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