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