maintenance/dumpIterator.php

   1 <?php
   2 /**
   3  * Take page text out of an XML dump file and perform some operation on it.
   4  * Used as a base class for CompareParsers and PreprocessDump.
   5  * We implement below the simple task of searching inside a dump.
   6  *
   7  * Copyright © 2011 Platonides
   8  * https://www.mediawiki.org/
   9  *
  10  * This program is free software; you can redistribute it and/or modify
  11  * it under the terms of the GNU General Public License as published by
  12  * the Free Software Foundation; either version 2 of the License, or
  13  * (at your option) any later version.
  14  *
  15  * This program is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  18  * GNU General Public License for more details.
  19  *
  20  * You should have received a copy of the GNU General Public License along
  21  * with this program; if not, write to the Free Software Foundation, Inc.,
  22  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  23  * http://www.gnu.org/copyleft/gpl.html
  24  *
  25  * @file
  26  * @ingroup Maintenance
  27  */
  28
  29 require_once __DIR__ . '/Maintenance.php';
  30
  31 /**
  32  * Base class for interating over a dump.
  33  *
  34  * @ingroup Maintenance
  35  */
  36 abstract class DumpIterator extends Maintenance {
  37
  38         private $count = 0;
  39         private $startTime;
  40
  41         public function __construct() {
  42                 parent::__construct();
  43                 $this->addDescription( 'Does something with a dump' );
  44                 $this->addOption( 'file', 'File with text to run.', false, true );
  45                 $this->addOption( 'dump', 'XML dump to execute all revisions.', false, true );
  46                 $this->addOption( 'from', 'Article from XML dump to start from.', false, true );
  47         }
  48
  49         public function execute() {
  50                 if ( !( $this->hasOption( 'file' ) ^ $this->hasOption( 'dump' ) ) ) {
  51                         $this->fatalError( "You must provide a file or dump" );
  52                 }
  53
  54                 $this->checkOptions();
  55
  56                 if ( $this->hasOption( 'file' ) ) {
  57                         $revision = new WikiRevision( $this->getConfig() );
  58
  59                         $revision->setText( file_get_contents( $this->getOption( 'file' ) ) );
  60                         $revision->setTitle( Title::newFromText(
  61                                 rawurldecode( basename( $this->getOption( 'file' ), '.txt' ) )
  62                         ) );
  63                         $this->handleRevision( $revision );
  64
  65                         return;
  66                 }
  67
  68                 $this->startTime = microtime( true );
  69
  70                 if ( $this->getOption( 'dump' ) == '-' ) {
  71                         $source = new ImportStreamSource( $this->getStdin() );
  72                 } else {
  73                         $this->fatalError( "Sorry, I don't support dump filenames yet. "
  74                                 . "Use - and provide it on stdin on the meantime." );
  75                 }
  76                 $importer = new WikiImporter( $source, $this->getConfig() );
  77
  78                 $importer->setRevisionCallback(
  79                         [ $this, 'handleRevision' ] );
  80
  81                 $this->from = $this->getOption( 'from', null );
  82                 $this->count = 0;
  83                 $importer->doImport();
  84
  85                 $this->conclusions();
  86
  87                 $delta = microtime( true ) - $this->startTime;
  88                 $this->error( "Done {$this->count} revisions in " . round( $delta, 2 ) . " seconds " );
  89                 if ( $delta > 0 ) {
  90                         $this->error( round( $this->count / $delta, 2 ) . " pages/sec" );
  91                 }
  92
  93                 # Perform the memory_get_peak_usage() when all the other data has been
  94                 # output so there's no damage if it dies. It is only available since
  95                 # 5.2.0 (since 5.2.1 if you haven't compiled with --enable-memory-limit)
  96                 $this->error( "Memory peak usage of " . memory_get_peak_usage() . " bytes\n" );
  97         }
  98
  99         public function finalSetup() {
 100                 parent::finalSetup();
 101
 102                 if ( $this->getDbType() == Maintenance::DB_NONE ) {
 103                         global $wgUseDatabaseMessages, $wgLocalisationCacheConf, $wgHooks;
 104                         $wgUseDatabaseMessages = false;
 105                         $wgLocalisationCacheConf['storeClass'] = 'LCStoreNull';
 106                         $wgHooks['InterwikiLoadPrefix'][] = 'DumpIterator::disableInterwikis';
 107                 }
 108         }
 109
 110         static function disableInterwikis( $prefix, &$data ) {
 111                 # Title::newFromText will check on each namespaced article if it's an interwiki.
 112                 # We always answer that it is not.
 113
 114                 return false;
 115         }
 116
 117         /**
 118          * Callback function for each revision, child classes should override
 119          * processRevision instead.
 120          * @param WikiRevision $rev
 121          */
 122         public function handleRevision( $rev ) {
 123                 $title = $rev->getTitle();
 124                 if ( !$title ) {
 125                         $this->error( "Got bogus revision with null title!" );
 126
 127                         return;
 128                 }
 129
 130                 $this->count++;
 131                 if ( isset( $this->from ) ) {
 132                         if ( $this->from != $title ) {
 133                                 return;
 134                         }
 135                         $this->output( "Skipped " . ( $this->count - 1 ) . " pages\n" );
 136
 137                         $this->count = 1;
 138                         $this->from = null;
 139                 }
 140
 141                 $this->processRevision( $rev );
 142         }
 143
 144         /* Stub function for processing additional options */
 145         public function checkOptions() {
 146                 return;
 147         }
 148
 149         /* Stub function for giving data about what was computed */
 150         public function conclusions() {
 151                 return;
 152         }
 153
 154         /* Core function which does whatever the maintenance script is designed to do */
 155         abstract public function processRevision( $rev );
 156 }
 157
 158 /**
 159  * Maintenance script that runs a regex in the revisions from a dump.
 160  *
 161  * @ingroup Maintenance
 162  */
 163 class SearchDump extends DumpIterator {
 164
 165         public function __construct() {
 166                 parent::__construct();
 167                 $this->addDescription( 'Runs a regex in the revisions from a dump' );
 168                 $this->addOption( 'regex', 'Searching regex', true, true );
 169         }
 170
 171         public function getDbType() {
 172                 return Maintenance::DB_NONE;
 173         }
 174
 175         /**
 176          * @param Revision $rev
 177          */
 178         public function processRevision( $rev ) {
 179                 if ( preg_match( $this->getOption( 'regex' ), $rev->getContent()->getTextForSearchIndex() ) ) {
 180                         $this->output( $rev->getTitle() . " matches at edit from " . $rev->getTimestamp() . "\n" );
 181                 }
 182         }
 183 }
 184
 185 $maintClass = "SearchDump";
 186 require_once RUN_MAINTENANCE_IF_MAIN;