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 Brooke Vibber <bvibber@wikimedia.org>
  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 Brooke Vibber <bvibber@wikimedia.org>
  30  * @ingroup Maintenance
  31  */
  32
  33 // @codeCoverageIgnoreStart
  34 require_once __DIR__ . '/Maintenance.php';
  35 // @codeCoverageIgnoreEnd
  36
  37 use MediaWiki\FileRepo\File\FileSelectQueryBuilder;
  38 use MediaWiki\Maintenance\Maintenance;
  39 use MediaWiki\Specials\SpecialUpload;
  40 use MediaWiki\User\User;
  41 use Wikimedia\Rdbms\IMaintainableDatabase;
  42
  43 /**
  44  * Maintenance script to update image metadata records.
  45  *
  46  * @ingroup Maintenance
  47  */
  48 class ImageBuilder extends Maintenance {
  49         /**
  50          * @var IMaintainableDatabase
  51          */
  52         protected $dbw;
  53
  54         /** @var bool */
  55         private $dryrun;
  56
  57         /** @var LocalRepo|null */
  58         private $repo;
  59
  60         /** @var int */
  61         private $updated;
  62
  63         /** @var int */
  64         private $processed;
  65
  66         /** @var int */
  67         private $count;
  68
  69         /** @var float */
  70         private $startTime;
  71
  72         /** @var string */
  73         private $table;
  74
  75         public function __construct() {
  76                 parent::__construct();
  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->getPrimaryDB();
  85                 $this->dryrun = $this->hasOption( 'dry-run' );
  86                 if ( $this->dryrun ) {
  87                         $this->getServiceContainer()->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 = $this->getServiceContainer()->getRepoGroup()
 104                                 ->newCustomLocalRepo( [
 105                                         // make sure to update old, but compatible img_metadata fields.
 106                                         'updateCompatibleMetadata' => true
 107                                 ] );
 108                 }
 109
 110                 return $this->repo;
 111         }
 112
 113         private function build() {
 114                 $this->buildImage();
 115                 $this->buildOldImage();
 116         }
 117
 118         /**
 119          * @param int $count
 120          * @param string $table
 121          */
 122         private function init( $count, $table ) {
 123                 $this->processed = 0;
 124                 $this->updated = 0;
 125                 $this->count = $count;
 126                 $this->startTime = microtime( true );
 127                 $this->table = $table;
 128         }
 129
 130         private function progress( $updated ) {
 131                 $this->updated += $updated;
 132                 $this->processed++;
 133                 if ( $this->processed % 100 != 0 ) {
 134                         return;
 135                 }
 136                 $portion = $this->processed / $this->count;
 137                 $updateRate = $this->updated / $this->processed;
 138
 139                 $now = microtime( true );
 140                 $delta = $now - $this->startTime;
 141                 $estimatedTotalTime = $delta / $portion;
 142                 $eta = $this->startTime + $estimatedTotalTime;
 143                 $rate = $this->processed / $delta;
 144
 145                 $this->output( sprintf( "%s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
 146                         wfTimestamp( TS_DB, intval( $now ) ),
 147                         $portion * 100.0,
 148                         $this->table,
 149                         wfTimestamp( TS_DB, intval( $eta ) ),
 150                         $this->processed,
 151                         $this->count,
 152                         $rate,
 153                         $updateRate * 100.0 ) );
 154                 flush();
 155         }
 156
 157         private function buildTable( $table, $queryBuilder, $callback ) {
 158                 $count = $this->dbw->newSelectQueryBuilder()
 159                         ->select( 'count(*)' )
 160                         ->from( $table )
 161                         ->caller( __METHOD__ )->fetchField();
 162                 $this->init( $count, $table );
 163                 $this->output( "Processing $table...\n" );
 164
 165                 $result = $queryBuilder->caller( __METHOD__ )->fetchResultSet();
 166
 167                 foreach ( $result as $row ) {
 168                         $update = call_user_func( $callback, $row );
 169                         if ( $update ) {
 170                                 $this->progress( 1 );
 171                         } else {
 172                                 $this->progress( 0 );
 173                         }
 174                 }
 175                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 176         }
 177
 178         private function buildImage() {
 179                 $callback = [ $this, 'imageCallback' ];
 180                 $this->buildTable( 'image', FileSelectQueryBuilder::newForFile( $this->getReplicaDB() ), $callback );
 181         }
 182
 183         private function imageCallback( $row ) {
 184                 // Create a File object from the row
 185                 // This will also upgrade it
 186                 $file = $this->getRepo()->newFileFromRow( $row );
 187
 188                 return $file->getUpgraded();
 189         }
 190
 191         private function buildOldImage() {
 192                 $this->buildTable( 'oldimage', FileSelectQueryBuilder::newForOldFile( $this->getReplicaDB() ),
 193                         [ $this, 'oldimageCallback' ] );
 194         }
 195
 196         private function oldimageCallback( $row ) {
 197                 // Create a File object from the row
 198                 // This will also upgrade it
 199                 if ( $row->oi_archive_name == '' ) {
 200                         $this->output( "Empty oi_archive_name for oi_name={$row->oi_name}\n" );
 201
 202                         return false;
 203                 }
 204                 $file = $this->getRepo()->newFileFromRow( $row );
 205
 206                 return $file->getUpgraded();
 207         }
 208
 209         private function crawlMissing() {
 210                 $this->getRepo()->enumFiles( [ $this, 'checkMissingImage' ] );
 211         }
 212
 213         public function checkMissingImage( $fullpath ) {
 214                 $filename = wfBaseName( $fullpath );
 215                 $row = $this->dbw->newSelectQueryBuilder()
 216                         ->select( [ 'img_name' ] )
 217                         ->from( 'image' )
 218                         ->where( [ 'img_name' => $filename ] )
 219                         ->caller( __METHOD__ )->fetchRow();
 220
 221                 if ( !$row ) {
 222                         // file not registered
 223                         $this->addMissingImage( $filename, $fullpath );
 224                 }
 225         }
 226
 227         private function addMissingImage( $filename, $fullpath ) {
 228                 $timestamp = $this->dbw->timestamp( $this->getRepo()->getFileTimestamp( $fullpath ) );
 229                 $services = $this->getServiceContainer();
 230
 231                 $altname = $services->getContentLanguage()->checkTitleEncoding( $filename );
 232                 if ( $altname != $filename ) {
 233                         if ( $this->dryrun ) {
 234                                 $filename = $altname;
 235                                 $this->output( "Estimating transcoding... $altname\n" );
 236                         } else {
 237                                 // @fixme create renameFile()
 238                                 // @phan-suppress-next-line PhanUndeclaredMethod See comment above...
 239                                 $filename = $this->renameFile( $filename );
 240                         }
 241                 }
 242
 243                 if ( $filename == '' ) {
 244                         $this->output( "Empty filename for $fullpath\n" );
 245
 246                         return;
 247                 }
 248                 if ( !$this->dryrun ) {
 249                         $file = $services->getRepoGroup()->getLocalRepo()->newFile( $filename );
 250                         $pageText = SpecialUpload::getInitialPageText(
 251                                 '(recovered file, missing upload log entry)'
 252                         );
 253                         $user = User::newSystemUser( User::MAINTENANCE_SCRIPT_USER, [ 'steal' => true ] );
 254                         $status = $file->recordUpload3(
 255                                 '',
 256                                 '(recovered file, missing upload log entry)',
 257                                 $pageText,
 258                                 $user,
 259                                 false,
 260                                 $timestamp
 261                         );
 262                         if ( !$status->isOK() ) {
 263                                 $this->output( "Error uploading file $fullpath\n" );
 264
 265                                 return;
 266                         }
 267                 }
 268                 $this->output( $fullpath . "\n" );
 269         }
 270 }
 271
 272 // @codeCoverageIgnoreStart
 273 $maintClass = ImageBuilder::class;
 274 require_once RUN_MAINTENANCE_IF_MAIN;
 275 // @codeCoverageIgnoreEnd