maintenance/updateSearchIndex.php

   1 <?php
   2 /**
   3  * Script for periodic off-peak updating of the search index
   4  *
   5  * Usage: php updateSearchIndex.php [-s START] [-e END] [-p POSFILE] [-l LOCKTIME] [-q]
   6  * Where START is the starting timestamp
   7  * END is the ending timestamp
   8  * POSFILE is a file to load timestamps from and save them to, searchUpdate.WIKI_ID.pos by default
   9  * LOCKTIME is how long the searchindex and revision tables will be locked for
  10  * -q means quiet
  11  *
  12  * This program is free software; you can redistribute it and/or modify
  13  * it under the terms of the GNU General Public License as published by
  14  * the Free Software Foundation; either version 2 of the License, or
  15  * (at your option) any later version.
  16  *
  17  * This program is distributed in the hope that it will be useful,
  18  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  19  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  20  * GNU General Public License for more details.
  21  *
  22  * You should have received a copy of the GNU General Public License along
  23  * with this program; if not, write to the Free Software Foundation, Inc.,
  24  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  25  * http://www.gnu.org/copyleft/gpl.html
  26  *
  27  * @ingroup Maintenance
  28  */
  29
  30 require_once( dirname(__FILE__) . '/Maintenance.php' );
  31
  32 class UpdateSearchIndex extends Maintenance {
  33
  34         public function __construct() {
  35                 parent::__construct();
  36                 $this->mDescription = "Script for periodic off-peak updating of the search index";
  37                 $this->addOption( 's', 'starting timestamp', false, true );
  38                 $this->addOption( 'e', 'Ending timestamp', false, true );
  39                 $this->addOption( 'p', 'File for saving/loading timestamps, searchUpdate.WIKI_ID.pos by default', false, true );
  40                 $this->addOption( 'l', 'How long the searchindex and revision tables will be locked for', false, true );
  41         }
  42
  43         public function execute() {
  44                 $posFile = $this->getOption( 'p', 'searchUpdate.' . wfWikiId() . '.pos' );
  45                 $end = $this->getOption( 'e', wfTimestampNow() );
  46                 if ( $this->hasOption( 's' ) ) {
  47                         $start = $this->getOption('s');
  48                 } elseif( is_readable( 'searchUpdate.pos' ) ) {
  49                         # B/c to the old position file name which was hardcoded
  50                         # We can safely delete the file when we're done though.
  51                         $start = file_get_contents( 'searchUpdate.pos' );
  52                         unlink( 'searchUpdate.pos' );
  53                 } else {
  54                         $start = @file_get_contents( $posFile );
  55                         if ( !$start ) {
  56                                 $start = wfTimestamp( TS_MW, time() - 86400 );
  57                         }
  58                 }
  59                 $lockTime = $this->getOption( 'l', 20 );
  60
  61                 $this->doUpdateSearchIndex( $start, $end, $lockTime );
  62                 $file = fopen( $posFile, 'w' );
  63                 fwrite( $file, $end );
  64                 fclose( $file );
  65         }
  66
  67         private function doUpdateSearchIndex( $start, $end, $maxLockTime ) {
  68                 global $wgDisableSearchUpdate;
  69
  70                 $wgDisableSearchUpdate = false;
  71
  72                 $dbw = wfGetDB( DB_MASTER );
  73                 $recentchanges = $dbw->tableName( 'recentchanges' );
  74
  75                 $this->output( "Updating searchindex between $start and $end\n" );
  76
  77                 # Select entries from recentchanges which are on top and between the specified times
  78                 $start = $dbw->timestamp( $start );
  79                 $end = $dbw->timestamp( $end );
  80
  81                 $page = $dbw->tableName( 'page' );
  82                 $sql = "SELECT rc_cur_id,rc_type,rc_moved_to_ns,rc_moved_to_title FROM $recentchanges
  83                   JOIN $page ON rc_cur_id=page_id AND rc_this_oldid=page_latest
  84                   WHERE rc_timestamp BETWEEN '$start' AND '$end'
  85                   ";
  86                 $res = $dbw->query( $sql, __METHOD__ );
  87
  88
  89                 # Lock searchindex
  90                 if ( $maxLockTime ) {
  91                         $this->output( "   --- Waiting for lock ---" );
  92                         $this->lockSearchindex( $dbw );
  93                         $lockTime = time();
  94                         $this->output( "\n" );
  95                 }
  96
  97                 # Loop through the results and do a search update
  98                 while ( $row = $dbw->fetchObject( $res ) ) {
  99                         # Allow reads to be processed
 100                         if ( $maxLockTime && time() > $lockTime + $maxLockTime ) {
 101                                 $this->output( "    --- Relocking ---" );
 102                                 $this->relockSearchindex( $dbw );
 103                                 $lockTime = time();
 104                                 $this->output( "\n" );
 105                         }
 106                         if ( $row->rc_type == RC_LOG ) {
 107                                 continue;
 108                         } elseif ( $row->rc_type == RC_MOVE || $row->rc_type == RC_MOVE_OVER_REDIRECT ) {
 109                                 # Rename searchindex entry
 110                                 $titleObj = Title::makeTitle( $row->rc_moved_to_ns, $row->rc_moved_to_title );
 111                                 $title = $titleObj->getPrefixedDBkey();
 112                                 $this->output( "$title..." );
 113                                 $u = new SearchUpdate( $row->rc_cur_id, $title, false );
 114                                 $this->output( "\n" );
 115                         } else {
 116                                 // Get current revision
 117                                 $rev = Revision::loadFromPageId( $dbw, $row->rc_cur_id );
 118                                 if( $rev ) {
 119                                         $titleObj = $rev->getTitle();
 120                                         $title = $titleObj->getPrefixedDBkey();
 121                                         $this->output( $title );
 122                                         # Update searchindex
 123                                         $u = new SearchUpdate( $row->rc_cur_id, $titleObj->getText(), $rev->getText() );
 124                                         $u->doUpdate();
 125                                         $this->output( "\n" );
 126                                 }
 127                         }
 128                 }
 129
 130                 # Unlock searchindex
 131                 if ( $maxLockTime ) {
 132                         $this->output( "    --- Unlocking --" );
 133                         $this->unlockSearchindex( $dbw );
 134                         $this->output( "\n" );
 135                 }
 136                 $this->output( "Done\n" );
 137         }
 138
 139         /**
 140          * Lock the search index
 141          * @param &$db Database object
 142          */
 143         private function lockSearchindex( &$db ) {
 144                 $write = array( 'searchindex' );
 145                 $read = array( 'page', 'revision', 'text', 'interwiki' );
 146                 $items = array();
 147
 148                 foreach( $write as $table ) {
 149                         $items[] = $db->tableName( $table ) . ' LOW_PRIORITY WRITE';
 150                 }
 151                 foreach( $read as $table ) {
 152                         $items[] = $db->tableName( $table ) . ' READ';
 153                 }
 154                 $sql = "LOCK TABLES " . implode( ',', $items );
 155                 $db->query( $sql, 'updateSearchIndex.php ' . __METHOD__ );
 156         }
 157
 158         /**
 159          * Unlock the tables
 160          * @param &$db Database object
 161          */
 162         private function unlockSearchindex( &$db ) {
 163                 $db->query( "UNLOCK TABLES", 'updateSearchIndex.php ' . __METHOD__ );
 164         }
 165
 166         /**
 167          * Unlock and lock again
 168          * Since the lock is low-priority, queued reads will be able to complete
 169          * @param &$db Database object
 170          */
 171         private function relockSearchindex( &$db ) {
 172                 $this->unlockSearchindex( $db );
 173                 $this->lockSearchindex( $db );
 174         }
 175 }
 176
 177 $maintClass = "UpdateSearchIndex";
 178 require_once( DO_MAINTENANCE );