includes/content/ContentHandler.php

   1 <?php
   2
   3 use MediaWiki\Search\ParserOutputSearchDataExtractor;
   4
   5 /**
   6  * Base class for content handling.
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @since 1.21
  24  *
  25  * @file
  26  * @ingroup Content
  27  *
  28  * @author Daniel Kinzler
  29  */
  30 /**
  31  * A content handler knows how do deal with a specific type of content on a wiki
  32  * page. Content is stored in the database in a serialized form (using a
  33  * serialization format a.k.a. MIME type) and is unserialized into its native
  34  * PHP representation (the content model), which is wrapped in an instance of
  35  * the appropriate subclass of Content.
  36  *
  37  * ContentHandler instances are stateless singletons that serve, among other
  38  * things, as a factory for Content objects. Generally, there is one subclass
  39  * of ContentHandler and one subclass of Content for every type of content model.
  40  *
  41  * Some content types have a flat model, that is, their native representation
  42  * is the same as their serialized form. Examples would be JavaScript and CSS
  43  * code. As of now, this also applies to wikitext (MediaWiki's default content
  44  * type), but wikitext content may be represented by a DOM or AST structure in
  45  * the future.
  46  *
  47  * @ingroup Content
  48  */
  49 abstract class ContentHandler {
  50         /**
  51          * Convenience function for getting flat text from a Content object. This
  52          * should only be used in the context of backwards compatibility with code
  53          * that is not yet able to handle Content objects!
  54          *
  55          * If $content is null, this method returns the empty string.
  56          *
  57          * If $content is an instance of TextContent, this method returns the flat
  58          * text as returned by $content->getNativeData().
  59          *
  60          * If $content is not a TextContent object, the behavior of this method
  61          * depends on the global $wgContentHandlerTextFallback:
  62          * - If $wgContentHandlerTextFallback is 'fail' and $content is not a
  63          *   TextContent object, an MWException is thrown.
  64          * - If $wgContentHandlerTextFallback is 'serialize' and $content is not a
  65          *   TextContent object, $content->serialize() is called to get a string
  66          *   form of the content.
  67          * - If $wgContentHandlerTextFallback is 'ignore' and $content is not a
  68          *   TextContent object, this method returns null.
  69          * - otherwise, the behavior is undefined.
  70          *
  71          * @since 1.21
  72          *
  73          * @param Content $content
  74          *
  75          * @throws MWException If the content is not an instance of TextContent and
  76          * wgContentHandlerTextFallback was set to 'fail'.
  77          * @return string|null Textual form of the content, if available.
  78          */
  79         public static function getContentText( Content $content = null ) {
  80                 global $wgContentHandlerTextFallback;
  81
  82                 if ( is_null( $content ) ) {
  83                         return '';
  84                 }
  85
  86                 if ( $content instanceof TextContent ) {
  87                         return $content->getNativeData();
  88                 }
  89
  90                 wfDebugLog( 'ContentHandler', 'Accessing ' . $content->getModel() . ' content as text!' );
  91
  92                 if ( $wgContentHandlerTextFallback == 'fail' ) {
  93                         throw new MWException(
  94                                 "Attempt to get text from Content with model " .
  95                                 $content->getModel()
  96                         );
  97                 }
  98
  99                 if ( $wgContentHandlerTextFallback == 'serialize' ) {
 100                         return $content->serialize();
 101                 }
 102
 103                 return null;
 104         }
 105
 106         /**
 107          * Convenience function for creating a Content object from a given textual
 108          * representation.
 109          *
 110          * $text will be deserialized into a Content object of the model specified
 111          * by $modelId (or, if that is not given, $title->getContentModel()) using
 112          * the given format.
 113          *
 114          * @since 1.21
 115          *
 116          * @param string $text The textual representation, will be
 117          *    unserialized to create the Content object
 118          * @param Title $title The title of the page this text belongs to.
 119          *    Required if $modelId is not provided.
 120          * @param string $modelId The model to deserialize to. If not provided,
 121          *    $title->getContentModel() is used.
 122          * @param string $format The format to use for deserialization. If not
 123          *    given, the model's default format is used.
 124          *
 125          * @throws MWException If model ID or format is not supported or if the text can not be
 126          * unserialized using the format.
 127          * @return Content A Content object representing the text.
 128          */
 129         public static function makeContent( $text, Title $title = null,
 130                 $modelId = null, $format = null ) {
 131                 if ( is_null( $modelId ) ) {
 132                         if ( is_null( $title ) ) {
 133                                 throw new MWException( "Must provide a Title object or a content model ID." );
 134                         }
 135
 136                         $modelId = $title->getContentModel();
 137                 }
 138
 139                 $handler = self::getForModelID( $modelId );
 140
 141                 return $handler->unserializeContent( $text, $format );
 142         }
 143
 144         /**
 145          * Returns the name of the default content model to be used for the page
 146          * with the given title.
 147          *
 148          * Note: There should rarely be need to call this method directly.
 149          * To determine the actual content model for a given page, use
 150          * Title::getContentModel().
 151          *
 152          * Which model is to be used by default for the page is determined based
 153          * on several factors:
 154          * - The global setting $wgNamespaceContentModels specifies a content model
 155          *   per namespace.
 156          * - The hook ContentHandlerDefaultModelFor may be used to override the page's default
 157          *   model.
 158          * - Pages in NS_MEDIAWIKI and NS_USER default to the CSS or JavaScript
 159          *   model if they end in .js or .css, respectively.
 160          * - Pages in NS_MEDIAWIKI default to the wikitext model otherwise.
 161          * - The hook TitleIsCssOrJsPage may be used to force a page to use the CSS
 162          *   or JavaScript model. This is a compatibility feature. The ContentHandlerDefaultModelFor
 163          *   hook should be used instead if possible.
 164          * - The hook TitleIsWikitextPage may be used to force a page to use the
 165          *   wikitext model. This is a compatibility feature. The ContentHandlerDefaultModelFor
 166          *   hook should be used instead if possible.
 167          *
 168          * If none of the above applies, the wikitext model is used.
 169          *
 170          * Note: this is used by, and may thus not use, Title::getContentModel()
 171          *
 172          * @since 1.21
 173          *
 174          * @param Title $title
 175          *
 176          * @return string Default model name for the page given by $title
 177          */
 178         public static function getDefaultModelFor( Title $title ) {
 179                 // NOTE: this method must not rely on $title->getContentModel() directly or indirectly,
 180                 //       because it is used to initialize the mContentModel member.
 181
 182                 $ns = $title->getNamespace();
 183
 184                 $ext = false;
 185                 $m = null;
 186                 $model = MWNamespace::getNamespaceContentModel( $ns );
 187
 188                 // Hook can determine default model
 189                 if ( !Hooks::run( 'ContentHandlerDefaultModelFor', [ $title, &$model ] ) ) {
 190                         if ( !is_null( $model ) ) {
 191                                 return $model;
 192                         }
 193                 }
 194
 195                 // Could this page contain code based on the title?
 196                 $isCodePage = NS_MEDIAWIKI == $ns && preg_match( '!\.(css|js|json)$!u', $title->getText(), $m );
 197                 if ( $isCodePage ) {
 198                         $ext = $m[1];
 199                 }
 200
 201                 // Is this a user subpage containing code?
 202                 $isCodeSubpage = NS_USER == $ns
 203                         && !$isCodePage
 204                         && preg_match( "/\\/.*\\.(js|css|json)$/", $title->getText(), $m );
 205                 if ( $isCodeSubpage ) {
 206                         $ext = $m[1];
 207                 }
 208
 209                 // Is this wikitext, according to $wgNamespaceContentModels or the DefaultModelFor hook?
 210                 $isWikitext = is_null( $model ) || $model == CONTENT_MODEL_WIKITEXT;
 211                 $isWikitext = $isWikitext && !$isCodePage && !$isCodeSubpage;
 212
 213                 if ( !$isWikitext ) {
 214                         switch ( $ext ) {
 215                                 case 'js':
 216                                         return CONTENT_MODEL_JAVASCRIPT;
 217                                 case 'css':
 218                                         return CONTENT_MODEL_CSS;
 219                                 case 'json':
 220                                         return CONTENT_MODEL_JSON;
 221                                 default:
 222                                         return is_null( $model ) ? CONTENT_MODEL_TEXT : $model;
 223                         }
 224                 }
 225
 226                 // We established that it must be wikitext
 227
 228                 return CONTENT_MODEL_WIKITEXT;
 229         }
 230
 231         /**
 232          * Returns the appropriate ContentHandler singleton for the given title.
 233          *
 234          * @since 1.21
 235          *
 236          * @param Title $title
 237          *
 238          * @return ContentHandler
 239          */
 240         public static function getForTitle( Title $title ) {
 241                 $modelId = $title->getContentModel();
 242
 243                 return self::getForModelID( $modelId );
 244         }
 245
 246         /**
 247          * Returns the appropriate ContentHandler singleton for the given Content
 248          * object.
 249          *
 250          * @since 1.21
 251          *
 252          * @param Content $content
 253          *
 254          * @return ContentHandler
 255          */
 256         public static function getForContent( Content $content ) {
 257                 $modelId = $content->getModel();
 258
 259                 return self::getForModelID( $modelId );
 260         }
 261
 262         /**
 263          * @var array A Cache of ContentHandler instances by model id
 264          */
 265         protected static $handlers;
 266
 267         /**
 268          * Returns the ContentHandler singleton for the given model ID. Use the
 269          * CONTENT_MODEL_XXX constants to identify the desired content model.
 270          *
 271          * ContentHandler singletons are taken from the global $wgContentHandlers
 272          * array. Keys in that array are model names, the values are either
 273          * ContentHandler singleton objects, or strings specifying the appropriate
 274          * subclass of ContentHandler.
 275          *
 276          * If a class name is encountered when looking up the singleton for a given
 277          * model name, the class is instantiated and the class name is replaced by
 278          * the resulting singleton in $wgContentHandlers.
 279          *
 280          * If no ContentHandler is defined for the desired $modelId, the
 281          * ContentHandler may be provided by the ContentHandlerForModelID hook.
 282          * If no ContentHandler can be determined, an MWException is raised.
 283          *
 284          * @since 1.21
 285          *
 286          * @param string $modelId The ID of the content model for which to get a
 287          *    handler. Use CONTENT_MODEL_XXX constants.
 288          *
 289          * @throws MWException For internal errors and problems in the configuration.
 290          * @throws MWUnknownContentModelException If no handler is known for the model ID.
 291          * @return ContentHandler The ContentHandler singleton for handling the model given by the ID.
 292          */
 293         public static function getForModelID( $modelId ) {
 294                 global $wgContentHandlers;
 295
 296                 if ( isset( self::$handlers[$modelId] ) ) {
 297                         return self::$handlers[$modelId];
 298                 }
 299
 300                 if ( empty( $wgContentHandlers[$modelId] ) ) {
 301                         $handler = null;
 302
 303                         Hooks::run( 'ContentHandlerForModelID', [ $modelId, &$handler ] );
 304
 305                         if ( $handler === null ) {
 306                                 throw new MWUnknownContentModelException( $modelId );
 307                         }
 308
 309                         if ( !( $handler instanceof ContentHandler ) ) {
 310                                 throw new MWException( "ContentHandlerForModelID must supply a ContentHandler instance" );
 311                         }
 312                 } else {
 313                         $classOrCallback = $wgContentHandlers[$modelId];
 314
 315                         if ( is_callable( $classOrCallback ) ) {
 316                                 $handler = call_user_func( $classOrCallback, $modelId );
 317                         } else {
 318                                 $handler = new $classOrCallback( $modelId );
 319                         }
 320
 321                         if ( !( $handler instanceof ContentHandler ) ) {
 322                                 throw new MWException( "$classOrCallback from \$wgContentHandlers is not " .
 323                                         "compatible with ContentHandler" );
 324                         }
 325                 }
 326
 327                 wfDebugLog( 'ContentHandler', 'Created handler for ' . $modelId
 328                         . ': ' . get_class( $handler ) );
 329
 330                 self::$handlers[$modelId] = $handler;
 331
 332                 return self::$handlers[$modelId];
 333         }
 334
 335         /**
 336          * Returns the localized name for a given content model.
 337          *
 338          * Model names are localized using system messages. Message keys
 339          * have the form content-model-$name, where $name is getContentModelName( $id ).
 340          *
 341          * @param string $name The content model ID, as given by a CONTENT_MODEL_XXX
 342          *    constant or returned by Revision::getContentModel().
 343          * @param Language|null $lang The language to parse the message in (since 1.26)
 344          *
 345          * @throws MWException If the model ID isn't known.
 346          * @return string The content model's localized name.
 347          */
 348         public static function getLocalizedName( $name, Language $lang = null ) {
 349                 // Messages: content-model-wikitext, content-model-text,
 350                 // content-model-javascript, content-model-css
 351                 $key = "content-model-$name";
 352
 353                 $msg = wfMessage( $key );
 354                 if ( $lang ) {
 355                         $msg->inLanguage( $lang );
 356                 }
 357
 358                 return $msg->exists() ? $msg->plain() : $name;
 359         }
 360
 361         public static function getContentModels() {
 362                 global $wgContentHandlers;
 363
 364                 $models = array_keys( $wgContentHandlers );
 365                 Hooks::run( 'GetContentModels', [ &$models ] );
 366                 return $models;
 367         }
 368
 369         public static function getAllContentFormats() {
 370                 global $wgContentHandlers;
 371
 372                 $formats = [];
 373
 374                 foreach ( $wgContentHandlers as $model => $class ) {
 375                         $handler = self::getForModelID( $model );
 376                         $formats = array_merge( $formats, $handler->getSupportedFormats() );
 377                 }
 378
 379                 $formats = array_unique( $formats );
 380
 381                 return $formats;
 382         }
 383
 384         // ------------------------------------------------------------------------
 385
 386         /**
 387          * @var string
 388          */
 389         protected $mModelID;
 390
 391         /**
 392          * @var string[]
 393          */
 394         protected $mSupportedFormats;
 395
 396         /**
 397          * Constructor, initializing the ContentHandler instance with its model ID
 398          * and a list of supported formats. Values for the parameters are typically
 399          * provided as literals by subclass's constructors.
 400          *
 401          * @param string $modelId (use CONTENT_MODEL_XXX constants).
 402          * @param string[] $formats List for supported serialization formats
 403          *    (typically as MIME types)
 404          */
 405         public function __construct( $modelId, $formats ) {
 406                 $this->mModelID = $modelId;
 407                 $this->mSupportedFormats = $formats;
 408         }
 409
 410         /**
 411          * Serializes a Content object of the type supported by this ContentHandler.
 412          *
 413          * @since 1.21
 414          *
 415          * @param Content $content The Content object to serialize
 416          * @param string $format The desired serialization format
 417          *
 418          * @return string Serialized form of the content
 419          */
 420         abstract public function serializeContent( Content $content, $format = null );
 421
 422         /**
 423          * Applies transformations on export (returns the blob unchanged per default).
 424          * Subclasses may override this to perform transformations such as conversion
 425          * of legacy formats or filtering of internal meta-data.
 426          *
 427          * @param string $blob The blob to be exported
 428          * @param string|null $format The blob's serialization format
 429          *
 430          * @return string
 431          */
 432         public function exportTransform( $blob, $format = null ) {
 433                 return $blob;
 434         }
 435
 436         /**
 437          * Unserializes a Content object of the type supported by this ContentHandler.
 438          *
 439          * @since 1.21
 440          *
 441          * @param string $blob Serialized form of the content
 442          * @param string $format The format used for serialization
 443          *
 444          * @return Content The Content object created by deserializing $blob
 445          */
 446         abstract public function unserializeContent( $blob, $format = null );
 447
 448         /**
 449          * Apply import transformation (per default, returns $blob unchanged).
 450          * This gives subclasses an opportunity to transform data blobs on import.
 451          *
 452          * @since 1.24
 453          *
 454          * @param string $blob
 455          * @param string|null $format
 456          *
 457          * @return string
 458          */
 459         public function importTransform( $blob, $format = null ) {
 460                 return $blob;
 461         }
 462
 463         /**
 464          * Creates an empty Content object of the type supported by this
 465          * ContentHandler.
 466          *
 467          * @since 1.21
 468          *
 469          * @return Content
 470          */
 471         abstract public function makeEmptyContent();
 472
 473         /**
 474          * Creates a new Content object that acts as a redirect to the given page,
 475          * or null if redirects are not supported by this content model.
 476          *
 477          * This default implementation always returns null. Subclasses supporting redirects
 478          * must override this method.
 479          *
 480          * Note that subclasses that override this method to return a Content object
 481          * should also override supportsRedirects() to return true.
 482          *
 483          * @since 1.21
 484          *
 485          * @param Title $destination The page to redirect to.
 486          * @param string $text Text to include in the redirect, if possible.
 487          *
 488          * @return Content Always null.
 489          */
 490         public function makeRedirectContent( Title $destination, $text = '' ) {
 491                 return null;
 492         }
 493
 494         /**
 495          * Returns the model id that identifies the content model this
 496          * ContentHandler can handle. Use with the CONTENT_MODEL_XXX constants.
 497          *
 498          * @since 1.21
 499          *
 500          * @return string The model ID
 501          */
 502         public function getModelID() {
 503                 return $this->mModelID;
 504         }
 505
 506         /**
 507          * @since 1.21
 508          *
 509          * @param string $model_id The model to check
 510          *
 511          * @throws MWException If the model ID is not the ID of the content model supported by this
 512          * ContentHandler.
 513          */
 514         protected function checkModelID( $model_id ) {
 515                 if ( $model_id !== $this->mModelID ) {
 516                         throw new MWException( "Bad content model: " .
 517                                 "expected {$this->mModelID} " .
 518                                 "but got $model_id." );
 519                 }
 520         }
 521
 522         /**
 523          * Returns a list of serialization formats supported by the
 524          * serializeContent() and unserializeContent() methods of this
 525          * ContentHandler.
 526          *
 527          * @since 1.21
 528          *
 529          * @return string[] List of serialization formats as MIME type like strings
 530          */
 531         public function getSupportedFormats() {
 532                 return $this->mSupportedFormats;
 533         }
 534
 535         /**
 536          * The format used for serialization/deserialization by default by this
 537          * ContentHandler.
 538          *
 539          * This default implementation will return the first element of the array
 540          * of formats that was passed to the constructor.
 541          *
 542          * @since 1.21
 543          *
 544          * @return string The name of the default serialization format as a MIME type
 545          */
 546         public function getDefaultFormat() {
 547                 return $this->mSupportedFormats[0];
 548         }
 549
 550         /**
 551          * Returns true if $format is a serialization format supported by this
 552          * ContentHandler, and false otherwise.
 553          *
 554          * Note that if $format is null, this method always returns true, because
 555          * null means "use the default format".
 556          *
 557          * @since 1.21
 558          *
 559          * @param string $format The serialization format to check
 560          *
 561          * @return bool
 562          */
 563         public function isSupportedFormat( $format ) {
 564                 if ( !$format ) {
 565                         return true; // this means "use the default"
 566                 }
 567
 568                 return in_array( $format, $this->mSupportedFormats );
 569         }
 570
 571         /**
 572          * Convenient for checking whether a format provided as a parameter is actually supported.
 573          *
 574          * @param string $format The serialization format to check
 575          *
 576          * @throws MWException If the format is not supported by this content handler.
 577          */
 578         protected function checkFormat( $format ) {
 579                 if ( !$this->isSupportedFormat( $format ) ) {
 580                         throw new MWException(
 581                                 "Format $format is not supported for content model "
 582                                 . $this->getModelID()
 583                         );
 584                 }
 585         }
 586
 587         /**
 588          * Returns overrides for action handlers.
 589          * Classes listed here will be used instead of the default one when
 590          * (and only when) $wgActions[$action] === true. This allows subclasses
 591          * to override the default action handlers.
 592          *
 593          * @since 1.21
 594          *
 595          * @return array An array mapping action names (typically "view", "edit", "history" etc.) to
 596          *  either the full qualified class name of an Action class, a callable taking ( Page $page,
 597          *  IContextSource $context = null ) as parameters and returning an Action object, or an actual
 598          *  Action object. An empty array in this default implementation.
 599          *
 600          * @see Action::factory
 601          */
 602         public function getActionOverrides() {
 603                 return [];
 604         }
 605
 606         /**
 607          * Factory for creating an appropriate DifferenceEngine for this content model.
 608          *
 609          * @since 1.21
 610          *
 611          * @param IContextSource $context Context to use, anything else will be ignored.
 612          * @param int $old Revision ID we want to show and diff with.
 613          * @param int|string $new Either a revision ID or one of the strings 'cur', 'prev' or 'next'.
 614          * @param int $rcid FIXME: Deprecated, no longer used. Defaults to 0.
 615          * @param bool $refreshCache If set, refreshes the diff cache. Defaults to false.
 616          * @param bool $unhide If set, allow viewing deleted revs. Defaults to false.
 617          *
 618          * @return DifferenceEngine
 619          */
 620         public function createDifferenceEngine( IContextSource $context, $old = 0, $new = 0,
 621                 $rcid = 0, // FIXME: Deprecated, no longer used
 622                 $refreshCache = false, $unhide = false
 623         ) {
 624                 // hook: get difference engine
 625                 $differenceEngine = null;
 626                 if ( !Hooks::run( 'GetDifferenceEngine',
 627                         [ $context, $old, $new, $refreshCache, $unhide, &$differenceEngine ]
 628                 ) ) {
 629                         return $differenceEngine;
 630                 }
 631                 $diffEngineClass = $this->getDiffEngineClass();
 632                 return new $diffEngineClass( $context, $old, $new, $rcid, $refreshCache, $unhide );
 633         }
 634
 635         /**
 636          * Get the language in which the content of the given page is written.
 637          *
 638          * This default implementation just returns $wgContLang (except for pages
 639          * in the MediaWiki namespace)
 640          *
 641          * Note that the pages language is not cacheable, since it may in some
 642          * cases depend on user settings.
 643          *
 644          * Also note that the page language may or may not depend on the actual content of the page,
 645          * that is, this method may load the content in order to determine the language.
 646          *
 647          * @since 1.21
 648          *
 649          * @param Title $title The page to determine the language for.
 650          * @param Content $content The page's content, if you have it handy, to avoid reloading it.
 651          *
 652          * @return Language The page's language
 653          */
 654         public function getPageLanguage( Title $title, Content $content = null ) {
 655                 global $wgContLang, $wgLang;
 656                 $pageLang = $wgContLang;
 657
 658                 if ( $title->getNamespace() == NS_MEDIAWIKI ) {
 659                         // Parse mediawiki messages with correct target language
 660                         list( /* $unused */, $lang ) = MessageCache::singleton()->figureMessage( $title->getText() );
 661                         $pageLang = Language::factory( $lang );
 662                 }
 663
 664                 Hooks::run( 'PageContentLanguage', [ $title, &$pageLang, $wgLang ] );
 665
 666                 return wfGetLangObj( $pageLang );
 667         }
 668
 669         /**
 670          * Get the language in which the content of this page is written when
 671          * viewed by user. Defaults to $this->getPageLanguage(), but if the user
 672          * specified a preferred variant, the variant will be used.
 673          *
 674          * This default implementation just returns $this->getPageLanguage( $title, $content ) unless
 675          * the user specified a preferred variant.
 676          *
 677          * Note that the pages view language is not cacheable, since it depends on user settings.
 678          *
 679          * Also note that the page language may or may not depend on the actual content of the page,
 680          * that is, this method may load the content in order to determine the language.
 681          *
 682          * @since 1.21
 683          *
 684          * @param Title $title The page to determine the language for.
 685          * @param Content $content The page's content, if you have it handy, to avoid reloading it.
 686          *
 687          * @return Language The page's language for viewing
 688          */
 689         public function getPageViewLanguage( Title $title, Content $content = null ) {
 690                 $pageLang = $this->getPageLanguage( $title, $content );
 691
 692                 if ( $title->getNamespace() !== NS_MEDIAWIKI ) {
 693                         // If the user chooses a variant, the content is actually
 694                         // in a language whose code is the variant code.
 695                         $variant = $pageLang->getPreferredVariant();
 696                         if ( $pageLang->getCode() !== $variant ) {
 697                                 $pageLang = Language::factory( $variant );
 698                         }
 699                 }
 700
 701                 return $pageLang;
 702         }
 703
 704         /**
 705          * Determines whether the content type handled by this ContentHandler
 706          * can be used on the given page.
 707          *
 708          * This default implementation always returns true.
 709          * Subclasses may override this to restrict the use of this content model to specific locations,
 710          * typically based on the namespace or some other aspect of the title, such as a special suffix
 711          * (e.g. ".svg" for SVG content).
 712          *
 713          * @note this calls the ContentHandlerCanBeUsedOn hook which may be used to override which
 714          * content model can be used where.
 715          *
 716          * @param Title $title The page's title.
 717          *
 718          * @return bool True if content of this kind can be used on the given page, false otherwise.
 719          */
 720         public function canBeUsedOn( Title $title ) {
 721                 $ok = true;
 722
 723                 Hooks::run( 'ContentModelCanBeUsedOn', [ $this->getModelID(), $title, &$ok ] );
 724
 725                 return $ok;
 726         }
 727
 728         /**
 729          * Returns the name of the diff engine to use.
 730          *
 731          * @since 1.21
 732          *
 733          * @return string
 734          */
 735         protected function getDiffEngineClass() {
 736                 return DifferenceEngine::class;
 737         }
 738
 739         /**
 740          * Attempts to merge differences between three versions. Returns a new
 741          * Content object for a clean merge and false for failure or a conflict.
 742          *
 743          * This default implementation always returns false.
 744          *
 745          * @since 1.21
 746          *
 747          * @param Content $oldContent The page's previous content.
 748          * @param Content $myContent One of the page's conflicting contents.
 749          * @param Content $yourContent One of the page's conflicting contents.
 750          *
 751          * @return Content|bool Always false.
 752          */
 753         public function merge3( Content $oldContent, Content $myContent, Content $yourContent ) {
 754                 return false;
 755         }
 756
 757         /**
 758          * Return type of change if one exists for the given edit.
 759          *
 760          * @since 1.31
 761          *
 762          * @param Content|null $oldContent The previous text of the page.
 763          * @param Content|null $newContent The submitted text of the page.
 764          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 765          *
 766          * @return string|null String key representing type of change, or null.
 767          */
 768         private function getChangeType(
 769                 Content $oldContent = null,
 770                 Content $newContent = null,
 771                 $flags = 0
 772         ) {
 773                 $oldTarget = $oldContent !== null ? $oldContent->getRedirectTarget() : null;
 774                 $newTarget = $newContent !== null ? $newContent->getRedirectTarget() : null;
 775
 776                 // We check for the type of change in the given edit, and return string key accordingly
 777
 778                 // Blanking of a page
 779                 if ( $oldContent && $oldContent->getSize() > 0 &&
 780                         $newContent && $newContent->getSize() === 0
 781                 ) {
 782                         return 'blank';
 783                 }
 784
 785                 // Redirects
 786                 if ( $newTarget ) {
 787                         if ( !$oldTarget ) {
 788                                 // New redirect page (by creating new page or by changing content page)
 789                                 return 'new-redirect';
 790                         } elseif ( !$newTarget->equals( $oldTarget ) ||
 791                                 $oldTarget->getFragment() !== $newTarget->getFragment()
 792                         ) {
 793                                 // Redirect target changed
 794                                 return 'changed-redirect-target';
 795                         }
 796                 } elseif ( $oldTarget ) {
 797                         // Changing an existing redirect into a non-redirect
 798                         return 'removed-redirect';
 799                 }
 800
 801                 // New page created
 802                 if ( $flags & EDIT_NEW && $newContent ) {
 803                         if ( $newContent->getSize() === 0 ) {
 804                                 // New blank page
 805                                 return 'newblank';
 806                         } else {
 807                                 return 'newpage';
 808                         }
 809                 }
 810
 811                 // Removing more than 90% of the page
 812                 if ( $oldContent && $newContent && $oldContent->getSize() > 10 * $newContent->getSize() ) {
 813                         return 'replace';
 814                 }
 815
 816                 // Content model changed
 817                 if ( $oldContent && $newContent && $oldContent->getModel() !== $newContent->getModel() ) {
 818                         return 'contentmodelchange';
 819                 }
 820
 821                 return null;
 822         }
 823
 824         /**
 825          * Return an applicable auto-summary if one exists for the given edit.
 826          *
 827          * @since 1.21
 828          *
 829          * @param Content|null $oldContent The previous text of the page.
 830          * @param Content|null $newContent The submitted text of the page.
 831          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 832          *
 833          * @return string An appropriate auto-summary, or an empty string.
 834          */
 835         public function getAutosummary(
 836                 Content $oldContent = null,
 837                 Content $newContent = null,
 838                 $flags = 0
 839         ) {
 840                 $changeType = $this->getChangeType( $oldContent, $newContent, $flags );
 841
 842                 // There's no applicable auto-summary for our case, so our auto-summary is empty.
 843                 if ( !$changeType ) {
 844                         return '';
 845                 }
 846
 847                 // Decide what kind of auto-summary is needed.
 848                 switch ( $changeType ) {
 849                         case 'new-redirect':
 850                                 $newTarget = $newContent->getRedirectTarget();
 851                                 $truncatedtext = $newContent->getTextForSummary(
 852                                         250
 853                                         - strlen( wfMessage( 'autoredircomment' )->inContentLanguage()->text() )
 854                                         - strlen( $newTarget->getFullText() )
 855                                 );
 856
 857                                 return wfMessage( 'autoredircomment', $newTarget->getFullText() )
 858                                         ->plaintextParams( $truncatedtext )->inContentLanguage()->text();
 859                         case 'changed-redirect-target':
 860                                 $oldTarget = $oldContent->getRedirectTarget();
 861                                 $newTarget = $newContent->getRedirectTarget();
 862
 863                                 $truncatedtext = $newContent->getTextForSummary(
 864                                         250
 865                                         - strlen( wfMessage( 'autosumm-changed-redirect-target' )
 866                                                 ->inContentLanguage()->text() )
 867                                         - strlen( $oldTarget->getFullText() )
 868                                         - strlen( $newTarget->getFullText() )
 869                                 );
 870
 871                                 return wfMessage( 'autosumm-changed-redirect-target',
 872                                                 $oldTarget->getFullText(),
 873                                                 $newTarget->getFullText() )
 874                                         ->rawParams( $truncatedtext )->inContentLanguage()->text();
 875                         case 'removed-redirect':
 876                                 $oldTarget = $oldContent->getRedirectTarget();
 877                                 $truncatedtext = $newContent->getTextForSummary(
 878                                         250
 879                                         - strlen( wfMessage( 'autosumm-removed-redirect' )
 880                                                 ->inContentLanguage()->text() )
 881                                         - strlen( $oldTarget->getFullText() ) );
 882
 883                                 return wfMessage( 'autosumm-removed-redirect', $oldTarget->getFullText() )
 884                                         ->rawParams( $truncatedtext )->inContentLanguage()->text();
 885                         case 'newpage':
 886                                 // If they're making a new article, give its text, truncated, in the summary.
 887                                 $truncatedtext = $newContent->getTextForSummary(
 888                                         200 - strlen( wfMessage( 'autosumm-new' )->inContentLanguage()->text() ) );
 889
 890                                 return wfMessage( 'autosumm-new' )->rawParams( $truncatedtext )
 891                                         ->inContentLanguage()->text();
 892                         case 'blank':
 893                                 return wfMessage( 'autosumm-blank' )->inContentLanguage()->text();
 894                         case 'replace':
 895                                 $truncatedtext = $newContent->getTextForSummary(
 896                                         200 - strlen( wfMessage( 'autosumm-replace' )->inContentLanguage()->text() ) );
 897
 898                                 return wfMessage( 'autosumm-replace' )->rawParams( $truncatedtext )
 899                                         ->inContentLanguage()->text();
 900                         case 'newblank':
 901                                 return wfMessage( 'autosumm-newblank' )->inContentLanguage()->text();
 902                         default:
 903                                 return '';
 904                 }
 905         }
 906
 907         /**
 908          * Return an applicable tag if one exists for the given edit or return null.
 909          *
 910          * @since 1.31
 911          *
 912          * @param Content|null $oldContent The previous text of the page.
 913          * @param Content|null $newContent The submitted text of the page.
 914          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 915          *
 916          * @return string|null An appropriate tag, or null.
 917          */
 918         public function getChangeTag(
 919                 Content $oldContent = null,
 920                 Content $newContent = null,
 921                 $flags = 0
 922         ) {
 923                 $changeType = $this->getChangeType( $oldContent, $newContent, $flags );
 924
 925                 // There's no applicable tag for this change.
 926                 if ( !$changeType ) {
 927                         return null;
 928                 }
 929
 930                 // Core tags use the same keys as ones returned from $this->getChangeType()
 931                 // but prefixed with pseudo namespace 'mw-', so we add the prefix before checking
 932                 // if this type of change should be tagged
 933                 $tag = 'mw-' . $changeType;
 934
 935                 // Not all change types are tagged, so we check against the list of defined tags.
 936                 if ( in_array( $tag, ChangeTags::getSoftwareTags() ) ) {
 937                         return $tag;
 938                 }
 939
 940                 return null;
 941         }
 942
 943         /**
 944          * Auto-generates a deletion reason
 945          *
 946          * @since 1.21
 947          *
 948          * @param Title $title The page's title
 949          * @param bool &$hasHistory Whether the page has a history
 950          *
 951          * @return mixed String containing deletion reason or empty string, or
 952          *    boolean false if no revision occurred
 953          *
 954          * @todo &$hasHistory is extremely ugly, it's here because
 955          * WikiPage::getAutoDeleteReason() and Article::generateReason()
 956          * have it / want it.
 957          */
 958         public function getAutoDeleteReason( Title $title, &$hasHistory ) {
 959                 $dbr = wfGetDB( DB_REPLICA );
 960
 961                 // Get the last revision
 962                 $rev = Revision::newFromTitle( $title );
 963
 964                 if ( is_null( $rev ) ) {
 965                         return false;
 966                 }
 967
 968                 // Get the article's contents
 969                 $content = $rev->getContent();
 970                 $blank = false;
 971
 972                 // If the page is blank, use the text from the previous revision,
 973                 // which can only be blank if there's a move/import/protect dummy
 974                 // revision involved
 975                 if ( !$content || $content->isEmpty() ) {
 976                         $prev = $rev->getPrevious();
 977
 978                         if ( $prev ) {
 979                                 $rev = $prev;
 980                                 $content = $rev->getContent();
 981                                 $blank = true;
 982                         }
 983                 }
 984
 985                 $this->checkModelID( $rev->getContentModel() );
 986
 987                 // Find out if there was only one contributor
 988                 // Only scan the last 20 revisions
 989                 $res = $dbr->select( 'revision', 'rev_user_text',
 990                         [
 991                                 'rev_page' => $title->getArticleID(),
 992                                 $dbr->bitAnd( 'rev_deleted', Revision::DELETED_USER ) . ' = 0'
 993                         ],
 994                         __METHOD__,
 995                         [ 'LIMIT' => 20 ]
 996                 );
 997
 998                 if ( $res === false ) {
 999                         // This page has no revisions, which is very weird
1000                         return false;
1001                 }
1002
1003                 $hasHistory = ( $res->numRows() > 1 );
1004                 $row = $dbr->fetchObject( $res );
1005
1006                 if ( $row ) { // $row is false if the only contributor is hidden
1007                         $onlyAuthor = $row->rev_user_text;
1008                         // Try to find a second contributor
1009                         foreach ( $res as $row ) {
1010                                 if ( $row->rev_user_text != $onlyAuthor ) { // T24999
1011                                         $onlyAuthor = false;
1012                                         break;
1013                                 }
1014                         }
1015                 } else {
1016                         $onlyAuthor = false;
1017                 }
1018
1019                 // Generate the summary with a '$1' placeholder
1020                 if ( $blank ) {
1021                         // The current revision is blank and the one before is also
1022                         // blank. It's just not our lucky day
1023                         $reason = wfMessage( 'exbeforeblank', '$1' )->inContentLanguage()->text();
1024                 } else {
1025                         if ( $onlyAuthor ) {
1026                                 $reason = wfMessage(
1027                                         'excontentauthor',
1028                                         '$1',
1029                                         $onlyAuthor
1030                                 )->inContentLanguage()->text();
1031                         } else {
1032                                 $reason = wfMessage( 'excontent', '$1' )->inContentLanguage()->text();
1033                         }
1034                 }
1035
1036                 if ( $reason == '-' ) {
1037                         // Allow these UI messages to be blanked out cleanly
1038                         return '';
1039                 }
1040
1041                 // Max content length = max comment length - length of the comment (excl. $1)
1042                 $text = $content ? $content->getTextForSummary( 255 - ( strlen( $reason ) - 2 ) ) : '';
1043
1044                 // Now replace the '$1' placeholder
1045                 $reason = str_replace( '$1', $text, $reason );
1046
1047                 return $reason;
1048         }
1049
1050         /**
1051          * Get the Content object that needs to be saved in order to undo all revisions
1052          * between $undo and $undoafter. Revisions must belong to the same page,
1053          * must exist and must not be deleted.
1054          *
1055          * @since 1.21
1056          *
1057          * @param Revision $current The current text
1058          * @param Revision $undo The revision to undo
1059          * @param Revision $undoafter Must be an earlier revision than $undo
1060          *
1061          * @return mixed String on success, false on failure
1062          */
1063         public function getUndoContent( Revision $current, Revision $undo, Revision $undoafter ) {
1064                 $cur_content = $current->getContent();
1065
1066                 if ( empty( $cur_content ) ) {
1067                         return false; // no page
1068                 }
1069
1070                 $undo_content = $undo->getContent();
1071                 $undoafter_content = $undoafter->getContent();
1072
1073                 if ( !$undo_content || !$undoafter_content ) {
1074                         return false; // no content to undo
1075                 }
1076
1077                 try {
1078                         $this->checkModelID( $cur_content->getModel() );
1079                         $this->checkModelID( $undo_content->getModel() );
1080                         if ( $current->getId() !== $undo->getId() ) {
1081                                 // If we are undoing the most recent revision,
1082                                 // its ok to revert content model changes. However
1083                                 // if we are undoing a revision in the middle, then
1084                                 // doing that will be confusing.
1085                                 $this->checkModelID( $undoafter_content->getModel() );
1086                         }
1087                 } catch ( MWException $e ) {
1088                         // If the revisions have different content models
1089                         // just return false
1090                         return false;
1091                 }
1092
1093                 if ( $cur_content->equals( $undo_content ) ) {
1094                         // No use doing a merge if it's just a straight revert.
1095                         return $undoafter_content;
1096                 }
1097
1098                 $undone_content = $this->merge3( $undo_content, $undoafter_content, $cur_content );
1099
1100                 return $undone_content;
1101         }
1102
1103         /**
1104          * Get parser options suitable for rendering and caching the article
1105          *
1106          * @param IContextSource|User|string $context One of the following:
1107          *        - IContextSource: Use the User and the Language of the provided
1108          *                                            context
1109          *        - User: Use the provided User object and $wgLang for the language,
1110          *                                            so use an IContextSource object if possible.
1111          *        - 'canonical': Canonical options (anonymous user with default
1112          *                                            preferences and content language).
1113          *
1114          * @throws MWException
1115          * @return ParserOptions
1116          */
1117         public function makeParserOptions( $context ) {
1118                 global $wgContLang;
1119
1120                 if ( $context instanceof IContextSource ) {
1121                         $user = $context->getUser();
1122                         $lang = $context->getLanguage();
1123                 } elseif ( $context instanceof User ) { // settings per user (even anons)
1124                         $user = $context;
1125                         $lang = null;
1126                 } elseif ( $context === 'canonical' ) { // canonical settings
1127                         $user = new User;
1128                         $lang = $wgContLang;
1129                 } else {
1130                         throw new MWException( "Bad context for parser options: $context" );
1131                 }
1132
1133                 return ParserOptions::newCanonical( $user, $lang );
1134         }
1135
1136         /**
1137          * Returns true for content models that support caching using the
1138          * ParserCache mechanism. See WikiPage::shouldCheckParserCache().
1139          *
1140          * @since 1.21
1141          *
1142          * @return bool Always false.
1143          */
1144         public function isParserCacheSupported() {
1145                 return false;
1146         }
1147
1148         /**
1149          * Returns true if this content model supports sections.
1150          * This default implementation returns false.
1151          *
1152          * Content models that return true here should also implement
1153          * Content::getSection, Content::replaceSection, etc. to handle sections..
1154          *
1155          * @return bool Always false.
1156          */
1157         public function supportsSections() {
1158                 return false;
1159         }
1160
1161         /**
1162          * Returns true if this content model supports categories.
1163          * The default implementation returns true.
1164          *
1165          * @return bool Always true.
1166          */
1167         public function supportsCategories() {
1168                 return true;
1169         }
1170
1171         /**
1172          * Returns true if this content model supports redirects.
1173          * This default implementation returns false.
1174          *
1175          * Content models that return true here should also implement
1176          * ContentHandler::makeRedirectContent to return a Content object.
1177          *
1178          * @return bool Always false.
1179          */
1180         public function supportsRedirects() {
1181                 return false;
1182         }
1183
1184         /**
1185          * Return true if this content model supports direct editing, such as via EditPage.
1186          *
1187          * @return bool Default is false, and true for TextContent and it's derivatives.
1188          */
1189         public function supportsDirectEditing() {
1190                 return false;
1191         }
1192
1193         /**
1194          * Whether or not this content model supports direct editing via ApiEditPage
1195          *
1196          * @return bool Default is false, and true for TextContent and derivatives.
1197          */
1198         public function supportsDirectApiEditing() {
1199                 return $this->supportsDirectEditing();
1200         }
1201
1202         /**
1203          * Get fields definition for search index
1204          *
1205          * @todo Expose title, redirect, namespace, text, source_text, text_bytes
1206          *       field mappings here. (see T142670 and T143409)
1207          *
1208          * @param SearchEngine $engine
1209          * @return SearchIndexField[] List of fields this content handler can provide.
1210          * @since 1.28
1211          */
1212         public function getFieldsForSearchIndex( SearchEngine $engine ) {
1213                 $fields['category'] = $engine->makeSearchFieldMapping(
1214                         'category',
1215                         SearchIndexField::INDEX_TYPE_TEXT
1216                 );
1217                 $fields['category']->setFlag( SearchIndexField::FLAG_CASEFOLD );
1218
1219                 $fields['external_link'] = $engine->makeSearchFieldMapping(
1220                         'external_link',
1221                         SearchIndexField::INDEX_TYPE_KEYWORD
1222                 );
1223
1224                 $fields['outgoing_link'] = $engine->makeSearchFieldMapping(
1225                         'outgoing_link',
1226                         SearchIndexField::INDEX_TYPE_KEYWORD
1227                 );
1228
1229                 $fields['template'] = $engine->makeSearchFieldMapping(
1230                         'template',
1231                         SearchIndexField::INDEX_TYPE_KEYWORD
1232                 );
1233                 $fields['template']->setFlag( SearchIndexField::FLAG_CASEFOLD );
1234
1235                 $fields['content_model'] = $engine->makeSearchFieldMapping(
1236                         'content_model',
1237                         SearchIndexField::INDEX_TYPE_KEYWORD
1238                 );
1239
1240                 return $fields;
1241         }
1242
1243         /**
1244          * Add new field definition to array.
1245          * @param SearchIndexField[] &$fields
1246          * @param SearchEngine $engine
1247          * @param string $name
1248          * @param int $type
1249          * @return SearchIndexField[] new field defs
1250          * @since 1.28
1251          */
1252         protected function addSearchField( &$fields, SearchEngine $engine, $name, $type ) {
1253                 $fields[$name] = $engine->makeSearchFieldMapping( $name, $type );
1254                 return $fields;
1255         }
1256
1257         /**
1258          * Return fields to be indexed by search engine
1259          * as representation of this document.
1260          * Overriding class should call parent function or take care of calling
1261          * the SearchDataForIndex hook.
1262          * @param WikiPage $page Page to index
1263          * @param ParserOutput $output
1264          * @param SearchEngine $engine Search engine for which we are indexing
1265          * @return array Map of name=>value for fields
1266          * @since 1.28
1267          */
1268         public function getDataForSearchIndex(
1269                 WikiPage $page,
1270                 ParserOutput $output,
1271                 SearchEngine $engine
1272         ) {
1273                 $fieldData = [];
1274                 $content = $page->getContent();
1275
1276                 if ( $content ) {
1277                         $searchDataExtractor = new ParserOutputSearchDataExtractor();
1278
1279                         $fieldData['category'] = $searchDataExtractor->getCategories( $output );
1280                         $fieldData['external_link'] = $searchDataExtractor->getExternalLinks( $output );
1281                         $fieldData['outgoing_link'] = $searchDataExtractor->getOutgoingLinks( $output );
1282                         $fieldData['template'] = $searchDataExtractor->getTemplates( $output );
1283
1284                         $text = $content->getTextForSearchIndex();
1285
1286                         $fieldData['text'] = $text;
1287                         $fieldData['source_text'] = $text;
1288                         $fieldData['text_bytes'] = $content->getSize();
1289                         $fieldData['content_model'] = $content->getModel();
1290                 }
1291
1292                 Hooks::run( 'SearchDataForIndex', [ &$fieldData, $this, $page, $output, $engine ] );
1293                 return $fieldData;
1294         }
1295
1296         /**
1297          * Produce page output suitable for indexing.
1298          *
1299          * Specific content handlers may override it if they need different content handling.
1300          *
1301          * @param WikiPage $page
1302          * @param ParserCache $cache
1303          * @return ParserOutput
1304          */
1305         public function getParserOutputForIndexing( WikiPage $page, ParserCache $cache = null ) {
1306                 $parserOptions = $page->makeParserOptions( 'canonical' );
1307                 $revId = $page->getRevision()->getId();
1308                 if ( $cache ) {
1309                         $parserOutput = $cache->get( $page, $parserOptions );
1310                 }
1311                 if ( empty( $parserOutput ) ) {
1312                         $parserOutput =
1313                                 $page->getContent()->getParserOutput( $page->getTitle(), $revId, $parserOptions );
1314                         if ( $cache ) {
1315                                 $cache->save( $parserOutput, $page, $parserOptions );
1316                         }
1317                 }
1318                 return $parserOutput;
1319         }
1320
1321 }