maintenance/namespaceDupes.php

   1 <?php
   2 /**
   3  * Check for articles to fix after adding/deleting namespaces
   4  *
   5  * Copyright (C) 2005-2007 Brion Vibber <brion@pobox.com>
   6  * http://www.mediawiki.org/
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @ingroup Maintenance
  24  */
  25
  26 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  27
  28 class NamespaceConflictChecker extends Maintenance {
  29         public function __construct() {
  30                 parent::__construct();
  31                 $this->mDescription = "";
  32                 $this->addOption( 'fix', 'Attempt to automatically fix errors' );
  33                 $this->addOption( 'suffix', "Dupes will be renamed with correct namespace with " .
  34                                                                         "<text> appended after the article name", false, true );
  35                 $this->addOption( 'prefix', "Do an explicit check for the given title prefix " .
  36                                                                         "appended after the article name", false, true );
  37         }
  38
  39         public function execute() {
  40                 global $wgTitle;
  41
  42                 $this->db = wfGetDB( DB_MASTER );
  43                 $wgTitle = Title::newFromText( 'Namespace title conflict cleanup script' );
  44
  45                 $fix = $this->hasOption( 'fix' );
  46                 $suffix = $this->getOption( 'suffix', '' );
  47                 $prefix = $this->getOption( 'prefix', '' );
  48                 $key = intval( $this->getOption( 'key', 0 ) );
  49
  50                 if ( $prefix ) {
  51                         $retval = $this->checkPrefix( $key, $prefix, $fix, $suffix );
  52                 } else {
  53                         $retval = $this->checkAll( $fix, $suffix );
  54                 }
  55
  56                 if ( $retval ) {
  57                         $this->output( "\nLooks good!\n" );
  58                 } else {
  59                         $this->output( "\nOh noeees\n" );
  60                 }
  61         }
  62
  63         /**
  64          * @todo Document
  65          * @param $fix Boolean: whether or not to fix broken entries
  66          * @param $suffix String: suffix to append to renamed articles
  67          */
  68         private function checkAll( $fix, $suffix = '' ) {
  69                 global $wgContLang, $wgNamespaceAliases, $wgCapitalLinks;
  70
  71                 $spaces = array();
  72
  73                 // List interwikis first, so they'll be overridden
  74                 // by any conflicting local namespaces.
  75                 foreach ( $this->getInterwikiList() as $prefix ) {
  76                         $name = $wgContLang->ucfirst( $prefix );
  77                         $spaces[$name] = 0;
  78                 }
  79
  80                 // Now pull in all canonical and alias namespaces...
  81                 foreach ( MWNamespace::getCanonicalNamespaces() as $ns => $name ) {
  82                         // This includes $wgExtraNamespaces
  83                         if ( $name !== '' ) {
  84                                 $spaces[$name] = $ns;
  85                         }
  86                 }
  87                 foreach ( $wgContLang->getNamespaces() as $ns => $name ) {
  88                         if ( $name !== '' ) {
  89                                 $spaces[$name] = $ns;
  90                         }
  91                 }
  92                 foreach ( $wgNamespaceAliases as $name => $ns ) {
  93                         $spaces[$name] = $ns;
  94                 }
  95                 foreach ( $wgContLang->getNamespaceAliases() as $name => $ns ) {
  96                         $spaces[$name] = $ns;
  97                 }
  98
  99                 // We'll need to check for lowercase keys as well,
 100                 // since we're doing case-sensitive searches in the db.
 101                 foreach ( $spaces as $name => $ns ) {
 102                         $moreNames = array();
 103                         $moreNames[] = $wgContLang->uc( $name );
 104                         $moreNames[] = $wgContLang->ucfirst( $wgContLang->lc( $name ) );
 105                         $moreNames[] = $wgContLang->ucwords( $name );
 106                         $moreNames[] = $wgContLang->ucwords( $wgContLang->lc( $name ) );
 107                         $moreNames[] = $wgContLang->ucwordbreaks( $name );
 108                         $moreNames[] = $wgContLang->ucwordbreaks( $wgContLang->lc( $name ) );
 109                         if ( !$wgCapitalLinks ) {
 110                                 foreach ( $moreNames as $altName ) {
 111                                         $moreNames[] = $wgContLang->lcfirst( $altName );
 112                                 }
 113                                 $moreNames[] = $wgContLang->lcfirst( $name );
 114                         }
 115                         foreach ( array_unique( $moreNames ) as $altName ) {
 116                                 if ( $altName !== $name ) {
 117                                         $spaces[$altName] = $ns;
 118                                 }
 119                         }
 120                 }
 121
 122                 ksort( $spaces );
 123                 asort( $spaces );
 124
 125                 $ok = true;
 126                 foreach ( $spaces as $name => $ns ) {
 127                         $ok = $this->checkNamespace( $ns, $name, $fix, $suffix ) && $ok;
 128                 }
 129                 return $ok;
 130         }
 131
 132         /**
 133          * Get the interwiki list
 134          *
 135          * @todo Needs to respect interwiki cache!
 136          * @return Array
 137          */
 138         private function getInterwikiList() {
 139                 $result = $this->db->select( 'interwiki', array( 'iw_prefix' ) );
 140                 $prefixes = array();
 141                 foreach ( $result as $row ) {
 142                         $prefixes[] = $row->iw_prefix;
 143                 }
 144                 return $prefixes;
 145         }
 146
 147         /**
 148          * @todo Document
 149          * @param $ns Integer: a namespace id
 150          * @param $name String
 151          * @param $fix Boolean: whether to fix broken entries
 152          * @param $suffix String: suffix to append to renamed articles
 153          */
 154         private function checkNamespace( $ns, $name, $fix, $suffix = '' ) {
 155                 $conflicts = $this->getConflicts( $ns, $name );
 156                 $count = count( $conflicts );
 157                 if ( $count == 0 ) {
 158                         return true;
 159                 }
 160
 161                 $ok = true;
 162                 foreach ( $conflicts as $row ) {
 163                         $resolvable = $this->reportConflict( $row, $suffix );
 164                         $ok = $ok && $resolvable;
 165                         if ( $fix && ( $resolvable || $suffix != '' ) ) {
 166                                 $ok = $this->resolveConflict( $row, $resolvable, $suffix ) && $ok;
 167                         }
 168                 }
 169                 return $ok;
 170         }
 171
 172         /**
 173          * @todo: do this for reals
 174          */
 175         private function checkPrefix( $key, $prefix, $fix, $suffix = '' ) {
 176                 $this->output( "Checking prefix \"$prefix\" vs namespace $key\n" );
 177                 return $this->checkNamespace( $key, $prefix, $fix, $suffix );
 178         }
 179
 180         /**
 181          * Find pages in mainspace that have a prefix of the new namespace
 182          * so we know titles that will need migrating
 183          *
 184          * @param $ns Integer: namespace id (id for new namespace?)
 185          * @param $name String: prefix that is being made a namespace
 186          */
 187         private function getConflicts( $ns, $name ) {
 188                 $page  = 'page';
 189                 $table = $this->db->tableName( $page );
 190
 191                 $prefix     = $this->db->strencode( $name );
 192                 $encNamespace = $this->db->addQuotes( $ns );
 193
 194                 $titleSql = "TRIM(LEADING '$prefix:' FROM {$page}_title)";
 195                 if ( $ns == 0 ) {
 196                         // An interwiki; try an alternate encoding with '-' for ':'
 197                         $titleSql = $this->db->buildConcat( array( "'$prefix-'", $titleSql ) );
 198                 }
 199
 200                 $sql = "SELECT {$page}_id    AS id,
 201                                            {$page}_title AS oldtitle,
 202                                            $encNamespace + {$page}_namespace AS namespace,
 203                                    $titleSql     AS title,
 204                                    {$page}_namespace AS oldnamespace
 205                                   FROM {$table}
 206                                  WHERE ( {$page}_namespace=0 OR {$page}_namespace=1 )
 207                                    AND {$page}_title " . $this->db->buildLike( $name . ':', $this->db->anyString() );
 208
 209                 $result = $this->db->query( $sql, __METHOD__ );
 210
 211                 $set = array();
 212                 foreach ( $result as $row ) {
 213                         $set[] = $row;
 214                 }
 215                 return $set;
 216         }
 217
 218         /**
 219          * Report any conflicts we find
 220          */
 221         private function reportConflict( $row, $suffix ) {
 222                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 223                 if ( is_null( $newTitle ) || !$newTitle->canExist() ) {
 224                         // Title is also an illegal title...
 225                         // For the moment we'll let these slide to cleanupTitles or whoever.
 226                         $this->output( sprintf( "... %d (%d,\"%s\")\n",
 227                                 $row->id,
 228                                 $row->oldnamespace,
 229                                 $row->oldtitle ) );
 230                         $this->output( "...  *** cannot resolve automatically; illegal title ***\n" );
 231                         return false;
 232                 }
 233
 234                 $this->output( sprintf( "... %d (%d,\"%s\") -> (%d,\"%s\") [[%s]]\n",
 235                         $row->id,
 236                         $row->oldnamespace,
 237                         $row->oldtitle,
 238                         $newTitle->getNamespace(),
 239                         $newTitle->getDBkey(),
 240                         $newTitle->getPrefixedText() ) );
 241
 242                 $id = $newTitle->getArticleId();
 243                 if ( $id ) {
 244                         $this->output( "...  *** cannot resolve automatically; page exists with ID $id ***\n" );
 245                         return false;
 246                 } else {
 247                         return true;
 248                 }
 249         }
 250
 251         /**
 252          * Resolve any conflicts
 253          *
 254          * @param $row Object: row from the page table to fix
 255          * @param $resolvable Boolean
 256          * @param $suffix String: suffix to append to the fixed page
 257          */
 258         private function resolveConflict( $row, $resolvable, $suffix ) {
 259                 if ( !$resolvable ) {
 260                         $this->output( "...  *** old title {$row->title}\n" );
 261                         while ( true ) {
 262                                 $row->title .= $suffix;
 263                                 $this->output( "...  *** new title {$row->title}\n" );
 264                                 $title = Title::makeTitleSafe( $row->namespace, $row->title );
 265                                 if ( !$title ) {
 266                                         $this->output( "... !!! invalid title\n" );
 267                                         return false;
 268                                 }
 269                                 $id = $title->getArticleId();
 270                                 if ( $id ) {
 271                                         $this->output( "...  *** page exists with ID $id ***\n" );
 272                                 } else {
 273                                         break;
 274                                 }
 275                         }
 276                         $this->output( "...  *** using suffixed form [[" . $title->getPrefixedText() . "]] ***\n" );
 277                 }
 278                 $this->resolveConflictOn( $row, 'page', 'page' );
 279                 return true;
 280         }
 281
 282         /**
 283          * Resolve a given conflict
 284          *
 285          * @param $row Object: row from the old broken entry
 286          * @param $table String: table to update
 287          * @param $prefix String: prefix for column name, like page or ar
 288          */
 289         private function resolveConflictOn( $row, $table, $prefix ) {
 290                 $this->output( "... resolving on $table... " );
 291                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 292                 $this->db->update( $table,
 293                         array(
 294                                 "{$prefix}_namespace" => $newTitle->getNamespace(),
 295                                 "{$prefix}_title"     => $newTitle->getDBkey(),
 296                         ),
 297                         array(
 298                                 // "{$prefix}_namespace" => 0,
 299                                 // "{$prefix}_title"     => $row->oldtitle,
 300                                 "{$prefix}_id"           => $row->id,
 301                         ),
 302                         __METHOD__ );
 303                 $this->output( "ok.\n" );
 304                 return true;
 305         }
 306 }
 307
 308 $maintClass = "NamespaceConflictChecker";
 309 require_once( RUN_MAINTENANCE_IF_MAIN );