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] = $ns;
  76                 }
  77
  78                 if( !$wgCapitalLinks ) {
  79                         // We'll need to check for lowercase keys as well,
  80                         // since we're doing case-sensitive searches in the db.
  81                         foreach( array_values( $spaces ) as $name => $ns ) {
  82                                 $lcname = $wgContLang->lcfirst( $name );
  83                                 $spaces[$lcname] = $ns;
  84                         }
  85                 }
  86                 ksort( $spaces );
  87                 asort( $spaces );
  88
  89                 $ok = true;
  90                 foreach( $spaces as $name => $ns ) {
  91                         $ok = $this->checkNamespace( $ns, $name, $fix, $suffix ) && $ok;
  92                 }
  93                 return $ok;
  94         }
  95
  96         private function getInterwikiList() {
  97                 $result = $this->db->select( 'interwiki', array( 'iw_prefix' ) );
  98                 while( $row = $this->db->fetchObject( $result ) ) {
  99                         $prefixes[] = $row->iw_prefix;
 100                 }
 101                 $this->db->freeResult( $result );
 102                 return $prefixes;
 103         }
 104
 105         function checkNamespace( $ns, $name, $fix, $suffix = '' ) {
 106                 if( $ns == 0 ) {
 107                         $header = "Checking interwiki prefix: \"$name\"\n";
 108                 } else {
 109                         $header = "Checking namespace $ns: \"$name\"\n";
 110                 }
 111
 112                 $conflicts = $this->getConflicts( $ns, $name );
 113                 $count = count( $conflicts );
 114                 if( $count == 0 ) {
 115                         if( $this->verbose ) {
 116                                 echo $header;
 117                                 echo "... no conflicts detected!\n";
 118                         }
 119                         return true;
 120                 }
 121
 122                 echo $header;
 123                 echo "... $count conflicts detected:\n";
 124                 $ok = true;
 125                 foreach( $conflicts as $row ) {
 126                         $resolvable = $this->reportConflict( $row, $suffix );
 127                         $ok = $ok && $resolvable;
 128                         if( $fix && ( $resolvable || $suffix != '' ) ) {
 129                                 $ok = $this->resolveConflict( $row, $resolvable, $suffix ) && $ok;
 130                         }
 131                 }
 132                 return $ok;
 133         }
 134
 135         /**
 136          * @todo: do this for reals
 137          */
 138         function checkPrefix( $key, $prefix, $fix, $suffix = '' ) {
 139                 echo "Checking prefix \"$prefix\" vs namespace $key\n";
 140                 return $this->checkNamespace( $key, $prefix, $fix, $suffix );
 141         }
 142
 143         function getConflicts( $ns, $name ) {
 144                 $page  = 'page';
 145                 $table = $this->db->tableName( $page );
 146
 147                 $prefix     = $this->db->strencode( $name );
 148                 $likeprefix = str_replace( '_', '\\_', $prefix);
 149                 $encNamespace = $this->db->addQuotes( $ns );
 150
 151                 $titleSql = "TRIM(LEADING '$prefix:' FROM {$page}_title)";
 152                 if( $ns == 0 ) {
 153                         // An interwiki; try an alternate encoding with '-' for ':'
 154                         $titleSql = "CONCAT('$prefix-',$titleSql)";
 155                 }
 156
 157                 $sql = "SELECT {$page}_id    AS id,
 158                                {$page}_title AS oldtitle,
 159                                $encNamespace AS namespace,
 160                                $titleSql     AS title
 161                           FROM {$table}
 162                          WHERE {$page}_namespace=0
 163                            AND {$page}_title LIKE '$likeprefix:%'";
 164
 165                 $result = $this->db->query( $sql, 'NamespaceConflictChecker::getConflicts' );
 166
 167                 $set = array();
 168                 while( $row = $this->db->fetchObject( $result ) ) {
 169                         $set[] = $row;
 170                 }
 171                 $this->db->freeResult( $result );
 172
 173                 return $set;
 174         }
 175
 176         function reportConflict( $row, $suffix ) {
 177                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 178                 printf( "... %d (0,\"%s\") -> (%d,\"%s\") [[%s]]\n",
 179                         $row->id,
 180                         $row->oldtitle,
 181                         $newTitle->getNamespace(),
 182                         $newTitle->getDbKey(),
 183                         $newTitle->getPrefixedText() );
 184
 185                 $id = $newTitle->getArticleId();
 186                 if( $id ) {
 187                         echo "...  *** cannot resolve automatically; page exists with ID $id ***\n";
 188                         return false;
 189                 } else {
 190                         return true;
 191                 }
 192         }
 193
 194         function resolveConflict( $row, $resolvable, $suffix ) {
 195                 if( !$resolvable ) {
 196                         $row->title .= $suffix;
 197                         $title = Title::makeTitleSafe( $row->namespace, $row->title );
 198                         echo "...  *** using suffixed form [[" . $title->getPrefixedText() . "]] ***\n";
 199                 }
 200                 $tables = array( 'page' );
 201                 foreach( $tables as $table ) {
 202                         $this->resolveConflictOn( $row, $table );
 203                 }
 204                 return true;
 205         }
 206
 207         function resolveConflictOn( $row, $table ) {
 208                 $fname = 'NamespaceConflictChecker::resolveConflictOn';
 209                 echo "... resolving on $table... ";
 210                 $newTitle = Title::makeTitleSafe( $row->namespace, $row->title );
 211                 $this->db->update( $table,
 212                         array(
 213                                 "{$table}_namespace" => $newTitle->getNamespace(),
 214                                 "{$table}_title"     => $newTitle->getDbKey(),
 215                         ),
 216                         array(
 217                                 "{$table}_namespace" => 0,
 218                                 "{$table}_title"     => $row->oldtitle,
 219                         ),
 220                         $fname );
 221                 echo "ok.\n";
 222                 return true;
 223         }
 224 }
 225
 226
 227
 228
 229 $wgTitle = Title::newFromText( 'Namespace title conflict cleanup script' );
 230
 231 $verbose = isset( $options['verbose'] );
 232 $fix = isset( $options['fix'] );
 233 $suffix = isset( $options['suffix'] ) ? $options['suffix'] : '';
 234 $prefix = isset( $options['prefix'] ) ? $options['prefix'] : '';
 235 $key = isset( $options['key'] ) ? intval( $options['key'] ) : 0;
 236
 237 $dbw = wfGetDB( DB_MASTER );
 238 $duper = new NamespaceConflictChecker( $dbw, $verbose );
 239
 240 if( $prefix ) {
 241         $retval = $duper->checkPrefix( $key, $prefix, $fix, $suffix );
 242 } else {
 243         $retval = $duper->checkAll( $fix, $suffix );
 244 }
 245
 246 if( $retval ) {
 247         echo "\nLooks good!\n";
 248         exit( 0 );
 249 } else {
 250         echo "\nOh noeees\n";
 251         exit( -1 );
 252 }
 253
 254