maintenance/refreshImageMetadata.php

   1 <?php
   2 /**
   3  * 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 __DIR__ . '/Maintenance.php';
  31
  32 /**
  33  * Maintenance script to refresh image metadata fields.
  34  *
  35  * @ingroup Maintenance
  36  */
  37 class RefreshImageMetadata extends Maintenance {
  38
  39         /**
  40          * @var DatabaseBase
  41          */
  42         protected $dbw;
  43
  44         function __construct() {
  45                 parent::__construct();
  46
  47                 $this->mDescription = 'Script to update image metadata records';
  48                 $this->setBatchSize( 200 );
  49
  50                 $this->addOption( 'force', 'Reload metadata from file even if the metadata looks ok', false, false, 'f' );
  51                 $this->addOption( 'broken-only', 'Only fix really broken records, leave old but still compatible records alone.' );
  52                 $this->addOption( 'verbose', 'Output extra information about each upgraded/non-upgraded file.', false, false, 'v' );
  53                 $this->addOption( 'start', 'Name of file to start with', false, true );
  54                 $this->addOption( 'end', 'Name of file to end with', false, true );
  55
  56                 $this->addOption( 'mime', '(Inefficient!) Only refresh files with this mime type. Can accept wild-card image/*', false, true );
  57                 $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 );
  58
  59         }
  60
  61         public function execute() {
  62                 $force = $this->hasOption( 'force' );
  63                 $brokenOnly = $this->hasOption( 'broken-only' );
  64                 $verbose = $this->hasOption( 'verbose' );
  65                 $start = $this->getOption( 'start', false );
  66                 $this->setupParameters( $force, $brokenOnly );
  67
  68                 $upgraded = 0;
  69                 $leftAlone = 0;
  70                 $error = 0;
  71
  72                 $dbw = wfGetDB( DB_MASTER );
  73                 if ( $this->mBatchSize <= 0 ) {
  74                         $this->error( "Batch size is too low...", 12 );
  75                 }
  76
  77                 $repo = RepoGroup::singleton()->getLocalRepo();
  78                 $conds = $this->getConditions( $dbw );
  79
  80                 // For the WHERE img_name > 'foo' condition that comes after doing a batch
  81                 $conds2 = array();
  82                 if ( $start !== false ) {
  83                         $conds2[] = 'img_name >= ' . $dbw->addQuotes( $start );
  84                 }
  85
  86                 $options = array(
  87                         'LIMIT' => $this->mBatchSize,
  88                         'ORDER BY' => 'img_name ASC',
  89                 );
  90
  91                 do {
  92                         $res = $dbw->select(
  93                                 'image',
  94                                 '*',
  95                                 array_merge( $conds, $conds2 ),
  96                                 __METHOD__,
  97                                 $options
  98                         );
  99
 100                         if ( $res->numRows() > 0 ) {
 101                                 $row1 = $res->current();
 102                                 $this->output( "Processing next {$this->mBatchSize} rows starting with {$row1->img_name}.\n" );
 103                                 $res->rewind();
 104                         } else {
 105                                 $this->error( "No images to process.", 4 );
 106                         }
 107
 108                         foreach ( $res as $row ) {
 109                                 $file = $repo->newFileFromRow( $row );
 110                                 if ( $file->getUpgraded() ) {
 111                                         // File was upgraded.
 112                                         $upgraded++;
 113                                         $newLength = strlen( $file->getMetadata() );
 114                                         $oldLength = strlen( $row->img_metadata );
 115                                         if ( $newLength < $oldLength - 5 ) {
 116                                                 // If after updating, the metadata is smaller then
 117                                                 // what it was before, that's probably not a good thing
 118                                                 // because we extract more data with time, not less.
 119                                                 // Thus this probably indicates an error of some sort,
 120                                                 // or at the very least is suspicious. Have the - 5 just
 121                                                 // to weed out any inconsequential changes.
 122                                                 $error++;
 123                                                 $this->output( "Warning: File:{$row->img_name} used to have " .
 124                                                 "$oldLength bytes of metadata but now has $newLength bytes.\n" );
 125                                         } elseif ( $verbose ) {
 126                                                 $this->output( "Refreshed File:{$row->img_name}.\n" );
 127                                         }
 128                                 } else {
 129                                         $leftAlone++;
 130                                         if ( $force ) {
 131                                                 $file->upgradeRow();
 132                                                 $newLength = strlen( $file->getMetadata() );
 133                                                 $oldLength = strlen( $row->img_metadata );
 134                                                 if ( $newLength < $oldLength - 5 ) {
 135                                                         $error++;
 136                                                         $this->output( "Warning: File:{$row->img_name} used to have " .
 137                                                         "$oldLength bytes of metadata but now has $newLength bytes. (forced)\n" );
 138
 139                                                 }
 140                                                 if ( $verbose ) {
 141                                                         $this->output( "Forcibly refreshed File:{$row->img_name}.\n" );
 142                                                 }
 143                                         }
 144                                         else {
 145                                                 if ( $verbose ) {
 146                                                         $this->output( "Skipping File:{$row->img_name}.\n" );
 147                                                 }
 148                                         }
 149                                 }
 150
 151                         }
 152                         $conds2 = array( 'img_name > ' . $dbw->addQuotes( $row->img_name ) );
 153                         wfWaitForSlaves();
 154                 } while ( $res->numRows() === $this->mBatchSize );
 155
 156                 $total = $upgraded + $leftAlone;
 157                 if ( $force ) {
 158                         $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" );
 159                 } else {
 160                         $this->output( "\nFinished refreshing file metadata for $total files. $upgraded were refreshed, $leftAlone were already up to date, and $error refreshes were suspicious.\n" );
 161                 }
 162         }
 163
 164         /**
 165          * @param $dbw DatabaseBase
 166          * @return array
 167          */
 168         function getConditions( $dbw ) {
 169                 $conds = array();
 170
 171                 $end = $this->getOption( 'end', false );
 172                 $mime = $this->getOption( 'mime', false );
 173                 $like = $this->getOption( 'metadata-contains', false );
 174
 175                 if ( $end !== false ) {
 176                         $conds[] = 'img_name <= ' . $dbw->addQuotes( $end );
 177                 }
 178                 if ( $mime !== false ) {
 179                         list( $major, $minor ) = File::splitMime( $mime );
 180                         $conds['img_major_mime'] = $major;
 181                         if ( $minor !== '*' ) {
 182                                 $conds['img_minor_mime'] = $minor;
 183                         }
 184                 }
 185                 if ( $like ) {
 186                         $conds[] = 'img_metadata ' . $dbw->buildLike( $dbw->anyString(), $like, $dbw->anyString() );
 187                 }
 188                 return $conds;
 189         }
 190
 191         /**
 192          * @param $force bool
 193          * @param $brokenOnly bool
 194          */
 195         function setupParameters( $force, $brokenOnly ) {
 196                 global $wgUpdateCompatibleMetadata;
 197
 198                 if ( $brokenOnly ) {
 199                         $wgUpdateCompatibleMetadata = false;
 200                 } else {
 201                         $wgUpdateCompatibleMetadata = true;
 202                 }
 203
 204                 if ( $brokenOnly && $force ) {
 205                         $this->error( 'Cannot use --broken-only and --force together. ', 2 );
 206                 }
 207         }
 208 }
 209
 210
 211 $maintClass = 'RefreshImageMetadata';
 212 require_once RUN_MAINTENANCE_IF_MAIN;