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