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 /**
  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 RebuildInterwiki extends Maintenance {
  49         public function __construct() {
  50                 parent::__construct();
  51                 $this->mDescription = "Rebuild the interwiki table using the file on meta and the language list.";
  52                 $this->addOption( 'langlist', 'File with one language code per line', false, true );
  53                 $this->addOption( 'dblist', 'File with one db per line', false, true );
  54                 $this->addOption( 'd', 'Output folder', false, true );
  55         }
  56
  57         function execute() {
  58                 # List of language prefixes likely to be found in multi-language sites
  59                 $this->langlist = array_map( "trim", file( $this->getOption( 'langlist', "/home/wikipedia/common/langlist" ) ) );
  60
  61                 # List of all database names
  62                 $this->dblist = array_map( "trim", file( $this->getOption( 'dblist', "/home/wikipedia/common/all.dblist" ) ) );
  63
  64                 # Special-case databases
  65                 //$this->specials = array_flip( array_map( "trim", file( $this->getOption( 'specialdbs', "/home/wikipedia/common/special.dblist" ) ) ) );
  66
  67                 $this->makeInterwikiSQL( $this->getOption( 'd', '/home/wikipedia/conf/interwiki/sql' ) );
  68         }
  69
  70         function makeInterwikiSQL( $destDir ) {
  71                 $this->output( "Making new interwiki SQL files in $destDir\n" );
  72
  73                 # Multi-language sites
  74                 # db suffix => db suffix, iw prefix, hostname
  75                 $sites = array(
  76                         'wiki' => new Site( 'wiki', 'w', 'wikipedia.org' ),
  77                         'wiktionary' => new Site( 'wiktionary', 'wikt', 'wiktionary.org' ),
  78                         'wikiquote' => new Site( 'wikiquote', 'q', 'wikiquote.org' ),
  79                         'wikibooks' => new Site( 'wikibooks', 'b', 'wikibooks.org' ),
  80                         'wikinews' => new Site( 'wikinews', 'n', 'wikinews.org' ),
  81                         'wikisource' => new Site( 'wikisource', 's', 'wikisource.org' ),
  82                         'wikimedia' => new Site( 'wikimedia', 'chapter', 'wikimedia.org' ),
  83                         'wikiversity' => new Site( 'wikiversity', 'v', 'wikiversity.org' ),
  84                 );
  85
  86                 # Special-case hostnames
  87                 $this->specials = array(
  88                         'sourceswiki' => 'sources.wikipedia.org',
  89                         'quotewiki' => 'wikiquote.org',
  90                         'textbookwiki' => 'wikibooks.org',
  91                         'sep11wiki' => 'sep11.wikipedia.org',
  92                         'metawiki' => 'meta.wikimedia.org',
  93                         'commonswiki' => 'commons.wikimedia.org',
  94                         'specieswiki' => 'species.wikimedia.org',
  95                 );
  96
  97                 # Extra interwiki links that can't be in the intermap for some reason
  98                 $extraLinks = array(
  99                         array( 'm', 'http://meta.wikimedia.org/wiki/$1', 1 ),
 100                         array( 'meta', 'http://meta.wikimedia.org/wiki/$1', 1 ),
 101                         array( 'sep11', 'http://sep11.wikipedia.org/wiki/$1', 1 ),
 102                 );
 103
 104                 # Language aliases, usually configured as redirects to the real wiki in apache
 105                 # Interlanguage links are made directly to the real wiki
 106                 # Something horrible happens if you forget to list an alias here, I can't
 107                 #   remember what
 108                 $this->languageAliases = array(
 109                         'zh-cn' => 'zh',
 110                         'zh-tw' => 'zh',
 111                         'dk' => 'da',
 112                         'nb' => 'no',
 113                 );
 114
 115                 # Special case prefix rewrites, for the benefit of Swedish which uses s:t
 116                 # as an abbreviation for saint
 117                 $this->prefixRewrites = array(
 118                         'svwiki' => array( 's' => 'src' ),
 119                 );
 120
 121                 # Construct a list of reserved prefixes
 122                 $reserved = array();
 123                 foreach ( $this->langlist as $lang ) {
 124                         $reserved[$lang] = 1;
 125                 }
 126                 foreach ( $this->languageAliases as $alias => $lang ) {
 127                         $reserved[$alias] = 1;
 128                 }
 129                 foreach ( $sites as $site ) {
 130                         $reserved[$site->lateral] = 1;
 131                 }
 132
 133                 # Extract the intermap from meta
 134                 $intermap = Http::get( 'http://meta.wikimedia.org/w/index.php?title=Interwiki_map&action=raw', 30 );
 135                 $lines = array_map( 'trim', explode( "\n", trim( $intermap ) ) );
 136
 137                 if ( !$lines || count( $lines ) < 2 ) {
 138                         $this->error( "m:Interwiki_map not found", true );
 139                 }
 140
 141                 $iwArray = array();
 142
 143                 foreach ( $lines as $line ) {
 144                         $matches = array();
 145                         if ( preg_match( '/^\|\s*(.*?)\s*\|\|\s*(https?:\/\/.*?)\s*$/', $line, $matches ) ) {
 146                                 $prefix = strtolower( $matches[1] );
 147                                 $url = $matches[2];
 148                                 if ( preg_match( '/(wikipedia|wiktionary|wikisource|wikiquote|wikibooks|wikimedia)\.org/', $url ) ) {
 149                                         $local = 1;
 150                                 } else {
 151                                         $local = 0;
 152                                 }
 153
 154                                 if ( empty( $reserved[$prefix] ) ) {
 155                                         $iwArray[$prefix] = array( "iw_prefix" => $prefix, "iw_url" => $url, "iw_local" => $local );
 156                                 }
 157                         }
 158                 }
 159
 160                 foreach ( $this->dblist as $db ) {
 161                         $sql = "-- Generated by rebuildInterwiki.php";
 162                         if ( isset( $this->specials[$db] ) ) {
 163                                 # Special wiki
 164                                 # Has interwiki links and interlanguage links to wikipedia
 165
 166                                 $host = $this->specials[$db];
 167                                 $sql .= "\n--$host\n\n";
 168                                 $sql .= "USE $db;\n" .
 169                                                 "TRUNCATE TABLE interwiki;\n" .
 170                                                 "INSERT INTO interwiki (iw_prefix, iw_url, iw_local) VALUES \n";
 171                                 $first = true;
 172
 173                                 # Intermap links
 174                                 foreach ( $iwArray as $iwEntry ) {
 175                                         $sql .= $this->makeLink( $iwEntry, $first, $db );
 176                                 }
 177
 178                                 # Links to multilanguage sites
 179                                 foreach ( $sites as $targetSite ) {
 180                                         $sql .= $this->makeLink( array( $targetSite->lateral, $targetSite->getURL( 'en' ), 1 ), $first, $db );
 181                                 }
 182
 183                                 # Interlanguage links to wikipedia
 184                                 $sql .= $this->makeLanguageLinks( $sites['wiki'], $first, $db );
 185
 186                                 # Extra links
 187                                 foreach ( $extraLinks as $link ) {
 188                                         $sql .= $this->makeLink( $link, $first, $db );
 189                                 }
 190
 191                                 $sql .= ";\n";
 192                         } else {
 193                                 # Find out which site this DB belongs to
 194                                 $site = false;
 195                                 foreach ( $sites as $candidateSite ) {
 196                                         $suffix = $candidateSite->suffix;
 197                                         if ( preg_match( "/(.*)$suffix$/", $db, $matches ) ) {
 198                                                 $site = $candidateSite;
 199                                                 break;
 200                                         }
 201                                 }
 202                                 if ( !$site ) {
 203                                         print "Invalid database $db\n";
 204                                         continue;
 205                                 }
 206                                 $lang = $matches[1];
 207                                 $host = "$lang." . $site->url;
 208                                 $sql .= "\n--$host\n\n";
 209
 210                                 $sql .= "USE $db;\n" .
 211                                                 "TRUNCATE TABLE interwiki;\n" .
 212                                                 "INSERT INTO interwiki (iw_prefix,iw_url,iw_local) VALUES\n";
 213                                 $first = true;
 214
 215                                 # Intermap links
 216                                 foreach ( $iwArray as $iwEntry ) {
 217                                         # Suppress links with the same name as the site
 218                                         if ( ( $suffix == 'wiki' && $iwEntry['iw_prefix'] != 'wikipedia' ) ||
 219                                                 ( $suffix != 'wiki' && $suffix != $iwEntry['iw_prefix'] ) )
 220                                         {
 221                                                 $sql .= $this->makeLink( $iwEntry, $first, $db );
 222                                         }
 223                                 }
 224
 225                                 # Lateral links
 226                                 foreach ( $sites as $targetSite ) {
 227                                         # Suppress link to self
 228                                         if ( $targetSite->suffix != $site->suffix ) {
 229                                                 $sql .= $this->makeLink( array( $targetSite->lateral, $targetSite->getURL( $lang ), 1 ), $first, $db );
 230                                         }
 231                                 }
 232
 233                                 # Interlanguage links
 234                                 $sql .= $this->makeLanguageLinks( $site, $first, $db );
 235
 236                                 # w link within wikipedias
 237                                 # Other sites already have it as a lateral link
 238                                 if ( $site->suffix == "wiki" ) {
 239                                         $sql .= $this->makeLink( array( "w", "http://en.wikipedia.org/wiki/$1", 1 ), $first, $db );
 240                                 }
 241
 242                                 # Extra links
 243                                 foreach ( $extraLinks as $link ) {
 244                                                 $sql .= $this->makeLink( $link, $first, $db );
 245                                 }
 246                                 $sql .= ";\n";
 247                         }
 248                         file_put_contents( "$destDir/$db.sql", $sql );
 249                 }
 250         }
 251
 252         # ------------------------------------------------------------------------------------------
 253
 254         # Returns part of an INSERT statement, corresponding to all interlanguage links to a particular site
 255         function makeLanguageLinks( &$site, &$first, $source ) {
 256                 $sql = "";
 257
 258                 # Actual languages with their own databases
 259                 foreach ( $this->langlist as $targetLang ) {
 260                         $sql .= $this->makeLink( array( $targetLang, $site->getURL( $targetLang ), 1 ), $first, $source );
 261                 }
 262
 263                 # Language aliases
 264                 foreach ( $this->languageAliases as $alias => $lang ) {
 265                         $sql .= $this->makeLink( array( $alias, $site->getURL( $lang ), 1 ), $first, $source );
 266                 }
 267                 return $sql;
 268         }
 269
 270         # Make SQL for a single link from an array
 271         function makeLink( $entry, &$first, $source ) {
 272
 273                 if ( isset( $this->prefixRewrites[$source] ) && isset($entry[0]) && isset( $this->prefixRewrites[$source][$entry[0]] ) ) {
 274                         $entry[0] = $this->prefixRewrites[$source][$entry[0]];
 275                 }
 276
 277                 $sql = "";
 278                 # Add comma
 279                 if ( $first ) {
 280                         $first = false;
 281                 } else {
 282                         $sql .= ",\n";
 283                 }
 284                 $dbr = wfGetDB( DB_SLAVE );
 285                 $sql .= "(" . $dbr->makeList( $entry ) . ")";
 286                 return $sql;
 287         }
 288 }
 289
 290 $maintClass = "RebuildInterwiki";
 291 require_once( DO_MAINTENANCE );
 292