maintenance/cleanupTitles.php

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