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  * 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( __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         protected function processRow( $row ) {
  46                 global $wgContLang;
  47                 $display = Title::makeName( $row->page_namespace, $row->page_title );
  48                 $verified = $wgContLang->normalize( $display );
  49                 $title = Title::newFromText( $verified );
  50
  51                 if ( !is_null( $title )
  52                         && $title->canExist()
  53                         && $title->getNamespace() == $row->page_namespace
  54                         && $title->getDBkey() === $row->page_title )
  55                 {
  56                         return $this->progress( 0 );  // all is fine
  57                 }
  58
  59                 if ( $row->page_namespace == NS_FILE && $this->fileExists( $row->page_title ) ) {
  60                         $this->output( "file $row->page_title needs cleanup, please run cleanupImages.php.\n" );
  61                         return $this->progress( 0 );
  62                 } elseif ( is_null( $title ) ) {
  63                         $this->output( "page $row->page_id ($display) is illegal.\n" );
  64                         $this->moveIllegalPage( $row );
  65                         return $this->progress( 1 );
  66                 } else {
  67                         $this->output( "page $row->page_id ($display) doesn't match self.\n" );
  68                         $this->moveInconsistentPage( $row, $title );
  69                         return $this->progress( 1 );
  70                 }
  71         }
  72
  73         protected function fileExists( $name ) {
  74                 // XXX: Doesn't actually check for file existence, just presence of image record.
  75                 // This is reasonable, since cleanupImages.php only iterates over the image table.
  76                 $dbr = wfGetDB( DB_SLAVE );
  77                 $row = $dbr->selectRow( 'image', array( 'img_name' ), array( 'img_name' => $name ), __METHOD__ );
  78                 return $row !== false;
  79         }
  80
  81         protected function moveIllegalPage( $row ) {
  82                 $legal = 'A-Za-z0-9_/\\\\-';
  83                 $legalized = preg_replace_callback( "!([^$legal])!",
  84                         array( &$this, 'hexChar' ),
  85                         $row->page_title );
  86                 if ( $legalized == '.' ) $legalized = '(dot)';
  87                 if ( $legalized == '_' ) $legalized = '(space)';
  88                 $legalized = 'Broken/' . $legalized;
  89
  90                 $title = Title::newFromText( $legalized );
  91                 if ( is_null( $title ) ) {
  92                         $clean = 'Broken/id:' . $row->page_id;
  93                         $this->output( "Couldn't legalize; form '$legalized' still invalid; using '$clean'\n" );
  94                         $title = Title::newFromText( $clean );
  95                 } elseif ( $title->exists() ) {
  96                         $clean = 'Broken/id:' . $row->page_id;
  97                         $this->output( "Legalized for '$legalized' exists; using '$clean'\n" );
  98                         $title = Title::newFromText( $clean );
  99                 }
 100
 101                 $dest = $title->getDBkey();
 102                 if ( $this->dryrun ) {
 103                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 104                 } else {
 105                         $this->output( "renaming $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 106                         $dbw = wfGetDB( DB_MASTER );
 107                         $dbw->update( 'page',
 108                                 array( 'page_title' => $dest ),
 109                                 array( 'page_id' => $row->page_id ),
 110                                 __METHOD__ );
 111                 }
 112         }
 113
 114         protected function moveInconsistentPage( $row, $title ) {
 115                 if ( $title->exists() || $title->getInterwiki() || !$title->canExist() ) {
 116                         if ( $title->getInterwiki() || !$title->canExist() ) {
 117                                 $prior = $title->getPrefixedDbKey();
 118                         } else {
 119                                 $prior = $title->getDBkey();
 120                         }
 121
 122                         # Old cleanupTitles could move articles there. See bug 23147.
 123                         $ns = $row->page_namespace;
 124                         if ( $ns < 0 ) $ns = 0;
 125
 126                         $clean = 'Broken/' . $prior;
 127                         $verified = Title::makeTitleSafe( $ns, $clean );
 128                         if ( $verified->exists() ) {
 129                                 $blah = "Broken/id:" . $row->page_id;
 130                                 $this->output( "Couldn't legalize; form '$clean' exists; using '$blah'\n" );
 131                                 $verified = Title::makeTitleSafe( $ns, $blah );
 132                         }
 133                         $title = $verified;
 134                 }
 135                 if ( is_null( $title ) ) {
 136                         $this->error( "Something awry; empty title.", true );
 137                 }
 138                 $ns = $title->getNamespace();
 139                 $dest = $title->getDBkey();
 140
 141                 if ( $this->dryrun ) {
 142                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace,'$row->page_title') to ($ns,'$dest')\n" );
 143                 } else {
 144                         $this->output( "renaming $row->page_id ($row->page_namespace,'$row->page_title') to ($ns,'$dest')\n" );
 145                         $dbw = wfGetDB( DB_MASTER );
 146                         $dbw->update( 'page',
 147                                 array(
 148                                         'page_namespace' => $ns,
 149                                         'page_title' => $dest
 150                                 ),
 151                                 array( 'page_id' => $row->page_id ),
 152                                 __METHOD__ );
 153                         LinkCache::singleton()->clear();
 154                 }
 155         }
 156 }
 157
 158 $maintClass = "TitleCleanup";
 159 require_once( RUN_MAINTENANCE_IF_MAIN );