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 DatabaseBase
  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                                 $file = $repo->newFileFromRow( $row );
 143                                 if ( $file->getUpgraded() ) {
 144                                         // File was upgraded.
 145                                         $upgraded++;
 146                                         $newLength = strlen( $file->getMetadata() );
 147                                         $oldLength = strlen( $row->img_metadata );
 148                                         if ( $newLength < $oldLength - 5 ) {
 149                                                 // If after updating, the metadata is smaller then
 150                                                 // what it was before, that's probably not a good thing
 151                                                 // because we extract more data with time, not less.
 152                                                 // Thus this probably indicates an error of some sort,
 153                                                 // or at the very least is suspicious. Have the - 5 just
 154                                                 // to weed out any inconsequential changes.
 155                                                 $error++;
 156                                                 $this->output( "Warning: File:{$row->img_name} used to have " .
 157                                                         "$oldLength bytes of metadata but now has $newLength bytes.\n" );
 158                                         } elseif ( $verbose ) {
 159                                                 $this->output( "Refreshed File:{$row->img_name}.\n" );
 160                                         }
 161                                 } else {
 162                                         $leftAlone++;
 163                                         if ( $force ) {
 164                                                 $file->upgradeRow();
 165                                                 $newLength = strlen( $file->getMetadata() );
 166                                                 $oldLength = strlen( $row->img_metadata );
 167                                                 if ( $newLength < $oldLength - 5 ) {
 168                                                         $error++;
 169                                                         $this->output( "Warning: File:{$row->img_name} used to have " .
 170                                                                 "$oldLength bytes of metadata but now has $newLength bytes. (forced)\n" );
 171                                                 }
 172                                                 if ( $verbose ) {
 173                                                         $this->output( "Forcibly refreshed File:{$row->img_name}.\n" );
 174                                                 }
 175                                         } else {
 176                                                 if ( $verbose ) {
 177                                                         $this->output( "Skipping File:{$row->img_name}.\n" );
 178                                                 }
 179                                         }
 180                                 }
 181                         }
 182                         $conds2 = [ 'img_name > ' . $dbw->addQuotes( $row->img_name ) ];
 183                         wfWaitForSlaves();
 184                 } while ( $res->numRows() === $this->mBatchSize );
 185
 186                 $total = $upgraded + $leftAlone;
 187                 if ( $force ) {
 188                         $this->output( "\nFinished refreshing file metadata for $total files. "
 189                                 . "$upgraded needed to be refreshed, $leftAlone did not need to "
 190                                 . "be but were refreshed anyways, and $error refreshes were suspicious.\n" );
 191                 } else {
 192                         $this->output( "\nFinished refreshing file metadata for $total files. "
 193                                 . "$upgraded were refreshed, $leftAlone were already up to date, "
 194                                 . "and $error refreshes were suspicious.\n" );
 195                 }
 196         }
 197
 198         /**
 199          * @param DatabaseBase $dbw
 200          * @return array
 201          */
 202         function getConditions( $dbw ) {
 203                 $conds = [];
 204
 205                 $end = $this->getOption( 'end', false );
 206                 $mime = $this->getOption( 'mime', false );
 207                 $mediatype = $this->getOption( 'mediatype', false );
 208                 $like = $this->getOption( 'metadata-contains', false );
 209
 210                 if ( $end !== false ) {
 211                         $conds[] = 'img_name <= ' . $dbw->addQuotes( $end );
 212                 }
 213                 if ( $mime !== false ) {
 214                         list( $major, $minor ) = File::splitMime( $mime );
 215                         $conds['img_major_mime'] = $major;
 216                         if ( $minor !== '*' ) {
 217                                 $conds['img_minor_mime'] = $minor;
 218                         }
 219                 }
 220                 if ( $mediatype !== false ) {
 221                         $conds['img_media_type'] = $mediatype;
 222                 }
 223                 if ( $like ) {
 224                         $conds[] = 'img_metadata ' . $dbw->buildLike( $dbw->anyString(), $like, $dbw->anyString() );
 225                 }
 226
 227                 return $conds;
 228         }
 229
 230         /**
 231          * @param bool $force
 232          * @param bool $brokenOnly
 233          */
 234         function setupParameters( $force, $brokenOnly ) {
 235                 global $wgUpdateCompatibleMetadata;
 236
 237                 if ( $brokenOnly ) {
 238                         $wgUpdateCompatibleMetadata = false;
 239                 } else {
 240                         $wgUpdateCompatibleMetadata = true;
 241                 }
 242
 243                 if ( $brokenOnly && $force ) {
 244                         $this->error( 'Cannot use --broken-only and --force together. ', 2 );
 245                 }
 246         }
 247 }
 248
 249 $maintClass = 'RefreshImageMetadata';
 250 require_once RUN_MAINTENANCE_IF_MAIN;