maintenance/benchmarks/benchmarkParse.php

   1 <?php
   2 /**
   3  * Benchmark script for parse operations
   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  * @author Tim Starling <tstarling@wikimedia.org>
  22  * @ingroup Benchmark
  23  */
  24
  25 // @codeCoverageIgnoreStart
  26 require_once __DIR__ . '/../Maintenance.php';
  27 // @codeCoverageIgnoreEnd
  28
  29 use MediaWiki\Cache\LinkCache;
  30 use MediaWiki\Linker\LinkTarget;
  31 use MediaWiki\Maintenance\Maintenance;
  32 use MediaWiki\Revision\RevisionRecord;
  33 use MediaWiki\Revision\SlotRecord;
  34 use MediaWiki\Title\Title;
  35 use Wikimedia\Rdbms\SelectQueryBuilder;
  36
  37 /**
  38  * Maintenance script to benchmark how long it takes to parse a given title at an optionally
  39  * specified timestamp
  40  *
  41  * @since 1.23
  42  */
  43 class BenchmarkParse extends Maintenance {
  44         /** @var string MediaWiki concatenated string timestamp (YYYYMMDDHHMMSS) */
  45         private $templateTimestamp = null;
  46
  47         /** @var bool */
  48         private $clearLinkCache = false;
  49
  50         /**
  51          * @var LinkCache
  52          */
  53         private $linkCache;
  54
  55         /** @var array Cache that maps a Title DB key to revision ID for the requested timestamp */
  56         private $idCache = [];
  57
  58         public function __construct() {
  59                 parent::__construct();
  60                 $this->addDescription( 'Benchmark parse operation' );
  61                 $this->addArg( 'title', 'The name of the page to parse' );
  62                 $this->addOption( 'warmup', 'Repeat the parse operation this number of times to warm the cache',
  63                         false, true );
  64                 $this->addOption( 'loops', 'Number of times to repeat parse operation post-warmup',
  65                         false, true );
  66                 $this->addOption( 'page-time',
  67                         'Use the version of the page which was current at the given time',
  68                         false, true );
  69                 $this->addOption( 'tpl-time',
  70                         'Use templates which were current at the given time (except that moves and ' .
  71                         'deletes are not handled properly)',
  72                         false, true );
  73                 $this->addOption( 'reset-linkcache', 'Reset the LinkCache after every parse.',
  74                         false, false );
  75         }
  76
  77         public function execute() {
  78                 if ( $this->hasOption( 'tpl-time' ) ) {
  79                         $this->templateTimestamp = wfTimestamp( TS_MW, strtotime( $this->getOption( 'tpl-time' ) ) );
  80                         $hookContainer = $this->getHookContainer();
  81                         $hookContainer->register( 'BeforeParserFetchTemplateRevisionRecord', [ $this, 'onFetchTemplate' ] );
  82                 }
  83
  84                 $this->clearLinkCache = $this->hasOption( 'reset-linkcache' );
  85                 // Set as a member variable to avoid function calls when we're timing the parse
  86                 $this->linkCache = $this->getServiceContainer()->getLinkCache();
  87
  88                 $title = Title::newFromText( $this->getArg( 0 ) );
  89                 if ( !$title ) {
  90                         $this->fatalError( "Invalid title" );
  91                 }
  92
  93                 $revLookup = $this->getServiceContainer()->getRevisionLookup();
  94                 if ( $this->hasOption( 'page-time' ) ) {
  95                         $pageTimestamp = wfTimestamp( TS_MW, strtotime( $this->getOption( 'page-time' ) ) );
  96                         $id = $this->getRevIdForTime( $title, $pageTimestamp );
  97                         if ( !$id ) {
  98                                 $this->fatalError( "The page did not exist at that time" );
  99                         }
 100
 101                         $revision = $revLookup->getRevisionById( (int)$id );
 102                 } else {
 103                         $revision = $revLookup->getRevisionByTitle( $title );
 104                 }
 105
 106                 if ( !$revision ) {
 107                         $this->fatalError( "Unable to load revision, incorrect title?" );
 108                 }
 109
 110                 $warmup = $this->getOption( 'warmup', 1 );
 111                 for ( $i = 0; $i < $warmup; $i++ ) {
 112                         $this->runParser( $revision );
 113                 }
 114
 115                 $loops = $this->getOption( 'loops', 1 );
 116                 if ( $loops < 1 ) {
 117                         $this->fatalError( 'Invalid number of loops specified' );
 118                 }
 119                 $startUsage = getrusage();
 120                 $startTime = microtime( true );
 121                 for ( $i = 0; $i < $loops; $i++ ) {
 122                         $this->runParser( $revision );
 123                 }
 124                 $endUsage = getrusage();
 125                 $endTime = microtime( true );
 126
 127                 printf( "CPU time = %.3f s, wall clock time = %.3f s\n",
 128                         // CPU time
 129                         ( $endUsage['ru_utime.tv_sec'] + $endUsage['ru_utime.tv_usec'] * 1e-6
 130                         - $startUsage['ru_utime.tv_sec'] - $startUsage['ru_utime.tv_usec'] * 1e-6 ) / $loops,
 131                         // Wall clock time
 132                         ( $endTime - $startTime ) / $loops
 133                 );
 134         }
 135
 136         /**
 137          * Fetch the ID of the revision of a Title that occurred
 138          *
 139          * @param Title $title
 140          * @param string $timestamp
 141          * @return bool|string Revision ID, or false if not found or error
 142          */
 143         private function getRevIdForTime( Title $title, $timestamp ) {
 144                 $dbr = $this->getReplicaDB();
 145
 146                 $id = $dbr->newSelectQueryBuilder()
 147                         ->select( 'rev_id' )
 148                         ->from( 'revision' )
 149                         ->join( 'page', null, 'rev_page=page_id' )
 150                         ->where( [ 'page_namespace' => $title->getNamespace(), 'page_title' => $title->getDBkey() ] )
 151                         ->andWhere( $dbr->expr( 'rev_timestamp', '<=', $timestamp ) )
 152                         ->orderBy( 'rev_timestamp', SelectQueryBuilder::SORT_DESC )
 153                         ->caller( __METHOD__ )->fetchField();
 154
 155                 return $id;
 156         }
 157
 158         /**
 159          * Parse the text from a given RevisionRecord
 160          */
 161         private function runParser( RevisionRecord $revision ) {
 162                 $content = $revision->getContent( SlotRecord::MAIN );
 163                 $contentRenderer = $this->getServiceContainer()->getContentRenderer();
 164                 // @phan-suppress-next-line PhanTypeMismatchArgumentNullable getId does not return null here
 165                 $contentRenderer->getParserOutput( $content, $revision->getPage(), $revision->getId() );
 166                 if ( $this->clearLinkCache ) {
 167                         $this->linkCache->clear();
 168                 }
 169         }
 170
 171         /**
 172          * Hook into the parser's revision ID fetcher. Make sure that the parser only
 173          * uses revisions around the specified timestamp.
 174          *
 175          * @param ?LinkTarget $contextTitle
 176          * @param LinkTarget $titleTarget
 177          * @param bool &$skip
 178          * @param ?RevisionRecord &$revRecord
 179          * @return bool
 180          */
 181         private function onFetchTemplate(
 182                 ?LinkTarget $contextTitle,
 183                 LinkTarget $titleTarget,
 184                 bool &$skip,
 185                 ?RevisionRecord &$revRecord
 186         ): bool {
 187                 $title = Title::newFromLinkTarget( $titleTarget );
 188
 189                 $pdbk = $title->getPrefixedDBkey();
 190                 if ( !isset( $this->idCache[$pdbk] ) ) {
 191                         $proposedId = $this->getRevIdForTime( $title, $this->templateTimestamp );
 192                         $this->idCache[$pdbk] = $proposedId;
 193                 }
 194                 if ( $this->idCache[$pdbk] !== false ) {
 195                         $revLookup = $this->getServiceContainer()->getRevisionLookup();
 196                         $revRecord = $revLookup->getRevisionById( $this->idCache[$pdbk] );
 197                 }
 198
 199                 return true;
 200         }
 201 }
 202
 203 // @codeCoverageIgnoreStart
 204 $maintClass = BenchmarkParse::class;
 205 require_once RUN_MAINTENANCE_IF_MAIN;
 206 // @codeCoverageIgnoreEnd