maintenance/cleanupImages.php

   1 <?php
   2 /**
   3  * Clean up broken, unparseable upload filenames.
   4  *
   5  * Copyright © 2005-2006 Brion Vibber <brion@pobox.com>
   6  * https://www.mediawiki.org/
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @file
  24  * @author Brion Vibber <brion at pobox.com>
  25  * @ingroup Maintenance
  26  */
  27
  28 use MediaWiki\MediaWikiServices;
  29
  30 require_once __DIR__ . '/cleanupTable.inc';
  31
  32 /**
  33  * Maintenance script to clean up broken, unparseable upload filenames.
  34  *
  35  * @ingroup Maintenance
  36  */
  37 class CleanupImages extends TableCleanup {
  38         protected $defaultParams = [
  39                 'table' => 'image',
  40                 'conds' => [],
  41                 'index' => 'img_name',
  42                 'callback' => 'processRow',
  43         ];
  44
  45         /** @var LocalRepo|null */
  46         private $repo;
  47
  48         public function __construct() {
  49                 parent::__construct();
  50                 $this->addDescription( 'Script to clean up broken, unparseable upload filenames' );
  51         }
  52
  53         protected function processRow( $row ) {
  54                 $source = $row->img_name;
  55                 if ( $source == '' ) {
  56                         // Ye olde empty rows. Just kill them.
  57                         $this->killRow( $source );
  58
  59                         return $this->progress( 1 );
  60                 }
  61
  62                 $cleaned = $source;
  63
  64                 // About half of old bad image names have percent-codes
  65                 $cleaned = rawurldecode( $cleaned );
  66
  67                 // We also have some HTML entities there
  68                 $cleaned = Sanitizer::decodeCharReferences( $cleaned );
  69
  70                 $contLang = MediaWikiServices::getInstance()->getContentLanguage();
  71
  72                 // Some are old latin-1
  73                 $cleaned = $contLang->checkTitleEncoding( $cleaned );
  74
  75                 // Many of remainder look like non-normalized unicode
  76                 $cleaned = $contLang->normalize( $cleaned );
  77
  78                 $title = Title::makeTitleSafe( NS_FILE, $cleaned );
  79
  80                 if ( is_null( $title ) ) {
  81                         $this->output( "page $source ($cleaned) is illegal.\n" );
  82                         $safe = $this->buildSafeTitle( $cleaned );
  83                         if ( $safe === false ) {
  84                                 return $this->progress( 0 );
  85                         }
  86                         $this->pokeFile( $source, $safe );
  87
  88                         return $this->progress( 1 );
  89                 }
  90
  91                 if ( $title->getDBkey() !== $source ) {
  92                         $munged = $title->getDBkey();
  93                         $this->output( "page $source ($munged) doesn't match self.\n" );
  94                         $this->pokeFile( $source, $munged );
  95
  96                         return $this->progress( 1 );
  97                 }
  98
  99                 return $this->progress( 0 );
 100         }
 101
 102         /**
 103          * @param string $name
 104          */
 105         private function killRow( $name ) {
 106                 if ( $this->dryrun ) {
 107                         $this->output( "DRY RUN: would delete bogus row '$name'\n" );
 108                 } else {
 109                         $this->output( "deleting bogus row '$name'\n" );
 110                         $db = $this->getDB( DB_MASTER );
 111                         $db->delete( 'image',
 112                                 [ 'img_name' => $name ],
 113                                 __METHOD__ );
 114                 }
 115         }
 116
 117         /**
 118          * @param string $name
 119          * @return string
 120          */
 121         private function filePath( $name ) {
 122                 if ( $this->repo === null ) {
 123                         $this->repo = RepoGroup::singleton()->getLocalRepo();
 124                 }
 125
 126                 return $this->repo->getRootDirectory() . '/' . $this->repo->getHashPath( $name ) . $name;
 127         }
 128
 129         private function imageExists( $name, $db ) {
 130                 return $db->selectField( 'image', '1', [ 'img_name' => $name ], __METHOD__ );
 131         }
 132
 133         private function pageExists( $name, $db ) {
 134                 return $db->selectField(
 135                         'page',
 136                         '1',
 137                         [ 'page_namespace' => NS_FILE, 'page_title' => $name ],
 138                         __METHOD__
 139                 );
 140         }
 141
 142         private function pokeFile( $orig, $new ) {
 143                 $path = $this->filePath( $orig );
 144                 if ( !file_exists( $path ) ) {
 145                         $this->output( "missing file: $path\n" );
 146                         $this->killRow( $orig );
 147
 148                         return;
 149                 }
 150
 151                 $db = $this->getDB( DB_MASTER );
 152
 153                 /*
 154                  * To prevent key collisions in the update() statements below,
 155                  * if the target title exists in the image table, or if both the
 156                  * original and target titles exist in the page table, append
 157                  * increasing version numbers until the target title exists in
 158                  * neither.  (See also T18916.)
 159                  */
 160                 $version = 0;
 161                 $final = $new;
 162                 $conflict = ( $this->imageExists( $final, $db ) ||
 163                         ( $this->pageExists( $orig, $db ) && $this->pageExists( $final, $db ) ) );
 164
 165                 while ( $conflict ) {
 166                         $this->output( "Rename conflicts with '$final'...\n" );
 167                         $version++;
 168                         $final = $this->appendTitle( $new, "_$version" );
 169                         $conflict = ( $this->imageExists( $final, $db ) || $this->pageExists( $final, $db ) );
 170                 }
 171
 172                 $finalPath = $this->filePath( $final );
 173
 174                 if ( $this->dryrun ) {
 175                         $this->output( "DRY RUN: would rename $path to $finalPath\n" );
 176                 } else {
 177                         $this->output( "renaming $path to $finalPath\n" );
 178                         // @todo FIXME: Should this use File::move()?
 179                         $this->beginTransaction( $db, __METHOD__ );
 180                         $db->update( 'image',
 181                                 [ 'img_name' => $final ],
 182                                 [ 'img_name' => $orig ],
 183                                 __METHOD__ );
 184                         $db->update( 'oldimage',
 185                                 [ 'oi_name' => $final ],
 186                                 [ 'oi_name' => $orig ],
 187                                 __METHOD__ );
 188                         $db->update( 'page',
 189                                 [ 'page_title' => $final ],
 190                                 [ 'page_title' => $orig, 'page_namespace' => NS_FILE ],
 191                                 __METHOD__ );
 192                         $dir = dirname( $finalPath );
 193                         if ( !file_exists( $dir ) ) {
 194                                 if ( !wfMkdirParents( $dir, null, __METHOD__ ) ) {
 195                                         $this->output( "RENAME FAILED, COULD NOT CREATE $dir" );
 196                                         $this->rollbackTransaction( $db, __METHOD__ );
 197
 198                                         return;
 199                                 }
 200                         }
 201                         if ( rename( $path, $finalPath ) ) {
 202                                 $this->commitTransaction( $db, __METHOD__ );
 203                         } else {
 204                                 $this->error( "RENAME FAILED" );
 205                                 $this->rollbackTransaction( $db, __METHOD__ );
 206                         }
 207                 }
 208         }
 209
 210         private function appendTitle( $name, $suffix ) {
 211                 return preg_replace( '/^(.*)(\..*?)$/',
 212                         "\\1$suffix\\2", $name );
 213         }
 214
 215         private function buildSafeTitle( $name ) {
 216                 $x = preg_replace_callback(
 217                         '/([^' . Title::legalChars() . ']|~)/',
 218                         [ $this, 'hexChar' ],
 219                         $name );
 220
 221                 $test = Title::makeTitleSafe( NS_FILE, $x );
 222                 if ( is_null( $test ) || $test->getDBkey() !== $x ) {
 223                         $this->error( "Unable to generate safe title from '$name', got '$x'" );
 224
 225                         return false;
 226                 }
 227
 228                 return $x;
 229         }
 230 }
 231
 232 $maintClass = CleanupImages::class;
 233 require_once RUN_MAINTENANCE_IF_MAIN;