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[$modelId];
 333         }
 334
 335         /**
 336          * Clean up handlers cache.
 337          */
 338         public static function cleanupHandlersCache() {
 339                 self::$handlers = [];
 340         }
 341
 342         /**
 343          * Returns the localized name for a given content model.
 344          *
 345          * Model names are localized using system messages. Message keys
 346          * have the form content-model-$name, where $name is getContentModelName( $id ).
 347          *
 348          * @param string $name The content model ID, as given by a CONTENT_MODEL_XXX
 349          *    constant or returned by Revision::getContentModel().
 350          * @param Language|null $lang The language to parse the message in (since 1.26)
 351          *
 352          * @throws MWException If the model ID isn't known.
 353          * @return string The content model's localized name.
 354          */
 355         public static function getLocalizedName( $name, Language $lang = null ) {
 356                 // Messages: content-model-wikitext, content-model-text,
 357                 // content-model-javascript, content-model-css
 358                 $key = "content-model-$name";
 359
 360                 $msg = wfMessage( $key );
 361                 if ( $lang ) {
 362                         $msg->inLanguage( $lang );
 363                 }
 364
 365                 return $msg->exists() ? $msg->plain() : $name;
 366         }
 367
 368         public static function getContentModels() {
 369                 global $wgContentHandlers;
 370
 371                 $models = array_keys( $wgContentHandlers );
 372                 Hooks::run( 'GetContentModels', [ &$models ] );
 373                 return $models;
 374         }
 375
 376         public static function getAllContentFormats() {
 377                 global $wgContentHandlers;
 378
 379                 $formats = [];
 380
 381                 foreach ( $wgContentHandlers as $model => $class ) {
 382                         $handler = self::getForModelID( $model );
 383                         $formats = array_merge( $formats, $handler->getSupportedFormats() );
 384                 }
 385
 386                 $formats = array_unique( $formats );
 387
 388                 return $formats;
 389         }
 390
 391         // ------------------------------------------------------------------------
 392
 393         /**
 394          * @var string
 395          */
 396         protected $mModelID;
 397
 398         /**
 399          * @var string[]
 400          */
 401         protected $mSupportedFormats;
 402
 403         /**
 404          * Constructor, initializing the ContentHandler instance with its model ID
 405          * and a list of supported formats. Values for the parameters are typically
 406          * provided as literals by subclass's constructors.
 407          *
 408          * @param string $modelId (use CONTENT_MODEL_XXX constants).
 409          * @param string[] $formats List for supported serialization formats
 410          *    (typically as MIME types)
 411          */
 412         public function __construct( $modelId, $formats ) {
 413                 $this->mModelID = $modelId;
 414                 $this->mSupportedFormats = $formats;
 415         }
 416
 417         /**
 418          * Serializes a Content object of the type supported by this ContentHandler.
 419          *
 420          * @since 1.21
 421          *
 422          * @param Content $content The Content object to serialize
 423          * @param string $format The desired serialization format
 424          *
 425          * @return string Serialized form of the content
 426          */
 427         abstract public function serializeContent( Content $content, $format = null );
 428
 429         /**
 430          * Applies transformations on export (returns the blob unchanged per default).
 431          * Subclasses may override this to perform transformations such as conversion
 432          * of legacy formats or filtering of internal meta-data.
 433          *
 434          * @param string $blob The blob to be exported
 435          * @param string|null $format The blob's serialization format
 436          *
 437          * @return string
 438          */
 439         public function exportTransform( $blob, $format = null ) {
 440                 return $blob;
 441         }
 442
 443         /**
 444          * Unserializes a Content object of the type supported by this ContentHandler.
 445          *
 446          * @since 1.21
 447          *
 448          * @param string $blob Serialized form of the content
 449          * @param string $format The format used for serialization
 450          *
 451          * @return Content The Content object created by deserializing $blob
 452          */
 453         abstract public function unserializeContent( $blob, $format = null );
 454
 455         /**
 456          * Apply import transformation (per default, returns $blob unchanged).
 457          * This gives subclasses an opportunity to transform data blobs on import.
 458          *
 459          * @since 1.24
 460          *
 461          * @param string $blob
 462          * @param string|null $format
 463          *
 464          * @return string
 465          */
 466         public function importTransform( $blob, $format = null ) {
 467                 return $blob;
 468         }
 469
 470         /**
 471          * Creates an empty Content object of the type supported by this
 472          * ContentHandler.
 473          *
 474          * @since 1.21
 475          *
 476          * @return Content
 477          */
 478         abstract public function makeEmptyContent();
 479
 480         /**
 481          * Creates a new Content object that acts as a redirect to the given page,
 482          * or null if redirects are not supported by this content model.
 483          *
 484          * This default implementation always returns null. Subclasses supporting redirects
 485          * must override this method.
 486          *
 487          * Note that subclasses that override this method to return a Content object
 488          * should also override supportsRedirects() to return true.
 489          *
 490          * @since 1.21
 491          *
 492          * @param Title $destination The page to redirect to.
 493          * @param string $text Text to include in the redirect, if possible.
 494          *
 495          * @return Content Always null.
 496          */
 497         public function makeRedirectContent( Title $destination, $text = '' ) {
 498                 return null;
 499         }
 500
 501         /**
 502          * Returns the model id that identifies the content model this
 503          * ContentHandler can handle. Use with the CONTENT_MODEL_XXX constants.
 504          *
 505          * @since 1.21
 506          *
 507          * @return string The model ID
 508          */
 509         public function getModelID() {
 510                 return $this->mModelID;
 511         }
 512
 513         /**
 514          * @since 1.21
 515          *
 516          * @param string $model_id The model to check
 517          *
 518          * @throws MWException If the model ID is not the ID of the content model supported by this
 519          * ContentHandler.
 520          */
 521         protected function checkModelID( $model_id ) {
 522                 if ( $model_id !== $this->mModelID ) {
 523                         throw new MWException( "Bad content model: " .
 524                                 "expected {$this->mModelID} " .
 525                                 "but got $model_id." );
 526                 }
 527         }
 528
 529         /**
 530          * Returns a list of serialization formats supported by the
 531          * serializeContent() and unserializeContent() methods of this
 532          * ContentHandler.
 533          *
 534          * @since 1.21
 535          *
 536          * @return string[] List of serialization formats as MIME type like strings
 537          */
 538         public function getSupportedFormats() {
 539                 return $this->mSupportedFormats;
 540         }
 541
 542         /**
 543          * The format used for serialization/deserialization by default by this
 544          * ContentHandler.
 545          *
 546          * This default implementation will return the first element of the array
 547          * of formats that was passed to the constructor.
 548          *
 549          * @since 1.21
 550          *
 551          * @return string The name of the default serialization format as a MIME type
 552          */
 553         public function getDefaultFormat() {
 554                 return $this->mSupportedFormats[0];
 555         }
 556
 557         /**
 558          * Returns true if $format is a serialization format supported by this
 559          * ContentHandler, and false otherwise.
 560          *
 561          * Note that if $format is null, this method always returns true, because
 562          * null means "use the default format".
 563          *
 564          * @since 1.21
 565          *
 566          * @param string $format The serialization format to check
 567          *
 568          * @return bool
 569          */
 570         public function isSupportedFormat( $format ) {
 571                 if ( !$format ) {
 572                         return true; // this means "use the default"
 573                 }
 574
 575                 return in_array( $format, $this->mSupportedFormats );
 576         }
 577
 578         /**
 579          * Convenient for checking whether a format provided as a parameter is actually supported.
 580          *
 581          * @param string $format The serialization format to check
 582          *
 583          * @throws MWException If the format is not supported by this content handler.
 584          */
 585         protected function checkFormat( $format ) {
 586                 if ( !$this->isSupportedFormat( $format ) ) {
 587                         throw new MWException(
 588                                 "Format $format is not supported for content model "
 589                                 . $this->getModelID()
 590                         );
 591                 }
 592         }
 593
 594         /**
 595          * Returns overrides for action handlers.
 596          * Classes listed here will be used instead of the default one when
 597          * (and only when) $wgActions[$action] === true. This allows subclasses
 598          * to override the default action handlers.
 599          *
 600          * @since 1.21
 601          *
 602          * @return array An array mapping action names (typically "view", "edit", "history" etc.) to
 603          *  either the full qualified class name of an Action class, a callable taking ( Page $page,
 604          *  IContextSource $context = null ) as parameters and returning an Action object, or an actual
 605          *  Action object. An empty array in this default implementation.
 606          *
 607          * @see Action::factory
 608          */
 609         public function getActionOverrides() {
 610                 return [];
 611         }
 612
 613         /**
 614          * Factory for creating an appropriate DifferenceEngine for this content model.
 615          *
 616          * @since 1.21
 617          *
 618          * @param IContextSource $context Context to use, anything else will be ignored.
 619          * @param int $old Revision ID we want to show and diff with.
 620          * @param int|string $new Either a revision ID or one of the strings 'cur', 'prev' or 'next'.
 621          * @param int $rcid FIXME: Deprecated, no longer used. Defaults to 0.
 622          * @param bool $refreshCache If set, refreshes the diff cache. Defaults to false.
 623          * @param bool $unhide If set, allow viewing deleted revs. Defaults to false.
 624          *
 625          * @return DifferenceEngine
 626          */
 627         public function createDifferenceEngine( IContextSource $context, $old = 0, $new = 0,
 628                 $rcid = 0, // FIXME: Deprecated, no longer used
 629                 $refreshCache = false, $unhide = false
 630         ) {
 631                 // hook: get difference engine
 632                 $differenceEngine = null;
 633                 if ( !Hooks::run( 'GetDifferenceEngine',
 634                         [ $context, $old, $new, $refreshCache, $unhide, &$differenceEngine ]
 635                 ) ) {
 636                         return $differenceEngine;
 637                 }
 638                 $diffEngineClass = $this->getDiffEngineClass();
 639                 return new $diffEngineClass( $context, $old, $new, $rcid, $refreshCache, $unhide );
 640         }
 641
 642         /**
 643          * Get the language in which the content of the given page is written.
 644          *
 645          * This default implementation just returns $wgContLang (except for pages
 646          * in the MediaWiki namespace)
 647          *
 648          * Note that the pages language is not cacheable, since it may in some
 649          * cases depend on user settings.
 650          *
 651          * Also note that the page language may or may not depend on the actual content of the page,
 652          * that is, this method may load the content in order to determine the language.
 653          *
 654          * @since 1.21
 655          *
 656          * @param Title $title The page to determine the language for.
 657          * @param Content $content The page's content, if you have it handy, to avoid reloading it.
 658          *
 659          * @return Language The page's language
 660          */
 661         public function getPageLanguage( Title $title, Content $content = null ) {
 662                 global $wgContLang, $wgLang;
 663                 $pageLang = $wgContLang;
 664
 665                 if ( $title->getNamespace() == NS_MEDIAWIKI ) {
 666                         // Parse mediawiki messages with correct target language
 667                         list( /* $unused */, $lang ) = MessageCache::singleton()->figureMessage( $title->getText() );
 668                         $pageLang = Language::factory( $lang );
 669                 }
 670
 671                 Hooks::run( 'PageContentLanguage', [ $title, &$pageLang, $wgLang ] );
 672
 673                 return wfGetLangObj( $pageLang );
 674         }
 675
 676         /**
 677          * Get the language in which the content of this page is written when
 678          * viewed by user. Defaults to $this->getPageLanguage(), but if the user
 679          * specified a preferred variant, the variant will be used.
 680          *
 681          * This default implementation just returns $this->getPageLanguage( $title, $content ) unless
 682          * the user specified a preferred variant.
 683          *
 684          * Note that the pages view language is not cacheable, since it depends on user settings.
 685          *
 686          * Also note that the page language may or may not depend on the actual content of the page,
 687          * that is, this method may load the content in order to determine the language.
 688          *
 689          * @since 1.21
 690          *
 691          * @param Title $title The page to determine the language for.
 692          * @param Content $content The page's content, if you have it handy, to avoid reloading it.
 693          *
 694          * @return Language The page's language for viewing
 695          */
 696         public function getPageViewLanguage( Title $title, Content $content = null ) {
 697                 $pageLang = $this->getPageLanguage( $title, $content );
 698
 699                 if ( $title->getNamespace() !== NS_MEDIAWIKI ) {
 700                         // If the user chooses a variant, the content is actually
 701                         // in a language whose code is the variant code.
 702                         $variant = $pageLang->getPreferredVariant();
 703                         if ( $pageLang->getCode() !== $variant ) {
 704                                 $pageLang = Language::factory( $variant );
 705                         }
 706                 }
 707
 708                 return $pageLang;
 709         }
 710
 711         /**
 712          * Determines whether the content type handled by this ContentHandler
 713          * can be used on the given page.
 714          *
 715          * This default implementation always returns true.
 716          * Subclasses may override this to restrict the use of this content model to specific locations,
 717          * typically based on the namespace or some other aspect of the title, such as a special suffix
 718          * (e.g. ".svg" for SVG content).
 719          *
 720          * @note this calls the ContentHandlerCanBeUsedOn hook which may be used to override which
 721          * content model can be used where.
 722          *
 723          * @param Title $title The page's title.
 724          *
 725          * @return bool True if content of this kind can be used on the given page, false otherwise.
 726          */
 727         public function canBeUsedOn( Title $title ) {
 728                 $ok = true;
 729
 730                 Hooks::run( 'ContentModelCanBeUsedOn', [ $this->getModelID(), $title, &$ok ] );
 731
 732                 return $ok;
 733         }
 734
 735         /**
 736          * Returns the name of the diff engine to use.
 737          *
 738          * @since 1.21
 739          *
 740          * @return string
 741          */
 742         protected function getDiffEngineClass() {
 743                 return DifferenceEngine::class;
 744         }
 745
 746         /**
 747          * Attempts to merge differences between three versions. Returns a new
 748          * Content object for a clean merge and false for failure or a conflict.
 749          *
 750          * This default implementation always returns false.
 751          *
 752          * @since 1.21
 753          *
 754          * @param Content $oldContent The page's previous content.
 755          * @param Content $myContent One of the page's conflicting contents.
 756          * @param Content $yourContent One of the page's conflicting contents.
 757          *
 758          * @return Content|bool Always false.
 759          */
 760         public function merge3( Content $oldContent, Content $myContent, Content $yourContent ) {
 761                 return false;
 762         }
 763
 764         /**
 765          * Return type of change if one exists for the given edit.
 766          *
 767          * @since 1.31
 768          *
 769          * @param Content|null $oldContent The previous text of the page.
 770          * @param Content|null $newContent The submitted text of the page.
 771          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 772          *
 773          * @return string|null String key representing type of change, or null.
 774          */
 775         private function getChangeType(
 776                 Content $oldContent = null,
 777                 Content $newContent = null,
 778                 $flags = 0
 779         ) {
 780                 $oldTarget = $oldContent !== null ? $oldContent->getRedirectTarget() : null;
 781                 $newTarget = $newContent !== null ? $newContent->getRedirectTarget() : null;
 782
 783                 // We check for the type of change in the given edit, and return string key accordingly
 784
 785                 // Blanking of a page
 786                 if ( $oldContent && $oldContent->getSize() > 0 &&
 787                         $newContent && $newContent->getSize() === 0
 788                 ) {
 789                         return 'blank';
 790                 }
 791
 792                 // Redirects
 793                 if ( $newTarget ) {
 794                         if ( !$oldTarget ) {
 795                                 // New redirect page (by creating new page or by changing content page)
 796                                 return 'new-redirect';
 797                         } elseif ( !$newTarget->equals( $oldTarget ) ||
 798                                 $oldTarget->getFragment() !== $newTarget->getFragment()
 799                         ) {
 800                                 // Redirect target changed
 801                                 return 'changed-redirect-target';
 802                         }
 803                 } elseif ( $oldTarget ) {
 804                         // Changing an existing redirect into a non-redirect
 805                         return 'removed-redirect';
 806                 }
 807
 808                 // New page created
 809                 if ( $flags & EDIT_NEW && $newContent ) {
 810                         if ( $newContent->getSize() === 0 ) {
 811                                 // New blank page
 812                                 return 'newblank';
 813                         } else {
 814                                 return 'newpage';
 815                         }
 816                 }
 817
 818                 // Removing more than 90% of the page
 819                 if ( $oldContent && $newContent && $oldContent->getSize() > 10 * $newContent->getSize() ) {
 820                         return 'replace';
 821                 }
 822
 823                 // Content model changed
 824                 if ( $oldContent && $newContent && $oldContent->getModel() !== $newContent->getModel() ) {
 825                         return 'contentmodelchange';
 826                 }
 827
 828                 return null;
 829         }
 830
 831         /**
 832          * Return an applicable auto-summary if one exists for the given edit.
 833          *
 834          * @since 1.21
 835          *
 836          * @param Content|null $oldContent The previous text of the page.
 837          * @param Content|null $newContent The submitted text of the page.
 838          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 839          *
 840          * @return string An appropriate auto-summary, or an empty string.
 841          */
 842         public function getAutosummary(
 843                 Content $oldContent = null,
 844                 Content $newContent = null,
 845                 $flags = 0
 846         ) {
 847                 $changeType = $this->getChangeType( $oldContent, $newContent, $flags );
 848
 849                 // There's no applicable auto-summary for our case, so our auto-summary is empty.
 850                 if ( !$changeType ) {
 851                         return '';
 852                 }
 853
 854                 // Decide what kind of auto-summary is needed.
 855                 switch ( $changeType ) {
 856                         case 'new-redirect':
 857                                 $newTarget = $newContent->getRedirectTarget();
 858                                 $truncatedtext = $newContent->getTextForSummary(
 859                                         250
 860                                         - strlen( wfMessage( 'autoredircomment' )->inContentLanguage()->text() )
 861                                         - strlen( $newTarget->getFullText() )
 862                                 );
 863
 864                                 return wfMessage( 'autoredircomment', $newTarget->getFullText() )
 865                                         ->plaintextParams( $truncatedtext )->inContentLanguage()->text();
 866                         case 'changed-redirect-target':
 867                                 $oldTarget = $oldContent->getRedirectTarget();
 868                                 $newTarget = $newContent->getRedirectTarget();
 869
 870                                 $truncatedtext = $newContent->getTextForSummary(
 871                                         250
 872                                         - strlen( wfMessage( 'autosumm-changed-redirect-target' )
 873                                                 ->inContentLanguage()->text() )
 874                                         - strlen( $oldTarget->getFullText() )
 875                                         - strlen( $newTarget->getFullText() )
 876                                 );
 877
 878                                 return wfMessage( 'autosumm-changed-redirect-target',
 879                                                 $oldTarget->getFullText(),
 880                                                 $newTarget->getFullText() )
 881                                         ->rawParams( $truncatedtext )->inContentLanguage()->text();
 882                         case 'removed-redirect':
 883                                 $oldTarget = $oldContent->getRedirectTarget();
 884                                 $truncatedtext = $newContent->getTextForSummary(
 885                                         250
 886                                         - strlen( wfMessage( 'autosumm-removed-redirect' )
 887                                                 ->inContentLanguage()->text() )
 888                                         - strlen( $oldTarget->getFullText() ) );
 889
 890                                 return wfMessage( 'autosumm-removed-redirect', $oldTarget->getFullText() )
 891                                         ->rawParams( $truncatedtext )->inContentLanguage()->text();
 892                         case 'newpage':
 893                                 // If they're making a new article, give its text, truncated, in the summary.
 894                                 $truncatedtext = $newContent->getTextForSummary(
 895                                         200 - strlen( wfMessage( 'autosumm-new' )->inContentLanguage()->text() ) );
 896
 897                                 return wfMessage( 'autosumm-new' )->rawParams( $truncatedtext )
 898                                         ->inContentLanguage()->text();
 899                         case 'blank':
 900                                 return wfMessage( 'autosumm-blank' )->inContentLanguage()->text();
 901                         case 'replace':
 902                                 $truncatedtext = $newContent->getTextForSummary(
 903                                         200 - strlen( wfMessage( 'autosumm-replace' )->inContentLanguage()->text() ) );
 904
 905                                 return wfMessage( 'autosumm-replace' )->rawParams( $truncatedtext )
 906                                         ->inContentLanguage()->text();
 907                         case 'newblank':
 908                                 return wfMessage( 'autosumm-newblank' )->inContentLanguage()->text();
 909                         default:
 910                                 return '';
 911                 }
 912         }
 913
 914         /**
 915          * Return an applicable tag if one exists for the given edit or return null.
 916          *
 917          * @since 1.31
 918          *
 919          * @param Content|null $oldContent The previous text of the page.
 920          * @param Content|null $newContent The submitted text of the page.
 921          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 922          *
 923          * @return string|null An appropriate tag, or null.
 924          */
 925         public function getChangeTag(
 926                 Content $oldContent = null,
 927                 Content $newContent = null,
 928                 $flags = 0
 929         ) {
 930                 $changeType = $this->getChangeType( $oldContent, $newContent, $flags );
 931
 932                 // There's no applicable tag for this change.
 933                 if ( !$changeType ) {
 934                         return null;
 935                 }
 936
 937                 // Core tags use the same keys as ones returned from $this->getChangeType()
 938                 // but prefixed with pseudo namespace 'mw-', so we add the prefix before checking
 939                 // if this type of change should be tagged
 940                 $tag = 'mw-' . $changeType;
 941
 942                 // Not all change types are tagged, so we check against the list of defined tags.
 943                 if ( in_array( $tag, ChangeTags::getSoftwareTags() ) ) {
 944                         return $tag;
 945                 }
 946
 947                 return null;
 948         }
 949
 950         /**
 951          * Auto-generates a deletion reason
 952          *
 953          * @since 1.21
 954          *
 955          * @param Title $title The page's title
 956          * @param bool &$hasHistory Whether the page has a history
 957          *
 958          * @return mixed String containing deletion reason or empty string, or
 959          *    boolean false if no revision occurred
 960          *
 961          * @todo &$hasHistory is extremely ugly, it's here because
 962          * WikiPage::getAutoDeleteReason() and Article::generateReason()
 963          * have it / want it.
 964          */
 965         public function getAutoDeleteReason( Title $title, &$hasHistory ) {
 966                 $dbr = wfGetDB( DB_REPLICA );
 967
 968                 // Get the last revision
 969                 $rev = Revision::newFromTitle( $title );
 970
 971                 if ( is_null( $rev ) ) {
 972                         return false;
 973                 }
 974
 975                 // Get the article's contents
 976                 $content = $rev->getContent();
 977                 $blank = false;
 978
 979                 // If the page is blank, use the text from the previous revision,
 980                 // which can only be blank if there's a move/import/protect dummy
 981                 // revision involved
 982                 if ( !$content || $content->isEmpty() ) {
 983                         $prev = $rev->getPrevious();
 984
 985                         if ( $prev ) {
 986                                 $rev = $prev;
 987                                 $content = $rev->getContent();
 988                                 $blank = true;
 989                         }
 990                 }
 991
 992                 $this->checkModelID( $rev->getContentModel() );
 993
 994                 // Find out if there was only one contributor
 995                 // Only scan the last 20 revisions
 996                 $revQuery = Revision::getQueryInfo();
 997                 $res = $dbr->select(
 998                         $revQuery['tables'],
 999                         [ 'rev_user_text' => $revQuery['fields']['rev_user_text'] ],
1000                         [
1001                                 'rev_page' => $title->getArticleID(),
1002                                 $dbr->bitAnd( 'rev_deleted', Revision::DELETED_USER ) . ' = 0'
1003                         ],
1004                         __METHOD__,
1005                         [ 'LIMIT' => 20 ],
1006                         $revQuery['joins']
1007                 );
1008
1009                 if ( $res === false ) {
1010                         // This page has no revisions, which is very weird
1011                         return false;
1012                 }
1013
1014                 $hasHistory = ( $res->numRows() > 1 );
1015                 $row = $dbr->fetchObject( $res );
1016
1017                 if ( $row ) { // $row is false if the only contributor is hidden
1018                         $onlyAuthor = $row->rev_user_text;
1019                         // Try to find a second contributor
1020                         foreach ( $res as $row ) {
1021                                 if ( $row->rev_user_text != $onlyAuthor ) { // T24999
1022                                         $onlyAuthor = false;
1023                                         break;
1024                                 }
1025                         }
1026                 } else {
1027                         $onlyAuthor = false;
1028                 }
1029
1030                 // Generate the summary with a '$1' placeholder
1031                 if ( $blank ) {
1032                         // The current revision is blank and the one before is also
1033                         // blank. It's just not our lucky day
1034                         $reason = wfMessage( 'exbeforeblank', '$1' )->inContentLanguage()->text();
1035                 } else {
1036                         if ( $onlyAuthor ) {
1037                                 $reason = wfMessage(
1038                                         'excontentauthor',
1039                                         '$1',
1040                                         $onlyAuthor
1041                                 )->inContentLanguage()->text();
1042                         } else {
1043                                 $reason = wfMessage( 'excontent', '$1' )->inContentLanguage()->text();
1044                         }
1045                 }
1046
1047                 if ( $reason == '-' ) {
1048                         // Allow these UI messages to be blanked out cleanly
1049                         return '';
1050                 }
1051
1052                 // Max content length = max comment length - length of the comment (excl. $1)
1053                 $text = $content ? $content->getTextForSummary( 255 - ( strlen( $reason ) - 2 ) ) : '';
1054
1055                 // Now replace the '$1' placeholder
1056                 $reason = str_replace( '$1', $text, $reason );
1057
1058                 return $reason;
1059         }
1060
1061         /**
1062          * Get the Content object that needs to be saved in order to undo all revisions
1063          * between $undo and $undoafter. Revisions must belong to the same page,
1064          * must exist and must not be deleted.
1065          *
1066          * @since 1.21
1067          *
1068          * @param Revision $current The current text
1069          * @param Revision $undo The revision to undo
1070          * @param Revision $undoafter Must be an earlier revision than $undo
1071          *
1072          * @return mixed String on success, false on failure
1073          */
1074         public function getUndoContent( Revision $current, Revision $undo, Revision $undoafter ) {
1075                 $cur_content = $current->getContent();
1076
1077                 if ( empty( $cur_content ) ) {
1078                         return false; // no page
1079                 }
1080
1081                 $undo_content = $undo->getContent();
1082                 $undoafter_content = $undoafter->getContent();
1083
1084                 if ( !$undo_content || !$undoafter_content ) {
1085                         return false; // no content to undo
1086                 }
1087
1088                 try {
1089                         $this->checkModelID( $cur_content->getModel() );
1090                         $this->checkModelID( $undo_content->getModel() );
1091                         if ( $current->getId() !== $undo->getId() ) {
1092                                 // If we are undoing the most recent revision,
1093                                 // its ok to revert content model changes. However
1094                                 // if we are undoing a revision in the middle, then
1095                                 // doing that will be confusing.
1096                                 $this->checkModelID( $undoafter_content->getModel() );
1097                         }
1098                 } catch ( MWException $e ) {
1099                         // If the revisions have different content models
1100                         // just return false
1101                         return false;
1102                 }
1103
1104                 if ( $cur_content->equals( $undo_content ) ) {
1105                         // No use doing a merge if it's just a straight revert.
1106                         return $undoafter_content;
1107                 }
1108
1109                 $undone_content = $this->merge3( $undo_content, $undoafter_content, $cur_content );
1110
1111                 return $undone_content;
1112         }
1113
1114         /**
1115          * Get parser options suitable for rendering and caching the article
1116          *
1117          * @param IContextSource|User|string $context One of the following:
1118          *        - IContextSource: Use the User and the Language of the provided
1119          *                                            context
1120          *        - User: Use the provided User object and $wgLang for the language,
1121          *                                            so use an IContextSource object if possible.
1122          *        - 'canonical': Canonical options (anonymous user with default
1123          *                                            preferences and content language).
1124          *
1125          * @throws MWException
1126          * @return ParserOptions
1127          */
1128         public function makeParserOptions( $context ) {
1129                 global $wgContLang;
1130
1131                 if ( $context instanceof IContextSource ) {
1132                         $user = $context->getUser();
1133                         $lang = $context->getLanguage();
1134                 } elseif ( $context instanceof User ) { // settings per user (even anons)
1135                         $user = $context;
1136                         $lang = null;
1137                 } elseif ( $context === 'canonical' ) { // canonical settings
1138                         $user = new User;
1139                         $lang = $wgContLang;
1140                 } else {
1141                         throw new MWException( "Bad context for parser options: $context" );
1142                 }
1143
1144                 return ParserOptions::newCanonical( $user, $lang );
1145         }
1146
1147         /**
1148          * Returns true for content models that support caching using the
1149          * ParserCache mechanism. See WikiPage::shouldCheckParserCache().
1150          *
1151          * @since 1.21
1152          *
1153          * @return bool Always false.
1154          */
1155         public function isParserCacheSupported() {
1156                 return false;
1157         }
1158
1159         /**
1160          * Returns true if this content model supports sections.
1161          * This default implementation returns false.
1162          *
1163          * Content models that return true here should also implement
1164          * Content::getSection, Content::replaceSection, etc. to handle sections..
1165          *
1166          * @return bool Always false.
1167          */
1168         public function supportsSections() {
1169                 return false;
1170         }
1171
1172         /**
1173          * Returns true if this content model supports categories.
1174          * The default implementation returns true.
1175          *
1176          * @return bool Always true.
1177          */
1178         public function supportsCategories() {
1179                 return true;
1180         }
1181
1182         /**
1183          * Returns true if this content model supports redirects.
1184          * This default implementation returns false.
1185          *
1186          * Content models that return true here should also implement
1187          * ContentHandler::makeRedirectContent to return a Content object.
1188          *
1189          * @return bool Always false.
1190          */
1191         public function supportsRedirects() {
1192                 return false;
1193         }
1194
1195         /**
1196          * Return true if this content model supports direct editing, such as via EditPage.
1197          *
1198          * @return bool Default is false, and true for TextContent and it's derivatives.
1199          */
1200         public function supportsDirectEditing() {
1201                 return false;
1202         }
1203
1204         /**
1205          * Whether or not this content model supports direct editing via ApiEditPage
1206          *
1207          * @return bool Default is false, and true for TextContent and derivatives.
1208          */
1209         public function supportsDirectApiEditing() {
1210                 return $this->supportsDirectEditing();
1211         }
1212
1213         /**
1214          * Get fields definition for search index
1215          *
1216          * @todo Expose title, redirect, namespace, text, source_text, text_bytes
1217          *       field mappings here. (see T142670 and T143409)
1218          *
1219          * @param SearchEngine $engine
1220          * @return SearchIndexField[] List of fields this content handler can provide.
1221          * @since 1.28
1222          */
1223         public function getFieldsForSearchIndex( SearchEngine $engine ) {
1224                 $fields['category'] = $engine->makeSearchFieldMapping(
1225                         'category',
1226                         SearchIndexField::INDEX_TYPE_TEXT
1227                 );
1228                 $fields['category']->setFlag( SearchIndexField::FLAG_CASEFOLD );
1229
1230                 $fields['external_link'] = $engine->makeSearchFieldMapping(
1231                         'external_link',
1232                         SearchIndexField::INDEX_TYPE_KEYWORD
1233                 );
1234
1235                 $fields['outgoing_link'] = $engine->makeSearchFieldMapping(
1236                         'outgoing_link',
1237                         SearchIndexField::INDEX_TYPE_KEYWORD
1238                 );
1239
1240                 $fields['template'] = $engine->makeSearchFieldMapping(
1241                         'template',
1242                         SearchIndexField::INDEX_TYPE_KEYWORD
1243                 );
1244                 $fields['template']->setFlag( SearchIndexField::FLAG_CASEFOLD );
1245
1246                 $fields['content_model'] = $engine->makeSearchFieldMapping(
1247                         'content_model',
1248                         SearchIndexField::INDEX_TYPE_KEYWORD
1249                 );
1250
1251                 return $fields;
1252         }
1253
1254         /**
1255          * Add new field definition to array.
1256          * @param SearchIndexField[] &$fields
1257          * @param SearchEngine $engine
1258          * @param string $name
1259          * @param int $type
1260          * @return SearchIndexField[] new field defs
1261          * @since 1.28
1262          */
1263         protected function addSearchField( &$fields, SearchEngine $engine, $name, $type ) {
1264                 $fields[$name] = $engine->makeSearchFieldMapping( $name, $type );
1265                 return $fields;
1266         }
1267
1268         /**
1269          * Return fields to be indexed by search engine
1270          * as representation of this document.
1271          * Overriding class should call parent function or take care of calling
1272          * the SearchDataForIndex hook.
1273          * @param WikiPage $page Page to index
1274          * @param ParserOutput $output
1275          * @param SearchEngine $engine Search engine for which we are indexing
1276          * @return array Map of name=>value for fields
1277          * @since 1.28
1278          */
1279         public function getDataForSearchIndex(
1280                 WikiPage $page,
1281                 ParserOutput $output,
1282                 SearchEngine $engine
1283         ) {
1284                 $fieldData = [];
1285                 $content = $page->getContent();
1286
1287                 if ( $content ) {
1288                         $searchDataExtractor = new ParserOutputSearchDataExtractor();
1289
1290                         $fieldData['category'] = $searchDataExtractor->getCategories( $output );
1291                         $fieldData['external_link'] = $searchDataExtractor->getExternalLinks( $output );
1292                         $fieldData['outgoing_link'] = $searchDataExtractor->getOutgoingLinks( $output );
1293                         $fieldData['template'] = $searchDataExtractor->getTemplates( $output );
1294
1295                         $text = $content->getTextForSearchIndex();
1296
1297                         $fieldData['text'] = $text;
1298                         $fieldData['source_text'] = $text;
1299                         $fieldData['text_bytes'] = $content->getSize();
1300                         $fieldData['content_model'] = $content->getModel();
1301                 }
1302
1303                 Hooks::run( 'SearchDataForIndex', [ &$fieldData, $this, $page, $output, $engine ] );
1304                 return $fieldData;
1305         }
1306
1307         /**
1308          * Produce page output suitable for indexing.
1309          *
1310          * Specific content handlers may override it if they need different content handling.
1311          *
1312          * @param WikiPage $page
1313          * @param ParserCache $cache
1314          * @return ParserOutput
1315          */
1316         public function getParserOutputForIndexing( WikiPage $page, ParserCache $cache = null ) {
1317                 $parserOptions = $page->makeParserOptions( 'canonical' );
1318                 $revId = $page->getRevision()->getId();
1319                 if ( $cache ) {
1320                         $parserOutput = $cache->get( $page, $parserOptions );
1321                 }
1322                 if ( empty( $parserOutput ) ) {
1323                         $parserOutput =
1324                                 $page->getContent()->getParserOutput( $page->getTitle(), $revId, $parserOptions );
1325                         if ( $cache ) {
1326                                 $cache->save( $parserOutput, $page, $parserOptions );
1327                         }
1328                 }
1329                 return $parserOutput;
1330         }
1331
1332 }