maintenance/rebuildImages.php

   1 <?php
   2 /**
   3  * Script to 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  * http://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( dirname( __FILE__ ) . '/Maintenance.php' );
  34
  35 class ImageBuilder extends Maintenance {
  36
  37         /**
  38          * @var DatabaseBase
  39          */
  40         protected $dbw;
  41
  42         function __construct() {
  43                 parent::__construct();
  44
  45                 global $wgUpdateCompatibleMetadata;
  46                 //make sure to update old, but compatible img_metadata fields.
  47                 $wgUpdateCompatibleMetadata = true;
  48
  49                 $this->mDescription = 'Script to update image metadata records';
  50
  51                 $this->addOption( 'missing', 'Check for files without associated database record' );
  52                 $this->addOption( 'dry-run', 'Only report, don\'t update the database' );
  53         }
  54
  55         public function execute() {
  56                 $this->dbw = wfGetDB( DB_MASTER );
  57                 $this->maxLag = 10; # if slaves are lagged more than 10 secs, wait
  58                 $this->dryrun = $this->hasOption( 'dry-run' );
  59                 if ( $this->dryrun ) {
  60                         $GLOBALS['wgReadOnly'] = 'Dry run mode, image upgrades are suppressed';
  61                 }
  62
  63                 if ( $this->hasOption( 'missing' ) ) {
  64                         $this->crawlMissing();
  65                 } else {
  66                         $this->build();
  67                 }
  68         }
  69
  70         /**
  71          * @return FileRepo
  72          */
  73         function getRepo() {
  74                 if ( !isset( $this->repo ) ) {
  75                         $this->repo = RepoGroup::singleton()->getLocalRepo();
  76                 }
  77                 return $this->repo;
  78         }
  79
  80         function build() {
  81                 $this->buildImage();
  82                 $this->buildOldImage();
  83         }
  84
  85         function init( $count, $table ) {
  86                 $this->processed = 0;
  87                 $this->updated = 0;
  88                 $this->count = $count;
  89                 $this->startTime = wfTime();
  90                 $this->table = $table;
  91         }
  92
  93         function progress( $updated ) {
  94                 $this->updated += $updated;
  95                 $this->processed++;
  96                 if ( $this->processed % 100 != 0 ) {
  97                         return;
  98                 }
  99                 $portion = $this->processed / $this->count;
 100                 $updateRate = $this->updated / $this->processed;
 101
 102                 $now = wfTime();
 103                 $delta = $now - $this->startTime;
 104                 $estimatedTotalTime = $delta / $portion;
 105                 $eta = $this->startTime + $estimatedTotalTime;
 106                 $rate = $this->processed / $delta;
 107
 108                 $this->output( sprintf( "%s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
 109                         wfTimestamp( TS_DB, intval( $now ) ),
 110                         $portion * 100.0,
 111                         $this->table,
 112                         wfTimestamp( TS_DB, intval( $eta ) ),
 113                         $this->processed,
 114                         $this->count,
 115                         $rate,
 116                         $updateRate * 100.0 ) );
 117                 flush();
 118         }
 119
 120         function buildTable( $table, $key, $callback ) {
 121                 $count = $this->dbw->selectField( $table, 'count(*)', '', __METHOD__ );
 122                 $this->init( $count, $table );
 123                 $this->output( "Processing $table...\n" );
 124
 125                 $result = wfGetDB( DB_SLAVE )->select( $table, '*', array(), __METHOD__ );
 126
 127                 foreach ( $result as $row ) {
 128                         $update = call_user_func( $callback, $row, null );
 129                         if ( $update ) {
 130                                 $this->progress( 1 );
 131                         } else {
 132                                 $this->progress( 0 );
 133                         }
 134                 }
 135                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 136         }
 137
 138         function buildImage() {
 139                 $callback = array( $this, 'imageCallback' );
 140                 $this->buildTable( 'image', 'img_name', $callback );
 141         }
 142
 143         function imageCallback( $row, $copy ) {
 144                 // Create a File object from the row
 145                 // This will also upgrade it
 146                 $file = $this->getRepo()->newFileFromRow( $row );
 147                 return $file->getUpgraded();
 148         }
 149
 150         function buildOldImage() {
 151                 $this->buildTable( 'oldimage', 'oi_archive_name',
 152                         array( $this, 'oldimageCallback' ) );
 153         }
 154
 155         function oldimageCallback( $row, $copy ) {
 156                 // Create a File object from the row
 157                 // This will also upgrade it
 158                 if ( $row->oi_archive_name == '' ) {
 159                         $this->output( "Empty oi_archive_name for oi_name={$row->oi_name}\n" );
 160                         return false;
 161                 }
 162                 $file = $this->getRepo()->newFileFromRow( $row );
 163                 return $file->getUpgraded();
 164         }
 165
 166         function crawlMissing() {
 167                 $repo = RepoGroup::singleton()->getLocalRepo();
 168                 $repo->enumFilesInFS( array( $this, 'checkMissingImage' ) );
 169         }
 170
 171         function checkMissingImage( $fullpath ) {
 172                 $filename = wfBaseName( $fullpath );
 173                 if ( is_dir( $fullpath ) ) {
 174                         return;
 175                 }
 176                 if ( is_link( $fullpath ) ) {
 177                         $this->output( "skipping symlink at $fullpath\n" );
 178                         return;
 179                 }
 180                 $row = $this->dbw->selectRow( 'image',
 181                         array( 'img_name' ),
 182                         array( 'img_name' => $filename ),
 183                         __METHOD__ );
 184
 185                 if ( $row ) {
 186                         // already known, move on
 187                         return;
 188                 } else {
 189                         $this->addMissingImage( $filename, $fullpath );
 190                 }
 191         }
 192
 193         function addMissingImage( $filename, $fullpath ) {
 194                 $timestamp = $this->dbw->timestamp( filemtime( $fullpath ) );
 195
 196                 global $wgContLang;
 197                 $altname = $wgContLang->checkTitleEncoding( $filename );
 198                 if ( $altname != $filename ) {
 199                         if ( $this->dryrun ) {
 200                                 $filename = $altname;
 201                                 $this->output( "Estimating transcoding... $altname\n" );
 202                         } else {
 203                                 $filename = $this->renameFile( $filename );
 204                         }
 205                 }
 206
 207                 if ( $filename == '' ) {
 208                         $this->output( "Empty filename for $fullpath\n" );
 209                         return;
 210                 }
 211                 if ( !$this->dryrun ) {
 212                         $file = wfLocalFile( $filename );
 213                         if ( !$file->recordUpload( '', '(recovered file, missing upload log entry)', '', '', '',
 214                                 false, $timestamp ) )
 215                         {
 216                                 $this->output( "Error uploading file $fullpath\n" );
 217                                 return;
 218                         }
 219                 }
 220                 $this->output( $fullpath . "\n" );
 221         }
 222 }
 223
 224 $maintClass = 'ImageBuilder';
 225 require_once( RUN_MAINTENANCE_IF_MAIN );