maintenance/rebuildInterwiki.php

   1 <?php
   2 /**
   3  * Rebuild interwiki table using the file on meta and the language list
   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 RebuildInterwiki extends Maintenance {
  32         public function __construct() {
  33                 parent::__construct();
  34                 $this->mDescription = "Rebuild the interwiki table using the file on meta and the language list.";
  35                 $this->addOption( 'langlist', 'File with one language code per line', false, true );
  36                 $this->addOption( 'dblist', 'File with one db per line', false, true );
  37                 $this->addOption( 'd', 'Output folder', false, true );
  38         }
  39
  40         function execute() {
  41                 # List of language prefixes likely to be found in multi-language sites
  42                 $this->langlist = array_map( "trim", file( $this->getOption( 'langlist', "/home/wikipedia/common/langlist" ) ) );
  43
  44                 # List of all database names
  45                 $this->dblist = array_map( "trim", file( $this->getOption( 'dblist', "/home/wikipedia/common/all.dblist" ) ) );
  46
  47                 # Special-case databases
  48                 //$this->specials = array_flip( array_map( "trim", file( $this->getOption( 'specialdbs', "/home/wikipedia/common/special.dblist" ) ) ) );
  49
  50                 $this->makeInterwikiSQL( $this->getOption( 'd', '/home/wikipedia/conf/interwiki/sql' ) );
  51         }
  52
  53         function makeInterwikiSQL( $destDir ) {
  54                 $this->output( "Making new interwiki SQL files in $destDir\n" );
  55
  56                 # Multi-language sites
  57                 # db suffix => db suffix, iw prefix, hostname
  58                 $sites = array(
  59                         'wiki' => new Site( 'wiki', 'w', 'wikipedia.org' ),
  60                         'wiktionary' => new Site( 'wiktionary', 'wikt', 'wiktionary.org' ),
  61                         'wikiquote' => new Site( 'wikiquote', 'q', 'wikiquote.org' ),
  62                         'wikibooks' => new Site( 'wikibooks', 'b', 'wikibooks.org' ),
  63                         'wikinews' => new Site( 'wikinews', 'n', 'wikinews.org' ),
  64                         'wikisource' => new Site( 'wikisource', 's', 'wikisource.org' ),
  65                         'wikimedia' => new Site( 'wikimedia', 'chapter', 'wikimedia.org' ),
  66                         'wikiversity' => new Site( 'wikiversity', 'v', 'wikiversity.org' ),
  67                 );
  68
  69                 # Special-case hostnames
  70                 $this->specials = array(
  71                         'sourceswiki' => 'sources.wikipedia.org',
  72                         'quotewiki' => 'wikiquote.org',
  73                         'textbookwiki' => 'wikibooks.org',
  74                         'sep11wiki' => 'sep11.wikipedia.org',
  75                         'metawiki' => 'meta.wikimedia.org',
  76                         'commonswiki' => 'commons.wikimedia.org',
  77                         'specieswiki' => 'species.wikimedia.org',
  78                 );
  79
  80                 # Extra interwiki links that can't be in the intermap for some reason
  81                 $extraLinks = array(
  82                         array( 'm', 'http://meta.wikimedia.org/wiki/$1', 1 ),
  83                         array( 'meta', 'http://meta.wikimedia.org/wiki/$1', 1 ),
  84                         array( 'sep11', 'http://sep11.wikipedia.org/wiki/$1', 1 ),
  85                 );
  86
  87                 # Language aliases, usually configured as redirects to the real wiki in apache
  88                 # Interlanguage links are made directly to the real wiki
  89                 # Something horrible happens if you forget to list an alias here, I can't
  90                 #   remember what
  91                 $this->languageAliases = array(
  92                         'zh-cn' => 'zh',
  93                         'zh-tw' => 'zh',
  94                         'dk' => 'da',
  95                         'nb' => 'no',
  96                 );
  97
  98                 # Special case prefix rewrites, for the benefit of Swedish which uses s:t
  99                 # as an abbreviation for saint
 100                 $this->prefixRewrites = array(
 101                         'svwiki' => array( 's' => 'src' ),
 102                 );
 103
 104                 # Construct a list of reserved prefixes
 105                 $reserved = array();
 106                 foreach ( $this->langlist as $lang ) {
 107                         $reserved[$lang] = 1;
 108                 }
 109                 foreach ( $this->languageAliases as $alias => $lang ) {
 110                         $reserved[$alias] = 1;
 111                 }
 112                 foreach ( $sites as $site ) {
 113                         $reserved[$site->lateral] = 1;
 114                 }
 115
 116                 # Extract the intermap from meta
 117                 $intermap = Http::get( 'http://meta.wikimedia.org/w/index.php?title=Interwiki_map&action=raw', 30 );
 118                 $lines = array_map( 'trim', explode( "\n", trim( $intermap ) ) );
 119
 120                 if ( !$lines || count( $lines ) < 2 ) {
 121                         $this->error( "m:Interwiki_map not found", true );
 122                 }
 123
 124                 $iwArray = array();
 125
 126                 foreach ( $lines as $line ) {
 127                         $matches = array();
 128                         if ( preg_match( '/^\|\s*(.*?)\s*\|\|\s*(https?:\/\/.*?)\s*$/', $line, $matches ) ) {
 129                                 $prefix = strtolower( $matches[1] );
 130                                 $url = $matches[2];
 131                                 if ( preg_match( '/(wikipedia|wiktionary|wikisource|wikiquote|wikibooks|wikimedia)\.org/', $url ) ) {
 132                                         $local = 1;
 133                                 } else {
 134                                         $local = 0;
 135                                 }
 136
 137                                 if ( empty( $reserved[$prefix] ) ) {
 138                                         $iwArray[$prefix] = array( "iw_prefix" => $prefix, "iw_url" => $url, "iw_local" => $local );
 139                                 }
 140                         }
 141                 }
 142
 143                 foreach ( $this->dblist as $db ) {
 144                         $sql = "-- Generated by rebuildInterwiki.php";
 145                         if ( isset( $this->specials[$db] ) ) {
 146                                 # Special wiki
 147                                 # Has interwiki links and interlanguage links to wikipedia
 148
 149                                 $host = $this->specials[$db];
 150                                 $sql .= "\n--$host\n\n";
 151                                 $sql .= "USE $db;\n" .
 152                                                 "TRUNCATE TABLE interwiki;\n" .
 153                                                 "INSERT INTO interwiki (iw_prefix, iw_url, iw_local) VALUES \n";
 154                                 $first = true;
 155
 156                                 # Intermap links
 157                                 foreach ( $iwArray as $iwEntry ) {
 158                                         $sql .= $this->makeLink( $iwEntry, $first, $db );
 159                                 }
 160
 161                                 # Links to multilanguage sites
 162                                 foreach ( $sites as $targetSite ) {
 163                                         $sql .= $this->makeLink( array( $targetSite->lateral, $targetSite->getURL( 'en' ), 1 ), $first, $db );
 164                                 }
 165
 166                                 # Interlanguage links to wikipedia
 167                                 $sql .= $this->makeLanguageLinks( $sites['wiki'], $first, $db );
 168
 169                                 # Extra links
 170                                 foreach ( $extraLinks as $link ) {
 171                                         $sql .= $this->makeLink( $link, $first, $db );
 172                                 }
 173
 174                                 $sql .= ";\n";
 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 = $candidateSite;
 182                                                 break;
 183                                         }
 184                                 }
 185                                 if ( !$site ) {
 186                                         print "Invalid database $db\n";
 187                                         continue;
 188                                 }
 189                                 $lang = $matches[1];
 190                                 $host = "$lang." . $site->url;
 191                                 $sql .= "\n--$host\n\n";
 192
 193                                 $sql .= "USE $db;\n" .
 194                                                 "TRUNCATE TABLE interwiki;\n" .
 195                                                 "INSERT INTO interwiki (iw_prefix,iw_url,iw_local) VALUES\n";
 196                                 $first = true;
 197
 198                                 # Intermap links
 199                                 foreach ( $iwArray as $iwEntry ) {
 200                                         # Suppress links with the same name as the site
 201                                         if ( ( $suffix == 'wiki' && $iwEntry['iw_prefix'] != 'wikipedia' ) ||
 202                                                 ( $suffix != 'wiki' && $suffix != $iwEntry['iw_prefix'] ) )
 203                                         {
 204                                                 $sql .= $this->makeLink( $iwEntry, $first, $db );
 205                                         }
 206                                 }
 207
 208                                 # Lateral links
 209                                 foreach ( $sites as $targetSite ) {
 210                                         # Suppress link to self
 211                                         if ( $targetSite->suffix != $site->suffix ) {
 212                                                 $sql .= $this->makeLink( array( $targetSite->lateral, $targetSite->getURL( $lang ), 1 ), $first, $db );
 213                                         }
 214                                 }
 215
 216                                 # Interlanguage links
 217                                 $sql .= $this->makeLanguageLinks( $site, $first, $db );
 218
 219                                 # w link within wikipedias
 220                                 # Other sites already have it as a lateral link
 221                                 if ( $site->suffix == "wiki" ) {
 222                                         $sql .= $this->makeLink( array( "w", "http://en.wikipedia.org/wiki/$1", 1 ), $first, $db );
 223                                 }
 224
 225                                 # Extra links
 226                                 foreach ( $extraLinks as $link ) {
 227                                                 $sql .= $this->makeLink( $link, $first, $db );
 228                                 }
 229                                 $sql .= ";\n";
 230                         }
 231                         file_put_contents( "$destDir/$db.sql", $sql );
 232                 }
 233         }
 234
 235         # ------------------------------------------------------------------------------------------
 236
 237         # Returns part of an INSERT statement, corresponding to all interlanguage links to a particular site
 238         function makeLanguageLinks( &$site, &$first, $source ) {
 239                 $sql = "";
 240
 241                 # Actual languages with their own databases
 242                 foreach ( $this->langlist as $targetLang ) {
 243                         $sql .= $this->makeLink( array( $targetLang, $site->getURL( $targetLang ), 1 ), $first, $source );
 244                 }
 245
 246                 # Language aliases
 247                 foreach ( $this->languageAliases as $alias => $lang ) {
 248                         $sql .= $this->makeLink( array( $alias, $site->getURL( $lang ), 1 ), $first, $source );
 249                 }
 250                 return $sql;
 251         }
 252
 253         # Make SQL for a single link from an array
 254         function makeLink( $entry, &$first, $source ) {
 255
 256                 if ( isset( $this->prefixRewrites[$source] ) && isset($entry[0]) && isset( $this->prefixRewrites[$source][$entry[0]] ) ) {
 257                         $entry[0] = $this->prefixRewrites[$source][$entry[0]];
 258                 }
 259
 260                 $sql = "";
 261                 # Add comma
 262                 if ( $first ) {
 263                         $first = false;
 264                 } else {
 265                         $sql .= ",\n";
 266                 }
 267                 $dbr = wfGetDB( DB_SLAVE );
 268                 $sql .= "(" . $dbr->makeList( $entry ) . ")";
 269                 return $sql;
 270         }
 271 }
 272
 273 $maintClass = "RebuildInterwiki";
 274 require_once( RUN_MAINTENANCE_IF_MAIN );
 275