maintenance/cleanupTitles.php

   1 <?php
   2 /**
   3  * Clean up broken, unparseable titles.
   4  *
   5  * Usage: php cleanupTitles.php [--fix]
   6  * Options:
   7  *   --fix  Actually clean up titles; otherwise just checks for them
   8  *
   9  * Copyright © 2005 Brion Vibber <brion@pobox.com>
  10  * https://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 __DIR__ . '/cleanupTable.inc';
  33
  34 /**
  35  * Maintenance script to clean up broken, unparseable titles.
  36  *
  37  * @ingroup Maintenance
  38  */
  39 class TitleCleanup extends TableCleanup {
  40         public function __construct() {
  41                 parent::__construct();
  42                 $this->mDescription = "Script to clean up broken, unparseable titles";
  43         }
  44
  45         /**
  46          * @param object $row
  47          */
  48         protected function processRow( $row ) {
  49                 global $wgContLang;
  50                 $display = Title::makeName( $row->page_namespace, $row->page_title );
  51                 $verified = $wgContLang->normalize( $display );
  52                 $title = Title::newFromText( $verified );
  53
  54                 if ( !is_null( $title )
  55                         && $title->canExist()
  56                         && $title->getNamespace() == $row->page_namespace
  57                         && $title->getDBkey() === $row->page_title
  58                 ) {
  59                         $this->progress( 0 ); // all is fine
  60
  61                         return;
  62                 }
  63
  64                 if ( $row->page_namespace == NS_FILE && $this->fileExists( $row->page_title ) ) {
  65                         $this->output( "file $row->page_title needs cleanup, please run cleanupImages.php.\n" );
  66                         $this->progress( 0 );
  67                 } elseif ( is_null( $title ) ) {
  68                         $this->output( "page $row->page_id ($display) is illegal.\n" );
  69                         $this->moveIllegalPage( $row );
  70                         $this->progress( 1 );
  71                 } else {
  72                         $this->output( "page $row->page_id ($display) doesn't match self.\n" );
  73                         $this->moveInconsistentPage( $row, $title );
  74                         $this->progress( 1 );
  75                 }
  76         }
  77
  78         /**
  79          * @param string $name
  80          * @return bool
  81          */
  82         protected function fileExists( $name ) {
  83                 // XXX: Doesn't actually check for file existence, just presence of image record.
  84                 // This is reasonable, since cleanupImages.php only iterates over the image table.
  85                 $dbr = wfGetDB( DB_SLAVE );
  86                 $row = $dbr->selectRow( 'image', array( 'img_name' ), array( 'img_name' => $name ), __METHOD__ );
  87
  88                 return $row !== false;
  89         }
  90
  91         /**
  92          * @param object $row
  93          */
  94         protected function moveIllegalPage( $row ) {
  95                 $legal = 'A-Za-z0-9_/\\\\-';
  96                 $legalized = preg_replace_callback( "!([^$legal])!",
  97                         array( &$this, 'hexChar' ),
  98                         $row->page_title );
  99                 if ( $legalized == '.' ) {
 100                         $legalized = '(dot)';
 101                 }
 102                 if ( $legalized == '_' ) {
 103                         $legalized = '(space)';
 104                 }
 105                 $legalized = 'Broken/' . $legalized;
 106
 107                 $title = Title::newFromText( $legalized );
 108                 if ( is_null( $title ) ) {
 109                         $clean = 'Broken/id:' . $row->page_id;
 110                         $this->output( "Couldn't legalize; form '$legalized' still invalid; using '$clean'\n" );
 111                         $title = Title::newFromText( $clean );
 112                 } elseif ( $title->exists() ) {
 113                         $clean = 'Broken/id:' . $row->page_id;
 114                         $this->output( "Legalized for '$legalized' exists; using '$clean'\n" );
 115                         $title = Title::newFromText( $clean );
 116                 }
 117
 118                 $dest = $title->getDBkey();
 119                 if ( $this->dryrun ) {
 120                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace," .
 121                                 "'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 122                 } else {
 123                         $this->output( "renaming $row->page_id ($row->page_namespace," .
 124                                 "'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 125                         $dbw = wfGetDB( DB_MASTER );
 126                         $dbw->update( 'page',
 127                                 array( 'page_title' => $dest ),
 128                                 array( 'page_id' => $row->page_id ),
 129                                 __METHOD__ );
 130                 }
 131         }
 132
 133         /**
 134          * @param object $row
 135          * @param Title $title
 136          */
 137         protected function moveInconsistentPage( $row, $title ) {
 138                 if ( $title->exists() || $title->getInterwiki() || !$title->canExist() ) {
 139                         if ( $title->getInterwiki() || !$title->canExist() ) {
 140                                 $prior = $title->getPrefixedDBkey();
 141                         } else {
 142                                 $prior = $title->getDBkey();
 143                         }
 144
 145                         # Old cleanupTitles could move articles there. See bug 23147.
 146                         $ns = $row->page_namespace;
 147                         if ( $ns < 0 ) {
 148                                 $ns = 0;
 149                         }
 150
 151                         $clean = 'Broken/' . $prior;
 152                         $verified = Title::makeTitleSafe( $ns, $clean );
 153                         if ( $verified->exists() ) {
 154                                 $blah = "Broken/id:" . $row->page_id;
 155                                 $this->output( "Couldn't legalize; form '$clean' exists; using '$blah'\n" );
 156                                 $verified = Title::makeTitleSafe( $ns, $blah );
 157                         }
 158                         $title = $verified;
 159                 }
 160                 if ( is_null( $title ) ) {
 161                         $this->error( "Something awry; empty title.", true );
 162                 }
 163                 $ns = $title->getNamespace();
 164                 $dest = $title->getDBkey();
 165
 166                 if ( $this->dryrun ) {
 167                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace," .
 168                                 "'$row->page_title') to ($ns,'$dest')\n" );
 169                 } else {
 170                         $this->output( "renaming $row->page_id ($row->page_namespace," .
 171                                 "'$row->page_title') to ($ns,'$dest')\n" );
 172                         $dbw = wfGetDB( DB_MASTER );
 173                         $dbw->update( 'page',
 174                                 array(
 175                                         'page_namespace' => $ns,
 176                                         'page_title' => $dest
 177                                 ),
 178                                 array( 'page_id' => $row->page_id ),
 179                                 __METHOD__ );
 180                         LinkCache::singleton()->clear();
 181                 }
 182         }
 183 }
 184
 185 $maintClass = "TitleCleanup";
 186 require_once RUN_MAINTENANCE_IF_MAIN;