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 MediaWiki\MediaWikiServices;
  36 use Wikimedia\Rdbms\IMaintainableDatabase;
  37
  38 /**
  39  * Maintenance script to update image metadata records.
  40  *
  41  * @ingroup Maintenance
  42  */
  43 class ImageBuilder extends Maintenance {
  44
  45         /**
  46          * @var IMaintainableDatabase
  47          */
  48         protected $dbw;
  49
  50         function __construct() {
  51                 parent::__construct();
  52
  53                 global $wgUpdateCompatibleMetadata;
  54                 // make sure to update old, but compatible img_metadata fields.
  55                 $wgUpdateCompatibleMetadata = true;
  56
  57                 $this->addDescription( 'Script to update image metadata records' );
  58
  59                 $this->addOption( 'missing', 'Check for files without associated database record' );
  60                 $this->addOption( 'dry-run', 'Only report, don\'t update the database' );
  61         }
  62
  63         public function execute() {
  64                 $this->dbw = $this->getDB( DB_MASTER );
  65                 $this->dryrun = $this->hasOption( 'dry-run' );
  66                 if ( $this->dryrun ) {
  67                         MediaWiki\MediaWikiServices::getInstance()->getReadOnlyMode()
  68                                 ->setReason( 'Dry run mode, image upgrades are suppressed' );
  69                 }
  70
  71                 if ( $this->hasOption( 'missing' ) ) {
  72                         $this->crawlMissing();
  73                 } else {
  74                         $this->build();
  75                 }
  76         }
  77
  78         /**
  79          * @return LocalRepo
  80          */
  81         function getRepo() {
  82                 if ( !isset( $this->repo ) ) {
  83                         $this->repo = RepoGroup::singleton()->getLocalRepo();
  84                 }
  85
  86                 return $this->repo;
  87         }
  88
  89         function build() {
  90                 $this->buildImage();
  91                 $this->buildOldImage();
  92         }
  93
  94         /**
  95          * @param int $count
  96          * @param string $table
  97          */
  98         function init( $count, $table ) {
  99                 $this->processed = 0;
 100                 $this->updated = 0;
 101                 $this->count = $count;
 102                 $this->startTime = microtime( true );
 103                 $this->table = $table;
 104         }
 105
 106         function progress( $updated ) {
 107                 $this->updated += $updated;
 108                 $this->processed++;
 109                 if ( $this->processed % 100 != 0 ) {
 110                         return;
 111                 }
 112                 $portion = $this->processed / $this->count;
 113                 $updateRate = $this->updated / $this->processed;
 114
 115                 $now = microtime( true );
 116                 $delta = $now - $this->startTime;
 117                 $estimatedTotalTime = $delta / $portion;
 118                 $eta = $this->startTime + $estimatedTotalTime;
 119                 $rate = $this->processed / $delta;
 120
 121                 $this->output( sprintf( "%s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
 122                         wfTimestamp( TS_DB, intval( $now ) ),
 123                         $portion * 100.0,
 124                         $this->table,
 125                         wfTimestamp( TS_DB, intval( $eta ) ),
 126                         $this->processed,
 127                         $this->count,
 128                         $rate,
 129                         $updateRate * 100.0 ) );
 130                 flush();
 131         }
 132
 133         function buildTable( $table, $key, $queryInfo, $callback ) {
 134                 $count = $this->dbw->selectField( $table, 'count(*)', '', __METHOD__ );
 135                 $this->init( $count, $table );
 136                 $this->output( "Processing $table...\n" );
 137
 138                 $result = $this->getDB( DB_REPLICA )->select(
 139                         $queryInfo['tables'], $queryInfo['fields'], [], __METHOD__, [], $queryInfo['joins']
 140                 );
 141
 142                 foreach ( $result as $row ) {
 143                         $update = call_user_func( $callback, $row, null );
 144                         if ( $update ) {
 145                                 $this->progress( 1 );
 146                         } else {
 147                                 $this->progress( 0 );
 148                         }
 149                 }
 150                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 151         }
 152
 153         function buildImage() {
 154                 $callback = [ $this, 'imageCallback' ];
 155                 $this->buildTable( 'image', 'img_name', LocalFile::getQueryInfo(), $callback );
 156         }
 157
 158         function imageCallback( $row, $copy ) {
 159                 // Create a File object from the row
 160                 // This will also upgrade it
 161                 $file = $this->getRepo()->newFileFromRow( $row );
 162
 163                 return $file->getUpgraded();
 164         }
 165
 166         function buildOldImage() {
 167                 $this->buildTable( 'oldimage', 'oi_archive_name', OldLocalFile::getQueryInfo(),
 168                         [ $this, 'oldimageCallback' ] );
 169         }
 170
 171         function oldimageCallback( $row, $copy ) {
 172                 // Create a File object from the row
 173                 // This will also upgrade it
 174                 if ( $row->oi_archive_name == '' ) {
 175                         $this->output( "Empty oi_archive_name for oi_name={$row->oi_name}\n" );
 176
 177                         return false;
 178                 }
 179                 $file = $this->getRepo()->newFileFromRow( $row );
 180
 181                 return $file->getUpgraded();
 182         }
 183
 184         function crawlMissing() {
 185                 $this->getRepo()->enumFiles( [ $this, 'checkMissingImage' ] );
 186         }
 187
 188         function checkMissingImage( $fullpath ) {
 189                 $filename = wfBaseName( $fullpath );
 190                 $row = $this->dbw->selectRow( 'image',
 191                         [ 'img_name' ],
 192                         [ 'img_name' => $filename ],
 193                         __METHOD__ );
 194
 195                 if ( !$row ) { // file not registered
 196                         $this->addMissingImage( $filename, $fullpath );
 197                 }
 198         }
 199
 200         function addMissingImage( $filename, $fullpath ) {
 201                 $timestamp = $this->dbw->timestamp( $this->getRepo()->getFileTimestamp( $fullpath ) );
 202                 $services = MediaWikiServices::getInstance();
 203
 204                 $altname = $services->getContentLanguage()->checkTitleEncoding( $filename );
 205                 if ( $altname != $filename ) {
 206                         if ( $this->dryrun ) {
 207                                 $filename = $altname;
 208                                 $this->output( "Estimating transcoding... $altname\n" );
 209                         } else {
 210                                 // @fixme create renameFile()
 211                                 // @phan-suppress-next-line PhanUndeclaredMethod See comment above...
 212                                 $filename = $this->renameFile( $filename );
 213                         }
 214                 }
 215
 216                 if ( $filename == '' ) {
 217                         $this->output( "Empty filename for $fullpath\n" );
 218
 219                         return;
 220                 }
 221                 if ( !$this->dryrun ) {
 222                         $file = $services->getRepoGroup()->getLocalRepo()->newFile( $filename );
 223                         if ( !$file->recordUpload(
 224                                 '',
 225                                 '(recovered file, missing upload log entry)',
 226                                 '',
 227                                 '',
 228                                 '',
 229                                 false,
 230                                 $timestamp
 231                         ) ) {
 232                                 $this->output( "Error uploading file $fullpath\n" );
 233
 234                                 return;
 235                         }
 236                 }
 237                 $this->output( $fullpath . "\n" );
 238         }
 239 }
 240
 241 $maintClass = ImageBuilder::class;
 242 require_once RUN_MAINTENANCE_IF_MAIN;