maintenance/cleanupTable.inc

   1 <?php
   2 /**
   3  * Generic class to cleanup a database table.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Maintenance
  22  */
  23
  24 require_once __DIR__ . '/Maintenance.php';
  25
  26 /**
  27  * Generic class to cleanup a database table. Already subclasses Maintenance.
  28  *
  29  * @ingroup Maintenance
  30  */
  31 class TableCleanup extends Maintenance {
  32         protected $defaultParams = array(
  33                 'table' => 'page',
  34                 'conds' => array(),
  35                 'index' => 'page_id',
  36                 'callback' => 'processRow',
  37         );
  38
  39         protected $dryrun = false;
  40         protected $maxLag = 10; # if slaves are lagged more than 10 secs, wait
  41         public $batchSize = 100;
  42         public $reportInterval = 100;
  43
  44         public function __construct() {
  45                 parent::__construct();
  46                 $this->addOption( 'dry-run', 'Perform a dry run' );
  47         }
  48
  49         public function execute() {
  50                 global $wgUser;
  51                 $wgUser = User::newFromName( 'Conversion script' );
  52                 $this->dryrun = $this->hasOption( 'dry-run' );
  53                 if ( $this->dryrun ) {
  54                         $this->output( "Checking for bad titles...\n" );
  55                 } else {
  56                         $this->output( "Checking and fixing bad titles...\n" );
  57                 }
  58                 $this->runTable( $this->defaultParams );
  59         }
  60
  61         protected function init( $count, $table ) {
  62                 $this->processed = 0;
  63                 $this->updated = 0;
  64                 $this->count = $count;
  65                 $this->startTime = microtime( true );
  66                 $this->table = $table;
  67         }
  68
  69         protected function progress( $updated ) {
  70                 $this->updated += $updated;
  71                 $this->processed++;
  72                 if ( $this->processed % $this->reportInterval != 0 ) {
  73                         return;
  74                 }
  75                 $portion = $this->processed / $this->count;
  76                 $updateRate = $this->updated / $this->processed;
  77
  78                 $now = microtime( true );
  79                 $delta = $now - $this->startTime;
  80                 $estimatedTotalTime = $delta / $portion;
  81                 $eta = $this->startTime + $estimatedTotalTime;
  82
  83                 $this->output(
  84                         sprintf( "%s %s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
  85                                 wfWikiID(),
  86                                 wfTimestamp( TS_DB, intval( $now ) ),
  87                                 $portion * 100.0,
  88                                 $this->table,
  89                                 wfTimestamp( TS_DB, intval( $eta ) ),
  90                                 $this->processed,
  91                                 $this->count,
  92                                 $this->processed / $delta,
  93                                 $updateRate * 100.0
  94                         )
  95                 );
  96                 flush();
  97         }
  98
  99         public function runTable( $params ) {
 100                 $dbr = wfGetDB( DB_SLAVE );
 101
 102                 if ( array_diff( array_keys( $params ),
 103                         array( 'table', 'conds', 'index', 'callback' ) ) )
 104                 {
 105                         throw new MWException( __METHOD__ . ': Missing parameter ' . implode( ', ', $params ) );
 106                 }
 107
 108                 $table = $params['table'];
 109                 // count(*) would melt the DB for huge tables, we can estimate here
 110                 $count = $dbr->estimateRowCount( $table, '*', '', __METHOD__ );
 111                 $this->init( $count, $table );
 112                 $this->output( "Processing $table...\n" );
 113
 114
 115                 $index = (array)$params['index'];
 116                 $indexConds = array();
 117                 $options = array(
 118                         'ORDER BY' => implode( ',', $index ),
 119                         'LIMIT' => $this->batchSize
 120                 );
 121                 $callback = array( $this, $params['callback'] );
 122
 123                 while ( true ) {
 124                         $conds = array_merge( $params['conds'], $indexConds );
 125                         $res = $dbr->select( $table, '*', $conds, __METHOD__, $options );
 126                         if ( !$res->numRows() ) {
 127                                 // Done
 128                                 break;
 129                         }
 130
 131                         foreach ( $res as $row ) {
 132                                 call_user_func( $callback, $row );
 133                         }
 134
 135                         if ( $res->numRows() < $this->batchSize ) {
 136                                 // Done
 137                                 break;
 138                         }
 139
 140                         // Update the conditions to select the next batch.
 141                         // Construct a condition string by starting with the least significant part
 142                         // of the index, and adding more significant parts progressively to the left
 143                         // of the string.
 144                         $nextCond = '';
 145                         foreach ( array_reverse( $index ) as $field ) {
 146                                 $encValue = $dbr->addQuotes( $row->$field );
 147                                 if ( $nextCond === '' ) {
 148                                         $nextCond = "$field > $encValue";
 149                                 } else {
 150                                         $nextCond = "$field > $encValue OR ($field = $encValue AND ($nextCond))";
 151                                 }
 152                         }
 153                         $indexConds = array( $nextCond );
 154                 }
 155
 156                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 157         }
 158
 159         protected function hexChar( $matches ) {
 160                 return sprintf( "\\x%02x", ord( $matches[1] ) );
 161         }
 162 }
 163
 164 class TableCleanupTest extends TableCleanup {
 165         function processRow( $row ) {
 166                 $this->progress( mt_rand( 0, 1 ) );
 167         }
 168 }
 169