maintenance/cleanupImages.php

   1 <?php
   2 /*
   3  * Script to clean up broken, unparseable upload filenames.
   4  *
   5  * Usage: php cleanupImages.php [--fix]
   6  * Options:
   7  *   --fix  Actually clean up titles; otherwise just checks for them
   8  *
   9  * Copyright (C) 2005-2006 Brion Vibber <brion@pobox.com>
  10  * http://www.mediawiki.org/
  11  *
  12  * This program is free software; you can redistribute it and/or modify
  13  * it under the terms of the GNU General Public License as published by
  14  * the Free Software Foundation; either version 2 of the License, or
  15  * (at your option) any later version.
  16  *
  17  * This program is distributed in the hope that it will be useful,
  18  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  19  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  20  * GNU General Public License for more details.
  21  *
  22  * You should have received a copy of the GNU General Public License along
  23  * with this program; if not, write to the Free Software Foundation, Inc.,
  24  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  25  * http://www.gnu.org/copyleft/gpl.html
  26  *
  27  * @file
  28  * @author Brion Vibber <brion at pobox.com>
  29  * @ingroup Maintenance
  30  */
  31
  32 require_once( 'commandLine.inc' );
  33 require_once( 'cleanupTable.inc' );
  34
  35 /**
  36  * @ingroup Maintenance
  37  */
  38 class ImageCleanup extends TableCleanup {
  39         function __construct( $dryrun = false ) {
  40                 parent::__construct( 'image', $dryrun );
  41         }
  42
  43         function processPage( $row ) {
  44                 global $wgContLang;
  45
  46                 $source = $row->img_name;
  47                 if( $source == '' ) {
  48                         // Ye olde empty rows. Just kill them.
  49                         $this->killRow( $source );
  50                         return $this->progress( 1 );
  51                 }
  52
  53                 $cleaned = $source;
  54
  55                 // About half of old bad image names have percent-codes
  56                 $cleaned = rawurldecode( $cleaned );
  57
  58                 // We also have some HTML entities there
  59                 $cleaned = Sanitizer::decodeCharReferences( $cleaned );
  60
  61                 // Some are old latin-1
  62                 $cleaned = $wgContLang->checkTitleEncoding( $cleaned );
  63
  64                 // Many of remainder look like non-normalized unicode
  65                 $cleaned = UtfNormal::cleanUp( $cleaned );
  66
  67                 $title = Title::makeTitleSafe( NS_FILE, $cleaned );
  68
  69                 if( is_null( $title ) ) {
  70                         $this->log( "page $source ($cleaned) is illegal." );
  71                         $safe = $this->buildSafeTitle( $cleaned );
  72                         if( $safe === false )
  73                                 return $this->progress( 0 );
  74                         $this->pokeFile( $source, $safe );
  75                         return $this->progress( 1 );
  76                 }
  77
  78                 if( $title->getDBkey() !== $source ) {
  79                         $munged = $title->getDBkey();
  80                         $this->log( "page $source ($munged) doesn't match self." );
  81                         $this->pokeFile( $source, $munged );
  82                         return $this->progress( 1 );
  83                 }
  84
  85                 $this->progress( 0 );
  86         }
  87
  88         function killRow( $name ) {
  89                 if( $this->dryrun ) {
  90                         $this->log( "DRY RUN: would delete bogus row '$name'" );
  91                 } else {
  92                         $this->log( "deleting bogus row '$name'" );
  93                         $db = wfGetDB( DB_MASTER );
  94                         $db->delete( 'image',
  95                                 array( 'img_name' => $name ),
  96                                 __METHOD__ );
  97                 }
  98         }
  99
 100         function filePath( $name ) {
 101                 if ( !isset( $this->repo ) ) {
 102                         $this->repo = RepoGroup::singleton()->getLocalRepo();
 103                 }
 104                 return $this->repo->getRootDirectory() . '/' . $this->repo->getHashPath( $name ) . $name;
 105         }
 106
 107         function pokeFile( $orig, $new ) {
 108                 $path = $this->filePath( $orig );
 109                 if( !file_exists( $path ) ) {
 110                         $this->log( "missing file: $path" );
 111                         return $this->killRow( $orig );
 112                 }
 113
 114                 $db = wfGetDB( DB_MASTER );
 115                 $version = 0;
 116                 $final = $new;
 117                 $has_page = Title::makeTitle( NS_FILE, $orig )->exists();
 118
 119                 while( true ) {
 120                         $test = $db->selectField( 'image', 'img_name', array( 'img_name' => $final ), __METHOD__ );
 121                         $conflict = ( $test !== false );
 122
 123                         if( !$conflict && ( $has_page || $version > 0 ) ) {
 124                                 // If the file has a description page, don't try to move it to a
 125                                 // title that already has one.  But if we have a misnamed file
 126                                 // with no corresponding page, and there just happens to exist a
 127                                 // page with no file at the corrected title, it's probably not a
 128                                 // coincidence and we shouldn't shy from reuniting them.
 129                                 $conflict = Title::makeTitle( NS_FILE, $final )->exists();
 130                         }
 131                         if( !$conflict ) {
 132                                 break;
 133                         }
 134                         $this->log( "Rename conflicts with '$final'..." );
 135                         $version++;
 136                         $final = $this->appendTitle( $new, "_$version" );
 137                 }
 138
 139                 $finalPath = $this->filePath( $final );
 140
 141                 if( $this->dryrun ) {
 142                         $this->log( "DRY RUN: would rename $path to $finalPath" );
 143                 } else {
 144                         $this->log( "renaming $path to $finalPath" );
 145                         // XXX: should this use File::move()?  FIXME?
 146                         $db->begin();
 147                         $db->update( 'image',
 148                                 array( 'img_name' => $final ),
 149                                 array( 'img_name' => $orig ),
 150                                 __METHOD__ );
 151                         $db->update( 'oldimage',
 152                                 array( 'oi_name' => $final ),
 153                                 array( 'oi_name' => $orig ),
 154                                 __METHOD__ );
 155                         $db->update( 'page',
 156                                 array( 'page_title' => $final ),
 157                                 array( 'page_title' => $orig, 'page_namespace' => NS_FILE ),
 158                                 __METHOD__ );
 159                         $dir = dirname( $finalPath );
 160                         if( !file_exists( $dir ) ) {
 161                                 if( !wfMkdirParents( $dir ) ) {
 162                                         $this->log( "RENAME FAILED, COULD NOT CREATE $dir" );
 163                                         $db->rollback();
 164                                         return;
 165                                 }
 166                         }
 167                         if( rename( $path, $finalPath ) ) {
 168                                 $db->commit();
 169                         } else {
 170                                 $this->log( "RENAME FAILED" );
 171                                 $db->rollback();
 172                         }
 173                 }
 174         }
 175
 176         function appendTitle( $name, $suffix ) {
 177                 return preg_replace( '/^(.*)(\..*?)$/',
 178                         "\\1$suffix\\2", $name );
 179         }
 180
 181         function buildSafeTitle( $name ) {
 182                 global $wgLegalTitleChars;
 183                 $x = preg_replace_callback(
 184                         "/([^$wgLegalTitleChars]|~)/",
 185                         array( $this, 'hexChar' ),
 186                         $name );
 187
 188                 $test = Title::makeTitleSafe( NS_FILE, $x );
 189                 if( is_null( $test ) || $test->getDBkey() !== $x ) {
 190                         $this->log( "Unable to generate safe title from '$name', got '$x'" );
 191                         return false;
 192                 }
 193
 194                 return $x;
 195         }
 196 }
 197
 198 $wgUser->setName( 'Conversion script' );
 199 $caps = new ImageCleanup( !isset( $options['fix'] ) );
 200 $caps->cleanup();
 201
 202