maintenance/updateCollation.php

   1 <?php
   2 /**
   3  * @file
   4  * @ingroup Maintenance
   5  * @author Aryeh Gregor (Simetrical)
   6  */
   7
   8 #$optionsWithArgs = array( 'begin', 'max-slave-lag' );
   9
  10 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  11
  12 class UpdateCollation extends Maintenance {
  13         const BATCH_SIZE = 1000;
  14
  15         public function __construct() {
  16                 parent::__construct();
  17
  18                 global $wgCollationVersion;
  19                 $this->mDescription = <<<TEXT
  20 This script will find all rows in the categorylinks table whose collation is
  21 out-of-date (cl_collation < $wgCollationVersion) and repopulate cl_sortkey
  22 using cl_raw_sortkey.  If everything's collation is up-to-date, it will do
  23 nothing.
  24 TEXT;
  25
  26                 #$this->addOption( 'force', 'Run on all rows, even if the collation is supposed to be up-to-date.' );
  27         }
  28
  29         public function execute() {
  30                 global $wgCollationVersion, $wgContLang;
  31
  32                 $dbw = wfGetDB( DB_MASTER );
  33                 $count = $dbw->estimateRowCount(
  34                         'categorylinks',
  35                         array( 'cl_from', 'cl_to', 'cl_raw_sortkey' ),
  36                         'cl_collation < ' . $dbw->addQuotes( $wgCollationVersion ),
  37                         __METHOD__
  38                 );
  39
  40                 $this->output( "Fixing around $count rows (estimate might be wrong).\n" );
  41
  42                 $count = 0;
  43                 do {
  44                         $res = $dbw->select(
  45                                 'categorylinks',
  46                                 array( 'cl_from', 'cl_to', 'cl_raw_sortkey' ),
  47                                 'cl_collation < ' . $dbw->addQuotes( $wgCollationVersion ),
  48                                 __METHOD__,
  49                                 array( 'LIMIT' => self::BATCH_SIZE )
  50                         );
  51
  52                         $dbw->begin();
  53                         foreach ( $res as $row ) {
  54                                 # TODO: Handle the case where cl_raw_sortkey is null.
  55                                 $dbw->update(
  56                                         'categorylinks',
  57                                         array(
  58                                                 'cl_sortkey' => $wgContLang->convertToSortkey( $row->cl_raw_sortkey ),
  59                                                 'cl_collation' => $wgCollationVersion
  60                                         ),
  61                                         array( 'cl_from' => $row->cl_from, 'cl_to' => $row->cl_to ),
  62                                         __METHOD__
  63                                 );
  64                         }
  65                         $dbw->commit();
  66
  67                         $count += self::BATCH_SIZE;
  68                         $this->output( "$count done.\n" );
  69                 } while ( $res->numRows() >= self::BATCH_SIZE );
  70         }
  71 }
  72
  73 $maintClass = "UpdateCollation";
  74 require_once( DO_MAINTENANCE );