maintenance/namespaceDupes.php

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