maintenance/cleanupTable.inc

   1 <?php
   2 /**
   3  * Generic table cleanup class. Already subclasses maintenance
   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  * @ingroup Maintenance
  21  */
  22
  23 require_once( dirname(__FILE__) . '/Maintenance.php' );
  24
  25 class TableCleanup extends Maintenance {
  26         protected $defaultParams = array(
  27                 'table' => 'page',
  28                 'conds' => array(),
  29                 'index' => 'page_id',
  30                 'callback' => 'processRow',
  31         );
  32
  33         protected $dryrun = false;
  34         protected $maxLag = 10; # if slaves are lagged more than 10 secs, wait
  35         public $batchSize = 100;
  36         public $reportInterval = 100;
  37
  38         public function __construct() {
  39                 parent::__construct();
  40                 $this->addOption( 'dry-run', 'Perform a dry run' );
  41         }
  42
  43         public function execute() {
  44                 global $wgUser;
  45                 $wgUser->setName( 'Conversion script' );
  46                 $this->dryrun = $this->hasOption( 'dry-run' );
  47                 if( $this->dryrun ) {
  48                         $this->output( "Checking for bad titles...\n" );
  49                 } else {
  50                         $this->output( "Checking and fixing bad titles...\n" );
  51                 }
  52                 $this->runTable( $this->defaultParams );
  53         }
  54
  55         protected function init( $count, $table ) {
  56                 $this->processed = 0;
  57                 $this->updated = 0;
  58                 $this->count = $count;
  59                 $this->startTime = wfTime();
  60                 $this->table = $table;
  61         }
  62
  63         protected function progress( $updated ) {
  64                 $this->updated += $updated;
  65                 $this->processed++;
  66                 if( $this->processed % $this->reportInterval != 0 ) {
  67                         return;
  68                 }
  69                 $portion = $this->processed / $this->count;
  70                 $updateRate = $this->updated / $this->processed;
  71
  72                 $now = wfTime();
  73                 $delta = $now - $this->startTime;
  74                 $estimatedTotalTime = $delta / $portion;
  75                 $eta = $this->startTime + $estimatedTotalTime;
  76
  77                 $this->output(
  78                         sprintf( "%s %s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n",
  79                                 wfWikiID(),
  80                                 wfTimestamp( TS_DB, intval( $now ) ),
  81                                 $portion * 100.0,
  82                                 $this->table,
  83                                 wfTimestamp( TS_DB, intval( $eta ) ),
  84                                 $this->processed,
  85                                 $this->count,
  86                                 $this->processed / $delta,
  87                                 $updateRate * 100.0
  88                         )
  89                 );
  90                 flush();
  91         }
  92
  93         public function runTable( $params ) {
  94                 $dbr = wfGetDB( DB_SLAVE );
  95
  96                 if ( array_diff( array_keys( $params ),
  97                         array( 'table', 'conds', 'index', 'callback' ) ) )
  98                 {
  99                         throw new MWException( __METHOD__.': Missing parameter ' . implode( ', ', $params ) );
 100                 }
 101
 102                 $table = $params['table'];
 103                 $count = $dbr->selectField( $table, 'count(*)', $params['conds'], __METHOD__ );
 104                 $this->init( $count, $table );
 105                 $this->output( "Processing $table...\n" );
 106
 107
 108                 $index = (array)$params['index'];
 109                 $indexConds = array();
 110                 $options = array(
 111                         'ORDER BY' => implode( ',', $index ),
 112                         'LIMIT' => $this->batchSize
 113                 );
 114                 $callback = array( $this, $params['callback'] );
 115
 116                 while ( true ) {
 117                         $conds = array_merge( $params['conds'], $indexConds );
 118                         $res = $dbr->select( $table, '*', $conds, __METHOD__, $options );
 119                         if ( !$res->numRows() ) {
 120                                 // Done
 121                                 break;
 122                         }
 123
 124                         foreach ( $res as $row ) {
 125                                 call_user_func( $callback, $row );
 126                         }
 127
 128                         if ( $res->numRows() < $this->batchSize ) {
 129                                 // Done
 130                                 break;
 131                         }
 132
 133                         // Update the conditions to select the next batch.
 134                         // Construct a condition string by starting with the least significant part
 135                         // of the index, and adding more significant parts progressively to the left
 136                         // of the string.
 137                         $nextCond = '';
 138                         foreach ( array_reverse( $index ) as $field ) {
 139                                 $encValue = $dbr->addQuotes( $row->$field );
 140                                 if ( $nextCond === '' ) {
 141                                         $nextCond = "$field > $encValue";
 142                                 } else {
 143                                         $nextCond = "$field > $encValue OR ($field = $encValue AND ($nextCond))";
 144                                 }
 145                         }
 146                         $indexConds = array( $nextCond );
 147                 }
 148
 149                 $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" );
 150         }
 151
 152         protected function hexChar( $matches ) {
 153                 return sprintf( "\\x%02x", ord( $matches[1] ) );
 154         }
 155 }
 156
 157 class TableCleanupTest extends TableCleanup {
 158         function processRow( $row ) {
 159                 $this->progress( mt_rand( 0, 1 ) );
 160         }
 161 }
 162