includes/cache/localisation/LocalisationCache.php

   1 <?php
   2 /**
   3  * Cache of the contents of localisation files.
   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  * @file
  21  */
  22
  23 use CLDRPluralRuleParser\Evaluator;
  24 use CLDRPluralRuleParser\Error as CLDRPluralRuleError;
  25 use MediaWiki\Logger\LoggerFactory;
  26 use MediaWiki\MediaWikiServices;
  27
  28 /**
  29  * Class for caching the contents of localisation files, Messages*.php
  30  * and *.i18n.php.
  31  *
  32  * An instance of this class is available using Language::getLocalisationCache().
  33  *
  34  * The values retrieved from here are merged, containing items from extension
  35  * files, core messages files and the language fallback sequence (e.g. zh-cn ->
  36  * zh-hans -> en ). Some common errors are corrected, for example namespace
  37  * names with spaces instead of underscores, but heavyweight processing, such
  38  * as grammatical transformation, is done by the caller.
  39  */
  40 class LocalisationCache {
  41         const VERSION = 4;
  42
  43         /** Configuration associative array */
  44         private $conf;
  45
  46         /**
  47          * True if recaching should only be done on an explicit call to recache().
  48          * Setting this reduces the overhead of cache freshness checking, which
  49          * requires doing a stat() for every extension i18n file.
  50          */
  51         private $manualRecache = false;
  52
  53         /**
  54          * True to treat all files as expired until they are regenerated by this object.
  55          */
  56         private $forceRecache = false;
  57
  58         /**
  59          * The cache data. 3-d array, where the first key is the language code,
  60          * the second key is the item key e.g. 'messages', and the third key is
  61          * an item specific subkey index. Some items are not arrays and so for those
  62          * items, there are no subkeys.
  63          */
  64         protected $data = [];
  65
  66         /**
  67          * The persistent store object. An instance of LCStore.
  68          *
  69          * @var LCStore
  70          */
  71         private $store;
  72
  73         /**
  74          * @var \Psr\Log\LoggerInterface
  75          */
  76         private $logger;
  77
  78         /**
  79          * A 2-d associative array, code/key, where presence indicates that the item
  80          * is loaded. Value arbitrary.
  81          *
  82          * For split items, if set, this indicates that all of the subitems have been
  83          * loaded.
  84          */
  85         private $loadedItems = [];
  86
  87         /**
  88          * A 3-d associative array, code/key/subkey, where presence indicates that
  89          * the subitem is loaded. Only used for the split items, i.e. messages.
  90          */
  91         private $loadedSubitems = [];
  92
  93         /**
  94          * An array where presence of a key indicates that that language has been
  95          * initialised. Initialisation includes checking for cache expiry and doing
  96          * any necessary updates.
  97          */
  98         private $initialisedLangs = [];
  99
 100         /**
 101          * An array mapping non-existent pseudo-languages to fallback languages. This
 102          * is filled by initShallowFallback() when data is requested from a language
 103          * that lacks a Messages*.php file.
 104          */
 105         private $shallowFallbacks = [];
 106
 107         /**
 108          * An array where the keys are codes that have been recached by this instance.
 109          */
 110         private $recachedLangs = [];
 111
 112         /**
 113          * All item keys
 114          */
 115         public static $allKeys = [
 116                 'fallback', 'namespaceNames', 'bookstoreList',
 117                 'magicWords', 'messages', 'rtl', 'capitalizeAllNouns',
 118                 'digitTransformTable', 'separatorTransformTable',
 119                 'minimumGroupingDigits', 'fallback8bitEncoding',
 120                 'linkPrefixExtension', 'linkTrail', 'linkPrefixCharset',
 121                 'namespaceAliases', 'dateFormats', 'datePreferences',
 122                 'datePreferenceMigrationMap', 'defaultDateFormat',
 123                 'specialPageAliases', 'imageFiles', 'preloadedMessages',
 124                 'namespaceGenderAliases', 'digitGroupingPattern', 'pluralRules',
 125                 'pluralRuleTypes', 'compiledPluralRules',
 126         ];
 127
 128         /**
 129          * Keys for items which consist of associative arrays, which may be merged
 130          * by a fallback sequence.
 131          */
 132         public static $mergeableMapKeys = [ 'messages', 'namespaceNames',
 133                 'namespaceAliases', 'dateFormats', 'imageFiles', 'preloadedMessages'
 134         ];
 135
 136         /**
 137          * Keys for items which are a numbered array.
 138          */
 139         public static $mergeableListKeys = [];
 140
 141         /**
 142          * Keys for items which contain an array of arrays of equivalent aliases
 143          * for each subitem. The aliases may be merged by a fallback sequence.
 144          */
 145         public static $mergeableAliasListKeys = [ 'specialPageAliases' ];
 146
 147         /**
 148          * Keys for items which contain an associative array, and may be merged if
 149          * the primary value contains the special array key "inherit". That array
 150          * key is removed after the first merge.
 151          */
 152         public static $optionalMergeKeys = [ 'bookstoreList' ];
 153
 154         /**
 155          * Keys for items that are formatted like $magicWords
 156          */
 157         public static $magicWordKeys = [ 'magicWords' ];
 158
 159         /**
 160          * Keys for items where the subitems are stored in the backend separately.
 161          */
 162         public static $splitKeys = [ 'messages' ];
 163
 164         /**
 165          * Keys which are loaded automatically by initLanguage()
 166          */
 167         public static $preloadedKeys = [ 'dateFormats', 'namespaceNames' ];
 168
 169         /**
 170          * Associative array of cached plural rules. The key is the language code,
 171          * the value is an array of plural rules for that language.
 172          */
 173         private $pluralRules = null;
 174
 175         /**
 176          * Associative array of cached plural rule types. The key is the language
 177          * code, the value is an array of plural rule types for that language. For
 178          * example, $pluralRuleTypes['ar'] = ['zero', 'one', 'two', 'few', 'many'].
 179          * The index for each rule type matches the index for the rule in
 180          * $pluralRules, thus allowing correlation between the two. The reason we
 181          * don't just use the type names as the keys in $pluralRules is because
 182          * Language::convertPlural applies the rules based on numeric order (or
 183          * explicit numeric parameter), not based on the name of the rule type. For
 184          * example, {{plural:count|wordform1|wordform2|wordform3}}, rather than
 185          * {{plural:count|one=wordform1|two=wordform2|many=wordform3}}.
 186          */
 187         private $pluralRuleTypes = null;
 188
 189         private $mergeableKeys = null;
 190
 191         /**
 192          * For constructor parameters, see the documentation in DefaultSettings.php
 193          * for $wgLocalisationCacheConf.
 194          *
 195          * @param array $conf
 196          * @throws MWException
 197          */
 198         function __construct( $conf ) {
 199                 global $wgCacheDirectory;
 200
 201                 $this->conf = $conf;
 202                 $this->logger = LoggerFactory::getInstance( 'localisation' );
 203
 204                 $directory = !empty( $conf['storeDirectory'] ) ? $conf['storeDirectory'] : $wgCacheDirectory;
 205                 $storeArg = [];
 206                 $storeArg['directory'] = $directory;
 207
 208                 if ( !empty( $conf['storeClass'] ) ) {
 209                         $storeClass = $conf['storeClass'];
 210                 } else {
 211                         switch ( $conf['store'] ) {
 212                                 case 'files':
 213                                 case 'file':
 214                                         $storeClass = LCStoreCDB::class;
 215                                         break;
 216                                 case 'db':
 217                                         $storeClass = LCStoreDB::class;
 218                                         $storeArg['server'] = $conf['storeServer'] ?? [];
 219                                         break;
 220                                 case 'array':
 221                                         $storeClass = LCStoreStaticArray::class;
 222                                         break;
 223                                 case 'detect':
 224                                         if ( $directory ) {
 225                                                 $storeClass = LCStoreCDB::class;
 226                                         } else {
 227                                                 $storeClass = LCStoreDB::class;
 228                                                 $storeArg['server'] = $conf['storeServer'] ?? [];
 229                                         }
 230                                         break;
 231                                 default:
 232                                         throw new MWException(
 233                                                 'Please set $wgLocalisationCacheConf[\'store\'] to something sensible.'
 234                                         );
 235                         }
 236                 }
 237                 $this->logger->debug( static::class . ": using store $storeClass" );
 238
 239                 $this->store = new $storeClass( $storeArg );
 240                 foreach ( [ 'manualRecache', 'forceRecache' ] as $var ) {
 241                         if ( isset( $conf[$var] ) ) {
 242                                 $this->$var = $conf[$var];
 243                         }
 244                 }
 245         }
 246
 247         /**
 248          * Returns true if the given key is mergeable, that is, if it is an associative
 249          * array which can be merged through a fallback sequence.
 250          * @param string $key
 251          * @return bool
 252          */
 253         public function isMergeableKey( $key ) {
 254                 if ( $this->mergeableKeys === null ) {
 255                         $this->mergeableKeys = array_flip( array_merge(
 256                                 self::$mergeableMapKeys,
 257                                 self::$mergeableListKeys,
 258                                 self::$mergeableAliasListKeys,
 259                                 self::$optionalMergeKeys,
 260                                 self::$magicWordKeys
 261                         ) );
 262                 }
 263
 264                 return isset( $this->mergeableKeys[$key] );
 265         }
 266
 267         /**
 268          * Get a cache item.
 269          *
 270          * Warning: this may be slow for split items (messages), since it will
 271          * need to fetch all of the subitems from the cache individually.
 272          * @param string $code
 273          * @param string $key
 274          * @return mixed
 275          */
 276         public function getItem( $code, $key ) {
 277                 if ( !isset( $this->loadedItems[$code][$key] ) ) {
 278                         $this->loadItem( $code, $key );
 279                 }
 280
 281                 if ( $key === 'fallback' && isset( $this->shallowFallbacks[$code] ) ) {
 282                         return $this->shallowFallbacks[$code];
 283                 }
 284
 285                 return $this->data[$code][$key];
 286         }
 287
 288         /**
 289          * Get a subitem, for instance a single message for a given language.
 290          * @param string $code
 291          * @param string $key
 292          * @param string $subkey
 293          * @return mixed|null
 294          */
 295         public function getSubitem( $code, $key, $subkey ) {
 296                 if ( !isset( $this->loadedSubitems[$code][$key][$subkey] ) &&
 297                         !isset( $this->loadedItems[$code][$key] )
 298                 ) {
 299                         $this->loadSubitem( $code, $key, $subkey );
 300                 }
 301
 302                 return $this->data[$code][$key][$subkey] ?? null;
 303         }
 304
 305         /**
 306          * Get the list of subitem keys for a given item.
 307          *
 308          * This is faster than array_keys($lc->getItem(...)) for the items listed in
 309          * self::$splitKeys.
 310          *
 311          * Will return null if the item is not found, or false if the item is not an
 312          * array.
 313          * @param string $code
 314          * @param string $key
 315          * @return bool|null|string|string[]
 316          */
 317         public function getSubitemList( $code, $key ) {
 318                 if ( in_array( $key, self::$splitKeys ) ) {
 319                         return $this->getSubitem( $code, 'list', $key );
 320                 } else {
 321                         $item = $this->getItem( $code, $key );
 322                         if ( is_array( $item ) ) {
 323                                 return array_keys( $item );
 324                         } else {
 325                                 return false;
 326                         }
 327                 }
 328         }
 329
 330         /**
 331          * Load an item into the cache.
 332          * @param string $code
 333          * @param string $key
 334          */
 335         protected function loadItem( $code, $key ) {
 336                 if ( !isset( $this->initialisedLangs[$code] ) ) {
 337                         $this->initLanguage( $code );
 338                 }
 339
 340                 // Check to see if initLanguage() loaded it for us
 341                 if ( isset( $this->loadedItems[$code][$key] ) ) {
 342                         return;
 343                 }
 344
 345                 if ( isset( $this->shallowFallbacks[$code] ) ) {
 346                         $this->loadItem( $this->shallowFallbacks[$code], $key );
 347
 348                         return;
 349                 }
 350
 351                 if ( in_array( $key, self::$splitKeys ) ) {
 352                         $subkeyList = $this->getSubitem( $code, 'list', $key );
 353                         foreach ( $subkeyList as $subkey ) {
 354                                 if ( isset( $this->data[$code][$key][$subkey] ) ) {
 355                                         continue;
 356                                 }
 357                                 $this->data[$code][$key][$subkey] = $this->getSubitem( $code, $key, $subkey );
 358                         }
 359                 } else {
 360                         $this->data[$code][$key] = $this->store->get( $code, $key );
 361                 }
 362
 363                 $this->loadedItems[$code][$key] = true;
 364         }
 365
 366         /**
 367          * Load a subitem into the cache
 368          * @param string $code
 369          * @param string $key
 370          * @param string $subkey
 371          */
 372         protected function loadSubitem( $code, $key, $subkey ) {
 373                 if ( !in_array( $key, self::$splitKeys ) ) {
 374                         $this->loadItem( $code, $key );
 375
 376                         return;
 377                 }
 378
 379                 if ( !isset( $this->initialisedLangs[$code] ) ) {
 380                         $this->initLanguage( $code );
 381                 }
 382
 383                 // Check to see if initLanguage() loaded it for us
 384                 if ( isset( $this->loadedItems[$code][$key] ) ||
 385                         isset( $this->loadedSubitems[$code][$key][$subkey] )
 386                 ) {
 387                         return;
 388                 }
 389
 390                 if ( isset( $this->shallowFallbacks[$code] ) ) {
 391                         $this->loadSubitem( $this->shallowFallbacks[$code], $key, $subkey );
 392
 393                         return;
 394                 }
 395
 396                 $value = $this->store->get( $code, "$key:$subkey" );
 397                 $this->data[$code][$key][$subkey] = $value;
 398                 $this->loadedSubitems[$code][$key][$subkey] = true;
 399         }
 400
 401         /**
 402          * Returns true if the cache identified by $code is missing or expired.
 403          *
 404          * @param string $code
 405          *
 406          * @return bool
 407          */
 408         public function isExpired( $code ) {
 409                 if ( $this->forceRecache && !isset( $this->recachedLangs[$code] ) ) {
 410                         $this->logger->debug( __METHOD__ . "($code): forced reload" );
 411
 412                         return true;
 413                 }
 414
 415                 $deps = $this->store->get( $code, 'deps' );
 416                 $keys = $this->store->get( $code, 'list' );
 417                 $preload = $this->store->get( $code, 'preload' );
 418                 // Different keys may expire separately for some stores
 419                 if ( $deps === null || $keys === null || $preload === null ) {
 420                         $this->logger->debug( __METHOD__ . "($code): cache missing, need to make one" );
 421
 422                         return true;
 423                 }
 424
 425                 foreach ( $deps as $dep ) {
 426                         // Because we're unserializing stuff from cache, we
 427                         // could receive objects of classes that don't exist
 428                         // anymore (e.g. uninstalled extensions)
 429                         // When this happens, always expire the cache
 430                         if ( !$dep instanceof CacheDependency || $dep->isExpired() ) {
 431                                 $this->logger->debug( __METHOD__ . "($code): cache for $code expired due to " .
 432                                         get_class( $dep ) );
 433
 434                                 return true;
 435                         }
 436                 }
 437
 438                 return false;
 439         }
 440
 441         /**
 442          * Initialise a language in this object. Rebuild the cache if necessary.
 443          * @param string $code
 444          * @throws MWException
 445          */
 446         protected function initLanguage( $code ) {
 447                 if ( isset( $this->initialisedLangs[$code] ) ) {
 448                         return;
 449                 }
 450
 451                 $this->initialisedLangs[$code] = true;
 452
 453                 # If the code is of the wrong form for a Messages*.php file, do a shallow fallback
 454                 if ( !Language::isValidBuiltInCode( $code ) ) {
 455                         $this->initShallowFallback( $code, 'en' );
 456
 457                         return;
 458                 }
 459
 460                 # Recache the data if necessary
 461                 if ( !$this->manualRecache && $this->isExpired( $code ) ) {
 462                         if ( Language::isSupportedLanguage( $code ) ) {
 463                                 $this->recache( $code );
 464                         } elseif ( $code === 'en' ) {
 465                                 throw new MWException( 'MessagesEn.php is missing.' );
 466                         } else {
 467                                 $this->initShallowFallback( $code, 'en' );
 468                         }
 469
 470                         return;
 471                 }
 472
 473                 # Preload some stuff
 474                 $preload = $this->getItem( $code, 'preload' );
 475                 if ( $preload === null ) {
 476                         if ( $this->manualRecache ) {
 477                                 // No Messages*.php file. Do shallow fallback to en.
 478                                 if ( $code === 'en' ) {
 479                                         throw new MWException( 'No localisation cache found for English. ' .
 480                                                 'Please run maintenance/rebuildLocalisationCache.php.' );
 481                                 }
 482                                 $this->initShallowFallback( $code, 'en' );
 483
 484                                 return;
 485                         } else {
 486                                 throw new MWException( 'Invalid or missing localisation cache.' );
 487                         }
 488                 }
 489                 $this->data[$code] = $preload;
 490                 foreach ( $preload as $key => $item ) {
 491                         if ( in_array( $key, self::$splitKeys ) ) {
 492                                 foreach ( $item as $subkey => $subitem ) {
 493                                         $this->loadedSubitems[$code][$key][$subkey] = true;
 494                                 }
 495                         } else {
 496                                 $this->loadedItems[$code][$key] = true;
 497                         }
 498                 }
 499         }
 500
 501         /**
 502          * Create a fallback from one language to another, without creating a
 503          * complete persistent cache.
 504          * @param string $primaryCode
 505          * @param string $fallbackCode
 506          */
 507         public function initShallowFallback( $primaryCode, $fallbackCode ) {
 508                 $this->data[$primaryCode] =& $this->data[$fallbackCode];
 509                 $this->loadedItems[$primaryCode] =& $this->loadedItems[$fallbackCode];
 510                 $this->loadedSubitems[$primaryCode] =& $this->loadedSubitems[$fallbackCode];
 511                 $this->shallowFallbacks[$primaryCode] = $fallbackCode;
 512         }
 513
 514         /**
 515          * Read a PHP file containing localisation data.
 516          * @param string $_fileName
 517          * @param string $_fileType
 518          * @throws MWException
 519          * @return array
 520          */
 521         protected function readPHPFile( $_fileName, $_fileType ) {
 522                 include $_fileName;
 523
 524                 $data = [];
 525                 if ( $_fileType == 'core' || $_fileType == 'extension' ) {
 526                         foreach ( self::$allKeys as $key ) {
 527                                 // Not all keys are set in language files, so
 528                                 // check they exist first
 529                                 if ( isset( $$key ) ) {
 530                                         $data[$key] = $$key;
 531                                 }
 532                         }
 533                 } elseif ( $_fileType == 'aliases' ) {
 534                         if ( isset( $aliases ) ) {
 535                                 $data['aliases'] = $aliases;
 536                         }
 537                 } else {
 538                         throw new MWException( __METHOD__ . ": Invalid file type: $_fileType" );
 539                 }
 540
 541                 return $data;
 542         }
 543
 544         /**
 545          * Read a JSON file containing localisation messages.
 546          * @param string $fileName Name of file to read
 547          * @throws MWException If there is a syntax error in the JSON file
 548          * @return array Array with a 'messages' key, or empty array if the file doesn't exist
 549          */
 550         public function readJSONFile( $fileName ) {
 551                 if ( !is_readable( $fileName ) ) {
 552                         return [];
 553                 }
 554
 555                 $json = file_get_contents( $fileName );
 556                 if ( $json === false ) {
 557                         return [];
 558                 }
 559
 560                 $data = FormatJson::decode( $json, true );
 561                 if ( $data === null ) {
 562                         throw new MWException( __METHOD__ . ": Invalid JSON file: $fileName" );
 563                 }
 564
 565                 // Remove keys starting with '@', they're reserved for metadata and non-message data
 566                 foreach ( $data as $key => $unused ) {
 567                         if ( $key === '' || $key[0] === '@' ) {
 568                                 unset( $data[$key] );
 569                         }
 570                 }
 571
 572                 // The JSON format only supports messages, none of the other variables, so wrap the data
 573                 return [ 'messages' => $data ];
 574         }
 575
 576         /**
 577          * Get the compiled plural rules for a given language from the XML files.
 578          * @since 1.20
 579          * @param string $code
 580          * @return array|null
 581          */
 582         public function getCompiledPluralRules( $code ) {
 583                 $rules = $this->getPluralRules( $code );
 584                 if ( $rules === null ) {
 585                         return null;
 586                 }
 587                 try {
 588                         $compiledRules = Evaluator::compile( $rules );
 589                 } catch ( CLDRPluralRuleError $e ) {
 590                         $this->logger->debug( $e->getMessage() );
 591
 592                         return [];
 593                 }
 594
 595                 return $compiledRules;
 596         }
 597
 598         /**
 599          * Get the plural rules for a given language from the XML files.
 600          * Cached.
 601          * @since 1.20
 602          * @param string $code
 603          * @return array|null
 604          */
 605         public function getPluralRules( $code ) {
 606                 if ( $this->pluralRules === null ) {
 607                         $this->loadPluralFiles();
 608                 }
 609                 return $this->pluralRules[$code] ?? null;
 610         }
 611
 612         /**
 613          * Get the plural rule types for a given language from the XML files.
 614          * Cached.
 615          * @since 1.22
 616          * @param string $code
 617          * @return array|null
 618          */
 619         public function getPluralRuleTypes( $code ) {
 620                 if ( $this->pluralRuleTypes === null ) {
 621                         $this->loadPluralFiles();
 622                 }
 623                 return $this->pluralRuleTypes[$code] ?? null;
 624         }
 625
 626         /**
 627          * Load the plural XML files.
 628          */
 629         protected function loadPluralFiles() {
 630                 global $IP;
 631                 $cldrPlural = "$IP/languages/data/plurals.xml";
 632                 $mwPlural = "$IP/languages/data/plurals-mediawiki.xml";
 633                 // Load CLDR plural rules
 634                 $this->loadPluralFile( $cldrPlural );
 635                 if ( file_exists( $mwPlural ) ) {
 636                         // Override or extend
 637                         $this->loadPluralFile( $mwPlural );
 638                 }
 639         }
 640
 641         /**
 642          * Load a plural XML file with the given filename, compile the relevant
 643          * rules, and save the compiled rules in a process-local cache.
 644          *
 645          * @param string $fileName
 646          * @throws MWException
 647          */
 648         protected function loadPluralFile( $fileName ) {
 649                 // Use file_get_contents instead of DOMDocument::load (T58439)
 650                 $xml = file_get_contents( $fileName );
 651                 if ( !$xml ) {
 652                         throw new MWException( "Unable to read plurals file $fileName" );
 653                 }
 654                 $doc = new DOMDocument;
 655                 $doc->loadXML( $xml );
 656                 $rulesets = $doc->getElementsByTagName( "pluralRules" );
 657                 foreach ( $rulesets as $ruleset ) {
 658                         $codes = $ruleset->getAttribute( 'locales' );
 659                         $rules = [];
 660                         $ruleTypes = [];
 661                         $ruleElements = $ruleset->getElementsByTagName( "pluralRule" );
 662                         foreach ( $ruleElements as $elt ) {
 663                                 $ruleType = $elt->getAttribute( 'count' );
 664                                 if ( $ruleType === 'other' ) {
 665                                         // Don't record "other" rules, which have an empty condition
 666                                         continue;
 667                                 }
 668                                 $rules[] = $elt->nodeValue;
 669                                 $ruleTypes[] = $ruleType;
 670                         }
 671                         foreach ( explode( ' ', $codes ) as $code ) {
 672                                 $this->pluralRules[$code] = $rules;
 673                                 $this->pluralRuleTypes[$code] = $ruleTypes;
 674                         }
 675                 }
 676         }
 677
 678         /**
 679          * Read the data from the source files for a given language, and register
 680          * the relevant dependencies in the $deps array. If the localisation
 681          * exists, the data array is returned, otherwise false is returned.
 682          *
 683          * @param string $code
 684          * @param array &$deps
 685          * @return array
 686          */
 687         protected function readSourceFilesAndRegisterDeps( $code, &$deps ) {
 688                 global $IP;
 689
 690                 // This reads in the PHP i18n file with non-messages l10n data
 691                 $fileName = Language::getMessagesFileName( $code );
 692                 if ( !file_exists( $fileName ) ) {
 693                         $data = [];
 694                 } else {
 695                         $deps[] = new FileDependency( $fileName );
 696                         $data = $this->readPHPFile( $fileName, 'core' );
 697                 }
 698
 699                 # Load CLDR plural rules for JavaScript
 700                 $data['pluralRules'] = $this->getPluralRules( $code );
 701                 # And for PHP
 702                 $data['compiledPluralRules'] = $this->getCompiledPluralRules( $code );
 703                 # Load plural rule types
 704                 $data['pluralRuleTypes'] = $this->getPluralRuleTypes( $code );
 705
 706                 $deps['plurals'] = new FileDependency( "$IP/languages/data/plurals.xml" );
 707                 $deps['plurals-mw'] = new FileDependency( "$IP/languages/data/plurals-mediawiki.xml" );
 708
 709                 return $data;
 710         }
 711
 712         /**
 713          * Merge two localisation values, a primary and a fallback, overwriting the
 714          * primary value in place.
 715          * @param string $key
 716          * @param mixed &$value
 717          * @param mixed $fallbackValue
 718          */
 719         protected function mergeItem( $key, &$value, $fallbackValue ) {
 720                 if ( !is_null( $value ) ) {
 721                         if ( !is_null( $fallbackValue ) ) {
 722                                 if ( in_array( $key, self::$mergeableMapKeys ) ) {
 723                                         $value = $value + $fallbackValue;
 724                                 } elseif ( in_array( $key, self::$mergeableListKeys ) ) {
 725                                         // @phan-suppress-next-line PhanTypeMismatchArgumentInternal
 726                                         $value = array_unique( array_merge( $fallbackValue, $value ) );
 727                                 } elseif ( in_array( $key, self::$mergeableAliasListKeys ) ) {
 728                                         $value = array_merge_recursive( $value, $fallbackValue );
 729                                 } elseif ( in_array( $key, self::$optionalMergeKeys ) ) {
 730                                         if ( !empty( $value['inherit'] ) ) {
 731                                                 $value = array_merge( $fallbackValue, $value );
 732                                         }
 733
 734                                         if ( isset( $value['inherit'] ) ) {
 735                                                 unset( $value['inherit'] );
 736                                         }
 737                                 } elseif ( in_array( $key, self::$magicWordKeys ) ) {
 738                                         $this->mergeMagicWords( $value, $fallbackValue );
 739                                 }
 740                         }
 741                 } else {
 742                         $value = $fallbackValue;
 743                 }
 744         }
 745
 746         /**
 747          * @param mixed &$value
 748          * @param mixed $fallbackValue
 749          */
 750         protected function mergeMagicWords( &$value, $fallbackValue ) {
 751                 foreach ( $fallbackValue as $magicName => $fallbackInfo ) {
 752                         if ( !isset( $value[$magicName] ) ) {
 753                                 $value[$magicName] = $fallbackInfo;
 754                         } else {
 755                                 $oldSynonyms = array_slice( $fallbackInfo, 1 );
 756                                 $newSynonyms = array_slice( $value[$magicName], 1 );
 757                                 $synonyms = array_values( array_unique( array_merge(
 758                                         $newSynonyms, $oldSynonyms ) ) );
 759                                 $value[$magicName] = array_merge( [ $fallbackInfo[0] ], $synonyms );
 760                         }
 761                 }
 762         }
 763
 764         /**
 765          * Given an array mapping language code to localisation value, such as is
 766          * found in extension *.i18n.php files, iterate through a fallback sequence
 767          * to merge the given data with an existing primary value.
 768          *
 769          * Returns true if any data from the extension array was used, false
 770          * otherwise.
 771          * @param array $codeSequence
 772          * @param string $key
 773          * @param mixed &$value
 774          * @param mixed $fallbackValue
 775          * @return bool
 776          */
 777         protected function mergeExtensionItem( $codeSequence, $key, &$value, $fallbackValue ) {
 778                 $used = false;
 779                 foreach ( $codeSequence as $code ) {
 780                         if ( isset( $fallbackValue[$code] ) ) {
 781                                 $this->mergeItem( $key, $value, $fallbackValue[$code] );
 782                                 $used = true;
 783                         }
 784                 }
 785
 786                 return $used;
 787         }
 788
 789         /**
 790          * Gets the combined list of messages dirs from
 791          * core and extensions
 792          *
 793          * @since 1.25
 794          * @return array
 795          */
 796         public function getMessagesDirs() {
 797                 global $IP;
 798
 799                 $config = MediaWikiServices::getInstance()->getMainConfig();
 800                 $messagesDirs = $config->get( 'MessagesDirs' );
 801                 return [
 802                         'core' => "$IP/languages/i18n",
 803                         'exif' => "$IP/languages/i18n/exif",
 804                         'api' => "$IP/includes/api/i18n",
 805                         'oojs-ui' => "$IP/resources/lib/ooui/i18n",
 806                 ] + $messagesDirs;
 807         }
 808
 809         /**
 810          * Load localisation data for a given language for both core and extensions
 811          * and save it to the persistent cache store and the process cache
 812          * @param string $code
 813          * @throws MWException
 814          */
 815         public function recache( $code ) {
 816                 global $wgExtensionMessagesFiles;
 817
 818                 if ( !$code ) {
 819                         throw new MWException( "Invalid language code requested" );
 820                 }
 821                 $this->recachedLangs[ $code ] = true;
 822
 823                 # Initial values
 824                 $initialData = array_fill_keys( self::$allKeys, null );
 825                 $coreData = $initialData;
 826                 $deps = [];
 827
 828                 # Load the primary localisation from the source file
 829                 $data = $this->readSourceFilesAndRegisterDeps( $code, $deps );
 830                 $this->logger->debug( __METHOD__ . ": got localisation for $code from source" );
 831
 832                 # Merge primary localisation
 833                 foreach ( $data as $key => $value ) {
 834                         $this->mergeItem( $key, $coreData[ $key ], $value );
 835                 }
 836
 837                 # Fill in the fallback if it's not there already
 838                 if ( ( is_null( $coreData['fallback'] ) || $coreData['fallback'] === false ) && $code === 'en' ) {
 839                         $coreData['fallback'] = false;
 840                         $coreData['originalFallbackSequence'] = $coreData['fallbackSequence'] = [];
 841                 } else {
 842                         if ( !is_null( $coreData['fallback'] ) ) {
 843                                 $coreData['fallbackSequence'] = array_map( 'trim', explode( ',', $coreData['fallback'] ) );
 844                         } else {
 845                                 $coreData['fallbackSequence'] = [];
 846                         }
 847                         $len = count( $coreData['fallbackSequence'] );
 848
 849                         # Before we add the 'en' fallback for messages, keep a copy of
 850                         # the original fallback sequence
 851                         $coreData['originalFallbackSequence'] = $coreData['fallbackSequence'];
 852
 853                         # Ensure that the sequence ends at 'en' for messages
 854                         if ( !$len || $coreData['fallbackSequence'][$len - 1] !== 'en' ) {
 855                                 $coreData['fallbackSequence'][] = 'en';
 856                         }
 857                 }
 858
 859                 $codeSequence = array_merge( [ $code ], $coreData['fallbackSequence'] );
 860                 $messageDirs = $this->getMessagesDirs();
 861
 862                 # Load non-JSON localisation data for extensions
 863                 $extensionData = array_fill_keys( $codeSequence, $initialData );
 864                 foreach ( $wgExtensionMessagesFiles as $extension => $fileName ) {
 865                         if ( isset( $messageDirs[$extension] ) ) {
 866                                 # This extension has JSON message data; skip the PHP shim
 867                                 continue;
 868                         }
 869
 870                         $data = $this->readPHPFile( $fileName, 'extension' );
 871                         $used = false;
 872
 873                         foreach ( $data as $key => $item ) {
 874                                 foreach ( $codeSequence as $csCode ) {
 875                                         if ( isset( $item[$csCode] ) ) {
 876                                                 $this->mergeItem( $key, $extensionData[$csCode][$key], $item[$csCode] );
 877                                                 $used = true;
 878                                         }
 879                                 }
 880                         }
 881
 882                         if ( $used ) {
 883                                 $deps[] = new FileDependency( $fileName );
 884                         }
 885                 }
 886
 887                 # Load the localisation data for each fallback, then merge it into the full array
 888                 $allData = $initialData;
 889                 foreach ( $codeSequence as $csCode ) {
 890                         $csData = $initialData;
 891
 892                         # Load core messages and the extension localisations.
 893                         foreach ( $messageDirs as $dirs ) {
 894                                 foreach ( (array)$dirs as $dir ) {
 895                                         $fileName = "$dir/$csCode.json";
 896                                         $data = $this->readJSONFile( $fileName );
 897
 898                                         foreach ( $data as $key => $item ) {
 899                                                 $this->mergeItem( $key, $csData[$key], $item );
 900                                         }
 901
 902                                         $deps[] = new FileDependency( $fileName );
 903                                 }
 904                         }
 905
 906                         # Merge non-JSON extension data
 907                         if ( isset( $extensionData[$csCode] ) ) {
 908                                 foreach ( $extensionData[$csCode] as $key => $item ) {
 909                                         $this->mergeItem( $key, $csData[$key], $item );
 910                                 }
 911                         }
 912
 913                         if ( $csCode === $code ) {
 914                                 # Merge core data into extension data
 915                                 foreach ( $coreData as $key => $item ) {
 916                                         $this->mergeItem( $key, $csData[$key], $item );
 917                                 }
 918                         } else {
 919                                 # Load the secondary localisation from the source file to
 920                                 # avoid infinite cycles on cyclic fallbacks
 921                                 $fbData = $this->readSourceFilesAndRegisterDeps( $csCode, $deps );
 922                                 # Only merge the keys that make sense to merge
 923                                 foreach ( self::$allKeys as $key ) {
 924                                         if ( !isset( $fbData[ $key ] ) ) {
 925                                                 continue;
 926                                         }
 927
 928                                         if ( is_null( $coreData[ $key ] ) || $this->isMergeableKey( $key ) ) {
 929                                                 $this->mergeItem( $key, $csData[ $key ], $fbData[ $key ] );
 930                                         }
 931                                 }
 932                         }
 933
 934                         # Allow extensions an opportunity to adjust the data for this
 935                         # fallback
 936                         Hooks::run( 'LocalisationCacheRecacheFallback', [ $this, $csCode, &$csData ] );
 937
 938                         # Merge the data for this fallback into the final array
 939                         if ( $csCode === $code ) {
 940                                 $allData = $csData;
 941                         } else {
 942                                 foreach ( self::$allKeys as $key ) {
 943                                         if ( !isset( $csData[$key] ) ) {
 944                                                 continue;
 945                                         }
 946
 947                                         if ( is_null( $allData[$key] ) || $this->isMergeableKey( $key ) ) {
 948                                                 $this->mergeItem( $key, $allData[$key], $csData[$key] );
 949                                         }
 950                                 }
 951                         }
 952                 }
 953
 954                 # Add cache dependencies for any referenced globals
 955                 $deps['wgExtensionMessagesFiles'] = new GlobalDependency( 'wgExtensionMessagesFiles' );
 956                 // The 'MessagesDirs' config setting is used in LocalisationCache::getMessagesDirs().
 957                 // We use the key 'wgMessagesDirs' for historical reasons.
 958                 $deps['wgMessagesDirs'] = new MainConfigDependency( 'MessagesDirs' );
 959                 $deps['version'] = new ConstantDependency( 'LocalisationCache::VERSION' );
 960
 961                 # Add dependencies to the cache entry
 962                 $allData['deps'] = $deps;
 963
 964                 # Replace spaces with underscores in namespace names
 965                 $allData['namespaceNames'] = str_replace( ' ', '_', $allData['namespaceNames'] );
 966
 967                 # And do the same for special page aliases. $page is an array.
 968                 foreach ( $allData['specialPageAliases'] as &$page ) {
 969                         $page = str_replace( ' ', '_', $page );
 970                 }
 971                 # Decouple the reference to prevent accidental damage
 972                 unset( $page );
 973
 974                 # If there were no plural rules, return an empty array
 975                 if ( $allData['pluralRules'] === null ) {
 976                         $allData['pluralRules'] = [];
 977                 }
 978                 if ( $allData['compiledPluralRules'] === null ) {
 979                         $allData['compiledPluralRules'] = [];
 980                 }
 981                 # If there were no plural rule types, return an empty array
 982                 if ( $allData['pluralRuleTypes'] === null ) {
 983                         $allData['pluralRuleTypes'] = [];
 984                 }
 985
 986                 # Set the list keys
 987                 $allData['list'] = [];
 988                 foreach ( self::$splitKeys as $key ) {
 989                         $allData['list'][$key] = array_keys( $allData[$key] );
 990                 }
 991                 # Run hooks
 992                 $unused = true; // Used to be $purgeBlobs, removed in 1.34
 993                 Hooks::run( 'LocalisationCacheRecache', [ $this, $code, &$allData, &$unused ] );
 994
 995                 if ( is_null( $allData['namespaceNames'] ) ) {
 996                         throw new MWException( __METHOD__ . ': Localisation data failed sanity check! ' .
 997                                 'Check that your languages/messages/MessagesEn.php file is intact.' );
 998                 }
 999
1000                 # Set the preload key
1001                 $allData['preload'] = $this->buildPreload( $allData );
1002
1003                 # Save to the process cache and register the items loaded
1004                 $this->data[$code] = $allData;
1005                 foreach ( $allData as $key => $item ) {
1006                         $this->loadedItems[$code][$key] = true;
1007                 }
1008
1009                 # Save to the persistent cache
1010                 $this->store->startWrite( $code );
1011                 foreach ( $allData as $key => $value ) {
1012                         if ( in_array( $key, self::$splitKeys ) ) {
1013                                 foreach ( $value as $subkey => $subvalue ) {
1014                                         $this->store->set( "$key:$subkey", $subvalue );
1015                                 }
1016                         } else {
1017                                 $this->store->set( $key, $value );
1018                         }
1019                 }
1020                 $this->store->finishWrite();
1021
1022                 # Clear out the MessageBlobStore
1023                 # HACK: If using a null (i.e. disabled) storage backend, we
1024                 # can't write to the MessageBlobStore either
1025                 if ( !$this->store instanceof LCStoreNull ) {
1026                         $blobStore = MediaWikiServices::getInstance()->getResourceLoader()->getMessageBlobStore();
1027                         $blobStore->clear();
1028                 }
1029         }
1030
1031         /**
1032          * Build the preload item from the given pre-cache data.
1033          *
1034          * The preload item will be loaded automatically, improving performance
1035          * for the commonly-requested items it contains.
1036          * @param array $data
1037          * @return array
1038          */
1039         protected function buildPreload( $data ) {
1040                 $preload = [ 'messages' => [] ];
1041                 foreach ( self::$preloadedKeys as $key ) {
1042                         $preload[$key] = $data[$key];
1043                 }
1044
1045                 foreach ( $data['preloadedMessages'] as $subkey ) {
1046                         $subitem = $data['messages'][$subkey] ?? null;
1047                         $preload['messages'][$subkey] = $subitem;
1048                 }
1049
1050                 return $preload;
1051         }
1052
1053         /**
1054          * Unload the data for a given language from the object cache.
1055          * Reduces memory usage.
1056          * @param string $code
1057          */
1058         public function unload( $code ) {
1059                 unset( $this->data[$code] );
1060                 unset( $this->loadedItems[$code] );
1061                 unset( $this->loadedSubitems[$code] );
1062                 unset( $this->initialisedLangs[$code] );
1063                 unset( $this->shallowFallbacks[$code] );
1064
1065                 foreach ( $this->shallowFallbacks as $shallowCode => $fbCode ) {
1066                         if ( $fbCode === $code ) {
1067                                 $this->unload( $shallowCode );
1068                         }
1069                 }
1070         }
1071
1072         /**
1073          * Unload all data
1074          */
1075         public function unloadAll() {
1076                 foreach ( $this->initialisedLangs as $lang => $unused ) {
1077                         $this->unload( $lang );
1078                 }
1079         }
1080
1081         /**
1082          * Disable the storage backend
1083          */
1084         public function disableBackend() {
1085                 $this->store = new LCStoreNull;
1086                 $this->manualRecache = false;
1087         }
1088
1089 }