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  * https://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 Database
  41          */
  42         protected $dbw;
  43
  44         function __construct() {
  45                 parent::__construct();
  46
  47                 $this->addDescription( 'Script to update image metadata records' );
  48                 $this->setBatchSize( 200 );
  49
  50                 $this->addOption(
  51                         'force',
  52                         'Reload metadata from file even if the metadata looks ok',
  53                         false,
  54                         false,
  55                         'f'
  56                 );
  57                 $this->addOption(
  58                         'broken-only',
  59                         'Only fix really broken records, leave old but still compatible records alone.'
  60                 );
  61                 $this->addOption(
  62                         'verbose',
  63                         'Output extra information about each upgraded/non-upgraded file.',
  64                         false,
  65                         false,
  66                         'v'
  67                 );
  68                 $this->addOption( 'start', 'Name of file to start with', false, true );
  69                 $this->addOption( 'end', 'Name of file to end with', false, true );
  70
  71                 $this->addOption(
  72                         'mediatype',
  73                         'Only refresh files with this media type, e.g. BITMAP, UNKNOWN etc.',
  74                         false,
  75                         true
  76                 );
  77                 $this->addOption(
  78                         'mime',
  79                         "Only refresh files with this MIME type. Can accept wild-card 'image/*'. "
  80                                 . "Potentially inefficient unless 'mediatype' is also specified",
  81                         false,
  82                         true
  83                 );
  84                 $this->addOption(
  85                         'metadata-contains',
  86                         '(Inefficient!) Only refresh files where the img_metadata field '
  87                                 . 'contains this string. Can be used if its known a specific '
  88                                 . 'property was being extracted incorrectly.',
  89                         false,
  90                         true
  91                 );
  92         }
  93
  94         public function execute() {
  95                 $force = $this->hasOption( 'force' );
  96                 $brokenOnly = $this->hasOption( 'broken-only' );
  97                 $verbose = $this->hasOption( 'verbose' );
  98                 $start = $this->getOption( 'start', false );
  99                 $this->setupParameters( $force, $brokenOnly );
 100
 101                 $upgraded = 0;
 102                 $leftAlone = 0;
 103                 $error = 0;
 104
 105                 $dbw = $this->getDB( DB_MASTER );
 106                 if ( $this->mBatchSize <= 0 ) {
 107                         $this->error( "Batch size is too low...", 12 );
 108                 }
 109
 110                 $repo = RepoGroup::singleton()->getLocalRepo();
 111                 $conds = $this->getConditions( $dbw );
 112
 113                 // For the WHERE img_name > 'foo' condition that comes after doing a batch
 114                 $conds2 = [];
 115                 if ( $start !== false ) {
 116                         $conds2[] = 'img_name >= ' . $dbw->addQuotes( $start );
 117                 }
 118
 119                 $options = [
 120                         'LIMIT' => $this->mBatchSize,
 121                         'ORDER BY' => 'img_name ASC',
 122                 ];
 123
 124                 do {
 125                         $res = $dbw->select(
 126                                 'image',
 127                                 '*',
 128                                 array_merge( $conds, $conds2 ),
 129                                 __METHOD__,
 130                                 $options
 131                         );
 132
 133                         if ( $res->numRows() > 0 ) {
 134                                 $row1 = $res->current();
 135                                 $this->output( "Processing next {$this->mBatchSize} rows starting with {$row1->img_name}.\n" );
 136                                 $res->rewind();
 137                         } else {
 138                                 $this->error( "No images to process.", 4 );
 139                         }
 140
 141                         foreach ( $res as $row ) {
 142                                 try {
 143                                         // LocalFile will upgrade immediately here if obsolete
 144                                         $file = $repo->newFileFromRow( $row );
 145                                         if ( $file->getUpgraded() ) {
 146                                                 // File was upgraded.
 147                                                 $upgraded++;
 148                                                 $newLength = strlen( $file->getMetadata() );
 149                                                 $oldLength = strlen( $row->img_metadata );
 150                                                 if ( $newLength < $oldLength - 5 ) {
 151                                                         // If after updating, the metadata is smaller then
 152                                                         // what it was before, that's probably not a good thing
 153                                                         // because we extract more data with time, not less.
 154                                                         // Thus this probably indicates an error of some sort,
 155                                                         // or at the very least is suspicious. Have the - 5 just
 156                                                         // to weed out any inconsequential changes.
 157                                                         $error++;
 158                                                         $this->output(
 159                                                                 "Warning: File:{$row->img_name} used to have " .
 160                                                                 "$oldLength bytes of metadata but now has $newLength bytes.\n"
 161                                                         );
 162                                                 } elseif ( $verbose ) {
 163                                                         $this->output( "Refreshed File:{$row->img_name}.\n" );
 164                                                 }
 165                                         } else {
 166                                                 $leftAlone++;
 167                                                 if ( $force ) {
 168                                                         $file->upgradeRow();
 169                                                         $newLength = strlen( $file->getMetadata() );
 170                                                         $oldLength = strlen( $row->img_metadata );
 171                                                         if ( $newLength < $oldLength - 5 ) {
 172                                                                 $error++;
 173                                                                 $this->output(
 174                                                                         "Warning: File:{$row->img_name} used to have " .
 175                                                                         "$oldLength bytes of metadata but now has $newLength bytes. (forced)\n"
 176                                                                 );
 177                                                         }
 178                                                         if ( $verbose ) {
 179                                                                 $this->output( "Forcibly refreshed File:{$row->img_name}.\n" );
 180                                                         }
 181                                                 } else {
 182                                                         if ( $verbose ) {
 183                                                                 $this->output( "Skipping File:{$row->img_name}.\n" );
 184                                                         }
 185                                                 }
 186                                         }
 187                                 } catch ( Exception $e ) {
 188                                         $this->output( "{$row->img_name} failed. {$e->getMessage()}\n" );
 189                                 }
 190                         }
 191                         $conds2 = [ 'img_name > ' . $dbw->addQuotes( $row->img_name ) ];
 192                         wfWaitForSlaves();
 193                 } while ( $res->numRows() === $this->mBatchSize );
 194
 195                 $total = $upgraded + $leftAlone;
 196                 if ( $force ) {
 197                         $this->output( "\nFinished refreshing file metadata for $total files. "
 198                                 . "$upgraded needed to be refreshed, $leftAlone did not need to "
 199                                 . "be but were refreshed anyways, and $error refreshes were suspicious.\n" );
 200                 } else {
 201                         $this->output( "\nFinished refreshing file metadata for $total files. "
 202                                 . "$upgraded were refreshed, $leftAlone were already up to date, "
 203                                 . "and $error refreshes were suspicious.\n" );
 204                 }
 205         }
 206
 207         /**
 208          * @param Database $dbw
 209          * @return array
 210          */
 211         function getConditions( $dbw ) {
 212                 $conds = [];
 213
 214                 $end = $this->getOption( 'end', false );
 215                 $mime = $this->getOption( 'mime', false );
 216                 $mediatype = $this->getOption( 'mediatype', false );
 217                 $like = $this->getOption( 'metadata-contains', false );
 218
 219                 if ( $end !== false ) {
 220                         $conds[] = 'img_name <= ' . $dbw->addQuotes( $end );
 221                 }
 222                 if ( $mime !== false ) {
 223                         list( $major, $minor ) = File::splitMime( $mime );
 224                         $conds['img_major_mime'] = $major;
 225                         if ( $minor !== '*' ) {
 226                                 $conds['img_minor_mime'] = $minor;
 227                         }
 228                 }
 229                 if ( $mediatype !== false ) {
 230                         $conds['img_media_type'] = $mediatype;
 231                 }
 232                 if ( $like ) {
 233                         $conds[] = 'img_metadata ' . $dbw->buildLike( $dbw->anyString(), $like, $dbw->anyString() );
 234                 }
 235
 236                 return $conds;
 237         }
 238
 239         /**
 240          * @param bool $force
 241          * @param bool $brokenOnly
 242          */
 243         function setupParameters( $force, $brokenOnly ) {
 244                 global $wgUpdateCompatibleMetadata;
 245
 246                 if ( $brokenOnly ) {
 247                         $wgUpdateCompatibleMetadata = false;
 248                 } else {
 249                         $wgUpdateCompatibleMetadata = true;
 250                 }
 251
 252                 if ( $brokenOnly && $force ) {
 253                         $this->error( 'Cannot use --broken-only and --force together. ', 2 );
 254                 }
 255         }
 256 }
 257
 258 $maintClass = 'RefreshImageMetadata';
 259 require_once RUN_MAINTENANCE_IF_MAIN;