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 © 2011 Platonides
  10  * https://www.mediawiki.org/
  11  *
  12  * This program is free software; you can redistribute it and/or modify
  13  * it under the terms of the GNU General Public License as published by
  14  * the Free Software Foundation; either version 2 of the License, or
  15  * (at your option) any later version.
  16  *
  17  * This program is distributed in the hope that it will be useful,
  18  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  19  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  20  * GNU General Public License for more details.
  21  *
  22  * You should have received a copy of the GNU General Public License along
  23  * with this program; if not, write to the Free Software Foundation, Inc.,
  24  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  25  * http://www.gnu.org/copyleft/gpl.html
  26  *
  27  * @file
  28  * @ingroup Maintenance
  29  */
  30
  31 require_once __DIR__ . '/dumpIterator.php';
  32
  33 /**
  34  * Maintenance script to take page text out of an XML dump file and render
  35  * basic HTML out to files.
  36  *
  37  * @ingroup Maintenance
  38  */
  39 class CompareParsers extends DumpIterator {
  40
  41         private $count = 0;
  42
  43         public function __construct() {
  44                 parent::__construct();
  45                 $this->saveFailed = false;
  46                 $this->addDescription( 'Run a file or dump with several parsers' );
  47                 $this->addOption( 'parser1', 'The first parser to compare.', true, true );
  48                 $this->addOption( 'parser2', 'The second parser to compare.', true, true );
  49                 $this->addOption( 'tidy', 'Run tidy on the articles.', false, false );
  50                 $this->addOption(
  51                         'save-failed',
  52                         'Folder in which articles which differ will be stored.',
  53                         false,
  54                         true
  55                 );
  56                 $this->addOption( 'show-diff', 'Show a diff of the two renderings.', false, false );
  57                 $this->addOption(
  58                         'diff-bin',
  59                         'Binary to use for diffing (can also be provided by DIFF env var).',
  60                         false,
  61                         false
  62                 );
  63                 $this->addOption(
  64                         'strip-parameters',
  65                         'Remove parameters of html tags to increase readability.',
  66                         false,
  67                         false
  68                 );
  69                 $this->addOption(
  70                         'show-parsed-output',
  71                         'Show the parsed html if both Parsers give the same output.',
  72                         false,
  73                         false
  74                 );
  75         }
  76
  77         public function checkOptions() {
  78                 if ( $this->hasOption( 'save-failed' ) ) {
  79                         $this->saveFailed = $this->getOption( 'save-failed' );
  80                 }
  81
  82                 $this->stripParametersEnabled = $this->hasOption( 'strip-parameters' );
  83                 $this->showParsedOutput = $this->hasOption( 'show-parsed-output' );
  84
  85                 $this->showDiff = $this->hasOption( 'show-diff' );
  86                 if ( $this->showDiff ) {
  87                         $bin = $this->getOption( 'diff-bin', getenv( 'DIFF' ) );
  88                         if ( $bin != '' ) {
  89                                 global $wgDiff;
  90                                 $wgDiff = $bin;
  91                         }
  92                 }
  93
  94                 $user = new User();
  95                 $this->options = ParserOptions::newFromUser( $user );
  96
  97                 if ( $this->hasOption( 'tidy' ) ) {
  98                         global $wgUseTidy;
  99                         if ( !$wgUseTidy ) {
 100                                 $this->error( 'Tidy was requested but $wgUseTidy is not set in LocalSettings.php', true );
 101                         }
 102                         $this->options->setTidy( true );
 103                 }
 104
 105                 $this->failed = 0;
 106         }
 107
 108         public function conclusions() {
 109                 $this->error( "{$this->failed} failed revisions out of {$this->count}" );
 110                 if ( $this->count > 0 ) {
 111                         $this->output( " (" . ( $this->failed / $this->count ) . "%)\n" );
 112                 }
 113         }
 114
 115         function stripParameters( $text ) {
 116                 if ( !$this->stripParametersEnabled ) {
 117                         return $text;
 118                 }
 119
 120                 return preg_replace( '/(<a) [^>]+>/', '$1>', $text );
 121         }
 122
 123         /**
 124          * Callback function for each revision, parse with both parsers and compare
 125          * @param Revision $rev
 126          */
 127         public function processRevision( $rev ) {
 128                 $title = $rev->getTitle();
 129
 130                 $parser1Name = $this->getOption( 'parser1' );
 131                 $parser2Name = $this->getOption( 'parser2' );
 132
 133                 self::checkParserLocally( $parser1Name );
 134                 self::checkParserLocally( $parser2Name );
 135
 136                 $parser1 = new $parser1Name();
 137                 $parser2 = new $parser2Name();
 138
 139                 $content = $rev->getContent();
 140
 141                 if ( $content->getModel() !== CONTENT_MODEL_WIKITEXT ) {
 142                         $this->error( "Page {$title->getPrefixedText()} does not contain wikitext "
 143                                 . "but {$content->getModel()}\n" );
 144
 145                         return;
 146                 }
 147
 148                 $text = strval( $content->getNativeData() );
 149
 150                 $output1 = $parser1->parse( $text, $title, $this->options );
 151                 $output2 = $parser2->parse( $text, $title, $this->options );
 152
 153                 if ( $output1->getText() != $output2->getText() ) {
 154                         $this->failed++;
 155                         $this->error( "Parsing for {$title->getPrefixedText()} differs\n" );
 156
 157                         if ( $this->saveFailed ) {
 158                                 file_put_contents(
 159                                         $this->saveFailed . '/' . rawurlencode( $title->getPrefixedText() ) . ".txt",
 160                                         $text
 161                                 );
 162                         }
 163                         if ( $this->showDiff ) {
 164                                 $this->output( wfDiff(
 165                                         $this->stripParameters( $output1->getText() ),
 166                                         $this->stripParameters( $output2->getText() ),
 167                                         ''
 168                                 ) );
 169                         }
 170                 } else {
 171                         $this->output( $title->getPrefixedText() . "\tOK\n" );
 172
 173                         if ( $this->showParsedOutput ) {
 174                                 $this->output( $this->stripParameters( $output1->getText() ) );
 175                         }
 176                 }
 177         }
 178
 179         private static function checkParserLocally( $parserName ) {
 180                 /* Look for the parser in a file appropiately named in the current folder */
 181                 if ( !class_exists( $parserName ) && file_exists( "$parserName.php" ) ) {
 182                         global $wgAutoloadClasses;
 183                         $wgAutoloadClasses[$parserName] = realpath( '.' ) . "/$parserName.php";
 184                 }
 185         }
 186 }
 187
 188 $maintClass = "CompareParsers";
 189 require_once RUN_MAINTENANCE_IF_MAIN;