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
  30         /**
  31          * @var DatabaseBase
  32          */
  33         protected $db;
  34
  35         public function __construct() {
  36                 parent::__construct();
  37                 $this->mDescription = "";
  38                 $this->addOption( 'fix', 'Attempt to automatically fix errors' );
  39                 $this->addOption( 'suffix', "Dupes will be renamed with correct namespace with " .
  40                                                                         "<text> appended after the article name", false, true );
  41                 $this->addOption( 'prefix', "Do an explicit check for the given title prefix " .
  42                                                                         "appended after the article name", false, true );
  43         }
  44
  45         public function execute() {
  46                 global $wgTitle;
  47
  48                 $this->db = wfGetDB( DB_MASTER );
  49                 $wgTitle = Title::newFromText( 'Namespace title conflict cleanup script' );
  50
  51                 $fix = $this->hasOption( 'fix' );
  52                 $suffix = $this->getOption( 'suffix', '' );
  53                 $prefix = $this->getOption( 'prefix', '' );
  54                 $key = intval( $this->getOption( 'key', 0 ) );
  55
  56                 if ( $prefix ) {
  57                         $retval = $this->checkPrefix( $key, $prefix, $fix, $suffix );
  58                 } else {
  59                         $retval = $this->checkAll( $fix, $suffix );
  60                 }
  61
  62                 if ( $retval ) {
  63                         $this->output( "\nLooks good!\n" );
  64                 } else {
  65                         $this->output( "\nOh noeees\n" );
  66                 }
  67         }
  68
  69         /**
  70          * @todo Document
  71          * @param $fix Boolean: whether or not to fix broken entries
  72          * @param $suffix String: suffix to append to renamed articles
  73          */
  74         private function checkAll( $fix, $suffix = '' ) {
  75                 global $wgContLang, $wgNamespaceAliases, $wgCapitalLinks;
  76
  77                 $spaces = array();
  78
  79                 // List interwikis first, so they'll be overridden
  80                 // by any conflicting local namespaces.
  81                 foreach ( $this->getInterwikiList() as $prefix ) {
  82                         $name = $wgContLang->ucfirst( $prefix );
  83                         $spaces[$name] = 0;
  84                 }
  85
  86                 // Now pull in all canonical and alias namespaces...
  87                 foreach ( MWNamespace::getCanonicalNamespaces() as $ns => $name ) {
  88                         // This includes $wgExtraNamespaces
  89                         if ( $name !== '' ) {
  90                                 $spaces[$name] = $ns;
  91                         }
  92                 }
  93                 foreach ( $wgContLang->getNamespaces() as $ns => $name ) {
  94                         if ( $name !== '' ) {
  95                                 $spaces[$name] = $ns;
  96                         }
  97                 }
  98                 foreach ( $wgNamespaceAliases as $name => $ns ) {
  99                         $spaces[$name] = $ns;
 100                 }
 101                 foreach ( $wgContLang->getNamespaceAliases() as $name => $ns ) {
 102                         $spaces[$name] = $ns;
 103                 }
 104
 105                 // We'll need to check for lowercase keys as well,
 106                 // since we're doing case-sensitive searches in the db.
 107                 foreach ( $spaces as $name => $ns ) {
 108                         $moreNames = array();
 109                         $moreNames[] = $wgContLang->uc( $name );
 110                         $moreNames[] = $wgContLang->ucfirst( $wgContLang->lc( $name ) );
 111                         $moreNames[] = $wgContLang->ucwords( $name );
 112                         $moreNames[] = $wgContLang->ucwords( $wgContLang->lc( $name ) );
 113                         $moreNames[] = $wgContLang->ucwordbreaks( $name );
 114                         $moreNames[] = $wgContLang->ucwordbreaks( $wgContLang->lc( $name ) );
 115                         if ( !$wgCapitalLinks ) {
 116                                 foreach ( $moreNames as $altName ) {
 117                                         $moreNames[] = $wgContLang->lcfirst( $altName );
 118                                 }
 119                                 $moreNames[] = $wgContLang->lcfirst( $name );
 120                         }
 121                         foreach ( array_unique( $moreNames ) as $altName ) {
 122                                 if ( $altName !== $name ) {
 123                                         $spaces[$altName] = $ns;
 124                                 }
 125                         }
 126                 }
 127
 128                 ksort( $spaces );
 129                 asort( $spaces );
 130
 131                 $ok = true;
 132                 foreach ( $spaces as $name => $ns ) {
 133                         $ok = $this->checkNamespace( $ns, $name, $fix, $suffix ) && $ok;
 134                 }
 135                 return $ok;
 136         }
 137
 138         /**
 139          * Get the interwiki list
 140          *
 141          * @todo Needs to respect interwiki cache!
 142          * @return Array
 143          */
 144         private function getInterwikiList() {
 145                 $result = $this->db->select( 'interwiki', array( 'iw_prefix' ) );
 146                 $prefixes = array();
 147                 foreach ( $result as $row ) {
 148                         $prefixes[] = $row->iw_prefix;
 149                 }
 150                 return $prefixes;
 151         }
 152
 153         /**
 154          * @todo Document
 155          * @param $ns Integer: a namespace id
 156          * @param $name String
 157          * @param $fix Boolean: whether to fix broken entries
 158          * @param $suffix String: suffix to append to renamed articles
 159          */
 160         private function checkNamespace( $ns, $name, $fix, $suffix = '' ) {
 161                 $conflicts = $this->getConflicts( $ns, $name );
 162                 $count = count( $conflicts );
 163                 if ( $count == 0 ) {
 164                         return true;
 165                 }
 166
 167                 $ok = true;
 168                 foreach ( $conflicts as $row ) {
 169                         $resolvable = $this->reportConflict( $row, $suffix );
 170                         $ok = $ok && $resolvable;
 171                         if ( $fix && ( $resolvable || $suffix != '' ) ) {
 172                                 $ok = $this->resolveConflict( $row, $resolvable, $suffix ) && $ok;
 173                         }
 174                 }
 175                 return $ok;
 176         }
 177
 178         /**
 179          * @todo: do this for reals
 180          */
 181         private function checkPrefix( $key, $prefix, $fix, $suffix = '' ) {
 182                 $this->output( "Checking prefix \"$prefix\" vs namespace $key\n" );
 183                 return $this->checkNamespace( $key, $prefix, $fix, $suffix );
 184         }
 185
 186         /**
 187          * Find pages in mainspace that have a prefix of the new namespace
 188          * so we know titles that will need migrating
 189          *
 190          * @param $ns Integer: namespace id (id for new namespace?)
 191          * @param $name String: prefix that is being made a namespace
 192          *
 193          * @return array
 194          */
 195         private function getConflicts( $ns, $name ) {
 196                 $page  = 'page';
 197                 $table = $this->db->tableName( $page );
 198
 199                 $prefix     = $this->db->strencode( $name );
 200                 $encNamespace = $this->db->addQuotes( $ns );
 201
 202                 $titleSql = "TRIM(LEADING '$prefix:' FROM {$page}_title)";
 203                 if ( $ns == 0 ) {
 204                         // An interwiki; try an alternate encoding with '-' for ':'
 205                         $titleSql = $this->db->buildConcat( array( "'$prefix-'", $titleSql ) );
 206                 }
 207
 208                 $sql = "SELECT {$page}_id    AS id,
 209                                            {$page}_title AS oldtitle,
 210                                            $encNamespace + {$page}_namespace AS namespace,
 211                                    $titleSql     AS title,
 212                                    {$page}_namespace AS oldnamespace
 213                                   FROM {$table}
 214                                  WHERE ( {$page}_namespace=0 OR {$page}_namespace=1 )
 215                                    AND {$page}_title " . $this->db->buildLike( $name . ':', $this->db->anyString() );
 216
 217                 $result = $this->db->query( $sql, __METHOD__ );
 218
 219                 $set = array();
 220                 foreach ( $result as $row ) {
 221                         $set[] = $row;
 222                 }
 223                 return $set;
 224         }
 225
 226         /**
 227          * Report any conflicts we find
 228          *
 229          * @return bool
 230          */
 231         private function reportConflict( $row, $suffix ) {
 232                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 233                 if ( is_null( $newTitle ) || !$newTitle->canExist() ) {
 234                         // Title is also an illegal title...
 235                         // For the moment we'll let these slide to cleanupTitles or whoever.
 236                         $this->output( sprintf( "... %d (%d,\"%s\")\n",
 237                                 $row->id,
 238                                 $row->oldnamespace,
 239                                 $row->oldtitle ) );
 240                         $this->output( "...  *** cannot resolve automatically; illegal title ***\n" );
 241                         return false;
 242                 }
 243
 244                 $this->output( sprintf( "... %d (%d,\"%s\") -> (%d,\"%s\") [[%s]]\n",
 245                         $row->id,
 246                         $row->oldnamespace,
 247                         $row->oldtitle,
 248                         $newTitle->getNamespace(),
 249                         $newTitle->getDBkey(),
 250                         $newTitle->getPrefixedText() ) );
 251
 252                 $id = $newTitle->getArticleId();
 253                 if ( $id ) {
 254                         $this->output( "...  *** cannot resolve automatically; page exists with ID $id ***\n" );
 255                         return false;
 256                 } else {
 257                         return true;
 258                 }
 259         }
 260
 261         /**
 262          * Resolve any conflicts
 263          *
 264          * @param $row Object: row from the page table to fix
 265          * @param $resolvable Boolean
 266          * @param $suffix String: suffix to append to the fixed page
 267          */
 268         private function resolveConflict( $row, $resolvable, $suffix ) {
 269                 if ( !$resolvable ) {
 270                         $this->output( "...  *** old title {$row->title}\n" );
 271                         while ( true ) {
 272                                 $row->title .= $suffix;
 273                                 $this->output( "...  *** new title {$row->title}\n" );
 274                                 $title = Title::makeTitleSafe( $row->namespace, $row->title );
 275                                 if ( !$title ) {
 276                                         $this->output( "... !!! invalid title\n" );
 277                                         return false;
 278                                 }
 279                                 $id = $title->getArticleId();
 280                                 if ( $id ) {
 281                                         $this->output( "...  *** page exists with ID $id ***\n" );
 282                                 } else {
 283                                         break;
 284                                 }
 285                         }
 286                         $this->output( "...  *** using suffixed form [[" . $title->getPrefixedText() . "]] ***\n" );
 287                 }
 288                 $this->resolveConflictOn( $row, 'page', 'page' );
 289                 return true;
 290         }
 291
 292         /**
 293          * Resolve a given conflict
 294          *
 295          * @param $row Object: row from the old broken entry
 296          * @param $table String: table to update
 297          * @param $prefix String: prefix for column name, like page or ar
 298          */
 299         private function resolveConflictOn( $row, $table, $prefix ) {
 300                 $this->output( "... resolving on $table... " );
 301                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 302                 $this->db->update( $table,
 303                         array(
 304                                 "{$prefix}_namespace" => $newTitle->getNamespace(),
 305                                 "{$prefix}_title"     => $newTitle->getDBkey(),
 306                         ),
 307                         array(
 308                                 // "{$prefix}_namespace" => 0,
 309                                 // "{$prefix}_title"     => $row->oldtitle,
 310                                 "{$prefix}_id"           => $row->id,
 311                         ),
 312                         __METHOD__ );
 313                 $this->output( "ok.\n" );
 314                 return true;
 315         }
 316 }
 317
 318 $maintClass = "NamespaceConflictChecker";
 319 require_once( RUN_MAINTENANCE_IF_MAIN );