3 * Cache of the contents of localisation files.
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.
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.
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
23 use Cdb\Exception
as CdbException
;
24 use Cdb\Reader
as CdbReader
;
25 use Cdb\Writer
as CdbWriter
;
26 use CLDRPluralRuleParser\Evaluator
;
29 * Class for caching the contents of localisation files, Messages*.php
32 * An instance of this class is available using Language::getLocalisationCache().
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.
40 class LocalisationCache
{
43 /** Configuration associative array */
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.
51 private $manualRecache = false;
54 * True to treat all files as expired until they are regenerated by this object.
56 private $forceRecache = false;
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.
64 protected $data = array();
67 * The persistent store object. An instance of LCStore.
74 * A 2-d associative array, code/key, where presence indicates that the item
75 * is loaded. Value arbitrary.
77 * For split items, if set, this indicates that all of the subitems have been
80 private $loadedItems = array();
83 * A 3-d associative array, code/key/subkey, where presence indicates that
84 * the subitem is loaded. Only used for the split items, i.e. messages.
86 private $loadedSubitems = array();
89 * An array where presence of a key indicates that that language has been
90 * initialised. Initialisation includes checking for cache expiry and doing
91 * any necessary updates.
93 private $initialisedLangs = array();
96 * An array mapping non-existent pseudo-languages to fallback languages. This
97 * is filled by initShallowFallback() when data is requested from a language
98 * that lacks a Messages*.php file.
100 private $shallowFallbacks = array();
103 * An array where the keys are codes that have been recached by this instance.
105 private $recachedLangs = array();
110 static public $allKeys = array(
111 'fallback', 'namespaceNames', 'bookstoreList',
112 'magicWords', 'messages', 'rtl', 'capitalizeAllNouns', 'digitTransformTable',
113 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
114 'linkTrail', 'linkPrefixCharset', 'namespaceAliases',
115 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
116 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
117 'imageFiles', 'preloadedMessages', 'namespaceGenderAliases',
118 'digitGroupingPattern', 'pluralRules', 'pluralRuleTypes', 'compiledPluralRules',
122 * Keys for items which consist of associative arrays, which may be merged
123 * by a fallback sequence.
125 static public $mergeableMapKeys = array( 'messages', 'namespaceNames',
126 'dateFormats', 'imageFiles', 'preloadedMessages'
130 * Keys for items which are a numbered array.
132 static public $mergeableListKeys = array( 'extraUserToggles' );
135 * Keys for items which contain an array of arrays of equivalent aliases
136 * for each subitem. The aliases may be merged by a fallback sequence.
138 static public $mergeableAliasListKeys = array( 'specialPageAliases' );
141 * Keys for items which contain an associative array, and may be merged if
142 * the primary value contains the special array key "inherit". That array
143 * key is removed after the first merge.
145 static public $optionalMergeKeys = array( 'bookstoreList' );
148 * Keys for items that are formatted like $magicWords
150 static public $magicWordKeys = array( 'magicWords' );
153 * Keys for items where the subitems are stored in the backend separately.
155 static public $splitKeys = array( 'messages' );
158 * Keys which are loaded automatically by initLanguage()
160 static public $preloadedKeys = array( 'dateFormats', 'namespaceNames' );
163 * Associative array of cached plural rules. The key is the language code,
164 * the value is an array of plural rules for that language.
166 private $pluralRules = null;
169 * Associative array of cached plural rule types. The key is the language
170 * code, the value is an array of plural rule types for that language. For
171 * example, $pluralRuleTypes['ar'] = ['zero', 'one', 'two', 'few', 'many'].
172 * The index for each rule type matches the index for the rule in
173 * $pluralRules, thus allowing correlation between the two. The reason we
174 * don't just use the type names as the keys in $pluralRules is because
175 * Language::convertPlural applies the rules based on numeric order (or
176 * explicit numeric parameter), not based on the name of the rule type. For
177 * example, {{plural:count|wordform1|wordform2|wordform3}}, rather than
178 * {{plural:count|one=wordform1|two=wordform2|many=wordform3}}.
180 private $pluralRuleTypes = null;
182 private $mergeableKeys = null;
186 * For constructor parameters, see the documentation in DefaultSettings.php
187 * for $wgLocalisationCacheConf.
190 * @throws MWException
192 function __construct( $conf ) {
193 global $wgCacheDirectory;
196 $storeConf = array();
197 if ( !empty( $conf['storeClass'] ) ) {
198 $storeClass = $conf['storeClass'];
200 switch ( $conf['store'] ) {
203 $storeClass = 'LCStoreCDB';
206 $storeClass = 'LCStoreDB';
209 $storeClass = 'LCStoreStaticArray';
212 $storeClass = $wgCacheDirectory ?
'LCStoreCDB' : 'LCStoreDB';
215 throw new MWException(
216 'Please set $wgLocalisationCacheConf[\'store\'] to something sensible.' );
220 wfDebugLog( 'caches', get_class( $this ) . ": using store $storeClass" );
221 if ( !empty( $conf['storeDirectory'] ) ) {
222 $storeConf['directory'] = $conf['storeDirectory'];
225 $this->store
= new $storeClass( $storeConf );
226 foreach ( array( 'manualRecache', 'forceRecache' ) as $var ) {
227 if ( isset( $conf[$var] ) ) {
228 $this->$var = $conf[$var];
234 * Returns true if the given key is mergeable, that is, if it is an associative
235 * array which can be merged through a fallback sequence.
239 public function isMergeableKey( $key ) {
240 if ( $this->mergeableKeys
=== null ) {
241 $this->mergeableKeys
= array_flip( array_merge(
242 self
::$mergeableMapKeys,
243 self
::$mergeableListKeys,
244 self
::$mergeableAliasListKeys,
245 self
::$optionalMergeKeys,
250 return isset( $this->mergeableKeys
[$key] );
256 * Warning: this may be slow for split items (messages), since it will
257 * need to fetch all of the subitems from the cache individually.
258 * @param string $code
262 public function getItem( $code, $key ) {
263 if ( !isset( $this->loadedItems
[$code][$key] ) ) {
264 $this->loadItem( $code, $key );
267 if ( $key === 'fallback' && isset( $this->shallowFallbacks
[$code] ) ) {
268 return $this->shallowFallbacks
[$code];
271 return $this->data
[$code][$key];
275 * Get a subitem, for instance a single message for a given language.
276 * @param string $code
278 * @param string $subkey
281 public function getSubitem( $code, $key, $subkey ) {
282 if ( !isset( $this->loadedSubitems
[$code][$key][$subkey] ) &&
283 !isset( $this->loadedItems
[$code][$key] )
285 $this->loadSubitem( $code, $key, $subkey );
288 if ( isset( $this->data
[$code][$key][$subkey] ) ) {
289 return $this->data
[$code][$key][$subkey];
296 * Get the list of subitem keys for a given item.
298 * This is faster than array_keys($lc->getItem(...)) for the items listed in
301 * Will return null if the item is not found, or false if the item is not an
303 * @param string $code
305 * @return bool|null|string
307 public function getSubitemList( $code, $key ) {
308 if ( in_array( $key, self
::$splitKeys ) ) {
309 return $this->getSubitem( $code, 'list', $key );
311 $item = $this->getItem( $code, $key );
312 if ( is_array( $item ) ) {
313 return array_keys( $item );
321 * Load an item into the cache.
322 * @param string $code
325 protected function loadItem( $code, $key ) {
326 if ( !isset( $this->initialisedLangs
[$code] ) ) {
327 $this->initLanguage( $code );
330 // Check to see if initLanguage() loaded it for us
331 if ( isset( $this->loadedItems
[$code][$key] ) ) {
335 if ( isset( $this->shallowFallbacks
[$code] ) ) {
336 $this->loadItem( $this->shallowFallbacks
[$code], $key );
341 if ( in_array( $key, self
::$splitKeys ) ) {
342 $subkeyList = $this->getSubitem( $code, 'list', $key );
343 foreach ( $subkeyList as $subkey ) {
344 if ( isset( $this->data
[$code][$key][$subkey] ) ) {
347 $this->data
[$code][$key][$subkey] = $this->getSubitem( $code, $key, $subkey );
350 $this->data
[$code][$key] = $this->store
->get( $code, $key );
353 $this->loadedItems
[$code][$key] = true;
357 * Load a subitem into the cache
358 * @param string $code
360 * @param string $subkey
362 protected function loadSubitem( $code, $key, $subkey ) {
363 if ( !in_array( $key, self
::$splitKeys ) ) {
364 $this->loadItem( $code, $key );
369 if ( !isset( $this->initialisedLangs
[$code] ) ) {
370 $this->initLanguage( $code );
373 // Check to see if initLanguage() loaded it for us
374 if ( isset( $this->loadedItems
[$code][$key] ) ||
375 isset( $this->loadedSubitems
[$code][$key][$subkey] )
380 if ( isset( $this->shallowFallbacks
[$code] ) ) {
381 $this->loadSubitem( $this->shallowFallbacks
[$code], $key, $subkey );
386 $value = $this->store
->get( $code, "$key:$subkey" );
387 $this->data
[$code][$key][$subkey] = $value;
388 $this->loadedSubitems
[$code][$key][$subkey] = true;
392 * Returns true if the cache identified by $code is missing or expired.
394 * @param string $code
398 public function isExpired( $code ) {
399 if ( $this->forceRecache
&& !isset( $this->recachedLangs
[$code] ) ) {
400 wfDebug( __METHOD__
. "($code): forced reload\n" );
405 $deps = $this->store
->get( $code, 'deps' );
406 $keys = $this->store
->get( $code, 'list' );
407 $preload = $this->store
->get( $code, 'preload' );
408 // Different keys may expire separately for some stores
409 if ( $deps === null ||
$keys === null ||
$preload === null ) {
410 wfDebug( __METHOD__
. "($code): cache missing, need to make one\n" );
415 foreach ( $deps as $dep ) {
416 // Because we're unserializing stuff from cache, we
417 // could receive objects of classes that don't exist
418 // anymore (e.g. uninstalled extensions)
419 // When this happens, always expire the cache
420 if ( !$dep instanceof CacheDependency ||
$dep->isExpired() ) {
421 wfDebug( __METHOD__
. "($code): cache for $code expired due to " .
422 get_class( $dep ) . "\n" );
432 * Initialise a language in this object. Rebuild the cache if necessary.
433 * @param string $code
434 * @throws MWException
436 protected function initLanguage( $code ) {
437 if ( isset( $this->initialisedLangs
[$code] ) ) {
441 $this->initialisedLangs
[$code] = true;
443 # If the code is of the wrong form for a Messages*.php file, do a shallow fallback
444 if ( !Language
::isValidBuiltInCode( $code ) ) {
445 $this->initShallowFallback( $code, 'en' );
450 # Recache the data if necessary
451 if ( !$this->manualRecache
&& $this->isExpired( $code ) ) {
452 if ( Language
::isSupportedLanguage( $code ) ) {
453 $this->recache( $code );
454 } elseif ( $code === 'en' ) {
455 throw new MWException( 'MessagesEn.php is missing.' );
457 $this->initShallowFallback( $code, 'en' );
464 $preload = $this->getItem( $code, 'preload' );
465 if ( $preload === null ) {
466 if ( $this->manualRecache
) {
467 // No Messages*.php file. Do shallow fallback to en.
468 if ( $code === 'en' ) {
469 throw new MWException( 'No localisation cache found for English. ' .
470 'Please run maintenance/rebuildLocalisationCache.php.' );
472 $this->initShallowFallback( $code, 'en' );
476 throw new MWException( 'Invalid or missing localisation cache.' );
479 $this->data
[$code] = $preload;
480 foreach ( $preload as $key => $item ) {
481 if ( in_array( $key, self
::$splitKeys ) ) {
482 foreach ( $item as $subkey => $subitem ) {
483 $this->loadedSubitems
[$code][$key][$subkey] = true;
486 $this->loadedItems
[$code][$key] = true;
492 * Create a fallback from one language to another, without creating a
493 * complete persistent cache.
494 * @param string $primaryCode
495 * @param string $fallbackCode
497 public function initShallowFallback( $primaryCode, $fallbackCode ) {
498 $this->data
[$primaryCode] =& $this->data
[$fallbackCode];
499 $this->loadedItems
[$primaryCode] =& $this->loadedItems
[$fallbackCode];
500 $this->loadedSubitems
[$primaryCode] =& $this->loadedSubitems
[$fallbackCode];
501 $this->shallowFallbacks
[$primaryCode] = $fallbackCode;
505 * Read a PHP file containing localisation data.
506 * @param string $_fileName
507 * @param string $_fileType
508 * @throws MWException
511 protected function readPHPFile( $_fileName, $_fileType ) {
512 // Disable APC caching
513 MediaWiki\
suppressWarnings();
514 $_apcEnabled = ini_set( 'apc.cache_by_default', '0' );
515 MediaWiki\restoreWarnings
();
519 MediaWiki\
suppressWarnings();
520 ini_set( 'apc.cache_by_default', $_apcEnabled );
521 MediaWiki\restoreWarnings
();
523 if ( $_fileType == 'core' ||
$_fileType == 'extension' ) {
524 $data = compact( self
::$allKeys );
525 } elseif ( $_fileType == 'aliases' ) {
526 $data = compact( 'aliases' );
528 throw new MWException( __METHOD__
. ": Invalid file type: $_fileType" );
535 * Read a JSON file containing localisation messages.
536 * @param string $fileName Name of file to read
537 * @throws MWException If there is a syntax error in the JSON file
538 * @return array Array with a 'messages' key, or empty array if the file doesn't exist
540 public function readJSONFile( $fileName ) {
542 if ( !is_readable( $fileName ) ) {
546 $json = file_get_contents( $fileName );
547 if ( $json === false ) {
551 $data = FormatJson
::decode( $json, true );
552 if ( $data === null ) {
554 throw new MWException( __METHOD__
. ": Invalid JSON file: $fileName" );
557 // Remove keys starting with '@', they're reserved for metadata and non-message data
558 foreach ( $data as $key => $unused ) {
559 if ( $key === '' ||
$key[0] === '@' ) {
560 unset( $data[$key] );
564 // The JSON format only supports messages, none of the other variables, so wrap the data
565 return array( 'messages' => $data );
569 * Get the compiled plural rules for a given language from the XML files.
571 * @param string $code
574 public function getCompiledPluralRules( $code ) {
575 $rules = $this->getPluralRules( $code );
576 if ( $rules === null ) {
580 $compiledRules = Evaluator
::compile( $rules );
581 } catch ( CLDRPluralRuleError
$e ) {
582 wfDebugLog( 'l10n', $e->getMessage() );
587 return $compiledRules;
591 * Get the plural rules for a given language from the XML files.
594 * @param string $code
597 public function getPluralRules( $code ) {
598 if ( $this->pluralRules
=== null ) {
599 $this->loadPluralFiles();
601 if ( !isset( $this->pluralRules
[$code] ) ) {
604 return $this->pluralRules
[$code];
609 * Get the plural rule types for a given language from the XML files.
612 * @param string $code
615 public function getPluralRuleTypes( $code ) {
616 if ( $this->pluralRuleTypes
=== null ) {
617 $this->loadPluralFiles();
619 if ( !isset( $this->pluralRuleTypes
[$code] ) ) {
622 return $this->pluralRuleTypes
[$code];
627 * Load the plural XML files.
629 protected function loadPluralFiles() {
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 );
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.
645 * @param string $fileName
646 * @throws MWException
648 protected function loadPluralFile( $fileName ) {
649 // Use file_get_contents instead of DOMDocument::load (T58439)
650 $xml = file_get_contents( $fileName );
652 throw new MWException( "Unable to read plurals file $fileName" );
654 $doc = new DOMDocument
;
655 $doc->loadXML( $xml );
656 $rulesets = $doc->getElementsByTagName( "pluralRules" );
657 foreach ( $rulesets as $ruleset ) {
658 $codes = $ruleset->getAttribute( 'locales' );
660 $ruleTypes = array();
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
668 $rules[] = $elt->nodeValue
;
669 $ruleTypes[] = $ruleType;
671 foreach ( explode( ' ', $codes ) as $code ) {
672 $this->pluralRules
[$code] = $rules;
673 $this->pluralRuleTypes
[$code] = $ruleTypes;
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.
683 * @param string $code
687 protected function readSourceFilesAndRegisterDeps( $code, &$deps ) {
690 // This reads in the PHP i18n file with non-messages l10n data
691 $fileName = Language
::getMessagesFileName( $code );
692 if ( !file_exists( $fileName ) ) {
695 $deps[] = new FileDependency( $fileName );
696 $data = $this->readPHPFile( $fileName, 'core' );
699 # Load CLDR plural rules for JavaScript
700 $data['pluralRules'] = $this->getPluralRules( $code );
702 $data['compiledPluralRules'] = $this->getCompiledPluralRules( $code );
703 # Load plural rule types
704 $data['pluralRuleTypes'] = $this->getPluralRuleTypes( $code );
706 $deps['plurals'] = new FileDependency( "$IP/languages/data/plurals.xml" );
707 $deps['plurals-mw'] = new FileDependency( "$IP/languages/data/plurals-mediawiki.xml" );
713 * Merge two localisation values, a primary and a fallback, overwriting the
714 * primary value in place.
716 * @param mixed $value
717 * @param mixed $fallbackValue
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 $value = array_unique( array_merge( $fallbackValue, $value ) );
726 } elseif ( in_array( $key, self
::$mergeableAliasListKeys ) ) {
727 $value = array_merge_recursive( $value, $fallbackValue );
728 } elseif ( in_array( $key, self
::$optionalMergeKeys ) ) {
729 if ( !empty( $value['inherit'] ) ) {
730 $value = array_merge( $fallbackValue, $value );
733 if ( isset( $value['inherit'] ) ) {
734 unset( $value['inherit'] );
736 } elseif ( in_array( $key, self
::$magicWordKeys ) ) {
737 $this->mergeMagicWords( $value, $fallbackValue );
741 $value = $fallbackValue;
746 * @param mixed $value
747 * @param mixed $fallbackValue
749 protected function mergeMagicWords( &$value, $fallbackValue ) {
750 foreach ( $fallbackValue as $magicName => $fallbackInfo ) {
751 if ( !isset( $value[$magicName] ) ) {
752 $value[$magicName] = $fallbackInfo;
754 $oldSynonyms = array_slice( $fallbackInfo, 1 );
755 $newSynonyms = array_slice( $value[$magicName], 1 );
756 $synonyms = array_values( array_unique( array_merge(
757 $newSynonyms, $oldSynonyms ) ) );
758 $value[$magicName] = array_merge( array( $fallbackInfo[0] ), $synonyms );
764 * Given an array mapping language code to localisation value, such as is
765 * found in extension *.i18n.php files, iterate through a fallback sequence
766 * to merge the given data with an existing primary value.
768 * Returns true if any data from the extension array was used, false
770 * @param array $codeSequence
772 * @param mixed $value
773 * @param mixed $fallbackValue
776 protected function mergeExtensionItem( $codeSequence, $key, &$value, $fallbackValue ) {
778 foreach ( $codeSequence as $code ) {
779 if ( isset( $fallbackValue[$code] ) ) {
780 $this->mergeItem( $key, $value, $fallbackValue[$code] );
789 * Gets the combined list of messages dirs from
790 * core and extensions
795 public function getMessagesDirs() {
796 global $wgMessagesDirs, $IP;
798 'core' => "$IP/languages/i18n",
799 'api' => "$IP/includes/api/i18n",
800 'oojs-ui' => "$IP/resources/lib/oojs-ui/i18n",
805 * Load localisation data for a given language for both core and extensions
806 * and save it to the persistent cache store and the process cache
807 * @param string $code
808 * @throws MWException
810 public function recache( $code ) {
811 global $wgExtensionMessagesFiles;
814 throw new MWException( "Invalid language code requested" );
816 $this->recachedLangs
[$code] = true;
819 $initialData = array_combine(
821 array_fill( 0, count( self
::$allKeys ), null ) );
822 $coreData = $initialData;
825 # Load the primary localisation from the source file
826 $data = $this->readSourceFilesAndRegisterDeps( $code, $deps );
827 if ( $data === false ) {
828 wfDebug( __METHOD__
. ": no localisation file for $code, using fallback to en\n" );
829 $coreData['fallback'] = 'en';
831 wfDebug( __METHOD__
. ": got localisation for $code from source\n" );
833 # Merge primary localisation
834 foreach ( $data as $key => $value ) {
835 $this->mergeItem( $key, $coreData[$key], $value );
839 # Fill in the fallback if it's not there already
840 if ( is_null( $coreData['fallback'] ) ) {
841 $coreData['fallback'] = $code === 'en' ?
false : 'en';
843 if ( $coreData['fallback'] === false ) {
844 $coreData['fallbackSequence'] = array();
846 $coreData['fallbackSequence'] = array_map( 'trim', explode( ',', $coreData['fallback'] ) );
847 $len = count( $coreData['fallbackSequence'] );
849 # Ensure that the sequence ends at en
850 if ( $coreData['fallbackSequence'][$len - 1] !== 'en' ) {
851 $coreData['fallbackSequence'][] = 'en';
855 $codeSequence = array_merge( array( $code ), $coreData['fallbackSequence'] );
856 $messageDirs = $this->getMessagesDirs();
858 # Load non-JSON localisation data for extensions
859 $extensionData = array_combine(
861 array_fill( 0, count( $codeSequence ), $initialData ) );
862 foreach ( $wgExtensionMessagesFiles as $extension => $fileName ) {
863 if ( isset( $messageDirs[$extension] ) ) {
864 # This extension has JSON message data; skip the PHP shim
868 $data = $this->readPHPFile( $fileName, 'extension' );
871 foreach ( $data as $key => $item ) {
872 foreach ( $codeSequence as $csCode ) {
873 if ( isset( $item[$csCode] ) ) {
874 $this->mergeItem( $key, $extensionData[$csCode][$key], $item[$csCode] );
881 $deps[] = new FileDependency( $fileName );
885 # Load the localisation data for each fallback, then merge it into the full array
886 $allData = $initialData;
887 foreach ( $codeSequence as $csCode ) {
888 $csData = $initialData;
890 # Load core messages and the extension localisations.
891 foreach ( $messageDirs as $dirs ) {
892 foreach ( (array)$dirs as $dir ) {
893 $fileName = "$dir/$csCode.json";
894 $data = $this->readJSONFile( $fileName );
896 foreach ( $data as $key => $item ) {
897 $this->mergeItem( $key, $csData[$key], $item );
900 $deps[] = new FileDependency( $fileName );
904 # Merge non-JSON extension data
905 if ( isset( $extensionData[$csCode] ) ) {
906 foreach ( $extensionData[$csCode] as $key => $item ) {
907 $this->mergeItem( $key, $csData[$key], $item );
911 if ( $csCode === $code ) {
912 # Merge core data into extension data
913 foreach ( $coreData as $key => $item ) {
914 $this->mergeItem( $key, $csData[$key], $item );
917 # Load the secondary localisation from the source file to
918 # avoid infinite cycles on cyclic fallbacks
919 $fbData = $this->readSourceFilesAndRegisterDeps( $csCode, $deps );
920 if ( $fbData !== false ) {
921 # Only merge the keys that make sense to merge
922 foreach ( self
::$allKeys as $key ) {
923 if ( !isset( $fbData[$key] ) ) {
927 if ( is_null( $coreData[$key] ) ||
$this->isMergeableKey( $key ) ) {
928 $this->mergeItem( $key, $csData[$key], $fbData[$key] );
934 # Allow extensions an opportunity to adjust the data for this
936 Hooks
::run( 'LocalisationCacheRecacheFallback', array( $this, $csCode, &$csData ) );
938 # Merge the data for this fallback into the final array
939 if ( $csCode === $code ) {
942 foreach ( self
::$allKeys as $key ) {
943 if ( !isset( $csData[$key] ) ) {
947 if ( is_null( $allData[$key] ) ||
$this->isMergeableKey( $key ) ) {
948 $this->mergeItem( $key, $allData[$key], $csData[$key] );
954 # Add cache dependencies for any referenced globals
955 $deps['wgExtensionMessagesFiles'] = new GlobalDependency( 'wgExtensionMessagesFiles' );
956 // $wgMessagesDirs is used in LocalisationCache::getMessagesDirs()
957 $deps['wgMessagesDirs'] = new GlobalDependency( 'wgMessagesDirs' );
958 $deps['version'] = new ConstantDependency( 'LocalisationCache::VERSION' );
960 # Add dependencies to the cache entry
961 $allData['deps'] = $deps;
963 # Replace spaces with underscores in namespace names
964 $allData['namespaceNames'] = str_replace( ' ', '_', $allData['namespaceNames'] );
966 # And do the same for special page aliases. $page is an array.
967 foreach ( $allData['specialPageAliases'] as &$page ) {
968 $page = str_replace( ' ', '_', $page );
970 # Decouple the reference to prevent accidental damage
973 # If there were no plural rules, return an empty array
974 if ( $allData['pluralRules'] === null ) {
975 $allData['pluralRules'] = array();
977 if ( $allData['compiledPluralRules'] === null ) {
978 $allData['compiledPluralRules'] = array();
980 # If there were no plural rule types, return an empty array
981 if ( $allData['pluralRuleTypes'] === null ) {
982 $allData['pluralRuleTypes'] = array();
986 $allData['list'] = array();
987 foreach ( self
::$splitKeys as $key ) {
988 $allData['list'][$key] = array_keys( $allData[$key] );
992 Hooks
::run( 'LocalisationCacheRecache', array( $this, $code, &$allData, &$purgeBlobs ) );
994 if ( is_null( $allData['namespaceNames'] ) ) {
995 throw new MWException( __METHOD__
. ': Localisation data failed sanity check! ' .
996 'Check that your languages/messages/MessagesEn.php file is intact.' );
999 # Set the preload key
1000 $allData['preload'] = $this->buildPreload( $allData );
1002 # Save to the process cache and register the items loaded
1003 $this->data
[$code] = $allData;
1004 foreach ( $allData as $key => $item ) {
1005 $this->loadedItems
[$code][$key] = true;
1008 # Save to the persistent cache
1009 $this->store
->startWrite( $code );
1010 foreach ( $allData as $key => $value ) {
1011 if ( in_array( $key, self
::$splitKeys ) ) {
1012 foreach ( $value as $subkey => $subvalue ) {
1013 $this->store
->set( "$key:$subkey", $subvalue );
1016 $this->store
->set( $key, $value );
1019 $this->store
->finishWrite();
1021 # Clear out the MessageBlobStore
1022 # HACK: If using a null (i.e. disabled) storage backend, we
1023 # can't write to the MessageBlobStore either
1024 if ( $purgeBlobs && !$this->store
instanceof LCStoreNull
) {
1025 $blobStore = new MessageBlobStore();
1026 $blobStore->clear();
1032 * Build the preload item from the given pre-cache data.
1034 * The preload item will be loaded automatically, improving performance
1035 * for the commonly-requested items it contains.
1036 * @param array $data
1039 protected function buildPreload( $data ) {
1040 $preload = array( 'messages' => array() );
1041 foreach ( self
::$preloadedKeys as $key ) {
1042 $preload[$key] = $data[$key];
1045 foreach ( $data['preloadedMessages'] as $subkey ) {
1046 if ( isset( $data['messages'][$subkey] ) ) {
1047 $subitem = $data['messages'][$subkey];
1051 $preload['messages'][$subkey] = $subitem;
1058 * Unload the data for a given language from the object cache.
1059 * Reduces memory usage.
1060 * @param string $code
1062 public function unload( $code ) {
1063 unset( $this->data
[$code] );
1064 unset( $this->loadedItems
[$code] );
1065 unset( $this->loadedSubitems
[$code] );
1066 unset( $this->initialisedLangs
[$code] );
1067 unset( $this->shallowFallbacks
[$code] );
1069 foreach ( $this->shallowFallbacks
as $shallowCode => $fbCode ) {
1070 if ( $fbCode === $code ) {
1071 $this->unload( $shallowCode );
1079 public function unloadAll() {
1080 foreach ( $this->initialisedLangs
as $lang => $unused ) {
1081 $this->unload( $lang );
1086 * Disable the storage backend
1088 public function disableBackend() {
1089 $this->store
= new LCStoreNull
;
1090 $this->manualRecache
= false;
1095 * Interface for the persistence layer of LocalisationCache.
1097 * The persistence layer is two-level hierarchical cache. The first level
1098 * is the language, the second level is the item or subitem.
1100 * Since the data for a whole language is rebuilt in one operation, it needs
1101 * to have a fast and atomic method for deleting or replacing all of the
1102 * current data for a given language. The interface reflects this bulk update
1103 * operation. Callers writing to the cache must first call startWrite(), then
1104 * will call set() a couple of thousand times, then will call finishWrite()
1105 * to commit the operation. When finishWrite() is called, the cache is
1106 * expected to delete all data previously stored for that language.
1108 * The values stored are PHP variables suitable for serialize(). Implementations
1109 * of LCStore are responsible for serializing and unserializing.
1114 * @param string $code Language code
1115 * @param string $key Cache key
1117 function get( $code, $key );
1120 * Start a write transaction.
1121 * @param string $code Language code
1123 function startWrite( $code );
1126 * Finish a write transaction.
1128 function finishWrite();
1131 * Set a key to a given value. startWrite() must be called before this
1132 * is called, and finishWrite() must be called afterwards.
1133 * @param string $key
1134 * @param mixed $value
1136 function set( $key, $value );
1140 * LCStore implementation which uses the standard DB functions to store data.
1141 * This will work on any MediaWiki installation.
1143 class LCStoreDB
implements LCStore
{
1144 private $currentLang;
1145 private $writesDone = false;
1147 /** @var DatabaseBase */
1150 private $batch = array();
1152 private $readOnly = false;
1154 public function get( $code, $key ) {
1155 if ( $this->writesDone
) {
1156 $db = wfGetDB( DB_MASTER
);
1158 $db = wfGetDB( DB_SLAVE
);
1160 $row = $db->selectRow( 'l10n_cache', array( 'lc_value' ),
1161 array( 'lc_lang' => $code, 'lc_key' => $key ), __METHOD__
);
1163 return unserialize( $db->decodeBlob( $row->lc_value
) );
1169 public function startWrite( $code ) {
1170 if ( $this->readOnly
) {
1172 } elseif ( !$code ) {
1173 throw new MWException( __METHOD__
. ": Invalid language \"$code\"" );
1176 $this->dbw
= wfGetDB( DB_MASTER
);
1178 $this->currentLang
= $code;
1179 $this->batch
= array();
1182 public function finishWrite() {
1183 if ( $this->readOnly
) {
1185 } elseif ( is_null( $this->currentLang
) ) {
1186 throw new MWException( __CLASS__
. ': must call startWrite() before finishWrite()' );
1189 $this->dbw
->begin( __METHOD__
);
1191 $this->dbw
->delete( 'l10n_cache',
1192 array( 'lc_lang' => $this->currentLang
), __METHOD__
);
1193 foreach ( array_chunk( $this->batch
, 500 ) as $rows ) {
1194 $this->dbw
->insert( 'l10n_cache', $rows, __METHOD__
);
1196 $this->writesDone
= true;
1197 } catch ( DBQueryError
$e ) {
1198 if ( $this->dbw
->wasReadOnlyError() ) {
1199 $this->readOnly
= true; // just avoid site down time
1204 $this->dbw
->commit( __METHOD__
);
1206 $this->currentLang
= null;
1207 $this->batch
= array();
1210 public function set( $key, $value ) {
1211 if ( $this->readOnly
) {
1213 } elseif ( is_null( $this->currentLang
) ) {
1214 throw new MWException( __CLASS__
. ': must call startWrite() before set()' );
1217 $this->batch
[] = array(
1218 'lc_lang' => $this->currentLang
,
1220 'lc_value' => $this->dbw
->encodeBlob( serialize( $value ) ) );
1225 * LCStore implementation which stores data as a collection of CDB files in the
1226 * directory given by $wgCacheDirectory. If $wgCacheDirectory is not set, this
1227 * will throw an exception.
1229 * Profiling indicates that on Linux, this implementation outperforms MySQL if
1230 * the directory is on a local filesystem and there is ample kernel cache
1231 * space. The performance advantage is greater when the DBA extension is
1232 * available than it is with the PHP port.
1234 * See Cdb.php and http://cr.yp.to/cdb.html
1236 class LCStoreCDB
implements LCStore
{
1237 /** @var CdbReader[] */
1240 /** @var CdbWriter */
1243 /** @var string Current language code */
1244 private $currentLang;
1246 /** @var bool|string Cache directory. False if not set */
1249 function __construct( $conf = array() ) {
1250 global $wgCacheDirectory;
1252 if ( isset( $conf['directory'] ) ) {
1253 $this->directory
= $conf['directory'];
1255 $this->directory
= $wgCacheDirectory;
1259 public function get( $code, $key ) {
1260 if ( !isset( $this->readers
[$code] ) ) {
1261 $fileName = $this->getFileName( $code );
1263 $this->readers
[$code] = false;
1264 if ( file_exists( $fileName ) ) {
1266 $this->readers
[$code] = CdbReader
::open( $fileName );
1267 } catch ( CdbException
$e ) {
1268 wfDebug( __METHOD__
. ": unable to open cdb file for reading\n" );
1273 if ( !$this->readers
[$code] ) {
1278 $value = $this->readers
[$code]->get( $key );
1279 } catch ( CdbException
$e ) {
1280 wfDebug( __METHOD__
. ": CdbException caught, error message was "
1281 . $e->getMessage() . "\n" );
1283 if ( $value === false ) {
1287 return unserialize( $value );
1291 public function startWrite( $code ) {
1292 if ( !file_exists( $this->directory
) ) {
1293 if ( !wfMkdirParents( $this->directory
, null, __METHOD__
) ) {
1294 throw new MWException( "Unable to create the localisation store " .
1295 "directory \"{$this->directory}\"" );
1299 // Close reader to stop permission errors on write
1300 if ( !empty( $this->readers
[$code] ) ) {
1301 $this->readers
[$code]->close();
1305 $this->writer
= CdbWriter
::open( $this->getFileName( $code ) );
1306 } catch ( CdbException
$e ) {
1307 throw new MWException( $e->getMessage() );
1309 $this->currentLang
= $code;
1312 public function finishWrite() {
1315 $this->writer
->close();
1316 } catch ( CdbException
$e ) {
1317 throw new MWException( $e->getMessage() );
1319 $this->writer
= null;
1320 unset( $this->readers
[$this->currentLang
] );
1321 $this->currentLang
= null;
1324 public function set( $key, $value ) {
1325 if ( is_null( $this->writer
) ) {
1326 throw new MWException( __CLASS__
. ': must call startWrite() before calling set()' );
1329 $this->writer
->set( $key, serialize( $value ) );
1330 } catch ( CdbException
$e ) {
1331 throw new MWException( $e->getMessage() );
1335 protected function getFileName( $code ) {
1336 if ( strval( $code ) === '' ||
strpos( $code, '/' ) !== false ) {
1337 throw new MWException( __METHOD__
. ": Invalid language \"$code\"" );
1340 return "{$this->directory}/l10n_cache-$code.cdb";
1345 * Null store backend, used to avoid DB errors during install
1347 class LCStoreNull
implements LCStore
{
1348 public function get( $code, $key ) {
1352 public function startWrite( $code ) {
1355 public function finishWrite() {
1358 public function set( $key, $value ) {
1363 * A localisation cache optimised for loading large amounts of data for many
1364 * languages. Used by rebuildLocalisationCache.php.
1366 class LocalisationCacheBulkLoad
extends LocalisationCache
{
1368 * A cache of the contents of data files.
1369 * Core files are serialized to avoid using ~1GB of RAM during a recache.
1371 private $fileCache = array();
1374 * Most recently used languages. Uses the linked-list aspect of PHP hashtables
1375 * to keep the most recently used language codes at the end of the array, and
1376 * the language codes that are ready to be deleted at the beginning.
1378 private $mruLangs = array();
1381 * Maximum number of languages that may be loaded into $this->data
1383 private $maxLoadedLangs = 10;
1386 * @param string $fileName
1387 * @param string $fileType
1388 * @return array|mixed
1390 protected function readPHPFile( $fileName, $fileType ) {
1391 $serialize = $fileType === 'core';
1392 if ( !isset( $this->fileCache
[$fileName][$fileType] ) ) {
1393 $data = parent
::readPHPFile( $fileName, $fileType );
1396 $encData = serialize( $data );
1401 $this->fileCache
[$fileName][$fileType] = $encData;
1404 } elseif ( $serialize ) {
1405 return unserialize( $this->fileCache
[$fileName][$fileType] );
1407 return $this->fileCache
[$fileName][$fileType];
1412 * @param string $code
1413 * @param string $key
1416 public function getItem( $code, $key ) {
1417 unset( $this->mruLangs
[$code] );
1418 $this->mruLangs
[$code] = true;
1420 return parent
::getItem( $code, $key );
1424 * @param string $code
1425 * @param string $key
1426 * @param string $subkey
1429 public function getSubitem( $code, $key, $subkey ) {
1430 unset( $this->mruLangs
[$code] );
1431 $this->mruLangs
[$code] = true;
1433 return parent
::getSubitem( $code, $key, $subkey );
1437 * @param string $code
1439 public function recache( $code ) {
1440 parent
::recache( $code );
1441 unset( $this->mruLangs
[$code] );
1442 $this->mruLangs
[$code] = true;
1447 * @param string $code
1449 public function unload( $code ) {
1450 unset( $this->mruLangs
[$code] );
1451 parent
::unload( $code );
1455 * Unload cached languages until there are less than $this->maxLoadedLangs
1457 protected function trimCache() {
1458 while ( count( $this->data
) > $this->maxLoadedLangs
&& count( $this->mruLangs
) ) {
1459 reset( $this->mruLangs
);
1460 $code = key( $this->mruLangs
);
1461 wfDebug( __METHOD__
. ": unloading $code\n" );
1462 $this->unload( $code );