includes/Storage/NameTableStore.php

   1 <?php
   2 /**
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2 of the License, or
   6  * (at your option) any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License along
  14  * with this program; if not, write to the Free Software Foundation, Inc.,
  15  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  16  * http://www.gnu.org/copyleft/gpl.html
  17  *
  18  * @file
  19  */
  20
  21 namespace MediaWiki\Storage;
  22
  23 use IExpiringStore;
  24 use Psr\Log\LoggerInterface;
  25 use WANObjectCache;
  26 use Wikimedia\Assert\Assert;
  27 use Wikimedia\Rdbms\Database;
  28 use Wikimedia\Rdbms\IDatabase;
  29 use Wikimedia\Rdbms\LoadBalancer;
  30
  31 /**
  32  * @author Addshore
  33  * @since 1.31
  34  */
  35 class NameTableStore {
  36
  37         /** @var LoadBalancer */
  38         private $loadBalancer;
  39
  40         /** @var WANObjectCache */
  41         private $cache;
  42
  43         /** @var LoggerInterface */
  44         private $logger;
  45
  46         /** @var string[] */
  47         private $tableCache = null;
  48
  49         /** @var bool|string */
  50         private $wikiId = false;
  51
  52         /** @var int */
  53         private $cacheTTL;
  54
  55         /** @var string */
  56         private $table;
  57         /** @var string */
  58         private $idField;
  59         /** @var string */
  60         private $nameField;
  61         /** @var null|callable */
  62         private $normalizationCallback = null;
  63         /** @var null|callable */
  64         private $insertCallback = null;
  65
  66         /**
  67          * @param LoadBalancer $dbLoadBalancer A load balancer for acquiring database connections
  68          * @param WANObjectCache $cache A cache manager for caching data
  69          * @param LoggerInterface $logger
  70          * @param string $table
  71          * @param string $idField
  72          * @param string $nameField
  73          * @param callable $normalizationCallback Normalization to be applied to names before being
  74          * saved or queried. This should be a callback that accepts and returns a single string.
  75          * @param bool|string $wikiId The ID of the target wiki database. Use false for the local wiki.
  76          * @param callable $insertCallback Callback to change insert fields accordingly.
  77          * This parameter was introduced in 1.32
  78          */
  79         public function __construct(
  80                 LoadBalancer $dbLoadBalancer,
  81                 WANObjectCache $cache,
  82                 LoggerInterface $logger,
  83                 $table,
  84                 $idField,
  85                 $nameField,
  86                 callable $normalizationCallback = null,
  87                 $wikiId = false,
  88                 callable $insertCallback = null
  89         ) {
  90                 $this->loadBalancer = $dbLoadBalancer;
  91                 $this->cache = $cache;
  92                 $this->logger = $logger;
  93                 $this->table = $table;
  94                 $this->idField = $idField;
  95                 $this->nameField = $nameField;
  96                 $this->normalizationCallback = $normalizationCallback;
  97                 $this->wikiId = $wikiId;
  98                 $this->cacheTTL = IExpiringStore::TTL_MONTH;
  99                 $this->insertCallback = $insertCallback;
 100         }
 101
 102         /**
 103          * @param int $index A database index, like DB_MASTER or DB_REPLICA
 104          * @param int $flags Database connection flags
 105          *
 106          * @return IDatabase
 107          */
 108         private function getDBConnection( $index, $flags = 0 ) {
 109                 return $this->loadBalancer->getConnection( $index, [], $this->wikiId, $flags );
 110         }
 111
 112         /**
 113          * Gets the cache key for names.
 114          *
 115          * The cache key is constructed based on the wiki ID passed to the constructor, and allows
 116          * sharing of name tables cached for a specific database between wikis.
 117          *
 118          * @return string
 119          */
 120         private function getCacheKey() {
 121                 return $this->cache->makeGlobalKey(
 122                         'NameTableSqlStore',
 123                         $this->table,
 124                         $this->loadBalancer->resolveDomainID( $this->wikiId )
 125                 );
 126         }
 127
 128         /**
 129          * @param string $name
 130          * @return string
 131          */
 132         private function normalizeName( $name ) {
 133                 if ( $this->normalizationCallback === null ) {
 134                         return $name;
 135                 }
 136                 return call_user_func( $this->normalizationCallback, $name );
 137         }
 138
 139         /**
 140          * Acquire the id of the given name.
 141          * This creates a row in the table if it doesn't already exist.
 142          *
 143          * @param string $name
 144          * @throws NameTableAccessException
 145          * @return int
 146          */
 147         public function acquireId( $name ) {
 148                 Assert::parameterType( 'string', $name, '$name' );
 149                 $name = $this->normalizeName( $name );
 150
 151                 $table = $this->getTableFromCachesOrReplica();
 152                 $searchResult = array_search( $name, $table, true );
 153                 if ( $searchResult === false ) {
 154                         $id = $this->store( $name );
 155                         if ( $id === null ) {
 156                                 // RACE: $name was already in the db, probably just inserted, so load from master
 157                                 // Use DBO_TRX to avoid missing inserts due to other threads or REPEATABLE-READs
 158                                 $table = $this->loadTable(
 159                                         $this->getDBConnection( DB_MASTER, LoadBalancer::CONN_TRX_AUTOCOMMIT )
 160                                 );
 161                                 $searchResult = array_search( $name, $table, true );
 162                                 if ( $searchResult === false ) {
 163                                         // Insert failed due to IGNORE flag, but DB_MASTER didn't give us the data
 164                                         $m = "No insert possible but master didn't give us a record for " .
 165                                                 "'{$name}' in '{$this->table}'";
 166                                         $this->logger->error( $m );
 167                                         throw new NameTableAccessException( $m );
 168                                 }
 169                                 $this->purgeWANCache(
 170                                         function () {
 171                                                 $this->cache->reap( $this->getCacheKey(), INF );
 172                                         }
 173                                 );
 174                         } else {
 175                                 $table[$id] = $name;
 176                                 $searchResult = $id;
 177                                 // As store returned an ID we know we inserted so delete from WAN cache
 178                                 $this->purgeWANCache(
 179                                         function () {
 180                                                 $this->cache->delete( $this->getCacheKey() );
 181                                         }
 182                                 );
 183                         }
 184                         $this->tableCache = $table;
 185                 }
 186
 187                 return $searchResult;
 188         }
 189
 190         /**
 191          * Get the id of the given name.
 192          * If the name doesn't exist this will throw.
 193          * This should be used in cases where we believe the name already exists or want to check for
 194          * existence.
 195          *
 196          * @param string $name
 197          * @throws NameTableAccessException The name does not exist
 198          * @return int Id
 199          */
 200         public function getId( $name ) {
 201                 Assert::parameterType( 'string', $name, '$name' );
 202                 $name = $this->normalizeName( $name );
 203
 204                 $table = $this->getTableFromCachesOrReplica();
 205                 $searchResult = array_search( $name, $table, true );
 206
 207                 if ( $searchResult !== false ) {
 208                         return $searchResult;
 209                 }
 210
 211                 throw NameTableAccessException::newFromDetails( $this->table, 'name', $name );
 212         }
 213
 214         /**
 215          * Get the name of the given id.
 216          * If the id doesn't exist this will throw.
 217          * This should be used in cases where we believe the id already exists.
 218          *
 219          * Note: Calls to this method will result in a master select for non existing IDs.
 220          *
 221          * @param int $id
 222          * @throws NameTableAccessException The id does not exist
 223          * @return string name
 224          */
 225         public function getName( $id ) {
 226                 Assert::parameterType( 'integer', $id, '$id' );
 227
 228                 $table = $this->getTableFromCachesOrReplica();
 229                 if ( array_key_exists( $id, $table ) ) {
 230                         return $table[$id];
 231                 }
 232
 233                 $table = $this->cache->getWithSetCallback(
 234                         $this->getCacheKey(),
 235                         $this->cacheTTL,
 236                         function ( $oldValue, &$ttl, &$setOpts ) use ( $id ) {
 237                                 // Check if cached value is up-to-date enough to have $id
 238                                 if ( is_array( $oldValue ) && array_key_exists( $id, $oldValue ) ) {
 239                                         // Completely leave the cache key alone
 240                                         $ttl = WANObjectCache::TTL_UNCACHEABLE;
 241                                         // Use the old value
 242                                         return $oldValue;
 243                                 }
 244                                 // Regenerate from replica DB, and master DB if needed
 245                                 foreach ( [ DB_REPLICA, DB_MASTER ] as $source ) {
 246                                         // Log a fallback to master
 247                                         if ( $source === DB_MASTER ) {
 248                                                 $this->logger->info(
 249                                                         __METHOD__ . 'falling back to master select from ' .
 250                                                         $this->table . ' with id ' . $id
 251                                                 );
 252                                         }
 253                                         $db = $this->getDBConnection( $source );
 254                                         $cacheSetOpts = Database::getCacheSetOptions( $db );
 255                                         $table = $this->loadTable( $db );
 256                                         if ( array_key_exists( $id, $table ) ) {
 257                                                 break; // found it
 258                                         }
 259                                 }
 260                                 // Use the value from last source checked
 261                                 $setOpts += $cacheSetOpts;
 262
 263                                 return $table;
 264                         },
 265                         [ 'minAsOf' => INF ] // force callback run
 266                 );
 267
 268                 $this->tableCache = $table;
 269
 270                 if ( array_key_exists( $id, $table ) ) {
 271                         return $table[$id];
 272                 }
 273
 274                 throw NameTableAccessException::newFromDetails( $this->table, 'id', $id );
 275         }
 276
 277         /**
 278          * Get the whole table, in no particular order as a map of ids to names.
 279          * This method could be subject to DB or cache lag.
 280          *
 281          * @return string[] keys are the name ids, values are the names themselves
 282          *  Example: [ 1 => 'foo', 3 => 'bar' ]
 283          */
 284         public function getMap() {
 285                 return $this->getTableFromCachesOrReplica();
 286         }
 287
 288         /**
 289          * @return string[]
 290          */
 291         private function getTableFromCachesOrReplica() {
 292                 if ( $this->tableCache !== null ) {
 293                         return $this->tableCache;
 294                 }
 295
 296                 $table = $this->cache->getWithSetCallback(
 297                         $this->getCacheKey(),
 298                         $this->cacheTTL,
 299                         function ( $oldValue, &$ttl, &$setOpts ) {
 300                                 $dbr = $this->getDBConnection( DB_REPLICA );
 301                                 $setOpts += Database::getCacheSetOptions( $dbr );
 302                                 return $this->loadTable( $dbr );
 303                         }
 304                 );
 305
 306                 $this->tableCache = $table;
 307
 308                 return $table;
 309         }
 310
 311         /**
 312          * Reap the WANCache entry for this table.
 313          *
 314          * @param callable $purgeCallback callback to 'purge' the WAN cache
 315          */
 316         private function purgeWANCache( $purgeCallback ) {
 317                 // If the LB has no DB changes don't both with onTransactionPreCommitOrIdle
 318                 if ( !$this->loadBalancer->hasOrMadeRecentMasterChanges() ) {
 319                         $purgeCallback();
 320                         return;
 321                 }
 322
 323                 $this->getDBConnection( DB_MASTER )
 324                         ->onTransactionPreCommitOrIdle( $purgeCallback, __METHOD__ );
 325         }
 326
 327         /**
 328          * Gets the table from the db
 329          *
 330          * @param IDatabase $db
 331          *
 332          * @return string[]
 333          */
 334         private function loadTable( IDatabase $db ) {
 335                 $result = $db->select(
 336                         $this->table,
 337                         [
 338                                 'id' => $this->idField,
 339                                 'name' => $this->nameField
 340                         ],
 341                         [],
 342                         __METHOD__,
 343                         [ 'ORDER BY' => 'id' ]
 344                 );
 345
 346                 $assocArray = [];
 347                 foreach ( $result as $row ) {
 348                         $assocArray[$row->id] = $row->name;
 349                 }
 350
 351                 return $assocArray;
 352         }
 353
 354         /**
 355          * Stores the given name in the DB, returning the ID when an insert occurs.
 356          *
 357          * @param string $name
 358          * @return int|null int if we know the ID, null if we don't
 359          */
 360         private function store( $name ) {
 361                 Assert::parameterType( 'string', $name, '$name' );
 362                 Assert::parameter( $name !== '', '$name', 'should not be an empty string' );
 363                 // Note: this is only called internally so normalization of $name has already occurred.
 364
 365                 $dbw = $this->getDBConnection( DB_MASTER );
 366
 367                 $dbw->insert(
 368                         $this->table,
 369                         $this->getFieldsToStore( $name ),
 370                         __METHOD__,
 371                         [ 'IGNORE' ]
 372                 );
 373
 374                 if ( $dbw->affectedRows() === 0 ) {
 375                         $this->logger->info(
 376                                 'Tried to insert name into table ' . $this->table . ', but value already existed.'
 377                         );
 378                         return null;
 379                 }
 380
 381                 return $dbw->insertId();
 382         }
 383
 384         /**
 385          * @param string $name
 386          * @return array
 387          */
 388         private function getFieldsToStore( $name ) {
 389                 $fields = [ $this->nameField => $name ];
 390                 if ( $this->insertCallback !== null ) {
 391                         $fields = call_user_func( $this->insertCallback, $fields );
 392                 }
 393                 return $fields;
 394         }
 395
 396 }