includes/content/ContentHandler.php

   1 <?php
   2 /**
   3  * Base class for content handling.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @since 1.21
  21  *
  22  * @file
  23  * @ingroup Content
  24  *
  25  * @author Daniel Kinzler
  26  */
  27
  28 /**
  29  * Exception representing a failure to serialize or unserialize a content object.
  30  *
  31  * @ingroup Content
  32  */
  33 class MWContentSerializationException extends MWException {
  34 }
  35
  36 /**
  37  * Exception thrown when an unregistered content model is requested. This error
  38  * can be triggered by user input, so a separate exception class is provided so
  39  * callers can substitute a context-specific, internationalised error message.
  40  *
  41  * @ingroup Content
  42  * @since 1.27
  43  */
  44 class MWUnknownContentModelException extends MWException {
  45         /** @var string The name of the unknown content model */
  46         private $modelId;
  47
  48         /** @param string $modelId */
  49         function __construct( $modelId ) {
  50                 parent::__construct( "The content model '$modelId' is not registered on this wiki.\n" .
  51                         'See https://www.mediawiki.org/wiki/Content_handlers to find out which extensions ' .
  52                         'handle this content model.' );
  53                 $this->modelId = $modelId;
  54         }
  55
  56         /** @return string */
  57         public function getModelId() {
  58                 return $this->modelId;
  59         }
  60 }
  61
  62 /**
  63  * A content handler knows how do deal with a specific type of content on a wiki
  64  * page. Content is stored in the database in a serialized form (using a
  65  * serialization format a.k.a. MIME type) and is unserialized into its native
  66  * PHP representation (the content model), which is wrapped in an instance of
  67  * the appropriate subclass of Content.
  68  *
  69  * ContentHandler instances are stateless singletons that serve, among other
  70  * things, as a factory for Content objects. Generally, there is one subclass
  71  * of ContentHandler and one subclass of Content for every type of content model.
  72  *
  73  * Some content types have a flat model, that is, their native representation
  74  * is the same as their serialized form. Examples would be JavaScript and CSS
  75  * code. As of now, this also applies to wikitext (MediaWiki's default content
  76  * type), but wikitext content may be represented by a DOM or AST structure in
  77  * the future.
  78  *
  79  * @ingroup Content
  80  */
  81 abstract class ContentHandler {
  82         /**
  83          * Switch for enabling deprecation warnings. Used by ContentHandler::deprecated()
  84          * and ContentHandler::runLegacyHooks().
  85          *
  86          * Once the ContentHandler code has settled in a bit, this should be set to true to
  87          * make extensions etc. show warnings when using deprecated functions and hooks.
  88          */
  89         protected static $enableDeprecationWarnings = false;
  90
  91         /**
  92          * Convenience function for getting flat text from a Content object. This
  93          * should only be used in the context of backwards compatibility with code
  94          * that is not yet able to handle Content objects!
  95          *
  96          * If $content is null, this method returns the empty string.
  97          *
  98          * If $content is an instance of TextContent, this method returns the flat
  99          * text as returned by $content->getNativeData().
 100          *
 101          * If $content is not a TextContent object, the behavior of this method
 102          * depends on the global $wgContentHandlerTextFallback:
 103          * - If $wgContentHandlerTextFallback is 'fail' and $content is not a
 104          *   TextContent object, an MWException is thrown.
 105          * - If $wgContentHandlerTextFallback is 'serialize' and $content is not a
 106          *   TextContent object, $content->serialize() is called to get a string
 107          *   form of the content.
 108          * - If $wgContentHandlerTextFallback is 'ignore' and $content is not a
 109          *   TextContent object, this method returns null.
 110          * - otherwise, the behavior is undefined.
 111          *
 112          * @since 1.21
 113          *
 114          * @param Content $content
 115          *
 116          * @throws MWException If the content is not an instance of TextContent and
 117          * wgContentHandlerTextFallback was set to 'fail'.
 118          * @return string|null Textual form of the content, if available.
 119          */
 120         public static function getContentText( Content $content = null ) {
 121                 global $wgContentHandlerTextFallback;
 122
 123                 if ( is_null( $content ) ) {
 124                         return '';
 125                 }
 126
 127                 if ( $content instanceof TextContent ) {
 128                         return $content->getNativeData();
 129                 }
 130
 131                 wfDebugLog( 'ContentHandler', 'Accessing ' . $content->getModel() . ' content as text!' );
 132
 133                 if ( $wgContentHandlerTextFallback == 'fail' ) {
 134                         throw new MWException(
 135                                 "Attempt to get text from Content with model " .
 136                                 $content->getModel()
 137                         );
 138                 }
 139
 140                 if ( $wgContentHandlerTextFallback == 'serialize' ) {
 141                         return $content->serialize();
 142                 }
 143
 144                 return null;
 145         }
 146
 147         /**
 148          * Convenience function for creating a Content object from a given textual
 149          * representation.
 150          *
 151          * $text will be deserialized into a Content object of the model specified
 152          * by $modelId (or, if that is not given, $title->getContentModel()) using
 153          * the given format.
 154          *
 155          * @since 1.21
 156          *
 157          * @param string $text The textual representation, will be
 158          *    unserialized to create the Content object
 159          * @param Title $title The title of the page this text belongs to.
 160          *    Required if $modelId is not provided.
 161          * @param string $modelId The model to deserialize to. If not provided,
 162          *    $title->getContentModel() is used.
 163          * @param string $format The format to use for deserialization. If not
 164          *    given, the model's default format is used.
 165          *
 166          * @throws MWException If model ID or format is not supported or if the text can not be
 167          * unserialized using the format.
 168          * @return Content A Content object representing the text.
 169          */
 170         public static function makeContent( $text, Title $title = null,
 171                 $modelId = null, $format = null ) {
 172                 if ( is_null( $modelId ) ) {
 173                         if ( is_null( $title ) ) {
 174                                 throw new MWException( "Must provide a Title object or a content model ID." );
 175                         }
 176
 177                         $modelId = $title->getContentModel();
 178                 }
 179
 180                 $handler = ContentHandler::getForModelID( $modelId );
 181
 182                 return $handler->unserializeContent( $text, $format );
 183         }
 184
 185         /**
 186          * Returns the name of the default content model to be used for the page
 187          * with the given title.
 188          *
 189          * Note: There should rarely be need to call this method directly.
 190          * To determine the actual content model for a given page, use
 191          * Title::getContentModel().
 192          *
 193          * Which model is to be used by default for the page is determined based
 194          * on several factors:
 195          * - The global setting $wgNamespaceContentModels specifies a content model
 196          *   per namespace.
 197          * - The hook ContentHandlerDefaultModelFor may be used to override the page's default
 198          *   model.
 199          * - Pages in NS_MEDIAWIKI and NS_USER default to the CSS or JavaScript
 200          *   model if they end in .js or .css, respectively.
 201          * - Pages in NS_MEDIAWIKI default to the wikitext model otherwise.
 202          * - The hook TitleIsCssOrJsPage may be used to force a page to use the CSS
 203          *   or JavaScript model. This is a compatibility feature. The ContentHandlerDefaultModelFor
 204          *   hook should be used instead if possible.
 205          * - The hook TitleIsWikitextPage may be used to force a page to use the
 206          *   wikitext model. This is a compatibility feature. The ContentHandlerDefaultModelFor
 207          *   hook should be used instead if possible.
 208          *
 209          * If none of the above applies, the wikitext model is used.
 210          *
 211          * Note: this is used by, and may thus not use, Title::getContentModel()
 212          *
 213          * @since 1.21
 214          *
 215          * @param Title $title
 216          *
 217          * @return string Default model name for the page given by $title
 218          */
 219         public static function getDefaultModelFor( Title $title ) {
 220                 // NOTE: this method must not rely on $title->getContentModel() directly or indirectly,
 221                 //       because it is used to initialize the mContentModel member.
 222
 223                 $ns = $title->getNamespace();
 224
 225                 $ext = false;
 226                 $m = null;
 227                 $model = MWNamespace::getNamespaceContentModel( $ns );
 228
 229                 // Hook can determine default model
 230                 if ( !Hooks::run( 'ContentHandlerDefaultModelFor', [ $title, &$model ] ) ) {
 231                         if ( !is_null( $model ) ) {
 232                                 return $model;
 233                         }
 234                 }
 235
 236                 // Could this page contain code based on the title?
 237                 $isCodePage = NS_MEDIAWIKI == $ns && preg_match( '!\.(css|js|json)$!u', $title->getText(), $m );
 238                 if ( $isCodePage ) {
 239                         $ext = $m[1];
 240                 }
 241
 242                 // Hook can force JS/CSS
 243                 Hooks::run( 'TitleIsCssOrJsPage', [ $title, &$isCodePage ], '1.25' );
 244
 245                 // Is this a user subpage containing code?
 246                 $isCodeSubpage = NS_USER == $ns
 247                         && !$isCodePage
 248                         && preg_match( "/\\/.*\\.(js|css|json)$/", $title->getText(), $m );
 249                 if ( $isCodeSubpage ) {
 250                         $ext = $m[1];
 251                 }
 252
 253                 // Is this wikitext, according to $wgNamespaceContentModels or the DefaultModelFor hook?
 254                 $isWikitext = is_null( $model ) || $model == CONTENT_MODEL_WIKITEXT;
 255                 $isWikitext = $isWikitext && !$isCodePage && !$isCodeSubpage;
 256
 257                 // Hook can override $isWikitext
 258                 Hooks::run( 'TitleIsWikitextPage', [ $title, &$isWikitext ], '1.25' );
 259
 260                 if ( !$isWikitext ) {
 261                         switch ( $ext ) {
 262                                 case 'js':
 263                                         return CONTENT_MODEL_JAVASCRIPT;
 264                                 case 'css':
 265                                         return CONTENT_MODEL_CSS;
 266                                 case 'json':
 267                                         return CONTENT_MODEL_JSON;
 268                                 default:
 269                                         return is_null( $model ) ? CONTENT_MODEL_TEXT : $model;
 270                         }
 271                 }
 272
 273                 // We established that it must be wikitext
 274
 275                 return CONTENT_MODEL_WIKITEXT;
 276         }
 277
 278         /**
 279          * Returns the appropriate ContentHandler singleton for the given title.
 280          *
 281          * @since 1.21
 282          *
 283          * @param Title $title
 284          *
 285          * @return ContentHandler
 286          */
 287         public static function getForTitle( Title $title ) {
 288                 $modelId = $title->getContentModel();
 289
 290                 return ContentHandler::getForModelID( $modelId );
 291         }
 292
 293         /**
 294          * Returns the appropriate ContentHandler singleton for the given Content
 295          * object.
 296          *
 297          * @since 1.21
 298          *
 299          * @param Content $content
 300          *
 301          * @return ContentHandler
 302          */
 303         public static function getForContent( Content $content ) {
 304                 $modelId = $content->getModel();
 305
 306                 return ContentHandler::getForModelID( $modelId );
 307         }
 308
 309         /**
 310          * @var array A Cache of ContentHandler instances by model id
 311          */
 312         protected static $handlers;
 313
 314         /**
 315          * Returns the ContentHandler singleton for the given model ID. Use the
 316          * CONTENT_MODEL_XXX constants to identify the desired content model.
 317          *
 318          * ContentHandler singletons are taken from the global $wgContentHandlers
 319          * array. Keys in that array are model names, the values are either
 320          * ContentHandler singleton objects, or strings specifying the appropriate
 321          * subclass of ContentHandler.
 322          *
 323          * If a class name is encountered when looking up the singleton for a given
 324          * model name, the class is instantiated and the class name is replaced by
 325          * the resulting singleton in $wgContentHandlers.
 326          *
 327          * If no ContentHandler is defined for the desired $modelId, the
 328          * ContentHandler may be provided by the ContentHandlerForModelID hook.
 329          * If no ContentHandler can be determined, an MWException is raised.
 330          *
 331          * @since 1.21
 332          *
 333          * @param string $modelId The ID of the content model for which to get a
 334          *    handler. Use CONTENT_MODEL_XXX constants.
 335          *
 336          * @throws MWException For internal errors and problems in the configuration.
 337          * @throws MWUnknownContentModelException If no handler is known for the model ID.
 338          * @return ContentHandler The ContentHandler singleton for handling the model given by the ID.
 339          */
 340         public static function getForModelID( $modelId ) {
 341                 global $wgContentHandlers;
 342
 343                 if ( isset( ContentHandler::$handlers[$modelId] ) ) {
 344                         return ContentHandler::$handlers[$modelId];
 345                 }
 346
 347                 if ( empty( $wgContentHandlers[$modelId] ) ) {
 348                         $handler = null;
 349
 350                         Hooks::run( 'ContentHandlerForModelID', [ $modelId, &$handler ] );
 351
 352                         if ( $handler === null ) {
 353                                 throw new MWUnknownContentModelException( $modelId );
 354                         }
 355
 356                         if ( !( $handler instanceof ContentHandler ) ) {
 357                                 throw new MWException( "ContentHandlerForModelID must supply a ContentHandler instance" );
 358                         }
 359                 } else {
 360                         $classOrCallback = $wgContentHandlers[$modelId];
 361
 362                         if ( is_callable( $classOrCallback ) ) {
 363                                 $handler = call_user_func( $classOrCallback, $modelId );
 364                         } else {
 365                                 $handler = new $classOrCallback( $modelId );
 366                         }
 367
 368                         if ( !( $handler instanceof ContentHandler ) ) {
 369                                 throw new MWException( "$classOrCallback from \$wgContentHandlers is not " .
 370                                         "compatible with ContentHandler" );
 371                         }
 372                 }
 373
 374                 wfDebugLog( 'ContentHandler', 'Created handler for ' . $modelId
 375                         . ': ' . get_class( $handler ) );
 376
 377                 ContentHandler::$handlers[$modelId] = $handler;
 378
 379                 return ContentHandler::$handlers[$modelId];
 380         }
 381
 382         /**
 383          * Returns the localized name for a given content model.
 384          *
 385          * Model names are localized using system messages. Message keys
 386          * have the form content-model-$name, where $name is getContentModelName( $id ).
 387          *
 388          * @param string $name The content model ID, as given by a CONTENT_MODEL_XXX
 389          *    constant or returned by Revision::getContentModel().
 390          * @param Language|null $lang The language to parse the message in (since 1.26)
 391          *
 392          * @throws MWException If the model ID isn't known.
 393          * @return string The content model's localized name.
 394          */
 395         public static function getLocalizedName( $name, Language $lang = null ) {
 396                 // Messages: content-model-wikitext, content-model-text,
 397                 // content-model-javascript, content-model-css
 398                 $key = "content-model-$name";
 399
 400                 $msg = wfMessage( $key );
 401                 if ( $lang ) {
 402                         $msg->inLanguage( $lang );
 403                 }
 404
 405                 return $msg->exists() ? $msg->plain() : $name;
 406         }
 407
 408         public static function getContentModels() {
 409                 global $wgContentHandlers;
 410
 411                 return array_keys( $wgContentHandlers );
 412         }
 413
 414         public static function getAllContentFormats() {
 415                 global $wgContentHandlers;
 416
 417                 $formats = [];
 418
 419                 foreach ( $wgContentHandlers as $model => $class ) {
 420                         $handler = ContentHandler::getForModelID( $model );
 421                         $formats = array_merge( $formats, $handler->getSupportedFormats() );
 422                 }
 423
 424                 $formats = array_unique( $formats );
 425
 426                 return $formats;
 427         }
 428
 429         // ------------------------------------------------------------------------
 430
 431         /**
 432          * @var string
 433          */
 434         protected $mModelID;
 435
 436         /**
 437          * @var string[]
 438          */
 439         protected $mSupportedFormats;
 440
 441         /**
 442          * Constructor, initializing the ContentHandler instance with its model ID
 443          * and a list of supported formats. Values for the parameters are typically
 444          * provided as literals by subclass's constructors.
 445          *
 446          * @param string $modelId (use CONTENT_MODEL_XXX constants).
 447          * @param string[] $formats List for supported serialization formats
 448          *    (typically as MIME types)
 449          */
 450         public function __construct( $modelId, $formats ) {
 451                 $this->mModelID = $modelId;
 452                 $this->mSupportedFormats = $formats;
 453
 454                 $this->mModelName = preg_replace( '/(Content)?Handler$/', '', get_class( $this ) );
 455                 $this->mModelName = preg_replace( '/[_\\\\]/', '', $this->mModelName );
 456                 $this->mModelName = strtolower( $this->mModelName );
 457         }
 458
 459         /**
 460          * Serializes a Content object of the type supported by this ContentHandler.
 461          *
 462          * @since 1.21
 463          *
 464          * @param Content $content The Content object to serialize
 465          * @param string $format The desired serialization format
 466          *
 467          * @return string Serialized form of the content
 468          */
 469         abstract public function serializeContent( Content $content, $format = null );
 470
 471         /**
 472          * Applies transformations on export (returns the blob unchanged per default).
 473          * Subclasses may override this to perform transformations such as conversion
 474          * of legacy formats or filtering of internal meta-data.
 475          *
 476          * @param string $blob The blob to be exported
 477          * @param string|null $format The blob's serialization format
 478          *
 479          * @return string
 480          */
 481         public function exportTransform( $blob, $format = null ) {
 482                 return $blob;
 483         }
 484
 485         /**
 486          * Unserializes a Content object of the type supported by this ContentHandler.
 487          *
 488          * @since 1.21
 489          *
 490          * @param string $blob Serialized form of the content
 491          * @param string $format The format used for serialization
 492          *
 493          * @return Content The Content object created by deserializing $blob
 494          */
 495         abstract public function unserializeContent( $blob, $format = null );
 496
 497         /**
 498          * Apply import transformation (per default, returns $blob unchanged).
 499          * This gives subclasses an opportunity to transform data blobs on import.
 500          *
 501          * @since 1.24
 502          *
 503          * @param string $blob
 504          * @param string|null $format
 505          *
 506          * @return string
 507          */
 508         public function importTransform( $blob, $format = null ) {
 509                 return $blob;
 510         }
 511
 512         /**
 513          * Creates an empty Content object of the type supported by this
 514          * ContentHandler.
 515          *
 516          * @since 1.21
 517          *
 518          * @return Content
 519          */
 520         abstract public function makeEmptyContent();
 521
 522         /**
 523          * Creates a new Content object that acts as a redirect to the given page,
 524          * or null if redirects are not supported by this content model.
 525          *
 526          * This default implementation always returns null. Subclasses supporting redirects
 527          * must override this method.
 528          *
 529          * Note that subclasses that override this method to return a Content object
 530          * should also override supportsRedirects() to return true.
 531          *
 532          * @since 1.21
 533          *
 534          * @param Title $destination The page to redirect to.
 535          * @param string $text Text to include in the redirect, if possible.
 536          *
 537          * @return Content Always null.
 538          */
 539         public function makeRedirectContent( Title $destination, $text = '' ) {
 540                 return null;
 541         }
 542
 543         /**
 544          * Returns the model id that identifies the content model this
 545          * ContentHandler can handle. Use with the CONTENT_MODEL_XXX constants.
 546          *
 547          * @since 1.21
 548          *
 549          * @return string The model ID
 550          */
 551         public function getModelID() {
 552                 return $this->mModelID;
 553         }
 554
 555         /**
 556          * @since 1.21
 557          *
 558          * @param string $model_id The model to check
 559          *
 560          * @throws MWException If the model ID is not the ID of the content model supported by this
 561          * ContentHandler.
 562          */
 563         protected function checkModelID( $model_id ) {
 564                 if ( $model_id !== $this->mModelID ) {
 565                         throw new MWException( "Bad content model: " .
 566                                 "expected {$this->mModelID} " .
 567                                 "but got $model_id." );
 568                 }
 569         }
 570
 571         /**
 572          * Returns a list of serialization formats supported by the
 573          * serializeContent() and unserializeContent() methods of this
 574          * ContentHandler.
 575          *
 576          * @since 1.21
 577          *
 578          * @return string[] List of serialization formats as MIME type like strings
 579          */
 580         public function getSupportedFormats() {
 581                 return $this->mSupportedFormats;
 582         }
 583
 584         /**
 585          * The format used for serialization/deserialization by default by this
 586          * ContentHandler.
 587          *
 588          * This default implementation will return the first element of the array
 589          * of formats that was passed to the constructor.
 590          *
 591          * @since 1.21
 592          *
 593          * @return string The name of the default serialization format as a MIME type
 594          */
 595         public function getDefaultFormat() {
 596                 return $this->mSupportedFormats[0];
 597         }
 598
 599         /**
 600          * Returns true if $format is a serialization format supported by this
 601          * ContentHandler, and false otherwise.
 602          *
 603          * Note that if $format is null, this method always returns true, because
 604          * null means "use the default format".
 605          *
 606          * @since 1.21
 607          *
 608          * @param string $format The serialization format to check
 609          *
 610          * @return bool
 611          */
 612         public function isSupportedFormat( $format ) {
 613                 if ( !$format ) {
 614                         return true; // this means "use the default"
 615                 }
 616
 617                 return in_array( $format, $this->mSupportedFormats );
 618         }
 619
 620         /**
 621          * Convenient for checking whether a format provided as a parameter is actually supported.
 622          *
 623          * @param string $format The serialization format to check
 624          *
 625          * @throws MWException If the format is not supported by this content handler.
 626          */
 627         protected function checkFormat( $format ) {
 628                 if ( !$this->isSupportedFormat( $format ) ) {
 629                         throw new MWException(
 630                                 "Format $format is not supported for content model "
 631                                 . $this->getModelID()
 632                         );
 633                 }
 634         }
 635
 636         /**
 637          * Returns overrides for action handlers.
 638          * Classes listed here will be used instead of the default one when
 639          * (and only when) $wgActions[$action] === true. This allows subclasses
 640          * to override the default action handlers.
 641          *
 642          * @since 1.21
 643          *
 644          * @return array Always an empty array.
 645          */
 646         public function getActionOverrides() {
 647                 return [];
 648         }
 649
 650         /**
 651          * Factory for creating an appropriate DifferenceEngine for this content model.
 652          *
 653          * @since 1.21
 654          *
 655          * @param IContextSource $context Context to use, anything else will be ignored.
 656          * @param int $old Revision ID we want to show and diff with.
 657          * @param int|string $new Either a revision ID or one of the strings 'cur', 'prev' or 'next'.
 658          * @param int $rcid FIXME: Deprecated, no longer used. Defaults to 0.
 659          * @param bool $refreshCache If set, refreshes the diff cache. Defaults to false.
 660          * @param bool $unhide If set, allow viewing deleted revs. Defaults to false.
 661          *
 662          * @return DifferenceEngine
 663          */
 664         public function createDifferenceEngine( IContextSource $context, $old = 0, $new = 0,
 665                 $rcid = 0, // FIXME: Deprecated, no longer used
 666                 $refreshCache = false, $unhide = false ) {
 667
 668                 // hook: get difference engine
 669                 $differenceEngine = null;
 670                 if ( !Hooks::run( 'GetDifferenceEngine',
 671                         [ $context, $old, $new, $refreshCache, $unhide, &$differenceEngine ]
 672                 ) ) {
 673                         return $differenceEngine;
 674                 }
 675                 $diffEngineClass = $this->getDiffEngineClass();
 676                 return new $diffEngineClass( $context, $old, $new, $rcid, $refreshCache, $unhide );
 677         }
 678
 679         /**
 680          * Get the language in which the content of the given page is written.
 681          *
 682          * This default implementation just returns $wgContLang (except for pages
 683          * in the MediaWiki namespace)
 684          *
 685          * Note that the pages language is not cacheable, since it may in some
 686          * cases depend on user settings.
 687          *
 688          * Also note that the page language may or may not depend on the actual content of the page,
 689          * that is, this method may load the content in order to determine the language.
 690          *
 691          * @since 1.21
 692          *
 693          * @param Title $title The page to determine the language for.
 694          * @param Content $content The page's content, if you have it handy, to avoid reloading it.
 695          *
 696          * @return Language The page's language
 697          */
 698         public function getPageLanguage( Title $title, Content $content = null ) {
 699                 global $wgContLang, $wgLang;
 700                 $pageLang = $wgContLang;
 701
 702                 if ( $title->getNamespace() == NS_MEDIAWIKI ) {
 703                         // Parse mediawiki messages with correct target language
 704                         list( /* $unused */, $lang ) = MessageCache::singleton()->figureMessage( $title->getText() );
 705                         $pageLang = wfGetLangObj( $lang );
 706                 }
 707
 708                 Hooks::run( 'PageContentLanguage', [ $title, &$pageLang, $wgLang ] );
 709
 710                 return wfGetLangObj( $pageLang );
 711         }
 712
 713         /**
 714          * Get the language in which the content of this page is written when
 715          * viewed by user. Defaults to $this->getPageLanguage(), but if the user
 716          * specified a preferred variant, the variant will be used.
 717          *
 718          * This default implementation just returns $this->getPageLanguage( $title, $content ) unless
 719          * the user specified a preferred variant.
 720          *
 721          * Note that the pages view language is not cacheable, since it depends on user settings.
 722          *
 723          * Also note that the page language may or may not depend on the actual content of the page,
 724          * that is, this method may load the content in order to determine the language.
 725          *
 726          * @since 1.21
 727          *
 728          * @param Title $title The page to determine the language for.
 729          * @param Content $content The page's content, if you have it handy, to avoid reloading it.
 730          *
 731          * @return Language The page's language for viewing
 732          */
 733         public function getPageViewLanguage( Title $title, Content $content = null ) {
 734                 $pageLang = $this->getPageLanguage( $title, $content );
 735
 736                 if ( $title->getNamespace() !== NS_MEDIAWIKI ) {
 737                         // If the user chooses a variant, the content is actually
 738                         // in a language whose code is the variant code.
 739                         $variant = $pageLang->getPreferredVariant();
 740                         if ( $pageLang->getCode() !== $variant ) {
 741                                 $pageLang = Language::factory( $variant );
 742                         }
 743                 }
 744
 745                 return $pageLang;
 746         }
 747
 748         /**
 749          * Determines whether the content type handled by this ContentHandler
 750          * can be used on the given page.
 751          *
 752          * This default implementation always returns true.
 753          * Subclasses may override this to restrict the use of this content model to specific locations,
 754          * typically based on the namespace or some other aspect of the title, such as a special suffix
 755          * (e.g. ".svg" for SVG content).
 756          *
 757          * @note this calls the ContentHandlerCanBeUsedOn hook which may be used to override which
 758          * content model can be used where.
 759          *
 760          * @param Title $title The page's title.
 761          *
 762          * @return bool True if content of this kind can be used on the given page, false otherwise.
 763          */
 764         public function canBeUsedOn( Title $title ) {
 765                 $ok = true;
 766
 767                 Hooks::run( 'ContentModelCanBeUsedOn', [ $this->getModelID(), $title, &$ok ] );
 768
 769                 return $ok;
 770         }
 771
 772         /**
 773          * Returns the name of the diff engine to use.
 774          *
 775          * @since 1.21
 776          *
 777          * @return string
 778          */
 779         protected function getDiffEngineClass() {
 780                 return DifferenceEngine::class;
 781         }
 782
 783         /**
 784          * Attempts to merge differences between three versions. Returns a new
 785          * Content object for a clean merge and false for failure or a conflict.
 786          *
 787          * This default implementation always returns false.
 788          *
 789          * @since 1.21
 790          *
 791          * @param Content $oldContent The page's previous content.
 792          * @param Content $myContent One of the page's conflicting contents.
 793          * @param Content $yourContent One of the page's conflicting contents.
 794          *
 795          * @return Content|bool Always false.
 796          */
 797         public function merge3( Content $oldContent, Content $myContent, Content $yourContent ) {
 798                 return false;
 799         }
 800
 801         /**
 802          * Return an applicable auto-summary if one exists for the given edit.
 803          *
 804          * @since 1.21
 805          *
 806          * @param Content $oldContent The previous text of the page.
 807          * @param Content $newContent The submitted text of the page.
 808          * @param int $flags Bit mask: a bit mask of flags submitted for the edit.
 809          *
 810          * @return string An appropriate auto-summary, or an empty string.
 811          */
 812         public function getAutosummary( Content $oldContent = null, Content $newContent = null,
 813                 $flags ) {
 814                 // Decide what kind of auto-summary is needed.
 815
 816                 // Redirect auto-summaries
 817
 818                 /**
 819                  * @var $ot Title
 820                  * @var $rt Title
 821                  */
 822
 823                 $ot = !is_null( $oldContent ) ? $oldContent->getRedirectTarget() : null;
 824                 $rt = !is_null( $newContent ) ? $newContent->getRedirectTarget() : null;
 825
 826                 if ( is_object( $rt ) ) {
 827                         if ( !is_object( $ot )
 828                                 || !$rt->equals( $ot )
 829                                 || $ot->getFragment() != $rt->getFragment()
 830                         ) {
 831                                 $truncatedtext = $newContent->getTextForSummary(
 832                                         250
 833                                         - strlen( wfMessage( 'autoredircomment' )->inContentLanguage()->text() )
 834                                         - strlen( $rt->getFullText() ) );
 835
 836                                 return wfMessage( 'autoredircomment', $rt->getFullText() )
 837                                         ->rawParams( $truncatedtext )->inContentLanguage()->text();
 838                         }
 839                 }
 840
 841                 // New page auto-summaries
 842                 if ( $flags & EDIT_NEW && $newContent->getSize() > 0 ) {
 843                         // If they're making a new article, give its text, truncated, in
 844                         // the summary.
 845
 846                         $truncatedtext = $newContent->getTextForSummary(
 847                                 200 - strlen( wfMessage( 'autosumm-new' )->inContentLanguage()->text() ) );
 848
 849                         return wfMessage( 'autosumm-new' )->rawParams( $truncatedtext )
 850                                 ->inContentLanguage()->text();
 851                 }
 852
 853                 // Blanking auto-summaries
 854                 if ( !empty( $oldContent ) && $oldContent->getSize() > 0 && $newContent->getSize() == 0 ) {
 855                         return wfMessage( 'autosumm-blank' )->inContentLanguage()->text();
 856                 } elseif ( !empty( $oldContent )
 857                         && $oldContent->getSize() > 10 * $newContent->getSize()
 858                         && $newContent->getSize() < 500
 859                 ) {
 860                         // Removing more than 90% of the article
 861
 862                         $truncatedtext = $newContent->getTextForSummary(
 863                                 200 - strlen( wfMessage( 'autosumm-replace' )->inContentLanguage()->text() ) );
 864
 865                         return wfMessage( 'autosumm-replace' )->rawParams( $truncatedtext )
 866                                 ->inContentLanguage()->text();
 867                 }
 868
 869                 // New blank article auto-summary
 870                 if ( $flags & EDIT_NEW && $newContent->isEmpty() ) {
 871                         return wfMessage( 'autosumm-newblank' )->inContentLanguage()->text();
 872                 }
 873
 874                 // If we reach this point, there's no applicable auto-summary for our
 875                 // case, so our auto-summary is empty.
 876                 return '';
 877         }
 878
 879         /**
 880          * Auto-generates a deletion reason
 881          *
 882          * @since 1.21
 883          *
 884          * @param Title $title The page's title
 885          * @param bool &$hasHistory Whether the page has a history
 886          *
 887          * @return mixed String containing deletion reason or empty string, or
 888          *    boolean false if no revision occurred
 889          *
 890          * @todo &$hasHistory is extremely ugly, it's here because
 891          * WikiPage::getAutoDeleteReason() and Article::generateReason()
 892          * have it / want it.
 893          */
 894         public function getAutoDeleteReason( Title $title, &$hasHistory ) {
 895                 $dbw = wfGetDB( DB_MASTER );
 896
 897                 // Get the last revision
 898                 $rev = Revision::newFromTitle( $title );
 899
 900                 if ( is_null( $rev ) ) {
 901                         return false;
 902                 }
 903
 904                 // Get the article's contents
 905                 $content = $rev->getContent();
 906                 $blank = false;
 907
 908                 // If the page is blank, use the text from the previous revision,
 909                 // which can only be blank if there's a move/import/protect dummy
 910                 // revision involved
 911                 if ( !$content || $content->isEmpty() ) {
 912                         $prev = $rev->getPrevious();
 913
 914                         if ( $prev ) {
 915                                 $rev = $prev;
 916                                 $content = $rev->getContent();
 917                                 $blank = true;
 918                         }
 919                 }
 920
 921                 $this->checkModelID( $rev->getContentModel() );
 922
 923                 // Find out if there was only one contributor
 924                 // Only scan the last 20 revisions
 925                 $res = $dbw->select( 'revision', 'rev_user_text',
 926                         [
 927                                 'rev_page' => $title->getArticleID(),
 928                                 $dbw->bitAnd( 'rev_deleted', Revision::DELETED_USER ) . ' = 0'
 929                         ],
 930                         __METHOD__,
 931                         [ 'LIMIT' => 20 ]
 932                 );
 933
 934                 if ( $res === false ) {
 935                         // This page has no revisions, which is very weird
 936                         return false;
 937                 }
 938
 939                 $hasHistory = ( $res->numRows() > 1 );
 940                 $row = $dbw->fetchObject( $res );
 941
 942                 if ( $row ) { // $row is false if the only contributor is hidden
 943                         $onlyAuthor = $row->rev_user_text;
 944                         // Try to find a second contributor
 945                         foreach ( $res as $row ) {
 946                                 if ( $row->rev_user_text != $onlyAuthor ) { // Bug 22999
 947                                         $onlyAuthor = false;
 948                                         break;
 949                                 }
 950                         }
 951                 } else {
 952                         $onlyAuthor = false;
 953                 }
 954
 955                 // Generate the summary with a '$1' placeholder
 956                 if ( $blank ) {
 957                         // The current revision is blank and the one before is also
 958                         // blank. It's just not our lucky day
 959                         $reason = wfMessage( 'exbeforeblank', '$1' )->inContentLanguage()->text();
 960                 } else {
 961                         if ( $onlyAuthor ) {
 962                                 $reason = wfMessage(
 963                                         'excontentauthor',
 964                                         '$1',
 965                                         $onlyAuthor
 966                                 )->inContentLanguage()->text();
 967                         } else {
 968                                 $reason = wfMessage( 'excontent', '$1' )->inContentLanguage()->text();
 969                         }
 970                 }
 971
 972                 if ( $reason == '-' ) {
 973                         // Allow these UI messages to be blanked out cleanly
 974                         return '';
 975                 }
 976
 977                 // Max content length = max comment length - length of the comment (excl. $1)
 978                 $text = $content ? $content->getTextForSummary( 255 - ( strlen( $reason ) - 2 ) ) : '';
 979
 980                 // Now replace the '$1' placeholder
 981                 $reason = str_replace( '$1', $text, $reason );
 982
 983                 return $reason;
 984         }
 985
 986         /**
 987          * Get the Content object that needs to be saved in order to undo all revisions
 988          * between $undo and $undoafter. Revisions must belong to the same page,
 989          * must exist and must not be deleted.
 990          *
 991          * @since 1.21
 992          *
 993          * @param Revision $current The current text
 994          * @param Revision $undo The revision to undo
 995          * @param Revision $undoafter Must be an earlier revision than $undo
 996          *
 997          * @return mixed String on success, false on failure
 998          */
 999         public function getUndoContent( Revision $current, Revision $undo, Revision $undoafter ) {
1000                 $cur_content = $current->getContent();
1001
1002                 if ( empty( $cur_content ) ) {
1003                         return false; // no page
1004                 }
1005
1006                 $undo_content = $undo->getContent();
1007                 $undoafter_content = $undoafter->getContent();
1008
1009                 if ( !$undo_content || !$undoafter_content ) {
1010                         return false; // no content to undo
1011                 }
1012
1013                 $this->checkModelID( $cur_content->getModel() );
1014                 $this->checkModelID( $undo_content->getModel() );
1015                 $this->checkModelID( $undoafter_content->getModel() );
1016
1017                 if ( $cur_content->equals( $undo_content ) ) {
1018                         // No use doing a merge if it's just a straight revert.
1019                         return $undoafter_content;
1020                 }
1021
1022                 $undone_content = $this->merge3( $undo_content, $undoafter_content, $cur_content );
1023
1024                 return $undone_content;
1025         }
1026
1027         /**
1028          * Get parser options suitable for rendering and caching the article
1029          *
1030          * @param IContextSource|User|string $context One of the following:
1031          *        - IContextSource: Use the User and the Language of the provided
1032          *                                            context
1033          *        - User: Use the provided User object and $wgLang for the language,
1034          *                                            so use an IContextSource object if possible.
1035          *        - 'canonical': Canonical options (anonymous user with default
1036          *                                            preferences and content language).
1037          *
1038          * @throws MWException
1039          * @return ParserOptions
1040          */
1041         public function makeParserOptions( $context ) {
1042                 global $wgContLang, $wgEnableParserLimitReporting;
1043
1044                 if ( $context instanceof IContextSource ) {
1045                         $options = ParserOptions::newFromContext( $context );
1046                 } elseif ( $context instanceof User ) { // settings per user (even anons)
1047                         $options = ParserOptions::newFromUser( $context );
1048                 } elseif ( $context === 'canonical' ) { // canonical settings
1049                         $options = ParserOptions::newFromUserAndLang( new User, $wgContLang );
1050                 } else {
1051                         throw new MWException( "Bad context for parser options: $context" );
1052                 }
1053
1054                 $options->enableLimitReport( $wgEnableParserLimitReporting ); // show inclusion/loop reports
1055                 $options->setTidy( true ); // fix bad HTML
1056
1057                 return $options;
1058         }
1059
1060         /**
1061          * Returns true for content models that support caching using the
1062          * ParserCache mechanism. See WikiPage::shouldCheckParserCache().
1063          *
1064          * @since 1.21
1065          *
1066          * @return bool Always false.
1067          */
1068         public function isParserCacheSupported() {
1069                 return false;
1070         }
1071
1072         /**
1073          * Returns true if this content model supports sections.
1074          * This default implementation returns false.
1075          *
1076          * Content models that return true here should also implement
1077          * Content::getSection, Content::replaceSection, etc. to handle sections..
1078          *
1079          * @return bool Always false.
1080          */
1081         public function supportsSections() {
1082                 return false;
1083         }
1084
1085         /**
1086          * Returns true if this content model supports categories.
1087          * The default implementation returns true.
1088          *
1089          * @return bool Always true.
1090          */
1091         public function supportsCategories() {
1092                 return true;
1093         }
1094
1095         /**
1096          * Returns true if this content model supports redirects.
1097          * This default implementation returns false.
1098          *
1099          * Content models that return true here should also implement
1100          * ContentHandler::makeRedirectContent to return a Content object.
1101          *
1102          * @return bool Always false.
1103          */
1104         public function supportsRedirects() {
1105                 return false;
1106         }
1107
1108         /**
1109          * Return true if this content model supports direct editing, such as via EditPage.
1110          *
1111          * @return bool Default is false, and true for TextContent and it's derivatives.
1112          */
1113         public function supportsDirectEditing() {
1114                 return false;
1115         }
1116
1117         /**
1118          * Whether or not this content model supports direct editing via ApiEditPage
1119          *
1120          * @return bool Default is false, and true for TextContent and derivatives.
1121          */
1122         public function supportsDirectApiEditing() {
1123                 return $this->supportsDirectEditing();
1124         }
1125
1126         /**
1127          * Logs a deprecation warning, visible if $wgDevelopmentWarnings, but only if
1128          * self::$enableDeprecationWarnings is set to true.
1129          *
1130          * @param string $func The name of the deprecated function
1131          * @param string $version The version since the method is deprecated. Usually 1.21
1132          *   for ContentHandler related stuff.
1133          * @param string|bool $component : Component to which the function belongs.
1134          *   If false, it is assumed the function is in MediaWiki core.
1135          *
1136          * @see ContentHandler::$enableDeprecationWarnings
1137          * @see wfDeprecated
1138          */
1139         public static function deprecated( $func, $version, $component = false ) {
1140                 if ( self::$enableDeprecationWarnings ) {
1141                         wfDeprecated( $func, $version, $component, 3 );
1142                 }
1143         }
1144
1145         /**
1146          * Call a legacy hook that uses text instead of Content objects.
1147          * Will log a warning when a matching hook function is registered.
1148          * If the textual representation of the content is changed by the
1149          * hook function, a new Content object is constructed from the new
1150          * text.
1151          *
1152          * @param string $event Event name
1153          * @param array $args Parameters passed to hook functions
1154          * @param bool $warn Whether to log a warning.
1155          *                    Default to self::$enableDeprecationWarnings.
1156          *                    May be set to false for testing.
1157          *
1158          * @return bool True if no handler aborted the hook
1159          *
1160          * @see ContentHandler::$enableDeprecationWarnings
1161          */
1162         public static function runLegacyHooks( $event, $args = [],
1163                 $warn = null
1164         ) {
1165
1166                 if ( $warn === null ) {
1167                         $warn = self::$enableDeprecationWarnings;
1168                 }
1169
1170                 if ( !Hooks::isRegistered( $event ) ) {
1171                         return true; // nothing to do here
1172                 }
1173
1174                 if ( $warn ) {
1175                         // Log information about which handlers are registered for the legacy hook,
1176                         // so we can find and fix them.
1177
1178                         $handlers = Hooks::getHandlers( $event );
1179                         $handlerInfo = [];
1180
1181                         MediaWiki\suppressWarnings();
1182
1183                         foreach ( $handlers as $handler ) {
1184                                 if ( is_array( $handler ) ) {
1185                                         if ( is_object( $handler[0] ) ) {
1186                                                 $info = get_class( $handler[0] );
1187                                         } else {
1188                                                 $info = $handler[0];
1189                                         }
1190
1191                                         if ( isset( $handler[1] ) ) {
1192                                                 $info .= '::' . $handler[1];
1193                                         }
1194                                 } elseif ( is_object( $handler ) ) {
1195                                         $info = get_class( $handler[0] );
1196                                         $info .= '::on' . $event;
1197                                 } else {
1198                                         $info = $handler;
1199                                 }
1200
1201                                 $handlerInfo[] = $info;
1202                         }
1203
1204                         MediaWiki\restoreWarnings();
1205
1206                         wfWarn( "Using obsolete hook $event via ContentHandler::runLegacyHooks()! Handlers: " .
1207                                 implode( ', ', $handlerInfo ), 2 );
1208                 }
1209
1210                 // convert Content objects to text
1211                 $contentObjects = [];
1212                 $contentTexts = [];
1213
1214                 foreach ( $args as $k => $v ) {
1215                         if ( $v instanceof Content ) {
1216                                 /* @var Content $v */
1217
1218                                 $contentObjects[$k] = $v;
1219
1220                                 $v = $v->serialize();
1221                                 $contentTexts[$k] = $v;
1222                                 $args[$k] = $v;
1223                         }
1224                 }
1225
1226                 // call the hook functions
1227                 $ok = Hooks::run( $event, $args );
1228
1229                 // see if the hook changed the text
1230                 foreach ( $contentTexts as $k => $orig ) {
1231                         /* @var Content $content */
1232
1233                         $modified = $args[$k];
1234                         $content = $contentObjects[$k];
1235
1236                         if ( $modified !== $orig ) {
1237                                 // text was changed, create updated Content object
1238                                 $content = $content->getContentHandler()->unserializeContent( $modified );
1239                         }
1240
1241                         $args[$k] = $content;
1242                 }
1243
1244                 return $ok;
1245         }
1246 }