maintenance/dumpInterwiki.php

   1 <?php
   2 /**
   3  * Build constant slightly compact database of interwiki prefixes
   4  * Wikimedia specific!
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @file
  22  * @todo document
  23  * @ingroup Maintenance
  24  * @ingroup Wikimedia
  25  */
  26
  27 require_once( dirname( __FILE__ ) . '/Site.php' );
  28
  29 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  30
  31 class DumpInterwiki extends Maintenance {
  32
  33         public function __construct() {
  34                 parent::__construct();
  35                 $this->mDescription = "Build constant slightly compact database of interwiki prefixes.";
  36                 $this->addOption( 'langlist', 'File with one language code per line', false, true );
  37                 $this->addOption( 'dblist', 'File with one db per line', false, true );
  38                 $this->addOption( 'specialdbs', "File with one 'special' db per line", false, true );
  39                 $this->addOption( 'o', 'Cdb output file', false, true );
  40         }
  41
  42         function execute() {
  43                 # List of language prefixes likely to be found in multi-language sites
  44                 $this->langlist = array_map( "trim", file( $this->getOption( 'langlist', "/home/wikipedia/common/langlist" ) ) );
  45
  46                 # List of all database names
  47                 $this->dblist = array_map( "trim", file( $this->getOption( 'dblist', "/home/wikipedia/common/all.dblist" ) ) );
  48
  49                 # Special-case databases
  50                 $this->specials = array_flip(   array_map( "trim", file( $this->getOption( 'specialdbs', "/home/wikipedia/common/special.dblist" ) ) ) );
  51
  52                 if ( $this->hasOption( 'o' ) ) {
  53                         $this->dbFile = CdbWriter::open( $this->getOption( 'o' ) ) ;
  54                 } else {
  55                         $this->dbFile = false;
  56                 }
  57
  58                 $this->getRebuildInterwikiDump();
  59         }
  60
  61         function getRebuildInterwikiDump() {
  62                 global $wgContLang;
  63
  64                 # Multi-language sites
  65                 # db suffix => db suffix, iw prefix, hostname
  66                 $sites = array(
  67                         'wiki' => new Site( 'wiki', 'w', 'wikipedia.org' ),
  68                         'wiktionary' => new Site( 'wiktionary', 'wikt', 'wiktionary.org' ),
  69                         'wikiquote' => new Site( 'wikiquote', 'q', 'wikiquote.org' ),
  70                         'wikibooks' => new Site( 'wikibooks', 'b', 'wikibooks.org' ),
  71                         'wikinews' => new Site( 'wikinews', 'n', 'wikinews.org' ),
  72                         'wikisource' => new Site( 'wikisource', 's', 'wikisource.org' ),
  73                         'wikimedia' => new Site( 'wikimedia', 'chapter', 'wikimedia.org' ),
  74                         'wikiversity' => new Site( 'wikiversity', 'v', 'wikiversity.org' ),
  75                 );
  76
  77                 # Extra interwiki links that can't be in the intermap for some reason
  78                 $extraLinks = array(
  79                         array( 'm', 'http://meta.wikimedia.org/wiki/$1', 1 ),
  80                         array( 'meta', 'http://meta.wikimedia.org/wiki/$1', 1 ),
  81                         array( 'sep11', 'http://sep11.wikipedia.org/wiki/$1', 1 ),
  82                 );
  83
  84                 # Language aliases, usually configured as redirects to the real wiki in apache
  85                 # Interlanguage links are made directly to the real wiki
  86                 # Something horrible happens if you forget to list an alias here, I can't
  87                 #   remember what
  88                 $this->languageAliases = array(
  89                         'zh-cn' => 'zh',
  90                         'zh-tw' => 'zh',
  91                         'dk' => 'da',
  92                         'nb' => 'no',
  93                 );
  94
  95                 # Special case prefix rewrites, for the benefit of Swedish which uses s:t
  96                 # as an abbreviation for saint
  97                 $this->prefixRewrites = array(
  98                         'svwiki' => array( 's' => 'src' ),
  99                 );
 100
 101                 # Construct a list of reserved prefixes
 102                 $reserved = array();
 103                 foreach ( $this->langlist as $lang ) {
 104                         $reserved[$lang] = 1;
 105                 }
 106                 foreach ( $this->languageAliases as $alias => $lang ) {
 107                         $reserved[$alias] = 1;
 108                 }
 109                 foreach ( $sites as $site ) {
 110                         $reserved[$site->lateral] = 1;
 111                 }
 112
 113                 # Extract the intermap from meta
 114                 $intermap = Http::get( 'http://meta.wikimedia.org/w/index.php?title=Interwiki_map&action=raw', 30 );
 115                 $lines = array_map( 'trim', explode( "\n", trim( $intermap ) ) );
 116
 117                 if ( !$lines || count( $lines ) < 2 ) {
 118                         $this->error( "m:Interwiki_map not found", true );
 119                 }
 120
 121                 # Global iterwiki map
 122                 foreach ( $lines as $line ) {
 123                         if ( preg_match( '/^\|\s*(.*?)\s*\|\|\s*(.*?)\s*$/', $line, $matches ) ) {
 124                                 $prefix = $wgContLang->lc( $matches[1] );
 125                                 $prefix = str_replace( ' ', '_', $prefix );
 126
 127                                 $url = $matches[2];
 128                                 if ( preg_match( '/(wikipedia|wiktionary|wikisource|wikiquote|wikibooks|wikimedia)\.org/', $url ) ) {
 129                                         $local = 1;
 130                                 } else {
 131                                         $local = 0;
 132                                 }
 133
 134                                 if ( empty( $reserved[$prefix] ) ) {
 135                                         $imap  = array( "iw_prefix" => $prefix, "iw_url" => $url, "iw_local" => $local );
 136                                         $this->makeLink ( $imap, "__global" );
 137                                 }
 138                         }
 139                 }
 140
 141                 # Exclude Wikipedia for Wikipedia
 142                 $this->makeLink ( array ( 'iw_prefix' => 'wikipedia', 'is_url' => null ), "_wiki" );
 143
 144                 # Multilanguage sites
 145                 foreach ( $sites as $site ) {
 146                         $this->makeLanguageLinks ( $site, "_" . $site->suffix );
 147                 }
 148
 149                 foreach ( $this->dblist as $db ) {
 150                         if ( isset( $this->specials[$db] ) ) {
 151                                 # Special wiki
 152                                 # Has interwiki links and interlanguage links to wikipedia
 153
 154                                 $this->makeLink( array( 'iw_prefix' => $db, 'iw_url' => "wiki" ), "__sites" );
 155                                 # Links to multilanguage sites
 156                                 foreach ( $sites as $targetSite ) {
 157                                         $this->makeLink( array( 'iw_prefix' => $targetSite->lateral,
 158                                                 'iw_url' => $targetSite->getURL( 'en' ),
 159                                                 'iw_local' => 1 ), $db );
 160                                 }
 161                         } else {
 162                                 # Find out which site this DB belongs to
 163                                 $site = false;
 164                                 foreach ( $sites as $candidateSite ) {
 165                                         $suffix = $candidateSite->suffix;
 166                                         if ( preg_match( "/(.*)$suffix$/", $db, $matches ) ) {
 167                                                 $site = $candidateSite;
 168                                                 break;
 169                                         }
 170                                 }
 171                                 $this->makeLink( array( 'iw_prefix' => $db, 'iw_url' => $site->suffix ), "__sites" );
 172                                 if ( !$site ) {
 173                                         $this->error( "Invalid database $db\n" );
 174                                         continue;
 175                                 }
 176                                 $lang = $matches[1];
 177
 178                                 # Lateral links
 179                                 foreach ( $sites as $targetSite ) {
 180                                         if ( $targetSite->suffix != $site->suffix ) {
 181                                                 $this->makeLink( array( 'iw_prefix' => $targetSite->lateral,
 182                                                         'iw_url' => $targetSite->getURL( $lang ),
 183                                                         'iw_local' => 1 ), $db );
 184                                         }
 185                                 }
 186
 187                                 if ( $site->suffix == "wiki" ) {
 188                                         $this->makeLink( array( 'iw_prefix' => 'w',
 189                                                 'iw_url' => "http://en.wikipedia.org/wiki/$1",
 190                                                 'iw_local' => 1 ), $db );
 191                                 }
 192
 193                         }
 194                 }
 195                 foreach ( $extraLinks as $link ) {
 196                         $this->makeLink( $link, "__global" );
 197                 }
 198         }
 199
 200         # ------------------------------------------------------------------------------------------
 201
 202         # Executes part of an INSERT statement, corresponding to all interlanguage links to a particular site
 203         function makeLanguageLinks( &$site, $source ) {
 204                 # Actual languages with their own databases
 205                 foreach ( $this->langlist as $targetLang ) {
 206                         $this->makeLink( array( $targetLang, $site->getURL( $targetLang ), 1 ), $source );
 207                 }
 208
 209                 # Language aliases
 210                 foreach ( $this->languageAliases as $alias => $lang ) {
 211                         $this->makeLink( array( $alias, $site->getURL( $lang ), 1 ), $source );
 212                 }
 213         }
 214
 215         function makeLink( $entry, $source ) {
 216                 if ( isset( $this->prefixRewrites[$source] ) && isset( $this->prefixRewrites[$source][$entry[0]] ) )
 217                         $entry[0] = $this->prefixRewrites[$source][$entry[0]];
 218
 219                 if ( !array_key_exists( "iw_prefix", $entry ) ) {
 220                         $entry = array( "iw_prefix" => $entry[0], "iw_url" => $entry[1], "iw_local" => $entry[2] );
 221                 }
 222                 if ( array_key_exists( $source, $this->prefixRewrites ) &&
 223                                 array_key_exists( $entry['iw_prefix'], $this->prefixRewrites[$source] ) ) {
 224                         $entry['iw_prefix'] = $this->prefixRewrites[$source][$entry['iw_prefix']];
 225                 }
 226
 227                 if ( $this->dbFile ) {
 228                         $this->dbFile->set( "{$source}:{$entry['iw_prefix']}", trim( "{$entry['iw_local']} {$entry['iw_url']}" ) );
 229                 } else {
 230                         $this->output( "{$source}:{$entry['iw_prefix']} {$entry['iw_url']} {$entry['iw_local']}\n" );
 231                 }
 232         }
 233 }
 234
 235 $maintClass = "DumpInterwiki";
 236 require_once( RUN_MAINTENANCE_IF_MAIN );
 237