maintenance/populateContentModel.php

   1 <?php
   2 /**
   3  * Populate the page_content_model and {rev,ar}_content_{model,format} fields.
   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  * Usage:
  28  *  populateContentModel.php --ns=1 --table=page
  29  */
  30 class PopulateContentModel extends Maintenance {
  31         public function __construct() {
  32                 parent::__construct();
  33                 $this->mDescription = 'Populate the various content_* fields';
  34                 $this->addOption( 'ns', 'Namespace to run in, or "all" for all namespaces', true, true );
  35                 $this->addOption( 'table', 'Table to run in', true, true );
  36                 $this->setBatchSize( 100 );
  37         }
  38
  39         public function execute() {
  40                 $dbw = $this->getDB( DB_MASTER );
  41                 $ns = $this->getOption( 'ns' );
  42                 if ( !ctype_digit( $ns ) && $ns !== 'all' ) {
  43                         $this->error( 'Invalid namespace', 1 );
  44                 }
  45                 $ns = $ns === 'all' ? 'all' : (int)$ns;
  46                 $table = $this->getOption( 'table' );
  47                 switch ( $table ) {
  48                         case 'revision':
  49                         case 'archive':
  50                                 $this->populateRevisionOrArchive( $dbw, $table, $ns );
  51                                 break;
  52                         case 'page':
  53                                 $this->populatePage( $dbw, $ns );
  54                                 break;
  55                         default:
  56                                 $this->error( "Invalid table name: $table", 1 );
  57                 }
  58         }
  59
  60         private function updatePageRows( DatabaseBase $dbw, $pageIds, $model ) {
  61                 $count = count( $pageIds );
  62                 $this->output( "Setting $count rows to $model..." );
  63                 $dbw->update(
  64                         'page',
  65                         array( 'page_content_model' => $model ),
  66                         array( 'page_id' => $pageIds ),
  67                         __METHOD__
  68                 );
  69                 wfWaitForSlaves();
  70                 $this->output( "done.\n" );
  71         }
  72
  73         protected function populatePage( DatabaseBase $dbw, $ns ) {
  74                 $toSave = array();
  75                 $lastId = 0;
  76                 $nsCondition = $ns === 'all' ? array() : array( 'page_namespace' => $ns );
  77                 do {
  78                         $rows = $dbw->select(
  79                                 'page',
  80                                 array( 'page_namespace', 'page_title', 'page_id' ),
  81                                 array(
  82                                         'page_content_model' => null,
  83                                         'page_id > ' . $dbw->addQuotes( $lastId ),
  84                                 ) + $nsCondition,
  85                                 __METHOD__,
  86                                 array( 'LIMIT' => $this->mBatchSize, 'ORDER BY' => 'page_id ASC' )
  87                         );
  88                         $this->output( "Fetched {$rows->numRows()} rows.\n" );
  89                         foreach ( $rows as $row ) {
  90                                 $title = Title::newFromRow( $row );
  91                                 $model = ContentHandler::getDefaultModelFor( $title );
  92                                 $toSave[$model][] = $row->page_id;
  93                                 if ( count( $toSave[$model] ) >= $this->mBatchSize ) {
  94                                         $this->updatePageRows( $dbw, $toSave[$model], $model );
  95                                         unset( $toSave[$model] );
  96                                 }
  97                                 $lastId = $row->page_id;
  98                         }
  99                 } while ( $rows->numRows() >= $this->mBatchSize );
 100                 foreach ( $toSave as $model => $pages ) {
 101                         $this->updatePageRows( $dbw, $pages, $model );
 102                 }
 103         }
 104
 105         private function updateRevisionOrArchiveRows( DatabaseBase $dbw, $ids, $model, $table ) {
 106                 $prefix = $table === 'archive' ? 'ar' : 'rev';
 107                 $model_column = "{$prefix}_content_model";
 108                 $format_column = "{$prefix}_content_format";
 109                 $key = "{$prefix}_id";
 110
 111                 $count = count( $ids );
 112                 $format = ContentHandler::getForModelID( $model )->getDefaultFormat();
 113                 $this->output( "Setting $count rows to $model / $format..." );
 114                 $dbw->update(
 115                         $table,
 116                         array( $model_column => $model, $format_column => $format ),
 117                         array( $key => $ids ),
 118                         __METHOD__
 119                 );
 120                 $this->output( "done.\n" );
 121         }
 122
 123         protected function populateRevisionOrArchive( DatabaseBase $dbw, $table, $ns ) {
 124                 $prefix = $table === 'archive' ? 'ar' : 'rev';
 125                 $model_column = "{$prefix}_content_model";
 126                 $format_column = "{$prefix}_content_format";
 127                 $key = "{$prefix}_id";
 128                 if ( $table === 'archive' ) {
 129                         $selectTables = 'archive';
 130                         $fields = array( 'ar_namespace', 'ar_title' );
 131                         $join_conds = array();
 132                         $where = $ns === 'all' ? array() : array( 'ar_namespace' => $ns );
 133                 } else { // revision
 134                         $selectTables = array( 'revision', 'page' );
 135                         $fields = array( 'page_title', 'page_namespace' );
 136                         $join_conds = array( 'page' => array( 'INNER JOIN', 'rev_page=page_id' ) );
 137                         $where = $ns === 'all' ? array() : array( 'page_namespace' => $ns );
 138                 }
 139
 140                 $toSave = array();
 141                 $lastId = 0;
 142                 do {
 143                         $rows = $dbw->select(
 144                                 $selectTables,
 145                                 array_merge( $fields, array( $model_column, $format_column, $key ) ),
 146                                 // @todo support populating format if model is already set
 147                                 array(
 148                                         $model_column => null,
 149                                         "$key > " . $dbw->addQuotes( $lastId ),
 150                                 ) + $where,
 151                                 __METHOD__,
 152                                 array( 'LIMIT' => $this->mBatchSize, 'ORDER BY' => "$key ASC" ),
 153                                 $join_conds
 154                         );
 155                         $this->output( "Fetched {$rows->numRows()} rows.\n" );
 156                         foreach ( $rows as $row ) {
 157                                 if ( $table === 'archive' ) {
 158                                         $title = Title::makeTitle( $row->ar_namespace, $row->ar_title );
 159                                 } else {
 160                                         $title = Title::newFromRow( $row );
 161                                 }
 162                                 $lastId = $row->{$key};
 163                                 try {
 164                                         $handler = ContentHandler::getForTitle( $title );
 165                                 } catch ( MWException $e ) {
 166                                         $this->error( "Invalid content model for $title" );
 167                                         continue;
 168                                 }
 169                                 $defaultModel = $handler->getModelID();
 170                                 $defaultFormat = $handler->getDefaultFormat();
 171                                 $dbModel = $row->{$model_column};
 172                                 $dbFormat = $row->{$format_column};
 173                                 $id = $row->{$key};
 174                                 if ( $dbModel === null && $dbFormat === null ) {
 175                                         // Set the defaults
 176                                         $toSave[$defaultModel][] = $row->{$key};
 177                                 } else { // $dbModel === null, $dbFormat set.
 178                                         if ( $dbFormat === $defaultFormat ) {
 179                                                 $toSave[$defaultModel][] = $row->{$key};
 180                                         } else { // non-default format, just update now
 181                                                 $this->output( "Updating model to match format for $table $id of $title... " );
 182                                                 $dbw->update(
 183                                                         $table,
 184                                                         array( $model_column => $defaultModel ),
 185                                                         array( $key => $id ),
 186                                                         __METHOD__
 187                                                 );
 188                                                 wfWaitForSlaves();
 189                                                 $this->output( "done.\n" );
 190                                                 continue;
 191                                         }
 192                                 }
 193
 194                                 if ( count( $toSave[$defaultModel] ) >= $this->mBatchSize ) {
 195                                         $this->updateRevisionOrArchiveRows( $dbw, $toSave[$defaultModel], $defaultModel, $table );
 196                                         unset( $toSave[$defaultModel] );
 197                                 }
 198                         }
 199                 } while ( $rows->numRows() >= $this->mBatchSize );
 200                 foreach ( $toSave as $model => $ids ) {
 201                         $this->updateRevisionOrArchiveRows( $dbw, $ids, $model, $table );
 202                 }
 203         }
 204 }
 205
 206 $maintClass = 'PopulateContentModel';
 207 require_once RUN_MAINTENANCE_IF_MAIN;