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                 $this->addOption( 'protocolrelative', 'Output wikimedia interwiki urls as protocol relative', false, false );
  41         }
  42
  43         function execute() {
  44                 # List of language prefixes likely to be found in multi-language sites
  45                 $this->langlist = array_map( "trim", file( $this->getOption( 'langlist', "/home/wikipedia/common/langlist" ) ) );
  46
  47                 # List of all database names
  48                 $this->dblist = array_map( "trim", file( $this->getOption( 'dblist', "/home/wikipedia/common/all.dblist" ) ) );
  49
  50                 # Special-case databases
  51                 $this->specials = array_flip(   array_map( "trim", file( $this->getOption( 'specialdbs', "/home/wikipedia/common/special.dblist" ) ) ) );
  52
  53                 if ( $this->hasOption( 'o' ) ) {
  54                         $this->dbFile = CdbWriter::open( $this->getOption( 'o' ) ) ;
  55                 } else {
  56                         $this->dbFile = false;
  57                 }
  58
  59                 if ( $this->hasOption( 'protocolrelative' ) ) {
  60                         $this->urlprotocol = '';
  61                 } else {
  62                         $this->urlprotocol = 'http:';
  63                 }
  64
  65                 $this->getRebuildInterwikiDump();
  66         }
  67
  68         function getRebuildInterwikiDump() {
  69                 global $wgContLang;
  70
  71                 # Multi-language sites
  72                 # db suffix => db suffix, iw prefix, hostname
  73                 $sites = array(
  74                         'wiki' => new Site( 'wiki', 'w', 'wikipedia.org' ),
  75                         'wiktionary' => new Site( 'wiktionary', 'wikt', 'wiktionary.org' ),
  76                         'wikiquote' => new Site( 'wikiquote', 'q', 'wikiquote.org' ),
  77                         'wikibooks' => new Site( 'wikibooks', 'b', 'wikibooks.org' ),
  78                         'wikinews' => new Site( 'wikinews', 'n', 'wikinews.org' ),
  79                         'wikisource' => new Site( 'wikisource', 's', 'wikisource.org' ),
  80                         'wikimedia' => new Site( 'wikimedia', 'chapter', 'wikimedia.org' ),
  81                         'wikiversity' => new Site( 'wikiversity', 'v', 'wikiversity.org' ),
  82                 );
  83
  84                 # Extra interwiki links that can't be in the intermap for some reason
  85                 $extraLinks = array(
  86                         array( 'm', $this->urlprotocol . '//meta.wikimedia.org/wiki/$1', 1 ),
  87                         array( 'meta', $this->urlprotocol . '//meta.wikimedia.org/wiki/$1', 1 ),
  88                         array( 'sep11', $this->urlprotocol . '//sep11.wikipedia.org/wiki/$1', 1 ),
  89                 );
  90
  91                 # Language aliases, usually configured as redirects to the real wiki in apache
  92                 # Interlanguage links are made directly to the real wiki
  93                 # Something horrible happens if you forget to list an alias here, I can't
  94                 #   remember what
  95                 $this->languageAliases = array(
  96                         'zh-cn' => 'zh',
  97                         'zh-tw' => 'zh',
  98                         'dk' => 'da',
  99                         'nb' => 'no',
 100                 );
 101
 102                 # Special case prefix rewrites, for the benefit of Swedish which uses s:t
 103                 # as an abbreviation for saint
 104                 $this->prefixRewrites = array(
 105                         'svwiki' => array( 's' => 'src' ),
 106                 );
 107
 108                 # Construct a list of reserved prefixes
 109                 $reserved = array();
 110                 foreach ( $this->langlist as $lang ) {
 111                         $reserved[$lang] = 1;
 112                 }
 113                 foreach ( $this->languageAliases as $alias => $lang ) {
 114                         $reserved[$alias] = 1;
 115                 }
 116                 foreach ( $sites as $site ) {
 117                         $reserved[$site->lateral] = 1;
 118                 }
 119
 120                 # Extract the intermap from meta
 121                 $intermap = Http::get( 'http://meta.wikimedia.org/w/index.php?title=Interwiki_map&action=raw', 30 );
 122                 $lines = array_map( 'trim', explode( "\n", trim( $intermap ) ) );
 123
 124                 if ( !$lines || count( $lines ) < 2 ) {
 125                         $this->error( "m:Interwiki_map not found", true );
 126                 }
 127
 128                 # Global interwiki map
 129                 foreach ( $lines as $line ) {
 130                         if ( preg_match( '/^\|\s*(.*?)\s*\|\|\s*(.*?)\s*$/', $line, $matches ) ) {
 131                                 $prefix = $wgContLang->lc( $matches[1] );
 132                                 $prefix = str_replace( ' ', '_', $prefix );
 133
 134                                 $url = $matches[2];
 135                                 if ( preg_match( '/(wikipedia|wiktionary|wikisource|wikiquote|wikibooks|wikimedia)\.org/', $url ) ) {
 136                                         if ( $this->hasOption( 'protocolrelative' ) ) {
 137                                                 if ( substr( $url, 0, 5 ) == 'http:' ) {
 138                                                         $url = substr( $url, 5 );
 139                                                 } else if ( substr( $url, 0, 6 ) == 'https:' ) {
 140                                                         $url = substr( $url, 6 );
 141                                                 }
 142                                         }
 143                                         $local = 1;
 144                                 } else {
 145                                         $local = 0;
 146                                 }
 147
 148                                 if ( empty( $reserved[$prefix] ) ) {
 149                                         $imap  = array( "iw_prefix" => $prefix, "iw_url" => $url, "iw_local" => $local );
 150                                         $this->makeLink ( $imap, "__global" );
 151                                 }
 152                         }
 153                 }
 154
 155                 # Exclude Wikipedia for Wikipedia
 156                 $this->makeLink ( array ( 'iw_prefix' => 'wikipedia', 'is_url' => null ), "_wiki" );
 157
 158                 # Multilanguage sites
 159                 foreach ( $sites as $site ) {
 160                         $this->makeLanguageLinks ( $site, "_" . $site->suffix );
 161                 }
 162
 163                 foreach ( $this->dblist as $db ) {
 164                         if ( isset( $this->specials[$db] ) ) {
 165                                 # Special wiki
 166                                 # Has interwiki links and interlanguage links to wikipedia
 167
 168                                 $this->makeLink( array( 'iw_prefix' => $db, 'iw_url' => "wiki" ), "__sites" );
 169                                 # Links to multilanguage sites
 170                                 foreach ( $sites as $targetSite ) {
 171                                         $this->makeLink( array( 'iw_prefix' => $targetSite->lateral,
 172                                                 'iw_url' => $targetSite->getURL( 'en', $this->urlprotocol ),
 173                                                 'iw_local' => 1 ), $db );
 174                                 }
 175                         } else {
 176                                 # Find out which site this DB belongs to
 177                                 $site = false;
 178                                 foreach ( $sites as $candidateSite ) {
 179                                         $suffix = $candidateSite->suffix;
 180                                         if ( preg_match( "/(.*)$suffix$/", $db, $matches ) ) {
 181                                                 $site = $candidateSite;
 182                                                 break;
 183                                         }
 184                                 }
 185                                 $this->makeLink( array( 'iw_prefix' => $db, 'iw_url' => $site->suffix ), "__sites" );
 186                                 if ( !$site ) {
 187                                         $this->error( "Invalid database $db\n" );
 188                                         continue;
 189                                 }
 190                                 $lang = $matches[1];
 191
 192                                 # Lateral links
 193                                 foreach ( $sites as $targetSite ) {
 194                                         if ( $targetSite->suffix != $site->suffix ) {
 195                                                 $this->makeLink( array( 'iw_prefix' => $targetSite->lateral,
 196                                                         'iw_url' => $targetSite->getURL( $lang, $this->urlprotocol ),
 197                                                         'iw_local' => 1 ), $db );
 198                                         }
 199                                 }
 200
 201                                 if ( $site->suffix == "wiki" ) {
 202                                         $this->makeLink( array( 'iw_prefix' => 'w',
 203                                                 'iw_url' => $this->urlprotocol . "//en.wikipedia.org/wiki/$1",
 204                                                 'iw_local' => 1 ), $db );
 205                                 }
 206
 207                         }
 208                 }
 209                 foreach ( $extraLinks as $link ) {
 210                         $this->makeLink( $link, "__global" );
 211                 }
 212
 213                 # List prefixes for each source
 214                 foreach ( $this->prefixLists as $source => $hash ) {
 215                         $list = array_keys( $hash );
 216                         sort( $list );
 217                         if ( $this->dbFile ) {
 218                                 $this->dbFile->set( "__list:{$source}", implode( ' ', $list ) );
 219                         } else {
 220                                 print "__list:{$source} " . implode( ' ', $list ) . "\n";
 221                         }
 222                 }
 223         }
 224
 225         # ------------------------------------------------------------------------------------------
 226
 227         # Executes part of an INSERT statement, corresponding to all interlanguage links to a particular site
 228         function makeLanguageLinks( &$site, $source ) {
 229                 # Actual languages with their own databases
 230                 foreach ( $this->langlist as $targetLang ) {
 231                         $this->makeLink( array( $targetLang, $site->getURL( $targetLang, $this->urlprotocol ), 1 ), $source );
 232                 }
 233
 234                 # Language aliases
 235                 foreach ( $this->languageAliases as $alias => $lang ) {
 236                         $this->makeLink( array( $alias, $site->getURL( $lang, $this->urlprotocol ), 1 ), $source );
 237                 }
 238         }
 239
 240         function makeLink( $entry, $source ) {
 241                 if ( isset( $this->prefixRewrites[$source] ) && isset( $this->prefixRewrites[$source][$entry[0]] ) )
 242                         $entry[0] = $this->prefixRewrites[$source][$entry[0]];
 243
 244                 if ( !array_key_exists( "iw_prefix", $entry ) ) {
 245                         $entry = array( "iw_prefix" => $entry[0], "iw_url" => $entry[1], "iw_local" => $entry[2] );
 246                 }
 247                 if ( array_key_exists( $source, $this->prefixRewrites ) &&
 248                                 array_key_exists( $entry['iw_prefix'], $this->prefixRewrites[$source] ) ) {
 249                         $entry['iw_prefix'] = $this->prefixRewrites[$source][$entry['iw_prefix']];
 250                 }
 251
 252                 if ( $this->dbFile ) {
 253                         $this->dbFile->set( "{$source}:{$entry['iw_prefix']}", trim( "{$entry['iw_local']} {$entry['iw_url']}" ) );
 254                 } else {
 255                         $this->output( "{$source}:{$entry['iw_prefix']} {$entry['iw_url']} {$entry['iw_local']}\n" );
 256                 }
 257
 258                 # Add to list of prefixes
 259                 $this->prefixLists[$source][$entry['iw_prefix']] = 1;
 260         }
 261 }
 262
 263 $maintClass = "DumpInterwiki";
 264 require_once( RUN_MAINTENANCE_IF_MAIN );
 265