maintenance/compareParsers.php

   1 <?php
   2 /**
   3  * Take page text out of an XML dump file and render basic HTML out to files.
   4  * This is *NOT* suitable for publishing or offline use; it's intended for
   5  * running comparative tests of parsing behavior using real-world data.
   6  *
   7  * Templates etc are pulled from the local wiki database, not from the dump.
   8  *
   9  * Copyright (C) 2011 Platonides - http://www.mediawiki.org/
  10  *
  11  * This program is free software; you can redistribute it and/or modify
  12  * it under the terms of the GNU General Public License as published by
  13  * the Free Software Foundation; either version 2 of the License, or
  14  * (at your option) any later version.
  15  *
  16  * This program is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  19  * GNU General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU General Public License along
  22  * with this program; if not, write to the Free Software Foundation, Inc.,
  23  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  24  * http://www.gnu.org/copyleft/gpl.html
  25  *
  26  * @file
  27  * @ingroup Maintenance
  28  */
  29
  30 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  31
  32 class CompareParsers extends Maintenance {
  33
  34         private $count = 0;
  35         private $outputDirectory, $startTime;
  36
  37         public function __construct() {
  38                 parent::__construct();
  39                 $this->saveFailed = false;
  40                 $this->mDescription = "Run a file or dump with several parsers";
  41                 $this->addOption( 'parser1', 'The first parser to compare.', true, true );
  42                 $this->addOption( 'parser2', 'The second parser to compare.', true, true );
  43                 $this->addOption( 'file', 'File with text to run.', false, true );
  44                 $this->addOption( 'dump', 'XML dump to execute all revisions.', false, true );
  45                 $this->addOption( 'from', 'Article from XML dump to start from.', false, true );
  46                 $this->addOption( 'tidy', 'Run tidy on the articles.', false, false );
  47                 $this->addOption( 'save-failed', 'Folder in which articles which differ will be stored.', false, true );
  48                 $this->addOption( 'show-diff', 'Show a diff of the two renderings.', false, false );
  49                 $this->addOption( 'diff-bin', 'Binary to use for diffing (can also be provided by DIFF env var).', false, false );
  50                 $this->addOption( 'strip-parameters', 'Remove parameters of html tags to increase readability.', false, false );
  51                 $this->addOption( 'show-parsed-output', 'Show the parsed html if both Parsers give the same output.', false, false );
  52         }
  53
  54         public function execute() {
  55                 if (! ( $this->hasOption('file') ^ $this->hasOption('dump') ) ) {
  56                         $this->error("You must provide file or dump", true);
  57                 }
  58
  59                 if ( $this->hasOption('save-failed') ) {
  60                         $this->saveFailed = $this->getOption('save-failed');
  61                 }
  62
  63                 $this->stripParametersEnabled = $this->hasOption( 'strip-parameters' );
  64                 $this->showParsedOutput = $this->hasOption( 'show-parsed-output' );
  65
  66                 $this->showDiff = $this->hasOption( 'show-diff' );
  67                 if ( $this->showDiff ) {
  68                         $bin = $this->getOption( 'diff-bin', getenv( 'DIFF' ) );
  69                         if ( $bin != '' ) {
  70                                 global $wgDiff;
  71                                 $wgDiff = $bin;
  72                         }
  73                 }
  74
  75                 $user = new User();
  76                 $this->options = ParserOptions::newFromUser( $user );
  77
  78                 if ( $this->hasOption( 'tidy' ) ) {
  79                         global $wgUseTidy;
  80                         if ( !$wgUseTidy ) {
  81                                 $this->error( 'Tidy was requested but $wgUseTidy is not set in LocalSettings.php', true );
  82                         }
  83                         $this->options->setTidy( true );
  84                 }
  85
  86                 if ( $this->hasOption('file') ) {
  87                         $revision = new WikiRevision;
  88
  89                         $revision->setText( file_get_contents( $this->getOption('file') ) );
  90                         $revision->setTitle( Title::newFromText( rawurldecode( basename( $this->getOption('file'), '.txt' ) ) ) );
  91                         $this->handleRevision( $revision );
  92                         return;
  93                 }
  94
  95                 $this->startTime = wfTime();
  96
  97                 if ( $this->getOption('dump') == '-' ) {
  98                         $source = new ImportStreamSource( $this->getStdin() );
  99                 } else {
 100                         $this->error("Sorry, I don't support dump filenames yet. Use - and provide it on stdin on the meantime.", true);
 101                 }
 102                 $importer = new WikiImporter( $source );
 103
 104                 $importer->setRevisionCallback(
 105                         array( &$this, 'handleRevision' ) );
 106
 107                 $this->from = $this->getOption( 'from', null );
 108                 $this->count = 0;
 109                 $this->failed = 0;
 110                 $importer->doImport();
 111
 112                 $this->error( "{$this->failed} failed revisions out of {$this->count}" );
 113                 if ($this->count > 0)
 114                         $this->output( " (" . ( $this->failed / $this->count ) . "%)\n" );
 115
 116                 $delta = wfTime() - $this->startTime;
 117                 $this->error( "Compared {$this->count} pages in " . round($delta, 2) . " seconds " );
 118                 if ($delta > 0)
 119                         $this->error( round($this->count / $delta, 2) . " pages/sec" );
 120                 $this->error( "\n" );
 121         }
 122
 123         function stripParameters( $text ) {
 124                 if ( !$this->stripParametersEnabled ) {
 125                         return $text;
 126                 }
 127                 return preg_replace( '/(<a) [^>]+>/', '$1>', $text );
 128         }
 129
 130         /**
 131          * Callback function for each revision, parse with both parsers and compare
 132          * @param $rev Revision
 133          */
 134         public function handleRevision( $rev ) {
 135                 $title = $rev->getTitle();
 136                 if ( !$title ) {
 137                         $this->error( "Got bogus revision with null title!" );
 138                         return;
 139                 }
 140
 141                 $this->count++;
 142                 if ( isset( $this->from ) ) {
 143                         if ( $this->from != $title )
 144                                 return;
 145                         $this->output( "Skipped " . ($this->count - 1) . " pages\n" );
 146
 147                         $this->count = 1;
 148                         $this->from = null;
 149                 }
 150
 151
 152
 153                 $parser1Name = $this->getOption( 'parser1' );
 154                 $parser2Name = $this->getOption( 'parser2' );
 155
 156                 self::checkParserLocally( $parser1Name );
 157                 self::checkParserLocally( $parser2Name );
 158
 159                 $parser1 = new $parser1Name();
 160                 $parser2 = new $parser2Name();
 161
 162                 $output1 = $parser1->parse( $rev->getText(), $title, $this->options );
 163                 $output2 = $parser2->parse( $rev->getText(), $title, $this->options );
 164
 165                 if ( $output1->getText() != $output2->getText() ) {
 166                         $this->failed++;
 167                         $this->error( "Parsing for {$title->getPrefixedText()} differs\n" );
 168
 169                         if ( $this->saveFailed ) {
 170                                 file_put_contents( $this->saveFailed . '/' . rawurlencode( $title->getPrefixedText() ) . ".txt", $rev->getText());
 171                         }
 172                         if ( $this->showDiff ) {
 173                                 $this->output( wfDiff( $this->stripParameters( $output1->getText() ), $this->stripParameters( $output2->getText() ), '' ) );
 174                         }
 175                 } else {
 176                         $this->output( $title->getPrefixedText() . "\tOK\n" );
 177                         if ( $this->showParsedOutput ) {
 178                                 $this->output( $this->stripParameters( $output1->getText() ) );
 179                         }
 180                 }
 181         }
 182
 183         private static function checkParserLocally( $parserName ) {
 184                 /* Look for the parser in a file appropiately named in the current folder */
 185                 if ( !class_exists( $parserName ) && file_exists( "$parserName.php" ) ) {
 186                         global $wgAutoloadClasses;
 187                         $wgAutoloadClasses[ $parserName ] = realpath( '.' ) . "/$parserName.php";
 188                 }
 189         }
 190
 191 }
 192
 193 $maintClass = "CompareParsers";
 194 require_once( DO_MAINTENANCE );