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 (C) 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  * @author Brion Vibber <brion at pobox.com>
  28  * @ingroup Maintenance
  29  */
  30
  31 require_once( dirname( __FILE__ ) . '/cleanupTable.inc' );
  32
  33 class TitleCleanup extends TableCleanup {
  34         public function __construct() {
  35                 parent::__construct();
  36                 $this->mDescription = "Script to clean up broken, unparseable titles";
  37         }
  38
  39         protected function processRow( $row ) {
  40                 global $wgContLang;
  41                 $display = Title::makeName( $row->page_namespace, $row->page_title );
  42                 $verified = $wgContLang->normalize( $display );
  43                 $title = Title::newFromText( $verified );
  44
  45                 if ( !is_null( $title )
  46                         && $title->canExist()
  47                         && $title->getNamespace() == $row->page_namespace
  48                         && $title->getDBkey() === $row->page_title )
  49                 {
  50                         return $this->progress( 0 );  // all is fine
  51                 }
  52
  53                 if ( $row->page_namespace == NS_FILE && $this->fileExists( $row->page_title ) ) {
  54                         $this->output( "file $row->page_title needs cleanup, please run cleanupImages.php.\n" );
  55                         return $this->progress( 0 );
  56                 } elseif ( is_null( $title ) ) {
  57                         $this->output( "page $row->page_id ($display) is illegal.\n" );
  58                         $this->moveIllegalPage( $row );
  59                         return $this->progress( 1 );
  60                 } else {
  61                         $this->output( "page $row->page_id ($display) doesn't match self.\n" );
  62                         $this->moveInconsistentPage( $row, $title );
  63                         return $this->progress( 1 );
  64                 }
  65         }
  66
  67         protected function fileExists( $name ) {
  68                 // XXX: Doesn't actually check for file existence, just presence of image record.
  69                 // This is reasonable, since cleanupImages.php only iterates over the image table.
  70                 $dbr = wfGetDB( DB_SLAVE );
  71                 $row = $dbr->selectRow( 'image', array( 'img_name' ), array( 'img_name' => $name ), __METHOD__ );
  72                 return $row !== false;
  73         }
  74
  75         protected function moveIllegalPage( $row ) {
  76                 $legal = 'A-Za-z0-9_/\\\\-';
  77                 $legalized = preg_replace_callback( "!([^$legal])!",
  78                         array( &$this, 'hexChar' ),
  79                         $row->page_title );
  80                 if ( $legalized == '.' ) $legalized = '(dot)';
  81                 if ( $legalized == '_' ) $legalized = '(space)';
  82                 $legalized = 'Broken/' . $legalized;
  83
  84                 $title = Title::newFromText( $legalized );
  85                 if ( is_null( $title ) ) {
  86                         $clean = 'Broken/id:' . $row->page_id;
  87                         $this->output( "Couldn't legalize; form '$legalized' still invalid; using '$clean'\n" );
  88                         $title = Title::newFromText( $clean );
  89                 } elseif ( $title->exists() ) {
  90                         $clean = 'Broken/id:' . $row->page_id;
  91                         $this->output( "Legalized for '$legalized' exists; using '$clean'\n" );
  92                         $title = Title::newFromText( $clean );
  93                 }
  94
  95                 $dest = $title->getDBkey();
  96                 if ( $this->dryrun ) {
  97                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')\n" );
  98                 } else {
  99                         $this->output( "renaming $row->page_id ($row->page_namespace,'$row->page_title') to ($row->page_namespace,'$dest')\n" );
 100                         $dbw = wfGetDB( DB_MASTER );
 101                         $dbw->update( 'page',
 102                                 array( 'page_title' => $dest ),
 103                                 array( 'page_id' => $row->page_id ),
 104                                 __METHOD__ );
 105                 }
 106         }
 107
 108         protected function moveInconsistentPage( $row, $title ) {
 109                 if ( $title->exists() || $title->getInterwiki() || !$title->canExist() ) {
 110                         if ( $title->getInterwiki() || !$title->canExist() ) {
 111                                 $prior = $title->getPrefixedDbKey();
 112                         } else {
 113                                 $prior = $title->getDBkey();
 114                         }
 115
 116                         # Old cleanupTitles could move articles there. See bug 23147.
 117                         $ns = $row->page_namespace;
 118                         if ( $ns < 0 ) $ns = 0;
 119
 120                         $clean = 'Broken/' . $prior;
 121                         $verified = Title::makeTitleSafe( $ns, $clean );
 122                         if ( $verified->exists() ) {
 123                                 $blah = "Broken/id:" . $row->page_id;
 124                                 $this->output( "Couldn't legalize; form '$clean' exists; using '$blah'\n" );
 125                                 $verified = Title::makeTitleSafe( $ns, $blah );
 126                         }
 127                         $title = $verified;
 128                 }
 129                 if ( is_null( $title ) ) {
 130                         $this->error( "Something awry; empty title.", true );
 131                 }
 132                 $ns = $title->getNamespace();
 133                 $dest = $title->getDBkey();
 134
 135                 if ( $this->dryrun ) {
 136                         $this->output( "DRY RUN: would rename $row->page_id ($row->page_namespace,'$row->page_title') to ($ns,'$dest')\n" );
 137                 } else {
 138                         $this->output( "renaming $row->page_id ($row->page_namespace,'$row->page_title') to ($ns,'$dest')\n" );
 139                         $dbw = wfGetDB( DB_MASTER );
 140                         $dbw->update( 'page',
 141                                 array(
 142                                         'page_namespace' => $ns,
 143                                         'page_title' => $dest
 144                                 ),
 145                                 array( 'page_id' => $row->page_id ),
 146                                 __METHOD__ );
 147                         $linkCache = LinkCache::singleton();
 148                         $linkCache->clear();
 149                 }
 150         }
 151 }
 152
 153 $maintClass = "TitleCleanup";
 154 require_once( DO_MAINTENANCE );