includes/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 define( 'MW_LC_VERSION', 2 );
  24
  25 /**
  26  * Class for caching the contents of localisation files, Messages*.php
  27  * and *.i18n.php.
  28  *
  29  * An instance of this class is available using Language::getLocalisationCache().
  30  *
  31  * The values retrieved from here are merged, containing items from extension
  32  * files, core messages files and the language fallback sequence (e.g. zh-cn ->
  33  * zh-hans -> en ). Some common errors are corrected, for example namespace
  34  * names with spaces instead of underscores, but heavyweight processing, such
  35  * as grammatical transformation, is done by the caller.
  36  */
  37 class LocalisationCache {
  38         /** Configuration associative array */
  39         var $conf;
  40
  41         /**
  42          * True if recaching should only be done on an explicit call to recache().
  43          * Setting this reduces the overhead of cache freshness checking, which
  44          * requires doing a stat() for every extension i18n file.
  45          */
  46         var $manualRecache = false;
  47
  48         /**
  49          * True to treat all files as expired until they are regenerated by this object.
  50          */
  51         var $forceRecache = false;
  52
  53         /**
  54          * The cache data. 3-d array, where the first key is the language code,
  55          * the second key is the item key e.g. 'messages', and the third key is
  56          * an item specific subkey index. Some items are not arrays and so for those
  57          * items, there are no subkeys.
  58          */
  59         var $data = array();
  60
  61         /**
  62          * The persistent store object. An instance of LCStore.
  63          *
  64          * @var LCStore
  65          */
  66         var $store;
  67
  68         /**
  69          * A 2-d associative array, code/key, where presence indicates that the item
  70          * is loaded. Value arbitrary.
  71          *
  72          * For split items, if set, this indicates that all of the subitems have been
  73          * loaded.
  74          */
  75         var $loadedItems = array();
  76
  77         /**
  78          * A 3-d associative array, code/key/subkey, where presence indicates that
  79          * the subitem is loaded. Only used for the split items, i.e. messages.
  80          */
  81         var $loadedSubitems = array();
  82
  83         /**
  84          * An array where presence of a key indicates that that language has been
  85          * initialised. Initialisation includes checking for cache expiry and doing
  86          * any necessary updates.
  87          */
  88         var $initialisedLangs = array();
  89
  90         /**
  91          * An array mapping non-existent pseudo-languages to fallback languages. This
  92          * is filled by initShallowFallback() when data is requested from a language
  93          * that lacks a Messages*.php file.
  94          */
  95         var $shallowFallbacks = array();
  96
  97         /**
  98          * An array where the keys are codes that have been recached by this instance.
  99          */
 100         var $recachedLangs = array();
 101
 102         /**
 103          * All item keys
 104          */
 105         static public $allKeys = array(
 106                 'fallback', 'namespaceNames', 'bookstoreList',
 107                 'magicWords', 'messages', 'rtl', 'capitalizeAllNouns', 'digitTransformTable',
 108                 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
 109                 'linkTrail', 'namespaceAliases',
 110                 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
 111                 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
 112                 'imageFiles', 'preloadedMessages', 'namespaceGenderAliases',
 113                 'digitGroupingPattern', 'pluralRules', 'compiledPluralRules',
 114         );
 115
 116         /**
 117          * Keys for items which consist of associative arrays, which may be merged
 118          * by a fallback sequence.
 119          */
 120         static public $mergeableMapKeys = array( 'messages', 'namespaceNames',
 121                 'dateFormats', 'imageFiles', 'preloadedMessages'
 122         );
 123
 124         /**
 125          * Keys for items which are a numbered array.
 126          */
 127         static public $mergeableListKeys = array( 'extraUserToggles' );
 128
 129         /**
 130          * Keys for items which contain an array of arrays of equivalent aliases
 131          * for each subitem. The aliases may be merged by a fallback sequence.
 132          */
 133         static public $mergeableAliasListKeys = array( 'specialPageAliases' );
 134
 135         /**
 136          * Keys for items which contain an associative array, and may be merged if
 137          * the primary value contains the special array key "inherit". That array
 138          * key is removed after the first merge.
 139          */
 140         static public $optionalMergeKeys = array( 'bookstoreList' );
 141
 142         /**
 143          * Keys for items that are formatted like $magicWords
 144          */
 145         static public $magicWordKeys = array( 'magicWords' );
 146
 147         /**
 148          * Keys for items where the subitems are stored in the backend separately.
 149          */
 150         static public $splitKeys = array( 'messages' );
 151
 152         /**
 153          * Keys which are loaded automatically by initLanguage()
 154          */
 155         static public $preloadedKeys = array( 'dateFormats', 'namespaceNames' );
 156
 157         /**
 158          * Associative array of cached plural rules. The key is the language code,
 159          * the value is an array of plural rules for that language.
 160          */
 161         var $pluralRules = null;
 162
 163         var $mergeableKeys = null;
 164
 165         /**
 166          * Constructor.
 167          * For constructor parameters, see the documentation in DefaultSettings.php
 168          * for $wgLocalisationCacheConf.
 169          *
 170          * @param $conf Array
 171          */
 172         function __construct( $conf ) {
 173                 global $wgCacheDirectory;
 174
 175                 $this->conf = $conf;
 176                 $storeConf = array();
 177                 if ( !empty( $conf['storeClass'] ) ) {
 178                         $storeClass = $conf['storeClass'];
 179                 } else {
 180                         switch ( $conf['store'] ) {
 181                                 case 'files':
 182                                 case 'file':
 183                                         $storeClass = 'LCStore_CDB';
 184                                         break;
 185                                 case 'db':
 186                                         $storeClass = 'LCStore_DB';
 187                                         break;
 188                                 case 'accel':
 189                                         $storeClass = 'LCStore_Accel';
 190                                         break;
 191                                 case 'detect':
 192                                         $storeClass = $wgCacheDirectory ? 'LCStore_CDB' : 'LCStore_DB';
 193                                         break;
 194                                 default:
 195                                         throw new MWException(
 196                                                 'Please set $wgLocalisationCacheConf[\'store\'] to something sensible.' );
 197                         }
 198                 }
 199
 200                 wfDebug( get_class( $this ) . ": using store $storeClass\n" );
 201                 if ( !empty( $conf['storeDirectory'] ) ) {
 202                         $storeConf['directory'] = $conf['storeDirectory'];
 203                 }
 204
 205                 $this->store = new $storeClass( $storeConf );
 206                 foreach ( array( 'manualRecache', 'forceRecache' ) as $var ) {
 207                         if ( isset( $conf[$var] ) ) {
 208                                 $this->$var = $conf[$var];
 209                         }
 210                 }
 211         }
 212
 213         /**
 214          * Returns true if the given key is mergeable, that is, if it is an associative
 215          * array which can be merged through a fallback sequence.
 216          * @param $key
 217          * @return bool
 218          */
 219         public function isMergeableKey( $key ) {
 220                 if ( $this->mergeableKeys === null ) {
 221                         $this->mergeableKeys = array_flip( array_merge(
 222                                 self::$mergeableMapKeys,
 223                                 self::$mergeableListKeys,
 224                                 self::$mergeableAliasListKeys,
 225                                 self::$optionalMergeKeys,
 226                                 self::$magicWordKeys
 227                         ) );
 228                 }
 229                 return isset( $this->mergeableKeys[$key] );
 230         }
 231
 232         /**
 233          * Get a cache item.
 234          *
 235          * Warning: this may be slow for split items (messages), since it will
 236          * need to fetch all of the subitems from the cache individually.
 237          * @param $code
 238          * @param $key
 239          * @return mixed
 240          */
 241         public function getItem( $code, $key ) {
 242                 if ( !isset( $this->loadedItems[$code][$key] ) ) {
 243                         wfProfileIn( __METHOD__ . '-load' );
 244                         $this->loadItem( $code, $key );
 245                         wfProfileOut( __METHOD__ . '-load' );
 246                 }
 247
 248                 if ( $key === 'fallback' && isset( $this->shallowFallbacks[$code] ) ) {
 249                         return $this->shallowFallbacks[$code];
 250                 }
 251
 252                 return $this->data[$code][$key];
 253         }
 254
 255         /**
 256          * Get a subitem, for instance a single message for a given language.
 257          * @param $code
 258          * @param $key
 259          * @param $subkey
 260          * @return null
 261          */
 262         public function getSubitem( $code, $key, $subkey ) {
 263                 if ( !isset( $this->loadedSubitems[$code][$key][$subkey] ) &&
 264                          !isset( $this->loadedItems[$code][$key] ) ) {
 265                         wfProfileIn( __METHOD__ . '-load' );
 266                         $this->loadSubitem( $code, $key, $subkey );
 267                         wfProfileOut( __METHOD__ . '-load' );
 268                 }
 269
 270                 if ( isset( $this->data[$code][$key][$subkey] ) ) {
 271                         return $this->data[$code][$key][$subkey];
 272                 } else {
 273                         return null;
 274                 }
 275         }
 276
 277         /**
 278          * Get the list of subitem keys for a given item.
 279          *
 280          * This is faster than array_keys($lc->getItem(...)) for the items listed in
 281          * self::$splitKeys.
 282          *
 283          * Will return null if the item is not found, or false if the item is not an
 284          * array.
 285          * @param $code
 286          * @param $key
 287          * @return bool|null|string
 288          */
 289         public function getSubitemList( $code, $key ) {
 290                 if ( in_array( $key, self::$splitKeys ) ) {
 291                         return $this->getSubitem( $code, 'list', $key );
 292                 } else {
 293                         $item = $this->getItem( $code, $key );
 294                         if ( is_array( $item ) ) {
 295                                 return array_keys( $item );
 296                         } else {
 297                                 return false;
 298                         }
 299                 }
 300         }
 301
 302         /**
 303          * Load an item into the cache.
 304          * @param $code
 305          * @param $key
 306          */
 307         protected function loadItem( $code, $key ) {
 308                 if ( !isset( $this->initialisedLangs[$code] ) ) {
 309                         $this->initLanguage( $code );
 310                 }
 311
 312                 // Check to see if initLanguage() loaded it for us
 313                 if ( isset( $this->loadedItems[$code][$key] ) ) {
 314                         return;
 315                 }
 316
 317                 if ( isset( $this->shallowFallbacks[$code] ) ) {
 318                         $this->loadItem( $this->shallowFallbacks[$code], $key );
 319                         return;
 320                 }
 321
 322                 if ( in_array( $key, self::$splitKeys ) ) {
 323                         $subkeyList = $this->getSubitem( $code, 'list', $key );
 324                         foreach ( $subkeyList as $subkey ) {
 325                                 if ( isset( $this->data[$code][$key][$subkey] ) ) {
 326                                         continue;
 327                                 }
 328                                 $this->data[$code][$key][$subkey] = $this->getSubitem( $code, $key, $subkey );
 329                         }
 330                 } else {
 331                         $this->data[$code][$key] = $this->store->get( $code, $key );
 332                 }
 333
 334                 $this->loadedItems[$code][$key] = true;
 335         }
 336
 337         /**
 338          * Load a subitem into the cache
 339          * @param $code
 340          * @param $key
 341          * @param $subkey
 342          * @return
 343          */
 344         protected function loadSubitem( $code, $key, $subkey ) {
 345                 if ( !in_array( $key, self::$splitKeys ) ) {
 346                         $this->loadItem( $code, $key );
 347                         return;
 348                 }
 349
 350                 if ( !isset( $this->initialisedLangs[$code] ) ) {
 351                         $this->initLanguage( $code );
 352                 }
 353
 354                 // Check to see if initLanguage() loaded it for us
 355                 if ( isset( $this->loadedItems[$code][$key] ) ||
 356                          isset( $this->loadedSubitems[$code][$key][$subkey] ) ) {
 357                         return;
 358                 }
 359
 360                 if ( isset( $this->shallowFallbacks[$code] ) ) {
 361                         $this->loadSubitem( $this->shallowFallbacks[$code], $key, $subkey );
 362                         return;
 363                 }
 364
 365                 $value = $this->store->get( $code, "$key:$subkey" );
 366                 $this->data[$code][$key][$subkey] = $value;
 367                 $this->loadedSubitems[$code][$key][$subkey] = true;
 368         }
 369
 370         /**
 371          * Returns true if the cache identified by $code is missing or expired.
 372          * @return bool
 373          */
 374         public function isExpired( $code ) {
 375                 if ( $this->forceRecache && !isset( $this->recachedLangs[$code] ) ) {
 376                         wfDebug( __METHOD__ . "($code): forced reload\n" );
 377                         return true;
 378                 }
 379
 380                 $deps = $this->store->get( $code, 'deps' );
 381                 $keys = $this->store->get( $code, 'list', 'messages' );
 382                 $preload = $this->store->get( $code, 'preload' );
 383                 // Different keys may expire separately, at least in LCStore_Accel
 384                 if ( $deps === null || $keys === null || $preload === null ) {
 385                         wfDebug( __METHOD__ . "($code): cache missing, need to make one\n" );
 386                         return true;
 387                 }
 388
 389                 foreach ( $deps as $dep ) {
 390                         // Because we're unserializing stuff from cache, we
 391                         // could receive objects of classes that don't exist
 392                         // anymore (e.g. uninstalled extensions)
 393                         // When this happens, always expire the cache
 394                         if ( !$dep instanceof CacheDependency || $dep->isExpired() ) {
 395                                 wfDebug( __METHOD__ . "($code): cache for $code expired due to " .
 396                                         get_class( $dep ) . "\n" );
 397                                 return true;
 398                         }
 399                 }
 400
 401                 return false;
 402         }
 403
 404         /**
 405          * Initialise a language in this object. Rebuild the cache if necessary.
 406          * @param $code
 407          */
 408         protected function initLanguage( $code ) {
 409                 if ( isset( $this->initialisedLangs[$code] ) ) {
 410                         return;
 411                 }
 412
 413                 $this->initialisedLangs[$code] = true;
 414
 415                 # If the code is of the wrong form for a Messages*.php file, do a shallow fallback
 416                 if ( !Language::isValidBuiltInCode( $code ) ) {
 417                         $this->initShallowFallback( $code, 'en' );
 418                         return;
 419                 }
 420
 421                 # Recache the data if necessary
 422                 if ( !$this->manualRecache && $this->isExpired( $code ) ) {
 423                         if ( file_exists( Language::getMessagesFileName( $code ) ) ) {
 424                                 $this->recache( $code );
 425                         } elseif ( $code === 'en' ) {
 426                                 throw new MWException( 'MessagesEn.php is missing.' );
 427                         } else {
 428                                 $this->initShallowFallback( $code, 'en' );
 429                         }
 430                         return;
 431                 }
 432
 433                 # Preload some stuff
 434                 $preload = $this->getItem( $code, 'preload' );
 435                 if ( $preload === null ) {
 436                         if ( $this->manualRecache ) {
 437                                 // No Messages*.php file. Do shallow fallback to en.
 438                                 if ( $code === 'en' ) {
 439                                         throw new MWException( 'No localisation cache found for English. ' .
 440                                                 'Please run maintenance/rebuildLocalisationCache.php.' );
 441                                 }
 442                                 $this->initShallowFallback( $code, 'en' );
 443                                 return;
 444                         } else {
 445                                 throw new MWException( 'Invalid or missing localisation cache.' );
 446                         }
 447                 }
 448                 $this->data[$code] = $preload;
 449                 foreach ( $preload as $key => $item ) {
 450                         if ( in_array( $key, self::$splitKeys ) ) {
 451                                 foreach ( $item as $subkey => $subitem ) {
 452                                         $this->loadedSubitems[$code][$key][$subkey] = true;
 453                                 }
 454                         } else {
 455                                 $this->loadedItems[$code][$key] = true;
 456                         }
 457                 }
 458         }
 459
 460         /**
 461          * Create a fallback from one language to another, without creating a
 462          * complete persistent cache.
 463          * @param $primaryCode
 464          * @param $fallbackCode
 465          */
 466         public function initShallowFallback( $primaryCode, $fallbackCode ) {
 467                 $this->data[$primaryCode] =& $this->data[$fallbackCode];
 468                 $this->loadedItems[$primaryCode] =& $this->loadedItems[$fallbackCode];
 469                 $this->loadedSubitems[$primaryCode] =& $this->loadedSubitems[$fallbackCode];
 470                 $this->shallowFallbacks[$primaryCode] = $fallbackCode;
 471         }
 472
 473         /**
 474          * Read a PHP file containing localisation data.
 475          * @param $_fileName
 476          * @param $_fileType
 477          * @return array
 478          */
 479         protected function readPHPFile( $_fileName, $_fileType ) {
 480                 // Disable APC caching
 481                 $_apcEnabled = ini_set( 'apc.cache_by_default', '0' );
 482                 include( $_fileName );
 483                 ini_set( 'apc.cache_by_default', $_apcEnabled );
 484
 485                 if ( $_fileType == 'core' || $_fileType == 'extension' ) {
 486                         $data = compact( self::$allKeys );
 487                 } elseif ( $_fileType == 'aliases' ) {
 488                         $data = compact( 'aliases' );
 489                 } else {
 490                         throw new MWException( __METHOD__ . ": Invalid file type: $_fileType" );
 491                 }
 492                 return $data;
 493         }
 494
 495         /**
 496          * Get the compiled plural rules for a given language from the XML files.
 497          * @since 1.20
 498          */
 499         public function getCompiledPluralRules( $code ) {
 500                 $rules = $this->getPluralRules( $code );
 501                 if ( $rules === null ) {
 502                         return null;
 503                 }
 504                 try {
 505                         $compiledRules = CLDRPluralRuleEvaluator::compile( $rules );
 506                 } catch( CLDRPluralRuleError $e ) {
 507                         wfDebugLog( 'l10n', $e->getMessage() . "\n" );
 508                         return array();
 509                 }
 510                 return $compiledRules;
 511         }
 512
 513         /**
 514          * Get the plural rules for a given language from the XML files.
 515          * Cached.
 516          * @since 1.20
 517          */
 518         public function getPluralRules( $code ) {
 519                 if ( $this->pluralRules === null ) {
 520                         $cldrPlural = __DIR__ . "/../languages/data/plurals.xml";
 521                         $mwPlural = __DIR__ . "/../languages/data/plurals-mediawiki.xml";
 522                         // Load CLDR plural rules
 523                         $this->loadPluralFile( $cldrPlural );
 524                         if ( file_exists( $mwPlural ) ) {
 525                                 // Override or extend
 526                                 $this->loadPluralFile( $mwPlural );
 527                         }
 528                 }
 529                 if ( !isset( $this->pluralRules[$code] ) ) {
 530                         return null;
 531                 } else {
 532                         return $this->pluralRules[$code];
 533                 }
 534         }
 535
 536
 537         /**
 538          * Load a plural XML file with the given filename, compile the relevant
 539          * rules, and save the compiled rules in a process-local cache.
 540          */
 541         protected function loadPluralFile( $fileName ) {
 542                 $doc = new DOMDocument;
 543                 $doc->load( $fileName );
 544                 $rulesets = $doc->getElementsByTagName( "pluralRules" );
 545                 foreach ( $rulesets as $ruleset ) {
 546                         $codes = $ruleset->getAttribute( 'locales' );
 547                         $rules = array();
 548                         $ruleElements = $ruleset->getElementsByTagName( "pluralRule" );
 549                         foreach ( $ruleElements as $elt ) {
 550                                 $rules[] = $elt->nodeValue;
 551                         }
 552                         foreach ( explode( ' ', $codes ) as $code ) {
 553                                 $this->pluralRules[$code] = $rules;
 554                         }
 555                 }
 556         }
 557
 558         /**
 559          * Read the data from the source files for a given language, and register
 560          * the relevant dependencies in the $deps array. If the localisation
 561          * exists, the data array is returned, otherwise false is returned.
 562          */
 563         protected function readSourceFilesAndRegisterDeps( $code, &$deps ) {
 564                 $fileName = Language::getMessagesFileName( $code );
 565                 if ( !file_exists( $fileName ) ) {
 566                         return false;
 567                 }
 568
 569                 $deps[] = new FileDependency( $fileName );
 570                 $data = $this->readPHPFile( $fileName, 'core' );
 571
 572                 # Load CLDR plural rules for JavaScript
 573                 $data['pluralRules'] = $this->getPluralRules( $code );
 574                 # And for PHP
 575                 $data['compiledPluralRules'] = $this->getCompiledPluralRules( $code );
 576
 577                 $deps['plurals'] = new FileDependency( __DIR__ . "/../languages/data/plurals.xml" );
 578                 $deps['plurals-mw'] = new FileDependency( __DIR__ . "/../languages/data/plurals-mediawiki.xml" );
 579                 return $data;
 580         }
 581
 582         /**
 583          * Merge two localisation values, a primary and a fallback, overwriting the
 584          * primary value in place.
 585          * @param $key
 586          * @param $value
 587          * @param $fallbackValue
 588          */
 589         protected function mergeItem( $key, &$value, $fallbackValue ) {
 590                 if ( !is_null( $value ) ) {
 591                         if ( !is_null( $fallbackValue ) ) {
 592                                 if ( in_array( $key, self::$mergeableMapKeys ) ) {
 593                                         $value = $value + $fallbackValue;
 594                                 } elseif ( in_array( $key, self::$mergeableListKeys ) ) {
 595                                         $value = array_unique( array_merge( $fallbackValue, $value ) );
 596                                 } elseif ( in_array( $key, self::$mergeableAliasListKeys ) ) {
 597                                         $value = array_merge_recursive( $value, $fallbackValue );
 598                                 } elseif ( in_array( $key, self::$optionalMergeKeys ) ) {
 599                                         if ( !empty( $value['inherit'] ) )  {
 600                                                 $value = array_merge( $fallbackValue, $value );
 601                                         }
 602
 603                                         if ( isset( $value['inherit'] ) ) {
 604                                                 unset( $value['inherit'] );
 605                                         }
 606                                 } elseif ( in_array( $key, self::$magicWordKeys ) ) {
 607                                         $this->mergeMagicWords( $value, $fallbackValue );
 608                                 }
 609                         }
 610                 } else {
 611                         $value = $fallbackValue;
 612                 }
 613         }
 614
 615         /**
 616          * @param $value
 617          * @param $fallbackValue
 618          */
 619         protected function mergeMagicWords( &$value, $fallbackValue ) {
 620                 foreach ( $fallbackValue as $magicName => $fallbackInfo ) {
 621                         if ( !isset( $value[$magicName] ) ) {
 622                                 $value[$magicName] = $fallbackInfo;
 623                         } else {
 624                                 $oldSynonyms = array_slice( $fallbackInfo, 1 );
 625                                 $newSynonyms = array_slice( $value[$magicName], 1 );
 626                                 $synonyms = array_values( array_unique( array_merge(
 627                                         $newSynonyms, $oldSynonyms ) ) );
 628                                 $value[$magicName] = array_merge( array( $fallbackInfo[0] ), $synonyms );
 629                         }
 630                 }
 631         }
 632
 633         /**
 634          * Given an array mapping language code to localisation value, such as is
 635          * found in extension *.i18n.php files, iterate through a fallback sequence
 636          * to merge the given data with an existing primary value.
 637          *
 638          * Returns true if any data from the extension array was used, false
 639          * otherwise.
 640          * @param $codeSequence
 641          * @param $key
 642          * @param $value
 643          * @param $fallbackValue
 644          * @return bool
 645          */
 646         protected function mergeExtensionItem( $codeSequence, $key, &$value, $fallbackValue ) {
 647                 $used = false;
 648                 foreach ( $codeSequence as $code ) {
 649                         if ( isset( $fallbackValue[$code] ) ) {
 650                                 $this->mergeItem( $key, $value, $fallbackValue[$code] );
 651                                 $used = true;
 652                         }
 653                 }
 654
 655                 return $used;
 656         }
 657
 658         /**
 659          * Load localisation data for a given language for both core and extensions
 660          * and save it to the persistent cache store and the process cache
 661          * @param $code
 662          */
 663         public function recache( $code ) {
 664                 global $wgExtensionMessagesFiles;
 665                 wfProfileIn( __METHOD__ );
 666
 667                 if ( !$code ) {
 668                         throw new MWException( "Invalid language code requested" );
 669                 }
 670                 $this->recachedLangs[$code] = true;
 671
 672                 # Initial values
 673                 $initialData = array_combine(
 674                         self::$allKeys,
 675                         array_fill( 0, count( self::$allKeys ), null ) );
 676                 $coreData = $initialData;
 677                 $deps = array();
 678
 679                 # Load the primary localisation from the source file
 680                 $data = $this->readSourceFilesAndRegisterDeps( $code, $deps );
 681                 if ( $data === false ) {
 682                         wfDebug( __METHOD__ . ": no localisation file for $code, using fallback to en\n" );
 683                         $coreData['fallback'] = 'en';
 684                 } else {
 685                         wfDebug( __METHOD__ . ": got localisation for $code from source\n" );
 686
 687                         # Merge primary localisation
 688                         foreach ( $data as $key => $value ) {
 689                                 $this->mergeItem( $key, $coreData[$key], $value );
 690                         }
 691
 692                 }
 693
 694                 # Fill in the fallback if it's not there already
 695                 if ( is_null( $coreData['fallback'] ) ) {
 696                         $coreData['fallback'] = $code === 'en' ? false : 'en';
 697                 }
 698                 if ( $coreData['fallback'] === false ) {
 699                         $coreData['fallbackSequence'] = array();
 700                 } else {
 701                         $coreData['fallbackSequence'] = array_map( 'trim', explode( ',', $coreData['fallback'] ) );
 702                         $len = count( $coreData['fallbackSequence'] );
 703
 704                         # Ensure that the sequence ends at en
 705                         if ( $coreData['fallbackSequence'][$len - 1] !== 'en' ) {
 706                                 $coreData['fallbackSequence'][] = 'en';
 707                         }
 708
 709                         # Load the fallback localisation item by item and merge it
 710                         foreach ( $coreData['fallbackSequence'] as $fbCode ) {
 711                                 # Load the secondary localisation from the source file to
 712                                 # avoid infinite cycles on cyclic fallbacks
 713                                 $fbData = $this->readSourceFilesAndRegisterDeps( $fbCode, $deps );
 714                                 if ( $fbData === false ) {
 715                                         continue;
 716                                 }
 717
 718                                 foreach ( self::$allKeys as $key ) {
 719                                         if ( !isset( $fbData[$key] ) ) {
 720                                                 continue;
 721                                         }
 722
 723                                         if ( is_null( $coreData[$key] ) || $this->isMergeableKey( $key ) ) {
 724                                                 $this->mergeItem( $key, $coreData[$key], $fbData[$key] );
 725                                         }
 726                                 }
 727                         }
 728                 }
 729
 730                 $codeSequence = array_merge( array( $code ), $coreData['fallbackSequence'] );
 731
 732                 # Load the extension localisations
 733                 # This is done after the core because we know the fallback sequence now.
 734                 # But it has a higher precedence for merging so that we can support things
 735                 # like site-specific message overrides.
 736                 $allData = $initialData;
 737                 foreach ( $wgExtensionMessagesFiles as $fileName ) {
 738                         $data = $this->readPHPFile( $fileName, 'extension' );
 739                         $used = false;
 740
 741                         foreach ( $data as $key => $item ) {
 742                                 if ( $this->mergeExtensionItem( $codeSequence, $key, $allData[$key], $item ) ) {
 743                                         $used = true;
 744                                 }
 745                         }
 746
 747                         if ( $used ) {
 748                                 $deps[] = new FileDependency( $fileName );
 749                         }
 750                 }
 751
 752                 # Merge core data into extension data
 753                 foreach ( $coreData as $key => $item ) {
 754                         $this->mergeItem( $key, $allData[$key], $item );
 755                 }
 756
 757                 # Add cache dependencies for any referenced globals
 758                 $deps['wgExtensionMessagesFiles'] = new GlobalDependency( 'wgExtensionMessagesFiles' );
 759                 $deps['version'] = new ConstantDependency( 'MW_LC_VERSION' );
 760
 761                 # Add dependencies to the cache entry
 762                 $allData['deps'] = $deps;
 763
 764                 # Replace spaces with underscores in namespace names
 765                 $allData['namespaceNames'] = str_replace( ' ', '_', $allData['namespaceNames'] );
 766
 767                 # And do the same for special page aliases. $page is an array.
 768                 foreach ( $allData['specialPageAliases'] as &$page ) {
 769                         $page = str_replace( ' ', '_', $page );
 770                 }
 771                 # Decouple the reference to prevent accidental damage
 772                 unset( $page );
 773
 774                 # If there were no plural rules, return an empty array
 775                 if ( $allData['pluralRules'] === null ) {
 776                         $allData['pluralRules'] = array();
 777                 }
 778                 if ( $allData['compiledPluralRules'] === null ) {
 779                         $allData['compiledPluralRules'] = array();
 780                 }
 781
 782                 # Set the list keys
 783                 $allData['list'] = array();
 784                 foreach ( self::$splitKeys as $key ) {
 785                         $allData['list'][$key] = array_keys( $allData[$key] );
 786                 }
 787                 # Run hooks
 788                 wfRunHooks( 'LocalisationCacheRecache', array( $this, $code, &$allData ) );
 789
 790                 if ( is_null( $allData['namespaceNames'] ) ) {
 791                         throw new MWException( __METHOD__ . ': Localisation data failed sanity check! ' .
 792                                 'Check that your languages/messages/MessagesEn.php file is intact.' );
 793                 }
 794
 795                 # Set the preload key
 796                 $allData['preload'] = $this->buildPreload( $allData );
 797
 798                 # Save to the process cache and register the items loaded
 799                 $this->data[$code] = $allData;
 800                 foreach ( $allData as $key => $item ) {
 801                         $this->loadedItems[$code][$key] = true;
 802                 }
 803
 804                 # Save to the persistent cache
 805                 $this->store->startWrite( $code );
 806                 foreach ( $allData as $key => $value ) {
 807                         if ( in_array( $key, self::$splitKeys ) ) {
 808                                 foreach ( $value as $subkey => $subvalue ) {
 809                                         $this->store->set( "$key:$subkey", $subvalue );
 810                                 }
 811                         } else {
 812                                 $this->store->set( $key, $value );
 813                         }
 814                 }
 815                 $this->store->finishWrite();
 816
 817                 # Clear out the MessageBlobStore
 818                 # HACK: If using a null (i.e. disabled) storage backend, we
 819                 # can't write to the MessageBlobStore either
 820                 if ( !$this->store instanceof LCStore_Null ) {
 821                         MessageBlobStore::clear();
 822                 }
 823
 824                 wfProfileOut( __METHOD__ );
 825         }
 826
 827         /**
 828          * Build the preload item from the given pre-cache data.
 829          *
 830          * The preload item will be loaded automatically, improving performance
 831          * for the commonly-requested items it contains.
 832          * @param $data
 833          * @return array
 834          */
 835         protected function buildPreload( $data ) {
 836                 $preload = array( 'messages' => array() );
 837                 foreach ( self::$preloadedKeys as $key ) {
 838                         $preload[$key] = $data[$key];
 839                 }
 840
 841                 foreach ( $data['preloadedMessages'] as $subkey ) {
 842                         if ( isset( $data['messages'][$subkey] ) ) {
 843                                 $subitem = $data['messages'][$subkey];
 844                         } else {
 845                                 $subitem = null;
 846                         }
 847                         $preload['messages'][$subkey] = $subitem;
 848                 }
 849
 850                 return $preload;
 851         }
 852
 853         /**
 854          * Unload the data for a given language from the object cache.
 855          * Reduces memory usage.
 856          * @param $code
 857          */
 858         public function unload( $code ) {
 859                 unset( $this->data[$code] );
 860                 unset( $this->loadedItems[$code] );
 861                 unset( $this->loadedSubitems[$code] );
 862                 unset( $this->initialisedLangs[$code] );
 863
 864                 foreach ( $this->shallowFallbacks as $shallowCode => $fbCode ) {
 865                         if ( $fbCode === $code ) {
 866                                 $this->unload( $shallowCode );
 867                         }
 868                 }
 869         }
 870
 871         /**
 872          * Unload all data
 873          */
 874         public function unloadAll() {
 875                 foreach ( $this->initialisedLangs as $lang => $unused ) {
 876                         $this->unload( $lang );
 877                 }
 878         }
 879
 880         /**
 881          * Disable the storage backend
 882          */
 883         public function disableBackend() {
 884                 $this->store = new LCStore_Null;
 885                 $this->manualRecache = false;
 886         }
 887 }
 888
 889 /**
 890  * Interface for the persistence layer of LocalisationCache.
 891  *
 892  * The persistence layer is two-level hierarchical cache. The first level
 893  * is the language, the second level is the item or subitem.
 894  *
 895  * Since the data for a whole language is rebuilt in one operation, it needs
 896  * to have a fast and atomic method for deleting or replacing all of the
 897  * current data for a given language. The interface reflects this bulk update
 898  * operation. Callers writing to the cache must first call startWrite(), then
 899  * will call set() a couple of thousand times, then will call finishWrite()
 900  * to commit the operation. When finishWrite() is called, the cache is
 901  * expected to delete all data previously stored for that language.
 902  *
 903  * The values stored are PHP variables suitable for serialize(). Implementations
 904  * of LCStore are responsible for serializing and unserializing.
 905  */
 906 interface LCStore {
 907         /**
 908          * Get a value.
 909          * @param $code string Language code
 910          * @param $key string Cache key
 911          */
 912         function get( $code, $key );
 913
 914         /**
 915          * Start a write transaction.
 916          * @param $code Language code
 917          */
 918         function startWrite( $code );
 919
 920         /**
 921          * Finish a write transaction.
 922          */
 923         function finishWrite();
 924
 925         /**
 926          * Set a key to a given value. startWrite() must be called before this
 927          * is called, and finishWrite() must be called afterwards.
 928          * @param $key
 929          * @param $value
 930          */
 931         function set( $key, $value );
 932 }
 933
 934 /**
 935  * LCStore implementation which uses PHP accelerator to store data.
 936  * This will work if one of XCache, WinCache or APC cacher is configured.
 937  * (See ObjectCache.php)
 938  */
 939 class LCStore_Accel implements LCStore {
 940         var $currentLang;
 941         var $keys;
 942
 943         public function __construct() {
 944                 $this->cache = wfGetCache( CACHE_ACCEL );
 945         }
 946
 947         public function get( $code, $key ) {
 948                 $k = wfMemcKey( 'l10n', $code, 'k', $key );
 949                 $r = $this->cache->get( $k );
 950                 return $r === false ? null : $r;
 951         }
 952
 953         public function startWrite( $code ) {
 954                 $k = wfMemcKey( 'l10n', $code, 'l' );
 955                 $keys = $this->cache->get( $k );
 956                 if ( $keys ) {
 957                         foreach ( $keys as $k ) {
 958                                 $this->cache->delete( $k );
 959                         }
 960                 }
 961                 $this->currentLang = $code;
 962                 $this->keys = array();
 963         }
 964
 965         public function finishWrite() {
 966                 if ( $this->currentLang ) {
 967                         $k = wfMemcKey( 'l10n', $this->currentLang, 'l' );
 968                         $this->cache->set( $k, array_keys( $this->keys ) );
 969                 }
 970                 $this->currentLang = null;
 971                 $this->keys = array();
 972         }
 973
 974         public function set( $key, $value ) {
 975                 if ( $this->currentLang ) {
 976                         $k = wfMemcKey( 'l10n', $this->currentLang, 'k', $key );
 977                         $this->keys[$k] = true;
 978                         $this->cache->set( $k, $value );
 979                 }
 980         }
 981 }
 982
 983 /**
 984  * LCStore implementation which uses the standard DB functions to store data.
 985  * This will work on any MediaWiki installation.
 986  */
 987 class LCStore_DB implements LCStore {
 988         var $currentLang;
 989         var $writesDone = false;
 990
 991         /**
 992          * @var DatabaseBase
 993          */
 994         var $dbw;
 995         var $batch;
 996         var $readOnly = false;
 997
 998         public function get( $code, $key ) {
 999                 if ( $this->writesDone ) {
1000                         $db = wfGetDB( DB_MASTER );
1001                 } else {
1002                         $db = wfGetDB( DB_SLAVE );
1003                 }
1004                 $row = $db->selectRow( 'l10n_cache', array( 'lc_value' ),
1005                         array( 'lc_lang' => $code, 'lc_key' => $key ), __METHOD__ );
1006                 if ( $row ) {
1007                         return unserialize( $row->lc_value );
1008                 } else {
1009                         return null;
1010                 }
1011         }
1012
1013         public function startWrite( $code ) {
1014                 if ( $this->readOnly ) {
1015                         return;
1016                 }
1017
1018                 if ( !$code ) {
1019                         throw new MWException( __METHOD__ . ": Invalid language \"$code\"" );
1020                 }
1021
1022                 $this->dbw = wfGetDB( DB_MASTER );
1023                 try {
1024                         $this->dbw->begin( __METHOD__ );
1025                         $this->dbw->delete( 'l10n_cache', array( 'lc_lang' => $code ), __METHOD__ );
1026                 } catch ( DBQueryError $e ) {
1027                         if ( $this->dbw->wasReadOnlyError() ) {
1028                                 $this->readOnly = true;
1029                                 $this->dbw->rollback( __METHOD__ );
1030                                 $this->dbw->ignoreErrors( false );
1031                                 return;
1032                         } else {
1033                                 throw $e;
1034                         }
1035                 }
1036
1037                 $this->currentLang = $code;
1038                 $this->batch = array();
1039         }
1040
1041         public function finishWrite() {
1042                 if ( $this->readOnly ) {
1043                         return;
1044                 }
1045
1046                 if ( $this->batch ) {
1047                         $this->dbw->insert( 'l10n_cache', $this->batch, __METHOD__ );
1048                 }
1049
1050                 $this->dbw->commit( __METHOD__ );
1051                 $this->currentLang = null;
1052                 $this->dbw = null;
1053                 $this->batch = array();
1054                 $this->writesDone = true;
1055         }
1056
1057         public function set( $key, $value ) {
1058                 if ( $this->readOnly ) {
1059                         return;
1060                 }
1061
1062                 if ( is_null( $this->currentLang ) ) {
1063                         throw new MWException( __CLASS__ . ': must call startWrite() before calling set()' );
1064                 }
1065
1066                 $this->batch[] = array(
1067                         'lc_lang' => $this->currentLang,
1068                         'lc_key' => $key,
1069                         'lc_value' => serialize( $value ) );
1070
1071                 if ( count( $this->batch ) >= 100 ) {
1072                         $this->dbw->insert( 'l10n_cache', $this->batch, __METHOD__ );
1073                         $this->batch = array();
1074                 }
1075         }
1076 }
1077
1078 /**
1079  * LCStore implementation which stores data as a collection of CDB files in the
1080  * directory given by $wgCacheDirectory. If $wgCacheDirectory is not set, this
1081  * will throw an exception.
1082  *
1083  * Profiling indicates that on Linux, this implementation outperforms MySQL if
1084  * the directory is on a local filesystem and there is ample kernel cache
1085  * space. The performance advantage is greater when the DBA extension is
1086  * available than it is with the PHP port.
1087  *
1088  * See Cdb.php and http://cr.yp.to/cdb.html
1089  */
1090 class LCStore_CDB implements LCStore {
1091         var $readers, $writer, $currentLang, $directory;
1092
1093         function __construct( $conf = array() ) {
1094                 global $wgCacheDirectory;
1095
1096                 if ( isset( $conf['directory'] ) ) {
1097                         $this->directory = $conf['directory'];
1098                 } else {
1099                         $this->directory = $wgCacheDirectory;
1100                 }
1101         }
1102
1103         public function get( $code, $key ) {
1104                 if ( !isset( $this->readers[$code] ) ) {
1105                         $fileName = $this->getFileName( $code );
1106
1107                         if ( !file_exists( $fileName ) ) {
1108                                 $this->readers[$code] = false;
1109                         } else {
1110                                 $this->readers[$code] = CdbReader::open( $fileName );
1111                         }
1112                 }
1113
1114                 if ( !$this->readers[$code] ) {
1115                         return null;
1116                 } else {
1117                         $value = $this->readers[$code]->get( $key );
1118
1119                         if ( $value === false ) {
1120                                 return null;
1121                         }
1122                         return unserialize( $value );
1123                 }
1124         }
1125
1126         public function startWrite( $code ) {
1127                 if ( !file_exists( $this->directory ) ) {
1128                         if ( !wfMkdirParents( $this->directory, null, __METHOD__ ) ) {
1129                                 throw new MWException( "Unable to create the localisation store " .
1130                                         "directory \"{$this->directory}\"" );
1131                         }
1132                 }
1133
1134                 // Close reader to stop permission errors on write
1135                 if ( !empty( $this->readers[$code] ) ) {
1136                         $this->readers[$code]->close();
1137                 }
1138
1139                 $this->writer = CdbWriter::open( $this->getFileName( $code ) );
1140                 $this->currentLang = $code;
1141         }
1142
1143         public function finishWrite() {
1144                 // Close the writer
1145                 $this->writer->close();
1146                 $this->writer = null;
1147                 unset( $this->readers[$this->currentLang] );
1148                 $this->currentLang = null;
1149         }
1150
1151         public function set( $key, $value ) {
1152                 if ( is_null( $this->writer ) ) {
1153                         throw new MWException( __CLASS__ . ': must call startWrite() before calling set()' );
1154                 }
1155                 $this->writer->set( $key, serialize( $value ) );
1156         }
1157
1158         protected function getFileName( $code ) {
1159                 if ( !$code || strpos( $code, '/' ) !== false ) {
1160                         throw new MWException( __METHOD__ . ": Invalid language \"$code\"" );
1161                 }
1162                 return "{$this->directory}/l10n_cache-$code.cdb";
1163         }
1164 }
1165
1166 /**
1167  * Null store backend, used to avoid DB errors during install
1168  */
1169 class LCStore_Null implements LCStore {
1170         public function get( $code, $key ) {
1171                 return null;
1172         }
1173
1174         public function startWrite( $code ) {}
1175         public function finishWrite() {}
1176         public function set( $key, $value ) {}
1177 }
1178
1179 /**
1180  * A localisation cache optimised for loading large amounts of data for many
1181  * languages. Used by rebuildLocalisationCache.php.
1182  */
1183 class LocalisationCache_BulkLoad extends LocalisationCache {
1184         /**
1185          * A cache of the contents of data files.
1186          * Core files are serialized to avoid using ~1GB of RAM during a recache.
1187          */
1188         var $fileCache = array();
1189
1190         /**
1191          * Most recently used languages. Uses the linked-list aspect of PHP hashtables
1192          * to keep the most recently used language codes at the end of the array, and
1193          * the language codes that are ready to be deleted at the beginning.
1194          */
1195         var $mruLangs = array();
1196
1197         /**
1198          * Maximum number of languages that may be loaded into $this->data
1199          */
1200         var $maxLoadedLangs = 10;
1201
1202         /**
1203          * @param $fileName
1204          * @param $fileType
1205          * @return array|mixed
1206          */
1207         protected function readPHPFile( $fileName, $fileType ) {
1208                 $serialize = $fileType === 'core';
1209                 if ( !isset( $this->fileCache[$fileName][$fileType] ) ) {
1210                         $data = parent::readPHPFile( $fileName, $fileType );
1211
1212                         if ( $serialize ) {
1213                                 $encData = serialize( $data );
1214                         } else {
1215                                 $encData = $data;
1216                         }
1217
1218                         $this->fileCache[$fileName][$fileType] = $encData;
1219
1220                         return $data;
1221                 } elseif ( $serialize ) {
1222                         return unserialize( $this->fileCache[$fileName][$fileType] );
1223                 } else {
1224                         return $this->fileCache[$fileName][$fileType];
1225                 }
1226         }
1227
1228         /**
1229          * @param $code
1230          * @param $key
1231          * @return mixed
1232          */
1233         public function getItem( $code, $key ) {
1234                 unset( $this->mruLangs[$code] );
1235                 $this->mruLangs[$code] = true;
1236                 return parent::getItem( $code, $key );
1237         }
1238
1239         /**
1240          * @param $code
1241          * @param $key
1242          * @param $subkey
1243          * @return
1244          */
1245         public function getSubitem( $code, $key, $subkey ) {
1246                 unset( $this->mruLangs[$code] );
1247                 $this->mruLangs[$code] = true;
1248                 return parent::getSubitem( $code, $key, $subkey );
1249         }
1250
1251         /**
1252          * @param $code
1253          */
1254         public function recache( $code ) {
1255                 parent::recache( $code );
1256                 unset( $this->mruLangs[$code] );
1257                 $this->mruLangs[$code] = true;
1258                 $this->trimCache();
1259         }
1260
1261         /**
1262          * @param $code
1263          */
1264         public function unload( $code ) {
1265                 unset( $this->mruLangs[$code] );
1266                 parent::unload( $code );
1267         }
1268
1269         /**
1270          * Unload cached languages until there are less than $this->maxLoadedLangs
1271          */
1272         protected function trimCache() {
1273                 while ( count( $this->data ) > $this->maxLoadedLangs && count( $this->mruLangs ) ) {
1274                         reset( $this->mruLangs );
1275                         $code = key( $this->mruLangs );
1276                         wfDebug( __METHOD__ . ": unloading $code\n" );
1277                         $this->unload( $code );
1278                 }
1279         }
1280
1281 }