maintenance/cleanupTitles.php

   1 <?php
   2 /**
   3  * Clean up broken, unparseable titles.
   4  *
   5  * Copyright © 2005 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 require_once __DIR__ . '/cleanupTable.inc';
  29
  30 /**
  31  * Maintenance script to clean up broken, unparseable titles.
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class TitleCleanup extends TableCleanup {
  36         public function __construct() {
  37                 parent::__construct();
  38                 $this->addDescription( 'Script to clean up broken, unparseable titles' );
  39         }
  40
  41         /**
  42          * @param object $row
  43          */
  44         protected function processRow( $row ) {
  45                 global $wgContLang;
  46                 $display = Title::makeName( $row->page_namespace, $row->page_title );
  47                 $verified = $wgContLang->normalize( $display );
  48                 $title = Title::newFromText( $verified );
  49
  50                 if ( !is_null( $title )
  51                         && $title->canExist()
  52                         && $title->getNamespace() == $row->page_namespace
  53                         && $title->getDBkey() === $row->page_title
  54                 ) {
  55                         $this->progress( 0 ); // all is fine
  56
  57                         return;
  58                 }
  59
  60                 if ( $row->page_namespace == NS_FILE && $this->fileExists( $row->page_title ) ) {
  61                         $this->output( "file $row->page_title needs cleanup, please run cleanupImages.php.\n" );
  62                         $this->progress( 0 );
  63                 } elseif ( is_null( $title ) ) {
  64                         $this->output( "page $row->page_id ($display) is illegal.\n" );
  65                         $this->moveIllegalPage( $row );
  66                         $this->progress( 1 );
  67                 } else {
  68                         $this->output( "page $row->page_id ($display) doesn't match self.\n" );
  69                         $this->moveInconsistentPage( $row, $title );
  70                         $this->progress( 1 );
  71                 }
  72         }
  73
  74         /**
  75          * @param string $name
  76          * @return bool
  77          */
  78         protected function fileExists( $name ) {
  79                 // XXX: Doesn't actually check for file existence, just presence of image record.
  80                 // This is reasonable, since cleanupImages.php only iterates over the image table.
  81                 $dbr = $this->getDB( DB_REPLICA );
  82                 $row = $dbr->selectRow( 'image', [ 'img_name' ], [ 'img_name' => $name ], __METHOD__ );
  83
  84                 return $row !== false;
  85         }
  86
  87         /**
  88          * @param object $row
  89          */
  90         protected function moveIllegalPage( $row ) {
  91                 $legal = 'A-Za-z0-9_/\\\\-';
  92                 $legalized = preg_replace_callback( "!([^$legal])!",
  93                         [ $this, 'hexChar' ],
  94                         $row->page_title );
  95                 if ( $legalized == '.' ) {
  96                         $legalized = '(dot)';
  97                 }
  98                 if ( $legalized == '_' ) {
  99                         $legalized = '(space)';
 100                 }
 101                 $legalized = 'Broken/' . $legalized;
 102
 103                 $title = Title::newFromText( $legalized );
 104                 if ( is_null( $title ) ) {
 105                         $clean = 'Broken/id:' . $row->page_id;
 106                         $this->output( "Couldn't legalize; form '$legalized' still invalid; using '$clean'\n" );
 107                         $title = Title::newFromText( $clean );
 108                 } elseif ( $title->exists() ) {
 109                         $clean = 'Broken/id:' . $row->page_id;
 110                         $this->output( "Legalized for '$legalized' exists; using '$clean'\n" );
 111                         $title = Title::newFromText( $clean );
 112                 }
 113
 114                 $dest = $title->getDBkey();
 115                 if ( $this->dryrun ) {
 116                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace," .
 117                                 "'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 118                 } else {
 119                         $this->output( "renaming $row->page_id ($row->page_namespace," .
 120                                 "'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 121                         $dbw = $this->getDB( DB_MASTER );
 122                         $dbw->update( 'page',
 123                                 [ 'page_title' => $dest ],
 124                                 [ 'page_id' => $row->page_id ],
 125                                 __METHOD__ );
 126                 }
 127         }
 128
 129         /**
 130          * @param object $row
 131          * @param Title $title
 132          */
 133         protected function moveInconsistentPage( $row, $title ) {
 134                 if ( $title->exists() || $title->getInterwiki() || !$title->canExist() ) {
 135                         if ( $title->getInterwiki() || !$title->canExist() ) {
 136                                 $prior = $title->getPrefixedDBkey();
 137                         } else {
 138                                 $prior = $title->getDBkey();
 139                         }
 140
 141                         # Old cleanupTitles could move articles there. See bug 23147.
 142                         $ns = $row->page_namespace;
 143                         if ( $ns < 0 ) {
 144                                 $ns = 0;
 145                         }
 146
 147                         # Namespace which no longer exists. Put the page in the main namespace
 148                         # since we don't have any idea of the old namespace name. See bug 68501.
 149                         if ( !MWNamespace::exists( $ns ) ) {
 150                                 $ns = 0;
 151                         }
 152
 153                         $clean = 'Broken/' . $prior;
 154                         $verified = Title::makeTitleSafe( $ns, $clean );
 155                         if ( !$verified || $verified->exists() ) {
 156                                 $blah = "Broken/id:" . $row->page_id;
 157                                 $this->output( "Couldn't legalize; form '$clean' exists; using '$blah'\n" );
 158                                 $verified = Title::makeTitleSafe( $ns, $blah );
 159                         }
 160                         $title = $verified;
 161                 }
 162                 if ( is_null( $title ) ) {
 163                         $this->error( "Something awry; empty title.", true );
 164                 }
 165                 $ns = $title->getNamespace();
 166                 $dest = $title->getDBkey();
 167
 168                 if ( $this->dryrun ) {
 169                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace," .
 170                                 "'$row->page_title') to ($ns,'$dest')\n" );
 171                 } else {
 172                         $this->output( "renaming $row->page_id ($row->page_namespace," .
 173                                 "'$row->page_title') to ($ns,'$dest')\n" );
 174                         $dbw = $this->getDB( DB_MASTER );
 175                         $dbw->update( 'page',
 176                                 [
 177                                         'page_namespace' => $ns,
 178                                         'page_title' => $dest
 179                                 ],
 180                                 [ 'page_id' => $row->page_id ],
 181                                 __METHOD__ );
 182                         LinkCache::singleton()->clear();
 183                 }
 184         }
 185 }
 186
 187 $maintClass = "TitleCleanup";
 188 require_once RUN_MAINTENANCE_IF_MAIN;