[Summary view] [Print] [Text view]

   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->mDescription = '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              'mime',
  73              '(Inefficient!) Only refresh files with this MIME type. Can accept wild-card image/*',
  74              false,
  75              true
  76          );
  77          $this->addOption(
  78              'metadata-contains',
  79              '(Inefficient!) Only refresh files where the img_metadata field '
  80                  . 'contains this string. Can be used if its known a specific '
  81                  . 'property was being extracted incorrectly.',
  82              false,
  83              true
  84          );
  85      }
  86  
  87      public function execute() {
  88          $force = $this->hasOption( 'force' );
  89          $brokenOnly = $this->hasOption( 'broken-only' );
  90          $verbose = $this->hasOption( 'verbose' );
  91          $start = $this->getOption( 'start', false );
  92          $this->setupParameters( $force, $brokenOnly );
  93  
  94          $upgraded = 0;
  95          $leftAlone = 0;
  96          $error = 0;
  97  
  98          $dbw = wfGetDB( DB_MASTER );
  99          if ( $this->mBatchSize <= 0 ) {
 100              $this->error( "Batch size is too low...", 12 );
 101          }
 102  
 103          $repo = RepoGroup::singleton()->getLocalRepo();
 104          $conds = $this->getConditions( $dbw );
 105  
 106          // For the WHERE img_name > 'foo' condition that comes after doing a batch
 107          $conds2 = array();
 108          if ( $start !== false ) {
 109              $conds2[] = 'img_name >= ' . $dbw->addQuotes( $start );
 110          }
 111  
 112          $options = array(
 113              'LIMIT' => $this->mBatchSize,
 114              'ORDER BY' => 'img_name ASC',
 115          );
 116  
 117          do {
 118              $res = $dbw->select(
 119                  'image',
 120                  '*',
 121                  array_merge( $conds, $conds2 ),
 122                  __METHOD__,
 123                  $options
 124              );
 125  
 126              if ( $res->numRows() > 0 ) {
 127                  $row1 = $res->current();
 128                  $this->output( "Processing next {$this->mBatchSize} rows starting with {$row1->img_name}.\n" );
 129                  $res->rewind();
 130              } else {
 131                  $this->error( "No images to process.", 4 );
 132              }
 133  
 134              foreach ( $res as $row ) {
 135                  $file = $repo->newFileFromRow( $row );
 136                  if ( $file->getUpgraded() ) {
 137                      // File was upgraded.
 138                      $upgraded++;
 139                      $newLength = strlen( $file->getMetadata() );
 140                      $oldLength = strlen( $row->img_metadata );
 141                      if ( $newLength < $oldLength - 5 ) {
 142                          // If after updating, the metadata is smaller then
 143                          // what it was before, that's probably not a good thing
 144                          // because we extract more data with time, not less.
 145                          // Thus this probably indicates an error of some sort,
 146                          // or at the very least is suspicious. Have the - 5 just
 147                          // to weed out any inconsequential changes.
 148                          $error++;
 149                          $this->output( "Warning: File:{$row->img_name} used to have " .
 150                              "$oldLength bytes of metadata but now has $newLength bytes.\n" );
 151                      } elseif ( $verbose ) {
 152                          $this->output( "Refreshed File:{$row->img_name}.\n" );
 153                      }
 154                  } else {
 155                      $leftAlone++;
 156                      if ( $force ) {
 157                          $file->upgradeRow();
 158                          $newLength = strlen( $file->getMetadata() );
 159                          $oldLength = strlen( $row->img_metadata );
 160                          if ( $newLength < $oldLength - 5 ) {
 161                              $error++;
 162                              $this->output( "Warning: File:{$row->img_name} used to have " .
 163                                  "$oldLength bytes of metadata but now has $newLength bytes. (forced)\n" );
 164                          }
 165                          if ( $verbose ) {
 166                              $this->output( "Forcibly refreshed File:{$row->img_name}.\n" );
 167                          }
 168                      } else {
 169                          if ( $verbose ) {
 170                              $this->output( "Skipping File:{$row->img_name}.\n" );
 171                          }
 172                      }
 173                  }
 174              }
 175              $conds2 = array( 'img_name > ' . $dbw->addQuotes( $row->img_name ) );
 176              wfWaitForSlaves();
 177          } while ( $res->numRows() === $this->mBatchSize );
 178  
 179          $total = $upgraded + $leftAlone;
 180          if ( $force ) {
 181              $this->output( "\nFinished refreshing file metadata for $total files. "
 182                  . "$upgraded needed to be refreshed, $leftAlone did not need to "
 183                  . "be but were refreshed anyways, and $error refreshes were suspicious.\n" );
 184          } else {
 185              $this->output( "\nFinished refreshing file metadata for $total files. "
 186                  . "$upgraded were refreshed, $leftAlone were already up to date, "
 187                  . "and $error refreshes were suspicious.\n" );
 188          }
 189      }
 190  
 191      /**
 192       * @param DatabaseBase $dbw
 193       * @return array
 194       */
 195  	function getConditions( $dbw ) {
 196          $conds = array();
 197  
 198          $end = $this->getOption( 'end', false );
 199          $mime = $this->getOption( 'mime', false );
 200          $like = $this->getOption( 'metadata-contains', false );
 201  
 202          if ( $end !== false ) {
 203              $conds[] = 'img_name <= ' . $dbw->addQuotes( $end );
 204          }
 205          if ( $mime !== false ) {
 206              list( $major, $minor ) = File::splitMime( $mime );
 207              $conds['img_major_mime'] = $major;
 208              if ( $minor !== '*' ) {
 209                  $conds['img_minor_mime'] = $minor;
 210              }
 211          }
 212          if ( $like ) {
 213              $conds[] = 'img_metadata ' . $dbw->buildLike( $dbw->anyString(), $like, $dbw->anyString() );
 214          }
 215  
 216          return $conds;
 217      }
 218  
 219      /**
 220       * @param bool $force
 221       * @param bool $brokenOnly
 222       */
 223  	function setupParameters( $force, $brokenOnly ) {
 224          global $wgUpdateCompatibleMetadata;
 225  
 226          if ( $brokenOnly ) {
 227              $wgUpdateCompatibleMetadata = false;
 228          } else {
 229              $wgUpdateCompatibleMetadata = true;
 230          }
 231  
 232          if ( $brokenOnly && $force ) {
 233              $this->error( 'Cannot use --broken-only and --force together. ', 2 );
 234          }
 235      }
 236  }
 237  
 238  $maintClass = 'RefreshImageMetadata';
 239  require_once RUN_MAINTENANCE_IF_MAIN;