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  * 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  * @file
  24  * @ingroup Maintenance
  25  */
  26
  27 require_once( dirname( __FILE__ ) . '/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                 global $wgTitle;
  54
  55                 $this->db = wfGetDB( DB_MASTER );
  56                 $wgTitle = Title::newFromText( 'Namespace title conflict cleanup script' );
  57
  58                 $fix = $this->hasOption( 'fix' );
  59                 $suffix = $this->getOption( 'suffix', '' );
  60                 $prefix = $this->getOption( 'prefix', '' );
  61                 $key = intval( $this->getOption( 'key', 0 ) );
  62
  63                 if ( $prefix ) {
  64                         $retval = $this->checkPrefix( $key, $prefix, $fix, $suffix );
  65                 } else {
  66                         $retval = $this->checkAll( $fix, $suffix );
  67                 }
  68
  69                 if ( $retval ) {
  70                         $this->output( "\nLooks good!\n" );
  71                 } else {
  72                         $this->output( "\nOh noeees\n" );
  73                 }
  74         }
  75
  76         /**
  77          * @todo Document
  78          * @param $fix Boolean: whether or not to fix broken entries
  79          * @param $suffix String: suffix to append to renamed articles
  80          *
  81          * @return bool
  82          */
  83         private function checkAll( $fix, $suffix = '' ) {
  84                 global $wgContLang, $wgNamespaceAliases, $wgCapitalLinks;
  85
  86                 $spaces = array();
  87
  88                 // List interwikis first, so they'll be overridden
  89                 // by any conflicting local namespaces.
  90                 foreach ( $this->getInterwikiList() as $prefix ) {
  91                         $name = $wgContLang->ucfirst( $prefix );
  92                         $spaces[$name] = 0;
  93                 }
  94
  95                 // Now pull in all canonical and alias namespaces...
  96                 foreach ( MWNamespace::getCanonicalNamespaces() as $ns => $name ) {
  97                         // This includes $wgExtraNamespaces
  98                         if ( $name !== '' ) {
  99                                 $spaces[$name] = $ns;
 100                         }
 101                 }
 102                 foreach ( $wgContLang->getNamespaces() as $ns => $name ) {
 103                         if ( $name !== '' ) {
 104                                 $spaces[$name] = $ns;
 105                         }
 106                 }
 107                 foreach ( $wgNamespaceAliases as $name => $ns ) {
 108                         $spaces[$name] = $ns;
 109                 }
 110                 foreach ( $wgContLang->getNamespaceAliases() as $name => $ns ) {
 111                         $spaces[$name] = $ns;
 112                 }
 113
 114                 // We'll need to check for lowercase keys as well,
 115                 // since we're doing case-sensitive searches in the db.
 116                 foreach ( $spaces as $name => $ns ) {
 117                         $moreNames = array();
 118                         $moreNames[] = $wgContLang->uc( $name );
 119                         $moreNames[] = $wgContLang->ucfirst( $wgContLang->lc( $name ) );
 120                         $moreNames[] = $wgContLang->ucwords( $name );
 121                         $moreNames[] = $wgContLang->ucwords( $wgContLang->lc( $name ) );
 122                         $moreNames[] = $wgContLang->ucwordbreaks( $name );
 123                         $moreNames[] = $wgContLang->ucwordbreaks( $wgContLang->lc( $name ) );
 124                         if ( !$wgCapitalLinks ) {
 125                                 foreach ( $moreNames as $altName ) {
 126                                         $moreNames[] = $wgContLang->lcfirst( $altName );
 127                                 }
 128                                 $moreNames[] = $wgContLang->lcfirst( $name );
 129                         }
 130                         foreach ( array_unique( $moreNames ) as $altName ) {
 131                                 if ( $altName !== $name ) {
 132                                         $spaces[$altName] = $ns;
 133                                 }
 134                         }
 135                 }
 136
 137                 ksort( $spaces );
 138                 asort( $spaces );
 139
 140                 $ok = true;
 141                 foreach ( $spaces as $name => $ns ) {
 142                         $ok = $this->checkNamespace( $ns, $name, $fix, $suffix ) && $ok;
 143                 }
 144                 return $ok;
 145         }
 146
 147         /**
 148          * Get the interwiki list
 149          *
 150          * @todo Needs to respect interwiki cache!
 151          * @return Array
 152          */
 153         private function getInterwikiList() {
 154                 $result = $this->db->select( 'interwiki', array( 'iw_prefix' ) );
 155                 $prefixes = array();
 156                 foreach ( $result as $row ) {
 157                         $prefixes[] = $row->iw_prefix;
 158                 }
 159                 return $prefixes;
 160         }
 161
 162         /**
 163          * @todo Document
 164          * @param $ns Integer: a namespace id
 165          * @param $name String
 166          * @param $fix Boolean: whether to fix broken entries
 167          * @param $suffix String: suffix to append to renamed articles
 168          * @return bool
 169          */
 170         private function checkNamespace( $ns, $name, $fix, $suffix = '' ) {
 171                 $conflicts = $this->getConflicts( $ns, $name );
 172                 $count = count( $conflicts );
 173                 if ( $count == 0 ) {
 174                         return true;
 175                 }
 176
 177                 $ok = true;
 178                 foreach ( $conflicts as $row ) {
 179                         $resolvable = $this->reportConflict( $row, $suffix );
 180                         $ok = $ok && $resolvable;
 181                         if ( $fix && ( $resolvable || $suffix != '' ) ) {
 182                                 $ok = $this->resolveConflict( $row, $resolvable, $suffix ) && $ok;
 183                         }
 184                 }
 185                 return $ok;
 186         }
 187
 188         /**
 189          * @todo: do this for reals
 190          * @param $key
 191          * @param $prefix
 192          * @param $fix
 193          * @param $suffix string
 194          * @return bool
 195          */
 196         private function checkPrefix( $key, $prefix, $fix, $suffix = '' ) {
 197                 $this->output( "Checking prefix \"$prefix\" vs namespace $key\n" );
 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 $ns Integer: namespace id (id for new namespace?)
 206          * @param $name String: 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                 return $set;
 239         }
 240
 241         /**
 242          * Report any conflicts we find
 243          *
 244          * @return bool
 245          */
 246         private function reportConflict( $row, $suffix ) {
 247                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 248                 if ( is_null( $newTitle ) || !$newTitle->canExist() ) {
 249                         // Title is also an illegal title...
 250                         // For the moment we'll let these slide to cleanupTitles or whoever.
 251                         $this->output( sprintf( "... %d (%d,\"%s\")\n",
 252                                 $row->id,
 253                                 $row->oldnamespace,
 254                                 $row->oldtitle ) );
 255                         $this->output( "...  *** cannot resolve automatically; illegal title ***\n" );
 256                         return false;
 257                 }
 258
 259                 $this->output( sprintf( "... %d (%d,\"%s\") -> (%d,\"%s\") [[%s]]\n",
 260                         $row->id,
 261                         $row->oldnamespace,
 262                         $row->oldtitle,
 263                         $newTitle->getNamespace(),
 264                         $newTitle->getDBkey(),
 265                         $newTitle->getPrefixedText() ) );
 266
 267                 $id = $newTitle->getArticleID();
 268                 if ( $id ) {
 269                         $this->output( "...  *** cannot resolve automatically; page exists with ID $id ***\n" );
 270                         return false;
 271                 } else {
 272                         return true;
 273                 }
 274         }
 275
 276         /**
 277          * Resolve any conflicts
 278          *
 279          * @param $row Object: row from the page table to fix
 280          * @param $resolvable Boolean
 281          * @param $suffix String: suffix to append to the fixed page
 282          * @return bool
 283          */
 284         private function resolveConflict( $row, $resolvable, $suffix ) {
 285                 if ( !$resolvable ) {
 286                         $this->output( "...  *** old title {$row->title}\n" );
 287                         while ( true ) {
 288                                 $row->title .= $suffix;
 289                                 $this->output( "...  *** new title {$row->title}\n" );
 290                                 $title = Title::makeTitleSafe( $row->namespace, $row->title );
 291                                 if ( !$title ) {
 292                                         $this->output( "... !!! invalid title\n" );
 293                                         return false;
 294                                 }
 295                                 $id = $title->getArticleID();
 296                                 if ( $id ) {
 297                                         $this->output( "...  *** page exists with ID $id ***\n" );
 298                                 } else {
 299                                         break;
 300                                 }
 301                         }
 302                         $this->output( "...  *** using suffixed form [[" . $title->getPrefixedText() . "]] ***\n" );
 303                 }
 304                 $this->resolveConflictOn( $row, 'page', 'page' );
 305                 return true;
 306         }
 307
 308         /**
 309          * Resolve a given conflict
 310          *
 311          * @param $row Object: row from the old broken entry
 312          * @param $table String: table to update
 313          * @param $prefix String: prefix for column name, like page or ar
 314          * @return bool
 315          */
 316         private function resolveConflictOn( $row, $table, $prefix ) {
 317                 $this->output( "... resolving on $table... " );
 318                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 319                 $this->db->update( $table,
 320                         array(
 321                                 "{$prefix}_namespace" => $newTitle->getNamespace(),
 322                                 "{$prefix}_title"     => $newTitle->getDBkey(),
 323                         ),
 324                         array(
 325                                 // "{$prefix}_namespace" => 0,
 326                                 // "{$prefix}_title"     => $row->oldtitle,
 327                                 "{$prefix}_id"           => $row->id,
 328                         ),
 329                         __METHOD__ );
 330                 $this->output( "ok.\n" );
 331                 return true;
 332         }
 333 }
 334
 335 $maintClass = "NamespaceConflictChecker";
 336 require_once( RUN_MAINTENANCE_IF_MAIN );