includes/filerepo/LocalRepo.php

   1 <?php
   2 /**
   3  * Local repository that stores files in the local filesystem and registers them
   4  * in the wiki's own database.
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @file
  22  * @ingroup FileRepo
  23  */
  24
  25 /**
  26  * A repository that stores files in the local filesystem and registers them
  27  * in the wiki's own database. This is the most commonly used repository class.
  28  *
  29  * @ingroup FileRepo
  30  */
  31 class LocalRepo extends FileRepo {
  32         protected $fileFactory = array( 'LocalFile', 'newFromTitle' );
  33         protected $fileFactoryKey = array( 'LocalFile', 'newFromKey' );
  34         protected $fileFromRowFactory = array( 'LocalFile', 'newFromRow' );
  35         protected $oldFileFromRowFactory = array( 'OldLocalFile', 'newFromRow' );
  36         protected $oldFileFactory = array( 'OldLocalFile', 'newFromTitle' );
  37         protected $oldFileFactoryKey = array( 'OldLocalFile', 'newFromKey' );
  38
  39         /**
  40          * @throws MWException
  41          * @param $row
  42          * @return LocalFile
  43          */
  44         function newFileFromRow( $row ) {
  45                 if ( isset( $row->img_name ) ) {
  46                         return call_user_func( $this->fileFromRowFactory, $row, $this );
  47                 } elseif ( isset( $row->oi_name ) ) {
  48                         return call_user_func( $this->oldFileFromRowFactory, $row, $this );
  49                 } else {
  50                         throw new MWException( __METHOD__ . ': invalid row' );
  51                 }
  52         }
  53
  54         /**
  55          * @param $title
  56          * @param $archiveName
  57          * @return OldLocalFile
  58          */
  59         function newFromArchiveName( $title, $archiveName ) {
  60                 return OldLocalFile::newFromArchiveName( $title, $this, $archiveName );
  61         }
  62
  63         /**
  64          * Delete files in the deleted directory if they are not referenced in the
  65          * filearchive table. This needs to be done in the repo because it needs to
  66          * interleave database locks with file operations, which is potentially a
  67          * remote operation.
  68          *
  69          * @param $storageKeys array
  70          *
  71          * @return FileRepoStatus
  72          */
  73         function cleanupDeletedBatch( array $storageKeys ) {
  74                 $backend = $this->backend; // convenience
  75                 $root = $this->getZonePath( 'deleted' );
  76                 $dbw = $this->getMasterDB();
  77                 $status = $this->newGood();
  78                 $storageKeys = array_unique( $storageKeys );
  79                 foreach ( $storageKeys as $key ) {
  80                         $hashPath = $this->getDeletedHashPath( $key );
  81                         $path = "$root/$hashPath$key";
  82                         $dbw->begin( __METHOD__ );
  83                         // Check for usage in deleted/hidden files and pre-emptively
  84                         // lock the key to avoid any future use until we are finished.
  85                         $deleted = $this->deletedFileHasKey( $key, 'lock' );
  86                         $hidden = $this->hiddenFileHasKey( $key, 'lock' );
  87                         if ( !$deleted && !$hidden ) { // not in use now
  88                                 wfDebug( __METHOD__ . ": deleting $key\n" );
  89                                 $op = array( 'op' => 'delete', 'src' => $path );
  90                                 if ( !$backend->doOperation( $op )->isOK() ) {
  91                                         $status->error( 'undelete-cleanup-error', $path );
  92                                         $status->failCount++;
  93                                 }
  94                         } else {
  95                                 wfDebug( __METHOD__ . ": $key still in use\n" );
  96                                 $status->successCount++;
  97                         }
  98                         $dbw->commit( __METHOD__ );
  99                 }
 100
 101                 return $status;
 102         }
 103
 104         /**
 105          * Check if a deleted (filearchive) file has this sha1 key
 106          *
 107          * @param string $key File storage key (base-36 sha1 key with file extension)
 108          * @param string|null $lock Use "lock" to lock the row via FOR UPDATE
 109          * @return bool File with this key is in use
 110          */
 111         protected function deletedFileHasKey( $key, $lock = null ) {
 112                 $options = ( $lock === 'lock' ) ? array( 'FOR UPDATE' ) : array();
 113
 114                 $dbw = $this->getMasterDB();
 115
 116                 return (bool)$dbw->selectField( 'filearchive', '1',
 117                         array( 'fa_storage_group' => 'deleted', 'fa_storage_key' => $key ),
 118                         __METHOD__, $options
 119                 );
 120         }
 121
 122         /**
 123          * Check if a hidden (revision delete) file has this sha1 key
 124          *
 125          * @param string $key File storage key (base-36 sha1 key with file extension)
 126          * @param string|null $lock Use "lock" to lock the row via FOR UPDATE
 127          * @return bool File with this key is in use
 128          */
 129         protected function hiddenFileHasKey( $key, $lock = null ) {
 130                 $options = ( $lock === 'lock' ) ? array( 'FOR UPDATE' ) : array();
 131
 132                 $sha1 = self::getHashFromKey( $key );
 133                 $ext = File::normalizeExtension( substr( $key, strcspn( $key, '.' ) + 1 ) );
 134
 135                 $dbw = $this->getMasterDB();
 136
 137                 return (bool)$dbw->selectField( 'oldimage', '1',
 138                         array( 'oi_sha1' => $sha1,
 139                                 'oi_archive_name ' . $dbw->buildLike( $dbw->anyString(), ".$ext" ),
 140                                 $dbw->bitAnd( 'oi_deleted', File::DELETED_FILE ) => File::DELETED_FILE ),
 141                         __METHOD__, $options
 142                 );
 143         }
 144
 145         /**
 146          * Gets the SHA1 hash from a storage key
 147          *
 148          * @param string $key
 149          * @return string
 150          */
 151         public static function getHashFromKey( $key ) {
 152                 return strtok( $key, '.' );
 153         }
 154
 155         /**
 156          * Checks if there is a redirect named as $title
 157          *
 158          * @param $title Title of file
 159          * @return bool
 160          */
 161         function checkRedirect( Title $title ) {
 162                 global $wgMemc;
 163
 164                 $title = File::normalizeTitle( $title, 'exception' );
 165
 166                 $memcKey = $this->getSharedCacheKey( 'image_redirect', md5( $title->getDBkey() ) );
 167                 if ( $memcKey === false ) {
 168                         $memcKey = $this->getLocalCacheKey( 'image_redirect', md5( $title->getDBkey() ) );
 169                         $expiry = 300; // no invalidation, 5 minutes
 170                 } else {
 171                         $expiry = 86400; // has invalidation, 1 day
 172                 }
 173                 $cachedValue = $wgMemc->get( $memcKey );
 174                 if ( $cachedValue === ' ' || $cachedValue === '' ) {
 175                         // Does not exist
 176                         return false;
 177                 } elseif ( strval( $cachedValue ) !== '' && $cachedValue !== ' PURGED' ) {
 178                         return Title::newFromText( $cachedValue, NS_FILE );
 179                 } // else $cachedValue is false or null: cache miss
 180
 181                 $id = $this->getArticleID( $title );
 182                 if ( !$id ) {
 183                         $wgMemc->add( $memcKey, " ", $expiry );
 184
 185                         return false;
 186                 }
 187                 $dbr = $this->getSlaveDB();
 188                 $row = $dbr->selectRow(
 189                         'redirect',
 190                         array( 'rd_title', 'rd_namespace' ),
 191                         array( 'rd_from' => $id ),
 192                         __METHOD__
 193                 );
 194
 195                 if ( $row && $row->rd_namespace == NS_FILE ) {
 196                         $targetTitle = Title::makeTitle( $row->rd_namespace, $row->rd_title );
 197                         $wgMemc->add( $memcKey, $targetTitle->getDBkey(), $expiry );
 198
 199                         return $targetTitle;
 200                 } else {
 201                         $wgMemc->add( $memcKey, '', $expiry );
 202
 203                         return false;
 204                 }
 205         }
 206
 207         /**
 208          * Function link Title::getArticleID().
 209          * We can't say Title object, what database it should use, so we duplicate that function here.
 210          *
 211          * @param $title Title
 212          * @return bool|int|mixed
 213          */
 214         protected function getArticleID( $title ) {
 215                 if ( !$title instanceof Title ) {
 216                         return 0;
 217                 }
 218                 $dbr = $this->getSlaveDB();
 219                 $id = $dbr->selectField(
 220                         'page', // Table
 221                         'page_id', //Field
 222                         array( //Conditions
 223                                 'page_namespace' => $title->getNamespace(),
 224                                 'page_title' => $title->getDBkey(),
 225                         ),
 226                         __METHOD__ //Function name
 227                 );
 228
 229                 return $id;
 230         }
 231
 232         /**
 233          * Get an array or iterator of file objects for files that have a given
 234          * SHA-1 content hash.
 235          *
 236          * @param string $hash a sha1 hash to look for
 237          * @return Array
 238          */
 239         function findBySha1( $hash ) {
 240                 $dbr = $this->getSlaveDB();
 241                 $res = $dbr->select(
 242                         'image',
 243                         LocalFile::selectFields(),
 244                         array( 'img_sha1' => $hash ),
 245                         __METHOD__,
 246                         array( 'ORDER BY' => 'img_name' )
 247                 );
 248
 249                 $result = array();
 250                 foreach ( $res as $row ) {
 251                         $result[] = $this->newFileFromRow( $row );
 252                 }
 253                 $res->free();
 254
 255                 return $result;
 256         }
 257
 258         /**
 259          * Get an array of arrays or iterators of file objects for files that
 260          * have the given SHA-1 content hashes.
 261          *
 262          * Overrides generic implementation in FileRepo for performance reason
 263          *
 264          * @param array $hashes An array of hashes
 265          * @return array An Array of arrays or iterators of file objects and the hash as key
 266          */
 267         function findBySha1s( array $hashes ) {
 268                 if ( !count( $hashes ) ) {
 269                         return array(); //empty parameter
 270                 }
 271
 272                 $dbr = $this->getSlaveDB();
 273                 $res = $dbr->select(
 274                         'image',
 275                         LocalFile::selectFields(),
 276                         array( 'img_sha1' => $hashes ),
 277                         __METHOD__,
 278                         array( 'ORDER BY' => 'img_name' )
 279                 );
 280
 281                 $result = array();
 282                 foreach ( $res as $row ) {
 283                         $file = $this->newFileFromRow( $row );
 284                         $result[$file->getSha1()][] = $file;
 285                 }
 286                 $res->free();
 287
 288                 return $result;
 289         }
 290
 291         /**
 292          * Return an array of files where the name starts with $prefix.
 293          *
 294          * @param string $prefix The prefix to search for
 295          * @param int $limit The maximum amount of files to return
 296          * @return array
 297          */
 298         public function findFilesByPrefix( $prefix, $limit ) {
 299                 $selectOptions = array( 'ORDER BY' => 'img_name', 'LIMIT' => intval( $limit ) );
 300
 301                 // Query database
 302                 $dbr = $this->getSlaveDB();
 303                 $res = $dbr->select(
 304                         'image',
 305                         LocalFile::selectFields(),
 306                         'img_name ' . $dbr->buildLike( $prefix, $dbr->anyString() ),
 307                         __METHOD__,
 308                         $selectOptions
 309                 );
 310
 311                 // Build file objects
 312                 $files = array();
 313                 foreach ( $res as $row ) {
 314                         $files[] = $this->newFileFromRow( $row );
 315                 }
 316
 317                 return $files;
 318         }
 319
 320         /**
 321          * Get a connection to the slave DB
 322          * @return DatabaseBase
 323          */
 324         function getSlaveDB() {
 325                 return wfGetDB( DB_SLAVE );
 326         }
 327
 328         /**
 329          * Get a connection to the master DB
 330          * @return DatabaseBase
 331          */
 332         function getMasterDB() {
 333                 return wfGetDB( DB_MASTER );
 334         }
 335
 336         /**
 337          * Get a key on the primary cache for this repository.
 338          * Returns false if the repository's cache is not accessible at this site.
 339          * The parameters are the parts of the key, as for wfMemcKey().
 340          *
 341          * @return string
 342          */
 343         function getSharedCacheKey( /*...*/ ) {
 344                 $args = func_get_args();
 345
 346                 return call_user_func_array( 'wfMemcKey', $args );
 347         }
 348
 349         /**
 350          * Invalidates image redirect cache related to that image
 351          *
 352          * @param $title Title of page
 353          * @return void
 354          */
 355         function invalidateImageRedirect( Title $title ) {
 356                 global $wgMemc;
 357                 $memcKey = $this->getSharedCacheKey( 'image_redirect', md5( $title->getDBkey() ) );
 358                 if ( $memcKey ) {
 359                         // Set a temporary value for the cache key, to ensure
 360                         // that this value stays purged long enough so that
 361                         // it isn't refreshed with a stale value due to a
 362                         // lagged slave.
 363                         $wgMemc->set( $memcKey, ' PURGED', 12 );
 364                 }
 365         }
 366 }