maintenance/rebuildFileCache.php

   1 <?php
   2 /**
   3  * Build file cache for content pages
   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 use MediaWiki\Context\RequestContext;
  25 use MediaWiki\Debug\MWDebug;
  26 use MediaWiki\MainConfigNames;
  27 use MediaWiki\Maintenance\Maintenance;
  28 use MediaWiki\Settings\SettingsBuilder;
  29 use MediaWiki\Title\Title;
  30 use Wikimedia\AtEase\AtEase;
  31 use Wikimedia\Rdbms\SelectQueryBuilder;
  32
  33 // @codeCoverageIgnoreStart
  34 require_once __DIR__ . '/Maintenance.php';
  35 // @codeCoverageIgnoreEnd
  36
  37 /**
  38  * Maintenance script that builds the file cache.
  39  *
  40  * @ingroup Maintenance
  41  */
  42 class RebuildFileCache extends Maintenance {
  43         /** @var bool */
  44         private $enabled = true;
  45
  46         public function __construct() {
  47                 parent::__construct();
  48                 $this->addDescription( 'Build the file cache' );
  49                 $this->addOption( 'start', 'Page_id to start from', false, true );
  50                 $this->addOption( 'end', 'Page_id to end on', false, true );
  51                 $this->addOption( 'overwrite', 'Refresh page cache' );
  52                 $this->addOption( 'all', 'Build the file cache for pages in all namespaces, not just content pages' );
  53                 $this->setBatchSize( 100 );
  54         }
  55
  56         public function finalSetup( SettingsBuilder $settingsBuilder ) {
  57                 $this->enabled = $settingsBuilder->getConfig()->get( MainConfigNames::UseFileCache );
  58                 // Script will handle capturing output and saving it itself
  59                 $settingsBuilder->putConfigValue( MainConfigNames::UseFileCache, false );
  60
  61                 // Avoid DB writes (like enotif/counters)
  62                 $this->getServiceContainer()->getReadOnlyMode()
  63                         ->setReason( 'Building cache' );
  64
  65                 // Ensure no debug-specific logic ends up in the cache (must be after Setup.php)
  66                 MWDebug::deinit();
  67
  68                 parent::finalSetup( $settingsBuilder );
  69         }
  70
  71         public function execute() {
  72                 if ( !$this->enabled ) {
  73                         $this->fatalError( "Nothing to do -- \$wgUseFileCache is disabled." );
  74                 }
  75
  76                 $start = $this->getOption( 'start', "0" );
  77                 if ( !ctype_digit( $start ) ) {
  78                         $this->fatalError( "Invalid value for start parameter." );
  79                 }
  80                 $start = intval( $start );
  81
  82                 $end = $this->getOption( 'end', "0" );
  83                 if ( !ctype_digit( $end ) ) {
  84                         $this->fatalError( "Invalid value for end parameter." );
  85                 }
  86                 $end = intval( $end );
  87
  88                 $this->output( "Building page file cache from page_id {$start}!\n" );
  89
  90                 $dbr = $this->getReplicaDB();
  91                 $batchSize = $this->getBatchSize();
  92                 $overwrite = $this->hasOption( 'overwrite' );
  93                 $start = ( $start > 0 )
  94                         ? $start
  95                         : $dbr->newSelectQueryBuilder()
  96                                 ->select( 'MIN(page_id)' )
  97                                 ->from( 'page' )
  98                                 ->caller( __METHOD__ )->fetchField();
  99                 $end = ( $end > 0 )
 100                         ? $end
 101                         : $dbr->newSelectQueryBuilder()
 102                                 ->select( 'MAX(page_id)' )
 103                                 ->from( 'page' )
 104                                 ->caller( __METHOD__ )->fetchField();
 105                 if ( !$start ) {
 106                         $this->fatalError( "Nothing to do." );
 107                 }
 108
 109                 $where = [];
 110                 if ( !$this->getOption( 'all' ) ) {
 111                         // If 'all' isn't passed as an option, just fall back to previous behaviour
 112                         // of using content namespaces
 113                         $where['page_namespace'] =
 114                                 $this->getServiceContainer()->getNamespaceInfo()->getContentNamespaces();
 115                 }
 116
 117                 // Mock request (hack, no real client)
 118                 $_SERVER['HTTP_ACCEPT_ENCODING'] = 'bgzip';
 119
 120                 # Do remaining chunk
 121                 $end += $batchSize - 1;
 122                 $blockStart = $start;
 123                 $blockEnd = $start + $batchSize - 1;
 124
 125                 $dbw = $this->getPrimaryDB();
 126                 // Go through each page and save the output
 127                 while ( $blockEnd <= $end ) {
 128                         // Get the pages
 129                         $res = $dbr->newSelectQueryBuilder()
 130                                 ->select( [ 'page_namespace', 'page_title', 'page_id' ] )
 131                                 ->from( 'page' )
 132                                 ->useIndex( 'PRIMARY' )
 133                                 ->where( $where )
 134                                 ->andWhere( [
 135                                         $dbr->expr( 'page_id', '>=', (int)$blockStart ),
 136                                         $dbr->expr( 'page_id', '<=', (int)$blockEnd ),
 137                                 ] )
 138                                 ->orderBy( 'page_id', SelectQueryBuilder::SORT_ASC )
 139                                 ->caller( __METHOD__ )->fetchResultSet();
 140
 141                         $this->beginTransaction( $dbw, __METHOD__ ); // for any changes
 142                         foreach ( $res as $row ) {
 143                                 $rebuilt = false;
 144
 145                                 $title = Title::makeTitleSafe( $row->page_namespace, $row->page_title );
 146                                 if ( $title === null ) {
 147                                         $this->output( "Page {$row->page_id} has bad title\n" );
 148                                         continue; // broken title?
 149                                 }
 150
 151                                 $context = new RequestContext();
 152                                 $context->setTitle( $title );
 153                                 $article = Article::newFromTitle( $title, $context );
 154                                 $context->setWikiPage( $article->getPage() );
 155
 156                                 // Some extensions like FlaggedRevs while error out if this is unset
 157                                 RequestContext::getMain()->setTitle( $title );
 158
 159                                 // If the article is cacheable, then load it
 160                                 if ( $article->isFileCacheable( HTMLFileCache::MODE_REBUILD ) ) {
 161                                         $viewCache = new HTMLFileCache( $title, 'view' );
 162                                         $historyCache = new HTMLFileCache( $title, 'history' );
 163                                         if ( $viewCache->isCacheGood() && $historyCache->isCacheGood() ) {
 164                                                 if ( $overwrite ) {
 165                                                         $rebuilt = true;
 166                                                 } else {
 167                                                         $this->output( "Page '$title' (id {$row->page_id}) already cached\n" );
 168                                                         continue; // done already!
 169                                                 }
 170                                         }
 171
 172                                         AtEase::suppressWarnings(); // header notices
 173
 174                                         // 1. Cache ?action=view
 175                                         // Be sure to reset the mocked request time (T24852)
 176                                         $_SERVER['REQUEST_TIME_FLOAT'] = microtime( true );
 177                                         ob_start();
 178                                         $article->view();
 179                                         $context->getOutput()->output();
 180                                         $context->getOutput()->clearHTML();
 181                                         $viewHtml = ob_get_clean();
 182                                         $viewCache->saveToFileCache( $viewHtml );
 183
 184                                         // 2. Cache ?action=history
 185                                         // Be sure to reset the mocked request time (T24852)
 186                                         $_SERVER['REQUEST_TIME_FLOAT'] = microtime( true );
 187                                         ob_start();
 188                                         Action::factory( 'history', $article, $context )->show();
 189                                         $context->getOutput()->output();
 190                                         $context->getOutput()->clearHTML();
 191                                         $historyHtml = ob_get_clean();
 192                                         $historyCache->saveToFileCache( $historyHtml );
 193
 194                                         AtEase::restoreWarnings();
 195
 196                                         if ( $rebuilt ) {
 197                                                 $this->output( "Re-cached page '$title' (id {$row->page_id})..." );
 198                                         } else {
 199                                                 $this->output( "Cached page '$title' (id {$row->page_id})..." );
 200                                         }
 201                                         $this->output( "[view: " . strlen( $viewHtml ) . " bytes; " .
 202                                                 "history: " . strlen( $historyHtml ) . " bytes]\n" );
 203                                 } else {
 204                                         $this->output( "Page '$title' (id {$row->page_id}) not cacheable\n" );
 205                                 }
 206                         }
 207                         $this->commitTransaction( $dbw, __METHOD__ ); // commit any changes
 208
 209                         $blockStart += $batchSize;
 210                         $blockEnd += $batchSize;
 211                 }
 212                 $this->output( "Done!\n" );
 213         }
 214 }
 215
 216 // @codeCoverageIgnoreStart
 217 $maintClass = RebuildFileCache::class;
 218 require_once RUN_MAINTENANCE_IF_MAIN;
 219 // @codeCoverageIgnoreEnd