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          * @var IMaintainableDatabase
  46          */
  47         protected $dbw;
  48
  49         /** @var bool */
  50         private $dryrun;
  51
  52         /** @var LocalRepo|null */
  53         private $repo;
  54
  55         /** @var int */
  56         private $updated;
  57
  58         /** @var int */
  59         private $processed;
  60
  61         /** @var int */
  62         private $count;
  63
  64         /** @var int */
  65         private $startTime;
  66
  67         /** @var string */
  68         private $table;
  69
  70         public function __construct() {
  71                 parent::__construct();
  72
  73                 global $wgUpdateCompatibleMetadata;
  74                 // make sure to update old, but compatible img_metadata fields.
  75                 $wgUpdateCompatibleMetadata = true;
  76
  77                 $this->addDescription( 'Script to update image metadata records' );
  78
  79                 $this->addOption( 'missing', 'Check for files without associated database record' );
  80                 $this->addOption( 'dry-run', 'Only report, don\'t update the database' );
  81         }
  82
  83         public function execute() {
  84                 $this->dbw = $this->getDB( DB_MASTER );
  85                 $this->dryrun = $this->hasOption( 'dry-run' );
  86                 if ( $this->dryrun ) {
  87                         MediaWiki\MediaWikiServices::getInstance()->getReadOnlyMode()
  88                                 ->setReason( 'Dry run mode, image upgrades are suppressed' );
  89                 }
  90
  91                 if ( $this->hasOption( 'missing' ) ) {
  92                         $this->crawlMissing();
  93                 } else {
  94                         $this->build();
  95                 }
  96         }
  97
  98         /**
  99          * @return LocalRepo
 100          */
 101         private function getRepo() {
 102                 if ( $this->repo === null ) {
 103                         $this->repo = MediaWikiServices::getInstance()->getRepoGroup()->getLocalRepo();
 104                 }
 105
 106                 return $this->repo;
 107         }
 108
 109         private function build() {
 110                 $this->buildImage();
 111                 $this->buildOldImage();
 112         }
 113
 114         /**
 115          * @param int $count
 116          * @param string $table
 117          */
 118         private function init( $count, $table ) {
 119                 $this->processed = 0;
 120                 $this->updated = 0;
 121                 $this->count = $count;
 122                 $this->startTime = microtime( true );
 123                 $this->table = $table;
 124         }
 125
 126         private function progress( $updated ) {
 127                 $this->updated += $updated;
 128                 $this->processed++;
 129                 if ( $this->processed % 100 != 0 ) {
 130                         return;
 131                 }
 132                 $portion = $this->processed / $this->count;
 133                 $updateRate = $this->updated / $this->processed;
 134
 135                 $now = microtime( true );
 136                 $delta = $now - $this->startTime;
 137                 $estimatedTotalTime = $delta / $portion;
 138                 $eta = $this->startTime + $estimatedTotalTime;
 139                 $rate = $this->processed / $delta;
 140
 141                 $this->output( sprintf( "%s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
 142                         wfTimestamp( TS_DB, intval( $now ) ),
 143                         $portion * 100.0,
 144                         $this->table,
 145                         wfTimestamp( TS_DB, intval( $eta ) ),
 146                         $this->processed,
 147                         $this->count,
 148                         $rate,
 149                         $updateRate * 100.0 ) );
 150                 flush();
 151         }
 152
 153         private function buildTable( $table, $key, $queryInfo, $callback ) {
 154                 $count = $this->dbw->selectField( $table, 'count(*)', '', __METHOD__ );
 155                 $this->init( $count, $table );
 156                 $this->output( "Processing $table...\n" );
 157
 158                 $result = $this->getDB( DB_REPLICA )->select(
 159                         $queryInfo['tables'], $queryInfo['fields'], [], __METHOD__, [], $queryInfo['joins']
 160                 );
 161
 162                 foreach ( $result as $row ) {
 163                         $update = call_user_func( $callback, $row, null );
 164                         if ( $update ) {
 165                                 $this->progress( 1 );
 166                         } else {
 167                                 $this->progress( 0 );
 168                         }
 169                 }
 170                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 171         }
 172
 173         private function buildImage() {
 174                 $callback = [ $this, 'imageCallback' ];
 175                 $this->buildTable( 'image', 'img_name', LocalFile::getQueryInfo(), $callback );
 176         }
 177
 178         private function imageCallback( $row, $copy ) {
 179                 // Create a File object from the row
 180                 // This will also upgrade it
 181                 $file = $this->getRepo()->newFileFromRow( $row );
 182
 183                 return $file->getUpgraded();
 184         }
 185
 186         private function buildOldImage() {
 187                 $this->buildTable( 'oldimage', 'oi_archive_name', OldLocalFile::getQueryInfo(),
 188                         [ $this, 'oldimageCallback' ] );
 189         }
 190
 191         private function oldimageCallback( $row, $copy ) {
 192                 // Create a File object from the row
 193                 // This will also upgrade it
 194                 if ( $row->oi_archive_name == '' ) {
 195                         $this->output( "Empty oi_archive_name for oi_name={$row->oi_name}\n" );
 196
 197                         return false;
 198                 }
 199                 $file = $this->getRepo()->newFileFromRow( $row );
 200
 201                 return $file->getUpgraded();
 202         }
 203
 204         private function crawlMissing() {
 205                 $this->getRepo()->enumFiles( [ $this, 'checkMissingImage' ] );
 206         }
 207
 208         public function checkMissingImage( $fullpath ) {
 209                 $filename = wfBaseName( $fullpath );
 210                 $row = $this->dbw->selectRow( 'image',
 211                         [ 'img_name' ],
 212                         [ 'img_name' => $filename ],
 213                         __METHOD__ );
 214
 215                 if ( !$row ) {
 216                         // file not registered
 217                         $this->addMissingImage( $filename, $fullpath );
 218                 }
 219         }
 220
 221         private function addMissingImage( $filename, $fullpath ) {
 222                 $timestamp = $this->dbw->timestamp( $this->getRepo()->getFileTimestamp( $fullpath ) );
 223                 $services = MediaWikiServices::getInstance();
 224
 225                 $altname = $services->getContentLanguage()->checkTitleEncoding( $filename );
 226                 if ( $altname != $filename ) {
 227                         if ( $this->dryrun ) {
 228                                 $filename = $altname;
 229                                 $this->output( "Estimating transcoding... $altname\n" );
 230                         } else {
 231                                 // @fixme create renameFile()
 232                                 // @phan-suppress-next-line PhanUndeclaredMethod See comment above...
 233                                 $filename = $this->renameFile( $filename );
 234                         }
 235                 }
 236
 237                 if ( $filename == '' ) {
 238                         $this->output( "Empty filename for $fullpath\n" );
 239
 240                         return;
 241                 }
 242                 if ( !$this->dryrun ) {
 243                         $file = $services->getRepoGroup()->getLocalRepo()->newFile( $filename );
 244                         $pageText = SpecialUpload::getInitialPageText(
 245                                 '(recovered file, missing upload log entry)'
 246                         );
 247                         $user = User::newSystemUser( 'Maintenance script', [ 'steal' => true ] );
 248                         $status = $file->recordUpload3(
 249                                 '',
 250                                 '(recovered file, missing upload log entry)',
 251                                 $pageText,
 252                                 $user,
 253                                 false,
 254                                 $timestamp
 255                         );
 256                         if ( !$status->isOK() ) {
 257                                 $this->output( "Error uploading file $fullpath\n" );
 258
 259                                 return;
 260                         }
 261                 }
 262                 $this->output( $fullpath . "\n" );
 263         }
 264 }
 265
 266 $maintClass = ImageBuilder::class;
 267 require_once RUN_MAINTENANCE_IF_MAIN;