maintenance/preprocessDump.php

   1 <?php
   2 /**
   3  * Take page text out of an XML dump file and preprocess it to obj.
   4  * It may be useful for getting preprocessor statistics or filling the
   5  * preprocessor cache.
   6  *
   7  * Copyright (C) 2011 Platonides - http://www.mediawiki.org/
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License as published by
  11  * the Free Software Foundation; either version 2 of the License, or
  12  * (at your option) any later version.
  13  *
  14  * This program is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17  * GNU General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU General Public License along
  20  * with this program; if not, write to the Free Software Foundation, Inc.,
  21  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  22  * http://www.gnu.org/copyleft/gpl.html
  23  *
  24  * @file
  25  * @ingroup Maintenance
  26  */
  27
  28 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  29
  30 class PreprocessDump extends Maintenance {
  31
  32         private $count = 0;
  33         private $startTime;
  34
  35         /* Variables for dressing up as a parser */
  36         public $mTitle = 'PreprocessDump';
  37         public $mPPNodeCount = 0;
  38
  39         public function getStripList() {
  40                 global $wgParser;
  41                 return $wgParser->getStripList();
  42         }
  43
  44         public function __construct() {
  45                 parent::__construct();
  46                 $this->saveFailed = false;
  47                 $this->mDescription = "Run a file or dump with a preprocessor";
  48                 $this->addOption( 'file',  'File with text to run.', false, true );
  49                 $this->addOption( 'dump',  'XML dump to execute all revisions.', false, true );
  50                 $this->addOption( 'from',  'Article from XML dump to start from.', false, true );
  51                 $this->addOption( 'cache', 'Use and populate the preprocessor cache.', false, false );
  52                 $this->addOption( 'preprocessor', 'Preprocessor to use.', false, false );
  53         }
  54
  55         public function getDbType() {
  56                 return Maintenance::DB_NONE;
  57         }
  58
  59         public function finalSetup() {
  60                 parent::finalSetup();
  61
  62                 global $wgUseDatabaseMessages, $wgLocalisationCacheConf, $wgHooks;
  63                 $wgUseDatabaseMessages = false;
  64                 $wgLocalisationCacheConf['storeClass'] =  'LCStore_Null';
  65                 $wgHooks['InterwikiLoadPrefix'][] = 'PreprocessDump::disableInterwikis';
  66         }
  67
  68         static function disableInterwikis( $prefix, &$data ) {
  69                 # Title::newFromText will check on each namespaced article if it's an interwiki.
  70                 # We always answer that it is not.
  71
  72                 return false;
  73         }
  74
  75         public function execute() {
  76                 global $wgParser, $wgParserConf, $wgPreprocessorCacheThreshold;
  77
  78                 if (! ( $this->hasOption( 'file' ) ^ $this->hasOption( 'dump' ) ) ) {
  79                         $this->error("You must provide a file or dump", true);
  80                 }
  81
  82                 if ( !$this->hasOption( 'cache' ) ) {
  83                         $wgPreprocessorCacheThreshold = false;
  84                         $this->saveFailed = $this->getOption('save-failed');
  85                 }
  86
  87                 if ( $this->hasOption( 'preprocessor' ) ) {
  88                         $name = $this->getOption( 'preprocessor' );
  89                 } elseif ( isset( $wgParserConf['preprocessorClass'] ) ) {
  90                         $name = $wgParserConf['preprocessorClass'];
  91                 } else {
  92                         $name = 'Preprocessor_DOM';
  93                 }
  94
  95                 $wgParser->firstCallInit();
  96                 $this->mPreprocessor = new $name( $this );
  97
  98                 if ( $this->hasOption( 'file' ) ) {
  99                         $revision = new WikiRevision;
 100
 101                         $revision->setText( file_get_contents( $this->getOption( 'file' ) ) );
 102                         $revision->setTitle( Title::newFromText( rawurldecode( basename( $this->getOption('file'), '.txt' ) ) ) );
 103                         $this->handleRevision( $revision );
 104                         return;
 105                 }
 106
 107                 $this->startTime = wfTime();
 108
 109                 if ( $this->getOption('dump') == '-' ) {
 110                         $source = new ImportStreamSource( $this->getStdin() );
 111                 } else {
 112                         $this->error("Sorry, I don't support dump filenames yet. Use - and provide it on stdin on the meantime.", true);
 113                 }
 114                 $importer = new WikiImporter( $source );
 115
 116                 $importer->setRevisionCallback(
 117                         array( &$this, 'handleRevision' ) );
 118
 119                 $this->from = $this->getOption( 'from', null );
 120                 $this->count = 0;
 121                 $importer->doImport();
 122
 123                 $delta = wfTime() - $this->startTime;
 124                 $this->error( "{$this->count} revisions preprocessed in " . round($delta, 2) . " seconds " );
 125                 if ($delta > 0)
 126                         $this->error( round($this->count / $delta, 2) . " pages/sec" );
 127
 128                 # Perform the memory_get_peak_usage() when all the other data has been output so there's no damage if it dies.
 129                 # It is only available since 5.2.0 (since 5.2.1 if you haven't compiled with --enable-memory-limit)
 130                 $this->error( "Memory peak usage of " . memory_get_peak_usage() . " bytes\n" );
 131         }
 132
 133         /**
 134          * Callback function for each revision, preprocessToObj()
 135          * @param $rev Revision
 136          */
 137         public function handleRevision( $rev ) {
 138                 $title = $rev->getTitle();
 139                 if ( !$title ) {
 140                         $this->error( "Got bogus revision with null title!" );
 141                         return;
 142                 }
 143
 144                 $this->count++;
 145                 if ( isset( $this->from ) ) {
 146                         if ( $this->from != $title )
 147                                 return;
 148                         $this->output( "Skipped " . ($this->count - 1) . " pages\n" );
 149
 150                         $this->count = 1;
 151                         $this->from = null;
 152                 }
 153                 try {
 154                         $this->mPreprocessor->preprocessToObj( $rev->getText(), 0 );
 155                 }
 156                 catch(Exception $e) {
 157                         $this->error("Caught exception " . $e->getMessage() . " in " . $title-> getPrefixedText() );
 158                 }
 159         }
 160 }
 161
 162 $maintClass = "PreprocessDump";
 163 require_once( RUN_MAINTENANCE_IF_MAIN );
 164