maintenance/namespaceDupes.php

   1 <?php
   2 /**
   3  * Check for articles to fix after adding/deleting namespaces
   4  *
   5  * Copyright © 2005-2007 Brion Vibber <brion@pobox.com>
   6  * https://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  * @file
  24  * @ingroup Maintenance
  25  */
  26
  27 require_once __DIR__ . '/Maintenance.php';
  28
  29 /**
  30  * Maintenance script that checks for articles to fix after
  31  * adding/deleting namespaces.
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class NamespaceConflictChecker extends Maintenance {
  36
  37         /**
  38          * @var DatabaseBase
  39          */
  40         protected $db;
  41
  42         public function __construct() {
  43                 parent::__construct();
  44                 $this->mDescription = "";
  45                 $this->addOption( 'fix', 'Attempt to automatically fix errors' );
  46                 $this->addOption( 'suffix', "Dupes will be renamed with correct namespace with " .
  47                         "<text> appended after the article name", false, true );
  48                 $this->addOption( 'prefix', "Do an explicit check for the given title prefix " .
  49                         "appended after the article name", false, true );
  50         }
  51
  52         public function execute() {
  53                 $this->db = wfGetDB( DB_MASTER );
  54
  55                 $fix = $this->hasOption( 'fix' );
  56                 $suffix = $this->getOption( 'suffix', '' );
  57                 $prefix = $this->getOption( 'prefix', '' );
  58                 $key = intval( $this->getOption( 'key', 0 ) );
  59
  60                 if ( $prefix ) {
  61                         $retval = $this->checkPrefix( $key, $prefix, $fix, $suffix );
  62                 } else {
  63                         $retval = $this->checkAll( $fix, $suffix );
  64                 }
  65
  66                 if ( $retval ) {
  67                         $this->output( "\nLooks good!\n" );
  68                 } else {
  69                         $this->output( "\nOh noeees\n" );
  70                 }
  71         }
  72
  73         /**
  74          * @todo Document
  75          * @param bool $fix Whether or not to fix broken entries
  76          * @param string $suffix Suffix to append to renamed articles
  77          *
  78          * @return bool
  79          */
  80         private function checkAll( $fix, $suffix = '' ) {
  81                 global $wgContLang, $wgNamespaceAliases, $wgCapitalLinks;
  82
  83                 $spaces = array();
  84
  85                 // List interwikis first, so they'll be overridden
  86                 // by any conflicting local namespaces.
  87                 foreach ( $this->getInterwikiList() as $prefix ) {
  88                         $name = $wgContLang->ucfirst( $prefix );
  89                         $spaces[$name] = 0;
  90                 }
  91
  92                 // Now pull in all canonical and alias namespaces...
  93                 foreach ( MWNamespace::getCanonicalNamespaces() as $ns => $name ) {
  94                         // This includes $wgExtraNamespaces
  95                         if ( $name !== '' ) {
  96                                 $spaces[$name] = $ns;
  97                         }
  98                 }
  99                 foreach ( $wgContLang->getNamespaces() as $ns => $name ) {
 100                         if ( $name !== '' ) {
 101                                 $spaces[$name] = $ns;
 102                         }
 103                 }
 104                 foreach ( $wgNamespaceAliases as $name => $ns ) {
 105                         $spaces[$name] = $ns;
 106                 }
 107                 foreach ( $wgContLang->getNamespaceAliases() as $name => $ns ) {
 108                         $spaces[$name] = $ns;
 109                 }
 110
 111                 // We'll need to check for lowercase keys as well,
 112                 // since we're doing case-sensitive searches in the db.
 113                 foreach ( $spaces as $name => $ns ) {
 114                         $moreNames = array();
 115                         $moreNames[] = $wgContLang->uc( $name );
 116                         $moreNames[] = $wgContLang->ucfirst( $wgContLang->lc( $name ) );
 117                         $moreNames[] = $wgContLang->ucwords( $name );
 118                         $moreNames[] = $wgContLang->ucwords( $wgContLang->lc( $name ) );
 119                         $moreNames[] = $wgContLang->ucwordbreaks( $name );
 120                         $moreNames[] = $wgContLang->ucwordbreaks( $wgContLang->lc( $name ) );
 121                         if ( !$wgCapitalLinks ) {
 122                                 foreach ( $moreNames as $altName ) {
 123                                         $moreNames[] = $wgContLang->lcfirst( $altName );
 124                                 }
 125                                 $moreNames[] = $wgContLang->lcfirst( $name );
 126                         }
 127                         foreach ( array_unique( $moreNames ) as $altName ) {
 128                                 if ( $altName !== $name ) {
 129                                         $spaces[$altName] = $ns;
 130                                 }
 131                         }
 132                 }
 133
 134                 ksort( $spaces );
 135                 asort( $spaces );
 136
 137                 $ok = true;
 138                 foreach ( $spaces as $name => $ns ) {
 139                         $ok = $this->checkNamespace( $ns, $name, $fix, $suffix ) && $ok;
 140                 }
 141
 142                 return $ok;
 143         }
 144
 145         /**
 146          * Get the interwiki list
 147          *
 148          * @return array
 149          */
 150         private function getInterwikiList() {
 151                 $result = Interwiki::getAllPrefixes();
 152                 $prefixes = array();
 153                 foreach ( $result as $row ) {
 154                         $prefixes[] = $row['iw_prefix'];
 155                 }
 156
 157                 return $prefixes;
 158         }
 159
 160         /**
 161          * @todo Document
 162          * @param int $ns A namespace id
 163          * @param string $name
 164          * @param bool $fix Whether to fix broken entries
 165          * @param string $suffix Suffix to append to renamed articles
 166          * @return bool
 167          */
 168         private function checkNamespace( $ns, $name, $fix, $suffix = '' ) {
 169                 $conflicts = $this->getConflicts( $ns, $name );
 170                 $count = count( $conflicts );
 171                 if ( $count == 0 ) {
 172                         return true;
 173                 }
 174
 175                 $ok = true;
 176                 foreach ( $conflicts as $row ) {
 177                         $resolvable = $this->reportConflict( $row, $suffix );
 178                         $ok = $ok && $resolvable;
 179                         if ( $fix && ( $resolvable || $suffix != '' ) ) {
 180                                 $ok = $this->resolveConflict( $row, $resolvable, $suffix ) && $ok;
 181                         }
 182                 }
 183
 184                 return $ok;
 185         }
 186
 187         /**
 188          * @todo Do this for real
 189          * @param int $ns
 190          * @param string $name
 191          * @param bool $fix
 192          * @param string $suffix
 193          * @return bool
 194          */
 195         private function checkPrefix( $key, $prefix, $fix, $suffix = '' ) {
 196                 $this->output( "Checking prefix \"$prefix\" vs namespace $key\n" );
 197
 198                 return $this->checkNamespace( $key, $prefix, $fix, $suffix );
 199         }
 200
 201         /**
 202          * Find pages in mainspace that have a prefix of the new namespace
 203          * so we know titles that will need migrating
 204          *
 205          * @param int $ns Namespace id (id for new namespace?)
 206          * @param string $name Prefix that is being made a namespace
 207          *
 208          * @return array
 209          */
 210         private function getConflicts( $ns, $name ) {
 211                 $page = 'page';
 212                 $table = $this->db->tableName( $page );
 213
 214                 $prefix = $this->db->strencode( $name );
 215                 $encNamespace = $this->db->addQuotes( $ns );
 216
 217                 $titleSql = "TRIM(LEADING '$prefix:' FROM {$page}_title)";
 218                 if ( $ns == 0 ) {
 219                         // An interwiki; try an alternate encoding with '-' for ':'
 220                         $titleSql = $this->db->buildConcat( array( "'$prefix-'", $titleSql ) );
 221                 }
 222
 223                 $sql = "SELECT {$page}_id    AS id,
 224                                            {$page}_title AS oldtitle,
 225                                            $encNamespace + {$page}_namespace AS namespace,
 226                                    $titleSql     AS title,
 227                                    {$page}_namespace AS oldnamespace
 228                                   FROM {$table}
 229                                  WHERE ( {$page}_namespace=0 OR {$page}_namespace=1 )
 230                                    AND {$page}_title " . $this->db->buildLike( $name . ':', $this->db->anyString() );
 231
 232                 $result = $this->db->query( $sql, __METHOD__ );
 233
 234                 $set = array();
 235                 foreach ( $result as $row ) {
 236                         $set[] = $row;
 237                 }
 238
 239                 return $set;
 240         }
 241
 242         /**
 243          * Report any conflicts we find
 244          *
 245          * @param stdClass $row
 246          * @param string $suffix
 247          * @return bool
 248          */
 249         private function reportConflict( $row, $suffix ) {
 250                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 251                 if ( is_null( $newTitle ) || !$newTitle->canExist() ) {
 252                         // Title is also an illegal title...
 253                         // For the moment we'll let these slide to cleanupTitles or whoever.
 254                         $this->output( sprintf( "... %d (%d,\"%s\")\n",
 255                                 $row->id,
 256                                 $row->oldnamespace,
 257                                 $row->oldtitle ) );
 258                         $this->output( "...  *** cannot resolve automatically; illegal title ***\n" );
 259
 260                         return false;
 261                 }
 262
 263                 $this->output( sprintf( "... %d (%d,\"%s\") -> (%d,\"%s\") [[%s]]\n",
 264                         $row->id,
 265                         $row->oldnamespace,
 266                         $row->oldtitle,
 267                         $newTitle->getNamespace(),
 268                         $newTitle->getDBkey(),
 269                         $newTitle->getPrefixedText() ) );
 270
 271                 $id = $newTitle->getArticleID();
 272                 if ( $id ) {
 273                         $this->output( "...  *** cannot resolve automatically; page exists with ID $id ***\n" );
 274
 275                         return false;
 276                 } else {
 277                         return true;
 278                 }
 279         }
 280
 281         /**
 282          * Resolve any conflicts
 283          *
 284          * @param stClass $row Row from the page table to fix
 285          * @param bool $resolvable
 286          * @param string $suffix Suffix to append to the fixed page
 287          * @return bool
 288          */
 289         private function resolveConflict( $row, $resolvable, $suffix ) {
 290                 if ( !$resolvable ) {
 291                         $this->output( "...  *** old title {$row->title}\n" );
 292                         while ( true ) {
 293                                 $row->title .= $suffix;
 294                                 $this->output( "...  *** new title {$row->title}\n" );
 295                                 $title = Title::makeTitleSafe( $row->namespace, $row->title );
 296                                 if ( !$title ) {
 297                                         $this->output( "... !!! invalid title\n" );
 298
 299                                         return false;
 300                                 }
 301                                 $id = $title->getArticleID();
 302                                 if ( $id ) {
 303                                         $this->output( "...  *** page exists with ID $id ***\n" );
 304                                 } else {
 305                                         break;
 306                                 }
 307                         }
 308                         $this->output( "...  *** using suffixed form [[" . $title->getPrefixedText() . "]] ***\n" );
 309                 }
 310                 $this->resolveConflictOn( $row, 'page', 'page' );
 311
 312                 return true;
 313         }
 314
 315         /**
 316          * Resolve a given conflict
 317          *
 318          * @param stdClass $row Row from the old broken entry
 319          * @param string $table Table to update
 320          * @param string $prefix Prefix for column name, like page or ar
 321          * @return bool
 322          */
 323         private function resolveConflictOn( $row, $table, $prefix ) {
 324                 $this->output( "... resolving on $table... " );
 325                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 326                 $this->db->update( $table,
 327                         array(
 328                                 "{$prefix}_namespace" => $newTitle->getNamespace(),
 329                                 "{$prefix}_title" => $newTitle->getDBkey(),
 330                         ),
 331                         array(
 332                                 // "{$prefix}_namespace" => 0,
 333                                 // "{$prefix}_title" => $row->oldtitle,
 334                                 "{$prefix}_id" => $row->id,
 335                         ),
 336                         __METHOD__ );
 337                 $this->output( "ok.\n" );
 338
 339                 return true;
 340         }
 341 }
 342
 343 $maintClass = "NamespaceConflictChecker";
 344 require_once RUN_MAINTENANCE_IF_MAIN;