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 /**
  28  * @todo document
  29  * @ingroup Maintenance
  30  */
  31 class Site {
  32         var $suffix, $lateral, $url;
  33
  34         function __construct( $s, $l, $u ) {
  35                 $this->suffix = $s;
  36                 $this->lateral = $l;
  37                 $this->url = $u;
  38         }
  39
  40         function getURL( $lang ) {
  41                 $xlang = str_replace( '_', '-', $lang );
  42                 return "http://$xlang.{$this->url}/wiki/\$1";
  43         }
  44 }
  45
  46 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  47
  48 class DumpInterwiki extends Maintenance {
  49
  50         public function __construct() {
  51                 parent::__construct();
  52                 $this->mDescription = "Build constant slightly compact database of interwiki prefixes.";
  53                 $this->addOption( 'langlist', 'File with one language code per line', false, true );
  54                 $this->addOption( 'dblist', 'File with one db per line', false, true );
  55                 $this->addOption( 'specialdbs', "File with one 'special' db per line", false, true );
  56                 $this->addOption( 'o', 'Cdb output file', false, true );
  57         }
  58
  59         function execute() {
  60                 # List of language prefixes likely to be found in multi-language sites
  61                 $this->langlist = array_map( "trim", file( $this->getOption( 'langlist', "/home/wikipedia/common/langlist" ) ) );
  62
  63                 # List of all database names
  64                 $this->dblist = array_map( "trim", file( $this->getOption( 'dblist', "/home/wikipedia/common/all.dblist" ) ) );
  65
  66                 # Special-case databases
  67                 $this->specials = array_flip(   array_map( "trim", file( $this->getOption( 'specialdbs', "/home/wikipedia/common/special.dblist" ) ) ) );
  68
  69                 if ( $this->hasOption( 'o' ) ) {
  70                         $this->dbFile = CdbWriter::open( $this->getOption( 'o' ) ) ;
  71                 } else {
  72                         $this->dbFile = false;
  73                 }
  74
  75                 $this->getRebuildInterwikiDump();
  76         }
  77
  78         function getRebuildInterwikiDump() {
  79                 global $wgContLang;
  80
  81                 # Multi-language sites
  82                 # db suffix => db suffix, iw prefix, hostname
  83                 $sites = array(
  84                         'wiki' => new Site( 'wiki', 'w', 'wikipedia.org' ),
  85                         'wiktionary' => new Site( 'wiktionary', 'wikt', 'wiktionary.org' ),
  86                         'wikiquote' => new Site( 'wikiquote', 'q', 'wikiquote.org' ),
  87                         'wikibooks' => new Site( 'wikibooks', 'b', 'wikibooks.org' ),
  88                         'wikinews' => new Site( 'wikinews', 'n', 'wikinews.org' ),
  89                         'wikisource' => new Site( 'wikisource', 's', 'wikisource.org' ),
  90                         'wikimedia' => new Site( 'wikimedia', 'chapter', 'wikimedia.org' ),
  91                         'wikiversity' => new Site( 'wikiversity', 'v', 'wikiversity.org' ),
  92                 );
  93
  94                 # Extra interwiki links that can't be in the intermap for some reason
  95                 $extraLinks = array(
  96                         array( 'm', 'http://meta.wikimedia.org/wiki/$1', 1 ),
  97                         array( 'meta', 'http://meta.wikimedia.org/wiki/$1', 1 ),
  98                         array( 'sep11', 'http://sep11.wikipedia.org/wiki/$1', 1 ),
  99                 );
 100
 101                 # Language aliases, usually configured as redirects to the real wiki in apache
 102                 # Interlanguage links are made directly to the real wiki
 103                 # Something horrible happens if you forget to list an alias here, I can't
 104                 #   remember what
 105                 $this->languageAliases = array(
 106                         'zh-cn' => 'zh',
 107                         'zh-tw' => 'zh',
 108                         'dk' => 'da',
 109                         'nb' => 'no',
 110                 );
 111
 112                 # Special case prefix rewrites, for the benefit of Swedish which uses s:t
 113                 # as an abbreviation for saint
 114                 $this->prefixRewrites = array(
 115                         'svwiki' => array( 's' => 'src' ),
 116                 );
 117
 118                 # Construct a list of reserved prefixes
 119                 $reserved = array();
 120                 foreach ( $this->langlist as $lang ) {
 121                         $reserved[$lang] = 1;
 122                 }
 123                 foreach ( $this->languageAliases as $alias => $lang ) {
 124                         $reserved[$alias] = 1;
 125                 }
 126                 foreach ( $sites as $site ) {
 127                         $reserved[$site->lateral] = 1;
 128                 }
 129
 130                 # Extract the intermap from meta
 131                 $intermap = Http::get( 'http://meta.wikimedia.org/w/index.php?title=Interwiki_map&action=raw', 30 );
 132                 $lines = array_map( 'trim', explode( "\n", trim( $intermap ) ) );
 133
 134                 if ( !$lines || count( $lines ) < 2 ) {
 135                         $this->error( "m:Interwiki_map not found", true );
 136                 }
 137
 138                 # Global iterwiki map
 139                 foreach ( $lines as $line ) {
 140                         if ( preg_match( '/^\|\s*(.*?)\s*\|\|\s*(.*?)\s*$/', $line, $matches ) ) {
 141                                 $prefix = $wgContLang->lc( $matches[1] );
 142                                 $prefix = str_replace( ' ', '_', $prefix );
 143
 144                                 $url = $matches[2];
 145                                 if ( preg_match( '/(wikipedia|wiktionary|wikisource|wikiquote|wikibooks|wikimedia)\.org/', $url ) ) {
 146                                         $local = 1;
 147                                 } else {
 148                                         $local = 0;
 149                                 }
 150
 151                                 if ( empty( $reserved[$prefix] ) ) {
 152                                         $imap  = array( "iw_prefix" => $prefix, "iw_url" => $url, "iw_local" => $local );
 153                                         $this->makeLink ( $imap, "__global" );
 154                                 }
 155                         }
 156                 }
 157
 158                 # Exclude Wikipedia for Wikipedia
 159                 $this->makeLink ( array ( 'iw_prefix' => 'wikipedia', 'is_url' => null ), "_wiki" );
 160
 161                 # Multilanguage sites
 162                 foreach ( $sites as $site ) {
 163                         $this->makeLanguageLinks ( $site, "_" . $site->suffix );
 164                 }
 165
 166                 foreach ( $dblist as $db ) {
 167                         if ( isset( $this->specials[$db] ) ) {
 168                                 # Special wiki
 169                                 # Has interwiki links and interlanguage links to wikipedia
 170
 171                                 $this->makeLink( array( 'iw_prefix' => $db, 'iw_url' => "wiki" ), "__sites" );
 172                                 # Links to multilanguage sites
 173                                 foreach ( $sites as $targetSite ) {
 174                                         $this->makeLink( array( 'iw_prefix' => $targetSite->lateral,
 175                                                 'iw_url' => $targetSite->getURL( 'en' ),
 176                                                 'iw_local' => 1 ), $db );
 177                                 }
 178                         } else {
 179                                 # Find out which site this DB belongs to
 180                                 $site = false;
 181                                 foreach ( $sites as $candidateSite ) {
 182                                         $suffix = $candidateSite->suffix;
 183                                         if ( preg_match( "/(.*)$suffix$/", $db, $matches ) ) {
 184                                                 $site = $candidateSite;
 185                                                 break;
 186                                         }
 187                                 }
 188                                 $this->makeLink( array( 'iw_prefix' => $db, 'iw_url' => $site->suffix ), "__sites" );
 189                                 if ( !$site ) {
 190                                         $this->error( "Invalid database $db\n" );
 191                                         continue;
 192                                 }
 193                                 $lang = $matches[1];
 194
 195                                 # Lateral links
 196                                 foreach ( $sites as $targetSite ) {
 197                                         if ( $targetSite->suffix != $site->suffix ) {
 198                                                 $this->makeLink( array( 'iw_prefix' => $targetSite->lateral,
 199                                                         'iw_url' => $targetSite->getURL( $lang ),
 200                                                         'iw_local' => 1 ), $db );
 201                                         }
 202                                 }
 203
 204                                 if ( $site->suffix == "wiki" ) {
 205                                         $this->makeLink( array( 'iw_prefix' => 'w',
 206                                                 'iw_url' => "http://en.wikipedia.org/wiki/$1",
 207                                                 'iw_local' => 1 ), $db );
 208                                 }
 209
 210                         }
 211                 }
 212                 foreach ( $extraLinks as $link ) {
 213                         $this->makeLink( $link, "__global" );
 214                 }
 215         }
 216
 217         # ------------------------------------------------------------------------------------------
 218
 219         # Executes part of an INSERT statement, corresponding to all interlanguage links to a particular site
 220         function makeLanguageLinks( &$site, $source ) {
 221                 # Actual languages with their own databases
 222                 foreach ( $this->langlist as $targetLang ) {
 223                         $this->makeLink( array( $targetLang, $site->getURL( $targetLang ), 1 ), $source );
 224                 }
 225
 226                 # Language aliases
 227                 foreach ( $this->languageAliases as $alias => $lang ) {
 228                         $this->makeLink( array( $alias, $site->getURL( $lang ), 1 ), $source );
 229                 }
 230         }
 231
 232         function makeLink( $entry, $source ) {
 233                 if ( isset( $this->prefixRewrites[$source] ) && isset( $this->prefixRewrites[$source][$entry[0]] ) )
 234                         $entry[0] = $this->prefixRewrites[$source][$entry[0]];
 235
 236                 if ( !array_key_exists( "iw_prefix", $entry ) ) {
 237                         $entry = array( "iw_prefix" => $entry[0], "iw_url" => $entry[1], "iw_local" => $entry[2] );
 238                 }
 239                 if ( array_key_exists( $source, $this->prefixRewrites ) &&
 240                                 array_key_exists( $entry['iw_prefix'], $this->prefixRewrites[$source] ) ) {
 241                         $entry['iw_prefix'] = $this->prefixRewrites[$source][$entry['iw_prefix']];
 242                 }
 243
 244                 if ( $this->dbFile ) {
 245                         $this->dbFile->set( "{$source}:{$entry['iw_prefix']}", trim( "{$entry['iw_local']} {$entry['iw_url']}" ) );
 246                 } else {
 247                         $this->output( "{$source}:{$entry['iw_prefix']} {$entry['iw_url']} {$entry['iw_local']}\n" );
 248                 }
 249         }
 250 }
 251
 252 $maintClass = "DumpInterwiki";
 253 require_once( RUN_MAINTENANCE_IF_MAIN );
 254