $frame, |
| 388 | 'wikitext' => $wikitext, |
| 389 | 'srcOffsets' => $srcOffsets, |
| 390 | ] = PipelineUtils::preparePFragment( |
| 391 | $this->env, $this->frame, $pFragment, $opts |
| 392 | ); |
| 393 | $parseOpts = $opts['parseOpts'] ?? []; |
| 394 | $domFragment = PipelineUtils::processContentInPipeline( |
| 395 | $this->env, $frame, $wikitext, |
| 396 | [ |
| 397 | // Full pipeline for processing content |
| 398 | 'pipelineType' => 'wikitext-to-fragment', |
| 399 | 'pipelineOpts' => [ |
| 400 | 'expandTemplates' => true, |
| 401 | 'extTag' => $parseOpts['extTag'] ?? null, |
| 402 | 'extTagOpts' => $parseOpts['extTagOpts'] ?? null, |
| 403 | 'inTemplate' => $this->inTemplate(), |
| 404 | 'inlineContext' => ( $parseOpts['context'] ?? '' ) === 'inline', |
| 405 | ], |
| 406 | 'srcOffsets' => $srcOffsets, |
| 407 | 'sol' => $sol, |
| 408 | ] |
| 409 | ); |
| 410 | |
| 411 | if ( !empty( $opts['clearDSROffsets'] ) ) { |
| 412 | $dsrFn = static function ( DomSourceRange $dsr ) { |
| 413 | return null; |
| 414 | }; |
| 415 | } else { |
| 416 | $dsrFn = $opts['shiftDSRFn'] ?? null; |
| 417 | } |
| 418 | |
| 419 | if ( $dsrFn ) { |
| 420 | ContentUtils::shiftDSR( $this->env, $domFragment, $dsrFn, $this ); |
| 421 | } |
| 422 | } |
| 423 | return $domFragment; |
| 424 | } |
| 425 | |
| 426 | /** |
| 427 | * Parse extension tag to DOM. |
| 428 | * |
| 429 | * If a wrapper tag is requested, beyond parsing the contents of the extension tag, |
| 430 | * this method wraps the contents in a custom wrapper element (ex: <div>), sanitizes |
| 431 | * the arguments of the extension args and sets some content flags on the wrapper. |
| 432 | * |
| 433 | * @param array $extArgs Args sanitized and applied to wrapper |
| 434 | * @param string $wikitext Wikitext content of the tag |
| 435 | * @param array $opts |
| 436 | * - srcOffsets |
| 437 | * - wrapperTag (skip OR pass null to not add any wrapper tag) |
| 438 | * - processInNewFrame |
| 439 | * - clearDSROffsets |
| 440 | * - shiftDSRFn |
| 441 | * - parseOpts |
| 442 | * - extTag |
| 443 | * - extTagOpts |
| 444 | * - context |
| 445 | * @return DocumentFragment "prepared and loaded" |
| 446 | */ |
| 447 | public function extTagToDOM( |
| 448 | array $extArgs, string $wikitext, array $opts |
| 449 | ): DocumentFragment { |
| 450 | $extTagOffsets = $this->extTag->getOffsets(); |
| 451 | if ( !isset( $opts['srcOffsets'] ) ) { |
| 452 | $opts['srcOffsets'] = new SourceRange( |
| 453 | $extTagOffsets->innerStart(), |
| 454 | $extTagOffsets->innerEnd() |
| 455 | ); |
| 456 | } |
| 457 | |
| 458 | $domFragment = $this->wikitextToDOM( $wikitext, $opts, true /* sol */ ); |
| 459 | if ( !empty( $opts['wrapperTag'] ) ) { |
| 460 | // Create a wrapper and migrate content into the wrapper |
| 461 | $wrapper = $domFragment->ownerDocument->createElement( |
| 462 | $opts['wrapperTag'] |
| 463 | ); |
| 464 | DOMUtils::migrateChildren( $domFragment, $wrapper ); |
| 465 | $domFragment->appendChild( $wrapper ); |
| 466 | |
| 467 | // Sanitize args and set on the wrapper |
| 468 | Sanitizer::applySanitizedArgs( $this->env->getSiteConfig(), $wrapper, $extArgs ); |
| 469 | |
| 470 | // Mark empty content DOMs |
| 471 | if ( $wikitext === '' ) { |
| 472 | DOMDataUtils::getDataParsoid( $wrapper )->empty = true; |
| 473 | } |
| 474 | |
| 475 | if ( !empty( $this->extTag->isSelfClosed() ) ) { |
| 476 | DOMDataUtils::getDataParsoid( $wrapper )->selfClose = true; |
| 477 | } |
| 478 | } |
| 479 | |
| 480 | return $domFragment; |
| 481 | } |
| 482 | |
| 483 | /** |
| 484 | * Set temporary data into the DOM node that will be discarded |
| 485 | * when DOM is serialized |
| 486 | * |
| 487 | * Use the tag name as the key for TempData management |
| 488 | * |
| 489 | * @param Element $node |
| 490 | * @param mixed $data |
| 491 | */ |
| 492 | public function setTempNodeData( Element $node, $data ): void { |
| 493 | $dataParsoid = DOMDataUtils::getDataParsoid( $node ); |
| 494 | $tmpData = $dataParsoid->getTemp(); |
| 495 | $key = $this->extTag->getName(); |
| 496 | $tmpData->setTagData( $key, $data ); |
| 497 | } |
| 498 | |
| 499 | /** |
| 500 | * Get temporary data into the DOM node that will be discarded |
| 501 | * when DOM is serialized. |
| 502 | * |
| 503 | * This should only be used when the ExtensionTag is not available; otherwise access the newly created data |
| 504 | * directly. |
| 505 | * |
| 506 | * @param Element $node |
| 507 | * @param string $key to access TmpData |
| 508 | * @return mixed |
| 509 | * @unstable |
| 510 | */ |
| 511 | public function getTempNodeData( Element $node, string $key ) { |
| 512 | if ( $this->extTag ) { |
| 513 | throw new \RuntimeException( |
| 514 | 'ExtensionTag is available. Data should be available directly through the DOM.' |
| 515 | ); |
| 516 | } |
| 517 | $dataParsoid = DOMDataUtils::getDataParsoid( $node ); |
| 518 | $tmpData = $dataParsoid->getTemp(); |
| 519 | return $tmpData->getTagData( $key ); |
| 520 | } |
| 521 | |
| 522 | /** |
| 523 | * Process a specific extension arg as wikitext and return its DOM equivalent. |
| 524 | * By default, this method processes the argument value in inline context and normalizes |
| 525 | * every whitespace character to a single space. |
| 526 | * @param KV[] $extArgs |
| 527 | * @param string $key should be lower-case |
| 528 | * @param string $context |
| 529 | * @return ?DocumentFragment |
| 530 | */ |
| 531 | public function extArgToDOM( |
| 532 | array $extArgs, string $key, string $context = "inline" |
| 533 | ): ?DocumentFragment { |
| 534 | $argKV = KV::lookupKV( $extArgs, strtolower( $key ) ); |
| 535 | if ( $argKV === null || !$argKV->v ) { |
| 536 | return null; |
| 537 | } |
| 538 | |
| 539 | if ( $context === "inline" ) { |
| 540 | // `normalizeExtOptions` can mess up source offsets as well as the string |
| 541 | // that ought to be processed as wikitext. So, we do our own whitespace |
| 542 | // normalization of the original source here. |
| 543 | // |
| 544 | // If 'context' is 'inline' below, it ensures indent-pre / p-wrapping is suppressed. |
| 545 | // So, the normalization is primarily for HTML string parity. |
| 546 | $argVal = preg_replace( '/[\t\r\n ]/', ' ', $argKV->vsrc ); |
| 547 | } else { |
| 548 | $argVal = $argKV->vsrc; |
| 549 | } |
| 550 | |
| 551 | return $this->wikitextToDOM( |
| 552 | $argVal, |
| 553 | [ |
| 554 | 'parseOpts' => [ |
| 555 | 'extTag' => $this->extTag->getName(), |
| 556 | 'context' => $context, |
| 557 | ], |
| 558 | 'srcOffsets' => $argKV->valueOffset(), |
| 559 | ], |
| 560 | false // inline context => no sol state |
| 561 | ); |
| 562 | } |
| 563 | |
| 564 | /** |
| 565 | * Convert the ext args representation from an array of KV objects |
| 566 | * to a plain associative array mapping arg name strings to arg value strings. |
| 567 | * @param array<KV> $extArgs |
| 568 | * @return array<string,string> |
| 569 | */ |
| 570 | public function extArgsToArray( array $extArgs ): array { |
| 571 | return TokenUtils::kvToHash( $extArgs ); |
| 572 | } |
| 573 | |
| 574 | /** |
| 575 | * This method finds a requested arg by key name and return its current value. |
| 576 | * If a closure is passed in to update the current value, it is used to update the arg. |
| 577 | * |
| 578 | * @param KV[] &$extArgs Array of extension args |
| 579 | * @param string $key Argument key whose value needs an update |
| 580 | * @param ?Closure $updater $updater will get the existing string value |
| 581 | * for the arg and is expected to return an updated value. |
| 582 | * @return ?string |
| 583 | */ |
| 584 | public function findAndUpdateArg( |
| 585 | array &$extArgs, string $key, ?Closure $updater = null |
| 586 | ): ?string { |
| 587 | // FIXME: This code will get an overhaul when T250854 is resolved. |
| 588 | foreach ( $extArgs as $i => $kv ) { |
| 589 | $k = TokenUtils::tokensToString( $kv->k ); |
| 590 | if ( strtolower( trim( $k ) ) === strtolower( $key ) ) { |
| 591 | $val = $kv->v; |
| 592 | if ( $updater ) { |
| 593 | $kv = clone $kv; |
| 594 | $kv->v = $updater( TokenUtils::tokensToString( $val ) ); |
| 595 | $extArgs[$i] = $kv; |
| 596 | } |
| 597 | return $val; |
| 598 | } |
| 599 | } |
| 600 | |
| 601 | return null; |
| 602 | } |
| 603 | |
| 604 | /** |
| 605 | * Updates all arguments, similar to findAndUpdateArg, but applies the closure to all arguments |
| 606 | * @param KV[] &$extArgs Array of extension args |
| 607 | * @param ?Closure $updater $updater will get the existing string value |
| 608 | * for the key and value of the arg and is expected to return an updated value. |
| 609 | */ |
| 610 | public function updateAllArgs( array &$extArgs, ?Closure $updater = null ): void { |
| 611 | if ( !$updater ) { |
| 612 | return; |
| 613 | } |
| 614 | foreach ( $extArgs as $i => $kv ) { |
| 615 | $k = TokenUtils::tokensToString( $kv->k ); |
| 616 | $val = $kv->v; |
| 617 | $kv = clone $kv; |
| 618 | $kv->v = $updater( $k, TokenUtils::tokensToString( $val ) ); |
| 619 | $extArgs[$i] = $kv; |
| 620 | } |
| 621 | } |
| 622 | |
| 623 | /** |
| 624 | * Normalizes spaces from extension tag arguments, except for those keyed by values in $exceptions |
| 625 | * @param KV[] &$extArgs Array of extension args |
| 626 | * @param array[] $action array that is either empty or has one key, 'except' or 'only', which defines the |
| 627 | * attributes that should be respectively excluded or only included from the normalization |
| 628 | */ |
| 629 | public function normalizeWhiteSpaceInArgs( array &$extArgs, array $action = [] ) { |
| 630 | $except = $action['except'] ?? null; |
| 631 | $only = $action['only'] ?? null; |
| 632 | |
| 633 | if ( $except && $only ) { |
| 634 | $this->log( 'warn', 'normalizeWhiteSpaceInArgs should not have both except and only parameters' ); |
| 635 | return; |
| 636 | } |
| 637 | |
| 638 | if ( $except ) { |
| 639 | $closure = static function ( $key, $value ) use ( $except ) { |
| 640 | if ( in_array( strtolower( trim( $key ) ), $except, true ) ) { |
| 641 | return $value; |
| 642 | } else { |
| 643 | return trim( preg_replace( '/[\r\n\t ]+/', ' ', $value ) ); |
| 644 | } |
| 645 | }; |
| 646 | } elseif ( $only ) { |
| 647 | $closure = static function ( $key, $value ) use ( $only ) { |
| 648 | if ( in_array( strtolower( trim( $key ) ), $only, true ) ) { |
| 649 | return trim( preg_replace( '/[\r\n\t ]+/', ' ', $value ) ); |
| 650 | } else { |
| 651 | return $value; |
| 652 | } |
| 653 | }; |
| 654 | } else { |
| 655 | $closure = static function ( $key, $value ) { |
| 656 | return trim( preg_replace( '/[\r\n\t ]+/', ' ', $value ) ); |
| 657 | }; |
| 658 | } |
| 659 | |
| 660 | $this->updateAllArgs( $extArgs, $closure ); |
| 661 | } |
| 662 | |
| 663 | /** |
| 664 | * This method adds a new argument to the extension args array |
| 665 | * @param KV[] &$extArgs |
| 666 | * @param string $key |
| 667 | * @param string $value |
| 668 | */ |
| 669 | public function addNewArg( array &$extArgs, string $key, string $value ): void { |
| 670 | $extArgs[] = new KV( $key, $value ); |
| 671 | } |
| 672 | |
| 673 | // TODO: Provide support for extensions to register lints |
| 674 | // from their customized lint handlers. |
| 675 | |
| 676 | /** |
| 677 | * Forwards the logging request to the underlying logger |
| 678 | * @param string $prefix |
| 679 | * @param mixed ...$args |
| 680 | */ |
| 681 | public function log( string $prefix, ...$args ): void { |
| 682 | $this->env->log( $prefix, ...$args ); |
| 683 | } |
| 684 | |
| 685 | /** |
| 686 | * Extensions might be interested in examining (their) content embedded |
| 687 | * in attributes that don't otherwise show up in the DOM. |
| 688 | * |
| 689 | * Ex: inline media captions that aren't rendered, language variant markup, |
| 690 | * attributes that are transcluded. More scenarios might be added later. |
| 691 | * @deprecated |
| 692 | * Don't use this directly: use ::processAttributeEmbeddedDom(). |
| 693 | * This method may omit content which is embedded natively as |
| 694 | * DocumentFragments instead of as HTML strings. |
| 695 | * |
| 696 | * @param Element $elt The node whose data attributes need to be examined |
| 697 | * @param Closure $proc The processor that will process the embedded HTML |
| 698 | * Signature: (string) -> string |
| 699 | * This processor will be provided the HTML string as input |
| 700 | * and is expected to return a possibly modified string. |
| 701 | */ |
| 702 | public function processAttributeEmbeddedHTML( Element $elt, Closure $proc ): void { |
| 703 | // @phan-suppress-next-line PhanDeprecatedFunction |
| 704 | ContentUtils::processAttributeEmbeddedHTML( $this, $elt, $proc ); |
| 705 | } |
| 706 | |
| 707 | /** |
| 708 | * Extensions might be interested in examining (their) content embedded |
| 709 | * in attributes that don't otherwise show up in the DOM. |
| 710 | * |
| 711 | * Ex: inline media captions that aren't rendered, language variant markup, |
| 712 | * attributes that are transcluded. More scenarios might be added later. |
| 713 | * |
| 714 | * @param Element $elt The node whose data attributes need to be examined |
| 715 | * @param callable(DocumentFragment):bool $proc |
| 716 | * The processor that will process the embedded HTML. |
| 717 | * This processor will be provided a DocumentFragment |
| 718 | * and is expected to return true if that fragment was modified. |
| 719 | */ |
| 720 | public function processAttributeEmbeddedDom( Element $elt, callable $proc ): void { |
| 721 | ContentUtils::processAttributeEmbeddedDom( $this, $elt, $proc ); |
| 722 | } |
| 723 | |
| 724 | /** |
| 725 | * Copy $from->childNodes to $to and clone the data attributes of $from |
| 726 | * to $to. |
| 727 | * |
| 728 | * @param Element $from |
| 729 | * @param Element $to |
| 730 | */ |
| 731 | public static function migrateChildrenAndTransferWrapperDataAttribs( |
| 732 | Element $from, Element $to |
| 733 | ): void { |
| 734 | DOMUtils::migrateChildren( $from, $to ); |
| 735 | DOMDataUtils::setDataParsoid( |
| 736 | $to, clone DOMDataUtils::getDataParsoid( $from ) |
| 737 | ); |
| 738 | DOMDataUtils::setDataMw( |
| 739 | $to, clone DOMDataUtils::getDataMw( $from ) |
| 740 | ); |
| 741 | } |
| 742 | |
| 743 | /** |
| 744 | * Equivalent of 'preprocess' from Parser.php in core. |
| 745 | * - expands templates |
| 746 | * - replaces magic variables |
| 747 | * This does not run any hooks however since that would be unexpected. |
| 748 | * This also doesn't support replacing template args from a frame. |
| 749 | * |
| 750 | * @param string $wikitext |
| 751 | * @return array{error:bool,src?:string,fragment?:PFragment} |
| 752 | * - 'error' did we hit resource limits? |
| 753 | * - 'src' expanded wikitext OR error message to print |
| 754 | * FIXME: Maybe error message should be localizable |
| 755 | * - 'fragment' Optional fragment (wikitext plus strip state) |
| 756 | */ |
| 757 | public function preprocessWikitext( string $wikitext ) { |
| 758 | return Wikitext::preprocess( $this->env, $wikitext ); |
| 759 | } |
| 760 | |
| 761 | /** |
| 762 | * Parse input string into DOM. |
| 763 | * NOTE: This leaves the DOM in Parsoid-canonical state and is the preferred method |
| 764 | * to convert HTML to DOM that will be passed into Parsoid's processing code. |
| 765 | * |
| 766 | * @param string $html |
| 767 | * @param ?Document $doc XXX You probably don't want to be doing this |
| 768 | * @param ?array $options |
| 769 | * @return DocumentFragment |
| 770 | */ |
| 771 | public function htmlToDom( |
| 772 | string $html, ?Document $doc = null, ?array $options = [] |
| 773 | ): DocumentFragment { |
| 774 | return ContentUtils::createAndLoadDocumentFragment( |
| 775 | $doc ?? $this->getTopLevelDoc(), $html, $options |
| 776 | ); |
| 777 | } |
| 778 | |
| 779 | /** |
| 780 | * Serialize DOM element to string (inner/outer HTML is controlled by flag). |
| 781 | * If $releaseDom is set to true, the DOM will not be "prepared and loaded" |
| 782 | * and is not safe to use after this call. This is primarily a performance optimization. |
| 783 | * |
| 784 | * @param Node $node |
| 785 | * @param bool $innerHTML if true, inner HTML of the element will be returned |
| 786 | * This flag defaults to false |
| 787 | * @param bool $releaseDom if true, the DOM will not be in canonical form after this call |
| 788 | * This flag defaults to false |
| 789 | * @return string |
| 790 | */ |
| 791 | public function domToHtml( |
| 792 | Node $node, bool $innerHTML = false, bool $releaseDom = false |
| 793 | ): string { |
| 794 | // FIXME: This is going to drop any diff markers but since |
| 795 | // the dom differ doesn't traverse into extension content (right now), |
| 796 | // none should exist anyways. |
| 797 | // FIXME: This roundtrip discards data-parsoid.tmp data from $node. |
| 798 | // Maybe it meant to set keepTmp, but that flag is currently broken. |
| 799 | $html = ContentUtils::ppToXML( $node, [ 'innerXML' => $innerHTML, 'fragment' => true ] ); |
| 800 | if ( !$releaseDom ) { |
| 801 | DOMDataUtils::visitAndLoadDataAttribs( $node ); |
| 802 | } |
| 803 | return $html; |
| 804 | } |
| 805 | |
| 806 | /** |
| 807 | * Bit flags describing escaping / serializing context in html -> wt mode |
| 808 | */ |
| 809 | public const IN_SOL = 1; |
| 810 | public const IN_MEDIA = 2; |
| 811 | public const IN_LINK = 4; |
| 812 | public const IN_IMG_CAPTION = 8; |
| 813 | public const IN_OPTION = 16; |
| 814 | |
| 815 | /** |
| 816 | * FIXME: This is a bit broken - shouldn't be needed ideally |
| 817 | * @param string $flag |
| 818 | */ |
| 819 | public function setHtml2wtStateFlag( string $flag ) { |
| 820 | $this->serializerState->{$flag} = true; |
| 821 | } |
| 822 | |
| 823 | /** |
| 824 | * Emit the opening tag (including attributes) for the extension |
| 825 | * represented by this node. |
| 826 | * |
| 827 | * @param Element $node |
| 828 | * @return string |
| 829 | */ |
| 830 | public function extStartTagToWikitext( Element $node ): string { |
| 831 | $state = $this->serializerState; |
| 832 | return $state->serializer->serializeExtensionStartTag( $node, $state ); |
| 833 | } |
| 834 | |
| 835 | /** |
| 836 | * Convert the input DOM to wikitext. |
| 837 | * |
| 838 | * @param array $opts |
| 839 | * - extName: (string) Name of the extension whose body we are serializing |
| 840 | * - inPHPBlock: (bool) FIXME: This needs to be removed |
| 841 | * @param Element $node DOM to serialize |
| 842 | * @param bool $releaseDom If $releaseDom is set to true, the DOM will be left in |
| 843 | * non-canonical form and is not safe to use after this call. This is primarily a |
| 844 | * performance optimization. This flag defaults to false. |
| 845 | * @return mixed |
| 846 | */ |
| 847 | public function domToWikitext( array $opts, Element $node, bool $releaseDom = false ) { |
| 848 | // FIXME: WTS expects the input DOM to be a <body> element! |
| 849 | // Till that is fixed, we have to go through this round-trip! |
| 850 | // TODO: Move $node children to a fragment and call `$serializer->domToWikitext` |
| 851 | return $this->htmlToWikitext( $opts, $this->domToHtml( $node, $releaseDom ) ); |
| 852 | } |
| 853 | |
| 854 | /** |
| 855 | * Convert the HTML body of an extension to wikitext |
| 856 | * |
| 857 | * @param array $opts |
| 858 | * - extName: (string) Name of the extension whose body we are serializing |
| 859 | * - inPHPBlock: (bool) FIXME: This needs to be removed |
| 860 | * @param string $html HTML for the extension's body |
| 861 | * @return string |
| 862 | */ |
| 863 | public function htmlToWikitext( array $opts, string $html ): string { |
| 864 | // Type cast so phan has more information to ensure type safety |
| 865 | $state = $this->serializerState; |
| 866 | $opts['env'] = $this->env; |
| 867 | return $state->serializer->htmlToWikitext( $opts, $html ); |
| 868 | } |
| 869 | |
| 870 | /** |
| 871 | * Get the original source for an element. |
| 872 | * |
| 873 | * The callable, $checkIfOrigSrcReusable, is used to determine if the $elt |
| 874 | * is unedited and therefore valid to reuse source. This is assumed to be |
| 875 | * pretty specific to the callsite so no default is provided. |
| 876 | * |
| 877 | * @param Element $elt |
| 878 | * @param bool $inner |
| 879 | * @param callable $checkIfOrigSrcReusable |
| 880 | * @return string|null |
| 881 | */ |
| 882 | public function getOrigSrc( |
| 883 | Element $elt, bool $inner, callable $checkIfOrigSrcReusable |
| 884 | ): ?string { |
| 885 | $state = $this->serializerState; |
| 886 | if ( !$state->selserMode || $state->inInsertedContent ) { |
| 887 | return null; |
| 888 | } |
| 889 | $dsr = DOMDataUtils::getDataParsoid( $elt )->dsr ?? null; |
| 890 | if ( !Utils::isValidDSR( $dsr, $inner ) ) { |
| 891 | return null; |
| 892 | } |
| 893 | if ( $checkIfOrigSrcReusable( $elt ) ) { |
| 894 | return $state->getOrigSrc( |
| 895 | $inner ? $dsr->innerRange() : $dsr |
| 896 | ); |
| 897 | } else { |
| 898 | return null; |
| 899 | } |
| 900 | } |
| 901 | |
| 902 | /** |
| 903 | * @param Element $elt |
| 904 | * @param int $context OR-ed bit flags specifying escaping / serialization context |
| 905 | * @return string |
| 906 | */ |
| 907 | public function domChildrenToWikitext( Element $elt, int $context ): string { |
| 908 | $state = $this->serializerState; |
| 909 | if ( $context & self::IN_IMG_CAPTION ) { |
| 910 | if ( $context & self::IN_OPTION ) { |
| 911 | $escapeHandler = 'mediaOptionHandler'; // Escapes "|" as well |
| 912 | } else { |
| 913 | $escapeHandler = 'wikilinkHandler'; // image captions show up in wikilink syntax |
| 914 | } |
| 915 | $out = $state->serializeCaptionChildrenToString( $elt, |
| 916 | [ $state->serializer->wteHandlers, $escapeHandler ] ); |
| 917 | } else { |
| 918 | throw new \RuntimeException( 'Not yet supported!' ); |
| 919 | } |
| 920 | return $out; |
| 921 | } |
| 922 | |
| 923 | /** |
| 924 | * Escape any wikitext like constructs in a string so that when the output |
| 925 | * is parsed, it renders as a string. The escaping is sensitive to the context |
| 926 | * in which the string is embedded. For example, a "*" is not safe at the start |
| 927 | * of a line (since it will parse as a list item), but is safe if it is not in |
| 928 | * a start of line context. Similarly the "|" character is safe outside tables, |
| 929 | * links, and transclusions. |
| 930 | * |
| 931 | * @param string $str |
| 932 | * @param Node $node |
| 933 | * @param int $context OR-ed bit flags specifying escaping / serialization context |
| 934 | * @return string |
| 935 | */ |
| 936 | public function escapeWikitext( string $str, Node $node, int $context ): string { |
| 937 | if ( $context & ( self::IN_MEDIA | self::IN_LINK ) ) { |
| 938 | $state = $this->serializerState; |
| 939 | return $state->serializer->wteHandlers->escapeLinkContent( |
| 940 | $state, $str, |
| 941 | (bool)( $context & self::IN_SOL ), |
| 942 | $node, |
| 943 | (bool)( $context & self::IN_MEDIA ) |
| 944 | ); |
| 945 | } else { |
| 946 | throw new \RuntimeException( 'Not yet supported!' ); |
| 947 | } |
| 948 | } |
| 949 | |
| 950 | /** |
| 951 | * EXTAPI-FIXME: We have to figure out what it means to run a DOM pass |
| 952 | * (and what processors and what handlers apply) on content models that are |
| 953 | * not wikitext. For now, we are only storing data attribs back to the DOM |
| 954 | * and adding metadata to the page. |
| 955 | * |
| 956 | * @param Document $doc |
| 957 | */ |
| 958 | public function postProcessDOM( Document $doc ): void { |
| 959 | // Ugh! But, this whole method needs to go away anyway |
| 960 | ( new AddMetaData( null ) )->run( $this->env, DOMCompat::getBody( $doc ) ); |
| 961 | } |
| 962 | |
| 963 | /** |
| 964 | * Produce the HTML rendering of a title string and media options as the |
| 965 | * wikitext parser would for a wikilink in the file namespace |
| 966 | * |
| 967 | * @param string $titleStr Image title string |
| 968 | * @param array $imageOpts Array of a mix of strings or arrays, |
| 969 | * the latter of which can signify that the value came from source. |
| 970 | * Where, |
| 971 | * [0] is the fully-constructed image option |
| 972 | * [1] is the full wikitext source offset for it |
| 973 | * @param ?string &$error Error string is set when the return is null. |
| 974 | * @param ?bool $forceBlock Forces the media to be rendered in a figure as |
| 975 | * opposed to a span. |
| 976 | * @param ?bool $suppressMediaFormats If any media format is present in |
| 977 | * $imageOpts, it won't be applied and will result in a linting error. |
| 978 | * @return ?Element |
| 979 | */ |
| 980 | public function renderMedia( |
| 981 | string $titleStr, array $imageOpts, ?string &$error = null, |
| 982 | ?bool $forceBlock = false, ?bool $suppressMediaFormats = false |
| 983 | ): ?Element { |
| 984 | $extTagName = $this->extTag->getName(); |
| 985 | $extTagOpts = [ 'suppressMediaFormats' => $suppressMediaFormats ]; |
| 986 | |
| 987 | $fileNs = $this->getSiteConfig()->canonicalNamespaceId( 'file' ); |
| 988 | |
| 989 | $title = $this->makeTitle( $titleStr, 0 ); |
| 990 | if ( $title === null || $title->getNamespace() !== $fileNs ) { |
| 991 | $error = "{$extTagName}_no_image"; |
| 992 | return null; |
| 993 | } |
| 994 | |
| 995 | $pieces = [ '[[' ]; |
| 996 | // Since the above two chars aren't from source, the resulting figure |
| 997 | // won't have any dsr info, so we can omit an offset for the title as |
| 998 | // well. In any case, $titleStr may not necessarily be from source, |
| 999 | // see the special case in the gallery extension. |
| 1000 | $pieces[] = $titleStr; |
| 1001 | $pieces = array_merge( $pieces, $imageOpts ); |
| 1002 | |
| 1003 | if ( $forceBlock ) { |
| 1004 | // We add "none" here so that this renders in the block form |
| 1005 | // (ie. figure). It's a valid media option, so shouldn't turn into |
| 1006 | // a caption. And since it's first wins, it shouldn't interfere |
| 1007 | // with another horizontal alignment defined in $imageOpts. |
| 1008 | // We just have to remember to strip the class below. |
| 1009 | // NOTE: This will have to be adjusted with T305628 |
| 1010 | $pieces[] = '|none'; |
| 1011 | } |
| 1012 | |
| 1013 | $pieces[] = ']]'; |
| 1014 | |
| 1015 | $shiftOffset = static function ( int $offset ) use ( $pieces ): ?int { |
| 1016 | foreach ( $pieces as $p ) { |
| 1017 | if ( is_string( $p ) ) { |
| 1018 | $offset -= strlen( $p ); |
| 1019 | if ( $offset <= 0 ) { |
| 1020 | return null; |
| 1021 | } |
| 1022 | } else { |
| 1023 | if ( $offset <= strlen( $p[0] ) && isset( $p[1] ) ) { |
| 1024 | return $p[1] + $offset; |
| 1025 | } |
| 1026 | $offset -= strlen( $p[0] ); |
| 1027 | if ( $offset <= 0 ) { |
| 1028 | return null; |
| 1029 | } |
| 1030 | } |
| 1031 | } |
| 1032 | return null; |
| 1033 | }; |
| 1034 | |
| 1035 | $imageWt = ''; |
| 1036 | foreach ( $pieces as $p ) { |
| 1037 | $imageWt .= ( is_string( $p ) ? $p : $p[0] ); |
| 1038 | } |
| 1039 | |
| 1040 | $domFragment = $this->wikitextToDOM( |
| 1041 | $imageWt, |
| 1042 | [ |
| 1043 | 'parseOpts' => [ |
| 1044 | 'extTag' => $extTagName, |
| 1045 | 'extTagOpts' => $extTagOpts, |
| 1046 | 'context' => 'inline', |
| 1047 | ], |
| 1048 | // Create new frame, because $pieces doesn't literally appear |
| 1049 | // on the page, it has been hand-crafted here |
| 1050 | 'processInNewFrame' => true, |
| 1051 | // Shift the DSRs in the DOM by startOffset, and strip DSRs |
| 1052 | // for bits which aren't the caption or file, since they |
| 1053 | // don't refer to actual source wikitext |
| 1054 | 'shiftDSRFn' => static function ( DomSourceRange $dsr ) use ( $shiftOffset ) { |
| 1055 | $start = $dsr->start === null ? null : $shiftOffset( $dsr->start ); |
| 1056 | $end = $dsr->end === null ? null : $shiftOffset( $dsr->end ); |
| 1057 | // If either offset is newly-invalid, remove entire DSR |
| 1058 | if ( |
| 1059 | ( $dsr->start !== null && $start === null ) || |
| 1060 | ( $dsr->end !== null && $end === null ) |
| 1061 | ) { |
| 1062 | return null; |
| 1063 | } |
| 1064 | return new DomSourceRange( |
| 1065 | $start, $end, $dsr->openWidth, $dsr->closeWidth |
| 1066 | ); |
| 1067 | }, |
| 1068 | ], |
| 1069 | true // sol |
| 1070 | ); |
| 1071 | |
| 1072 | $thumb = $domFragment->firstChild; |
| 1073 | $validWrappers = [ 'figure' ]; |
| 1074 | // Downstream code expects a figcaption if we're forcing a block so we |
| 1075 | // validate that we did indeed parse a figure. It might not have |
| 1076 | // happened because $imageOpts has an unbalanced `]]` which closes |
| 1077 | // the wikilink syntax before we get in our `|none`. |
| 1078 | if ( !$forceBlock ) { |
| 1079 | $validWrappers[] = 'span'; |
| 1080 | } |
| 1081 | if ( !in_array( DOMCompat::nodeName( $thumb ), $validWrappers, true ) ) { |
| 1082 | $error = "{$extTagName}_invalid_image"; |
| 1083 | return null; |
| 1084 | } |
| 1085 | DOMUtils::assertElt( $thumb ); |
| 1086 | |
| 1087 | // Detach the $thumb since the $domFragment is going out of scope |
| 1088 | // See https://bugs.php.net/bug.php?id=39593 |
| 1089 | DOMCompat::remove( $thumb ); |
| 1090 | |
| 1091 | if ( $forceBlock ) { |
| 1092 | $dp = DOMDataUtils::getDataParsoid( $thumb ); |
| 1093 | array_pop( $dp->optList ); |
| 1094 | $explicitNone = false; |
| 1095 | foreach ( $dp->optList as $opt ) { |
| 1096 | if ( $opt['ck'] === 'none' ) { |
| 1097 | $explicitNone = true; |
| 1098 | } |
| 1099 | } |
| 1100 | if ( !$explicitNone ) { |
| 1101 | // FIXME: Should we worry about someone adding this with the |
| 1102 | // "class=" option? |
| 1103 | DOMCompat::getClassList( $thumb )->remove( 'mw-halign-none' ); |
| 1104 | } |
| 1105 | } |
| 1106 | |
| 1107 | return $thumb; |
| 1108 | } |
| 1109 | |
| 1110 | /** |
| 1111 | * Serialize a MediaStructure to a title and media options string. |
| 1112 | * The converse to ::renderMedia. |
| 1113 | * |
| 1114 | * @param MediaStructure $ms |
| 1115 | * @return array Where, |
| 1116 | * [0] is the media title string |
| 1117 | * [1] is the string of media options |
| 1118 | */ |
| 1119 | public function serializeMedia( MediaStructure $ms ): array { |
| 1120 | $ct = LinkHandlerUtils::figureToConstrainedText( $this->serializerState, $ms ); |
| 1121 | if ( $ct instanceof WikiLinkText ) { |
| 1122 | // Remove the opening and closing square brackets |
| 1123 | $text = substr( $ct->text, 2, -2 ); |
| 1124 | return array_pad( explode( '|', $text, 2 ), 2, '' ); |
| 1125 | } else { |
| 1126 | // Note that $ct could be an AutoURLLinkText, not just null |
| 1127 | return [ '', '' ]; |
| 1128 | } |
| 1129 | } |
| 1130 | |
| 1131 | /** |
| 1132 | * @param array $modules |
| 1133 | * @deprecated Use ::getMetadata()->appendOutputStrings( MODULE, ...) instead. |
| 1134 | */ |
| 1135 | public function addModules( array $modules ) { |
| 1136 | $this->getMetadata()->appendOutputStrings( CMCSS::MODULE, $modules ); |
| 1137 | } |
| 1138 | |
| 1139 | /** |
| 1140 | * @param array $modulestyles |
| 1141 | * @deprecated Use ::getMetadata()->appendOutputStrings(MODULE_STYLE, ...) instead. |
| 1142 | */ |
| 1143 | public function addModuleStyles( array $modulestyles ) { |
| 1144 | $this->getMetadata()->appendOutputStrings( CMCSS::MODULE_STYLE, $modulestyles ); |
| 1145 | } |
| 1146 | |
| 1147 | /** |
| 1148 | * Get an array of attributes to apply to an anchor linking to $url |
| 1149 | */ |
| 1150 | public function getExternalLinkAttribs( string $url ): array { |
| 1151 | return $this->env->getExternalLinkAttribs( $url ); |
| 1152 | } |
| 1153 | |
| 1154 | /** |
| 1155 | * Add a tracking category to the current page. |
| 1156 | * @param string $key Message key (not localized) |
| 1157 | */ |
| 1158 | public function addTrackingCategory( string $key ): void { |
| 1159 | $this->env->getDataAccess()->addTrackingCategory( |
| 1160 | $this->env->getPageConfig(), |
| 1161 | $this->env->getMetadata(), |
| 1162 | $key |
| 1163 | ); |
| 1164 | } |
| 1165 | } |