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