maintenance/rebuildImages.php

   1 <?php
   2 /**
   3  * Update image metadata records.
   4  *
   5  * Usage: php rebuildImages.php [--missing] [--dry-run]
   6  * Options:
   7  *   --missing  Crawl the uploads dir for images without records, and
   8  *              add them only.
   9  *
  10  * Copyright © 2005 Brion Vibber <brion@pobox.com>
  11  * https://www.mediawiki.org/
  12  *
  13  * This program is free software; you can redistribute it and/or modify
  14  * it under the terms of the GNU General Public License as published by
  15  * the Free Software Foundation; either version 2 of the License, or
  16  * (at your option) any later version.
  17  *
  18  * This program is distributed in the hope that it will be useful,
  19  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  20  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  21  * GNU General Public License for more details.
  22  *
  23  * You should have received a copy of the GNU General Public License along
  24  * with this program; if not, write to the Free Software Foundation, Inc.,
  25  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  26  * http://www.gnu.org/copyleft/gpl.html
  27  *
  28  * @file
  29  * @author Brion Vibber <brion at pobox.com>
  30  * @ingroup Maintenance
  31  */
  32
  33 require_once __DIR__ . '/Maintenance.php';
  34
  35 use Wikimedia\Rdbms\IMaintainableDatabase;
  36
  37 /**
  38  * Maintenance script to update image metadata records.
  39  *
  40  * @ingroup Maintenance
  41  */
  42 class ImageBuilder extends Maintenance {
  43
  44         /**
  45          * @var IMaintainableDatabase
  46          */
  47         protected $dbw;
  48
  49         function __construct() {
  50                 parent::__construct();
  51
  52                 global $wgUpdateCompatibleMetadata;
  53                 // make sure to update old, but compatible img_metadata fields.
  54                 $wgUpdateCompatibleMetadata = true;
  55
  56                 $this->addDescription( 'Script to update image metadata records' );
  57
  58                 $this->addOption( 'missing', 'Check for files without associated database record' );
  59                 $this->addOption( 'dry-run', 'Only report, don\'t update the database' );
  60         }
  61
  62         public function execute() {
  63                 $this->dbw = $this->getDB( DB_MASTER );
  64                 $this->dryrun = $this->hasOption( 'dry-run' );
  65                 if ( $this->dryrun ) {
  66                         MediaWiki\MediaWikiServices::getInstance()->getReadOnlyMode()
  67                                 ->setReason( 'Dry run mode, image upgrades are suppressed' );
  68                 }
  69
  70                 if ( $this->hasOption( 'missing' ) ) {
  71                         $this->crawlMissing();
  72                 } else {
  73                         $this->build();
  74                 }
  75         }
  76
  77         /**
  78          * @return FileRepo
  79          */
  80         function getRepo() {
  81                 if ( !isset( $this->repo ) ) {
  82                         $this->repo = RepoGroup::singleton()->getLocalRepo();
  83                 }
  84
  85                 return $this->repo;
  86         }
  87
  88         function build() {
  89                 $this->buildImage();
  90                 $this->buildOldImage();
  91         }
  92
  93         function init( $count, $table ) {
  94                 $this->processed = 0;
  95                 $this->updated = 0;
  96                 $this->count = $count;
  97                 $this->startTime = microtime( true );
  98                 $this->table = $table;
  99         }
 100
 101         function progress( $updated ) {
 102                 $this->updated += $updated;
 103                 $this->processed++;
 104                 if ( $this->processed % 100 != 0 ) {
 105                         return;
 106                 }
 107                 $portion = $this->processed / $this->count;
 108                 $updateRate = $this->updated / $this->processed;
 109
 110                 $now = microtime( true );
 111                 $delta = $now - $this->startTime;
 112                 $estimatedTotalTime = $delta / $portion;
 113                 $eta = $this->startTime + $estimatedTotalTime;
 114                 $rate = $this->processed / $delta;
 115
 116                 $this->output( sprintf( "%s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
 117                         wfTimestamp( TS_DB, intval( $now ) ),
 118                         $portion * 100.0,
 119                         $this->table,
 120                         wfTimestamp( TS_DB, intval( $eta ) ),
 121                         $this->processed,
 122                         $this->count,
 123                         $rate,
 124                         $updateRate * 100.0 ) );
 125                 flush();
 126         }
 127
 128         function buildTable( $table, $key, $queryInfo, $callback ) {
 129                 $count = $this->dbw->selectField( $table, 'count(*)', '', __METHOD__ );
 130                 $this->init( $count, $table );
 131                 $this->output( "Processing $table...\n" );
 132
 133                 $result = $this->getDB( DB_REPLICA )->select(
 134                         $queryInfo['tables'], $queryInfo['fields'], [], __METHOD__, [], $queryInfo['joins']
 135                 );
 136
 137                 foreach ( $result as $row ) {
 138                         $update = call_user_func( $callback, $row, null );
 139                         if ( $update ) {
 140                                 $this->progress( 1 );
 141                         } else {
 142                                 $this->progress( 0 );
 143                         }
 144                 }
 145                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 146         }
 147
 148         function buildImage() {
 149                 $callback = [ $this, 'imageCallback' ];
 150                 $this->buildTable( 'image', 'img_name', LocalFile::getQueryInfo(), $callback );
 151         }
 152
 153         function imageCallback( $row, $copy ) {
 154                 // Create a File object from the row
 155                 // This will also upgrade it
 156                 $file = $this->getRepo()->newFileFromRow( $row );
 157
 158                 return $file->getUpgraded();
 159         }
 160
 161         function buildOldImage() {
 162                 $this->buildTable( 'oldimage', 'oi_archive_name', OldLocalFile::getQueryInfo(),
 163                         [ $this, 'oldimageCallback' ] );
 164         }
 165
 166         function oldimageCallback( $row, $copy ) {
 167                 // Create a File object from the row
 168                 // This will also upgrade it
 169                 if ( $row->oi_archive_name == '' ) {
 170                         $this->output( "Empty oi_archive_name for oi_name={$row->oi_name}\n" );
 171
 172                         return false;
 173                 }
 174                 $file = $this->getRepo()->newFileFromRow( $row );
 175
 176                 return $file->getUpgraded();
 177         }
 178
 179         function crawlMissing() {
 180                 $this->getRepo()->enumFiles( [ $this, 'checkMissingImage' ] );
 181         }
 182
 183         function checkMissingImage( $fullpath ) {
 184                 $filename = wfBaseName( $fullpath );
 185                 $row = $this->dbw->selectRow( 'image',
 186                         [ 'img_name' ],
 187                         [ 'img_name' => $filename ],
 188                         __METHOD__ );
 189
 190                 if ( !$row ) { // file not registered
 191                         $this->addMissingImage( $filename, $fullpath );
 192                 }
 193         }
 194
 195         function addMissingImage( $filename, $fullpath ) {
 196                 global $wgContLang;
 197
 198                 $timestamp = $this->dbw->timestamp( $this->getRepo()->getFileTimestamp( $fullpath ) );
 199
 200                 $altname = $wgContLang->checkTitleEncoding( $filename );
 201                 if ( $altname != $filename ) {
 202                         if ( $this->dryrun ) {
 203                                 $filename = $altname;
 204                                 $this->output( "Estimating transcoding... $altname\n" );
 205                         } else {
 206                                 # @todo FIXME: create renameFile()
 207                                 $filename = $this->renameFile( $filename );
 208                         }
 209                 }
 210
 211                 if ( $filename == '' ) {
 212                         $this->output( "Empty filename for $fullpath\n" );
 213
 214                         return;
 215                 }
 216                 if ( !$this->dryrun ) {
 217                         $file = wfLocalFile( $filename );
 218                         if ( !$file->recordUpload(
 219                                 '',
 220                                 '(recovered file, missing upload log entry)',
 221                                 '',
 222                                 '',
 223                                 '',
 224                                 false,
 225                                 $timestamp
 226                         ) ) {
 227                                 $this->output( "Error uploading file $fullpath\n" );
 228
 229                                 return;
 230                         }
 231                 }
 232                 $this->output( $fullpath . "\n" );
 233         }
 234 }
 235
 236 $maintClass = ImageBuilder::class;
 237 require_once RUN_MAINTENANCE_IF_MAIN;