maintenance/refreshImageMetadata.php

   1 <?php
   2 /**
   3  * Script to refresh image metadata fields. See also rebuildImages.php
   4  *
   5  * Usage: php refreshImageMetadata.php
   6  *
   7  * Copyright © 2011 Brian Wolff
   8  * http://www.mediawiki.org/
   9  *
  10  * This program is free software; you can redistribute it and/or modify
  11  * it under the terms of the GNU General Public License as published by
  12  * the Free Software Foundation; either version 2 of the License, or
  13  * (at your option) any later version.
  14  *
  15  * This program is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  18  * GNU General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU General Public License along
  21  * with this program; if not, write to the Free Software Foundation, Inc.,
  22  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  23  * http://www.gnu.org/copyleft/gpl.html
  24  *
  25  * @file
  26  * @author Brian Wolff
  27  * @ingroup Maintenance
  28  */
  29
  30 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  31
  32 class RefreshImageMetadata extends Maintenance {
  33
  34         /**
  35          * @var DatabaseBase
  36          */
  37         protected $dbw;
  38
  39         function __construct() {
  40                 parent::__construct();
  41
  42                 $this->mDescription = 'Script to update image metadata records';
  43                 $this->setBatchSize( 200 );
  44
  45                 $this->addOption( 'force', 'Reload metadata from file even if the metadata looks ok', false, false, 'f' );
  46                 $this->addOption( 'broken-only', 'Only fix really broken records, leave old but still compatible records alone.' );
  47                 $this->addOption( 'verbose', 'Output extra information about each upgraded/non-upgraded file.', false, false, 'v' );
  48                 $this->addOption( 'start', 'Name of file to start with', false, true );
  49                 $this->addOption( 'end', 'Name of file to end with', false, true );
  50
  51                 $this->addOption( 'mime', '(Inefficient!) Only refresh files with this mime type. Can accept wild-card image/*' , false, true );
  52                 $this->addOption( 'metadata-contains', '(Inefficient!) Only refresh files where the img_metadata field contains this string. Can be used if its known a specific property was being extracted incorrectly.', false, true );
  53
  54         }
  55
  56         public function execute() {
  57                 $force = $this->hasOption( 'force' );
  58                 $brokenOnly = $this->hasOption( 'broken-only' );
  59                 $verbose = $this->hasOption( 'verbose' );
  60                 $start = $this->getOption( 'start', false );
  61                 $this->setupParameters( $force, $brokenOnly );
  62
  63                 $upgraded = 0;
  64                 $leftAlone = 0;
  65                 $error = 0;
  66
  67                 $dbw = wfGetDB( DB_MASTER );
  68                 if ( $this->mBatchSize <= 0 ) {
  69                         $this->error( "Batch size is too low...", 12 );
  70                 }
  71
  72                 $repo = RepoGroup::singleton()->getLocalRepo();
  73                 $conds = $this->getConditions( $dbw );
  74
  75                 // For the WHERE img_name > 'foo' condition that comes after doing a batch
  76                 $conds2 = array();
  77                 if ( $start !== false ) {
  78                         $conds2[] = 'img_name >= ' . $dbw->addQuotes( $start );
  79                 }
  80
  81                 $options = array(
  82                         'LIMIT' => $this->mBatchSize,
  83                         'ORDER BY' => 'img_name ASC',
  84                 );
  85
  86                 do {
  87                         $res = $dbw->select(
  88                                 'image',
  89                                 '*',
  90                                 array_merge( $conds, $conds2 ),
  91                                 __METHOD__,
  92                                 $options
  93                         );
  94
  95                         if ( $res->numRows() > 0 ) {
  96                                 $row1 = $res->current();
  97                                 $this->output( "Processing next {$this->mBatchSize} rows starting with {$row1->img_name}.\n");
  98                                 $res->rewind();
  99                         } else {
 100                                 $this->error( "No images to process.", 4 );
 101                         }
 102
 103                         foreach ( $res as $row ) {
 104                                 $file = $repo->newFileFromRow( $row );
 105                                 if ( $file->getUpgraded() ) {
 106                                         // File was upgraded.
 107                                         $upgraded++;
 108                                         $newLength = strlen( $file->getMetadata() );
 109                                         $oldLength = strlen( $row->img_metadata );
 110                                         if ( $newLength < $oldLength - 5 ) {
 111                                                 // If after updating, the metadata is smaller then
 112                                                 // what it was before, that's probably not a good thing
 113                                                 // because we extract more data with time, not less.
 114                                                 // Thus this probably indicates an error of some sort,
 115                                                 // or at the very least is suspicious. Have the - 5 just
 116                                                 // to weed out any inconsequential changes.
 117                                                 $error++;
 118                                                 $this->output( "Warning: File:{$row->img_name} used to have " .
 119                                                 "$oldLength bytes of metadata but now has $newLength bytes.\n" );
 120                                         } elseif ( $verbose ) {
 121                                                 $this->output("Refreshed File:{$row->img_name}.\n" );
 122                                         }
 123                                 } else {
 124                                         $leftAlone++;
 125                                         if ( $force ) {
 126                                                 $file->upgradeRow();
 127                                                 $newLength = strlen( $file->getMetadata() );
 128                                                 $oldLength = strlen( $row->img_metadata );
 129                                                 if ( $newLength < $oldLength - 5 ) {
 130                                                         $error++;
 131                                                         $this->output( "Warning: File:{$row->img_name} used to have " .
 132                                                         "$oldLength bytes of metadata but now has $newLength bytes. (forced)\n" );
 133
 134                                                 }
 135                                                 if ( $verbose ) {
 136                                                         $this->output("Forcibly refreshed File:{$row->img_name}.\n" );
 137                                                 }
 138                                         }
 139                                         else {
 140                                                 if ( $verbose ) {
 141                                                         $this->output( "Skipping File:{$row->img_name}.\n" );
 142                                                 }
 143                                         }
 144                                 }
 145
 146                         }
 147                         $conds2 = array( 'img_name > ' . $dbw->addQuotes( $row->img_name ) );
 148                         wfWaitForSlaves();
 149                 } while( $res->numRows() === $this->mBatchSize );
 150
 151                 $total = $upgraded + $leftAlone;
 152                 if ( $force ) {
 153                         $this->output( "\nFinished refreshing file metadata for $total files. $upgraded needed to be refreshed, $leftAlone did not need to be but were refreshed anyways, and $error refreshes were suspicious.\n" );
 154                 } else {
 155                         $this->output( "\nFinished refreshing file metadata for $total files. $upgraded were refreshed, $leftAlone were already up to date, and $error refreshes were suspicious.\n" );
 156                 }
 157         }
 158
 159         /**
 160          * @param $dbw DatabaseBase
 161          * @return array
 162          */
 163         function getConditions( $dbw ) {
 164                 $conds = array();
 165
 166                 $end = $this->getOption( 'end', false );
 167                 $mime = $this->getOption( 'mime', false );
 168                 $like = $this->getOption( 'metadata-contains', false );
 169
 170                 if ( $end !== false ) {
 171                         $conds[] = 'img_name <= ' . $dbw->addQuotes( $end ) ;
 172                 }
 173                 if ( $mime !== false ) {
 174                         list( $major, $minor ) = File::splitMime( $mime );
 175                         $conds['img_major_mime'] = $major;
 176                         if ( $minor !== '*' ) {
 177                                 $conds['img_minor_mime'] = $minor;
 178                         }
 179                 }
 180                 if ( $like ) {
 181                         $conds[] = 'img_metadata ' . $dbw->buildLike( $dbw->anyString(), $like, $dbw->anyString() );
 182                 }
 183                 return $conds;
 184         }
 185
 186         /**
 187          * @param $force bool
 188          * @param $brokenOnly bool
 189          */
 190         function setupParameters( $force, $brokenOnly ) {
 191                 global $wgUpdateCompatibleMetadata;
 192
 193                 if ( $brokenOnly ) {
 194                         $wgUpdateCompatibleMetadata = false;
 195                 } else {
 196                         $wgUpdateCompatibleMetadata = true;
 197                 }
 198
 199                 if ( $brokenOnly && $force ) {
 200                         $this->error( 'Cannot use --broken-only and --force together. ', 2 );
 201                 }
 202         }
 203 }
 204
 205
 206 $maintClass = 'RefreshImageMetadata';
 207 require_once( RUN_MAINTENANCE_IF_MAIN );