maintenance/importDump.php

   1 <?php
   2 /**
   3  * Copyright (C) 2005 Brion Vibber <brion@pobox.com>
   4  * http://www.mediawiki.org/
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @file
  22  * @ingroup Maintenance
  23  */
  24
  25 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  26
  27 /**
  28  * @ingroup Maintenance
  29  */
  30 class BackupReader extends Maintenance {
  31         var $reportingInterval = 100;
  32         var $pageCount = 0;
  33         var $revCount  = 0;
  34         var $dryRun    = false;
  35         var $uploads   = false;
  36         var $imageBasePath = false;
  37         var $nsFilter  = false;
  38
  39         function __construct() {
  40                 parent::__construct();
  41                 $gz = in_array('compress.zlib', stream_get_wrappers()) ? 'ok' : '(disabled; requires PHP zlib module)';
  42                 $bz2 = in_array('compress.bzip2', stream_get_wrappers()) ? 'ok' : '(disabled; requires PHP bzip2 module)';
  43
  44                 $this->mDescription = <<<TEXT
  45 This script reads pages from an XML file as produced from Special:Export or
  46 dumpBackup.php, and saves them into the current wiki.
  47
  48 Compressed XML files may be read directly:
  49   .gz $gz
  50   .bz2 $bz2
  51   .7z (if 7za executable is in PATH)
  52
  53 Note that for very large data sets, importDump.php may be slow; there are
  54 alternate methods which can be much faster for full site restoration:
  55 <http://www.mediawiki.org/wiki/Manual:Importing_XML_dumps>
  56 TEXT;
  57                 $this->stderr = fopen( "php://stderr", "wt" );
  58                 $this->addOption( 'report',
  59                         'Report position and speed after every n pages processed', false, true );
  60                 $this->addOption( 'namespaces',
  61                         'Import only the pages from namespaces belonging to the list of ' .
  62                         'pipe-separated namespace names or namespace indexes', false, true );
  63                 $this->addOption( 'dry-run', 'Parse dump without actually importing pages' );
  64                 $this->addOption( 'debug', 'Output extra verbose debug information' );
  65                 $this->addOption( 'uploads', 'Process file upload data if included (experimental)' );
  66                 $this->addOption( 'image-base-path', 'Import files from a specified path', false, true );
  67                 $this->addArg( 'file', 'Dump file to import [else use stdin]', false );
  68         }
  69
  70         public function execute() {
  71                 if( wfReadOnly() ) {
  72                         $this->error( "Wiki is in read-only mode; you'll need to disable it for import to work.", true );
  73                 }
  74
  75                 $this->reportingInterval = intval( $this->getOption( 'report', 100 ) );
  76                 $this->dryRun = $this->hasOption( 'dry-run' );
  77                 $this->uploads = $this->hasOption( 'uploads' ); // experimental!
  78                 if ( $this->hasOption( 'image-base-path' ) ) {
  79                         $this->imageBasePath = $this->getOption( 'image-base-path' );
  80                 }
  81                 if ( $this->hasOption( 'namespaces' ) ) {
  82                         $this->setNsfilter( explode( '|', $this->getOption( 'namespaces' ) ) );
  83                 }
  84
  85                 if( $this->hasArg() ) {
  86                         $this->importFromFile( $this->getArg() );
  87                 } else {
  88                         $this->importFromStdin();
  89                 }
  90
  91                 $this->output( "Done!\n" );
  92                 $this->output( "You might want to run rebuildrecentchanges.php to regenerate RecentChanges\n" );
  93         }
  94
  95         function setNsfilter( array $namespaces ) {
  96                 if ( count( $namespaces ) == 0 ) {
  97                         $this->nsFilter = false;
  98                         return;
  99                 }
 100                 $this->nsFilter = array_unique( array_map( array( $this, 'getNsIndex' ), $namespaces ) );
 101         }
 102
 103         private function getNsIndex( $namespace ) {
 104                 global $wgContLang;
 105                 if ( ( $result = $wgContLang->getNsIndex( $namespace ) ) !== false ) {
 106                         return $result;
 107                 }
 108                 $ns = intval( $namespace );
 109                 if ( strval( $ns ) === $namespace && $wgContLang->getNsText( $ns ) !== false ) {
 110                         return $ns;
 111                 }
 112                 $this->error( "Unknown namespace text / index specified: $namespace", true );
 113         }
 114
 115         private function skippedNamespace( $obj ) {
 116                 if ( $obj instanceof Title ) {
 117                         $ns = $obj->getNamespace();
 118                 } elseif ( $obj instanceof Revision ) {
 119                         $ns = $obj->getTitle()->getNamespace();
 120                 } elseif ( $obj instanceof WikiRevision ) {
 121                         $ns = $obj->title->getNamespace();
 122                 } else {
 123                         echo wfBacktrace();
 124                         $this->error( "Cannot get namespace of object in " . __METHOD__, true );
 125                 }
 126                 return is_array( $this->nsFilter ) && !in_array( $ns, $this->nsFilter );
 127         }
 128
 129         function reportPage( $page ) {
 130                 $this->pageCount++;
 131         }
 132
 133         function handleRevision( $rev ) {
 134                 $title = $rev->getTitle();
 135                 if ( !$title ) {
 136                         $this->progress( "Got bogus revision with null title!" );
 137                         return;
 138                 }
 139
 140                 if ( $this->skippedNamespace( $title ) ) {
 141                         return;
 142                 }
 143
 144                 $this->revCount++;
 145                 $this->report();
 146
 147                 if ( !$this->dryRun ) {
 148                         call_user_func( $this->importCallback, $rev );
 149                 }
 150         }
 151
 152         function handleUpload( $revision ) {
 153                 if ( $this->uploads ) {
 154                         if ( $this->skippedNamespace( $revision ) ) {
 155                                 return;
 156                         }
 157                         $this->uploadCount++;
 158                         // $this->report();
 159                         $this->progress( "upload: " . $revision->getFilename() );
 160
 161                         if ( !$this->dryRun ) {
 162                                 // bluuuh hack
 163                                 // call_user_func( $this->uploadCallback, $revision );
 164                                 $dbw = wfGetDB( DB_MASTER );
 165                                 return $dbw->deadlockLoop( array( $revision, 'importUpload' ) );
 166                         }
 167                 }
 168         }
 169
 170         function handleLogItem( $rev ) {
 171                 if ( $this->skippedNamespace( $rev ) ) {
 172                         return;
 173                 }
 174                 $this->revCount++;
 175                 $this->report();
 176
 177                 if ( !$this->dryRun ) {
 178                         call_user_func( $this->logItemCallback, $rev );
 179                 }
 180         }
 181
 182         function report( $final = false ) {
 183                 if ( $final xor ( $this->pageCount % $this->reportingInterval == 0 ) ) {
 184                         $this->showReport();
 185                 }
 186         }
 187
 188         function showReport() {
 189                 if ( $this->mQuiet ) {
 190                         $delta = wfTime() - $this->startTime;
 191                         if ( $delta ) {
 192                                 $rate = sprintf( "%.2f", $this->pageCount / $delta );
 193                                 $revrate = sprintf( "%.2f", $this->revCount / $delta );
 194                         } else {
 195                                 $rate = '-';
 196                                 $revrate = '-';
 197                         }
 198                         # Logs dumps don't have page tallies
 199                         if ( $this->pageCount ) {
 200                                 $this->progress( "$this->pageCount ($rate pages/sec $revrate revs/sec)" );
 201                         } else {
 202                                 $this->progress( "$this->revCount ($revrate revs/sec)" );
 203                         }
 204                 }
 205                 wfWaitForSlaves();
 206                 // XXX: Don't let deferred jobs array get absurdly large (bug 24375)
 207                 wfDoUpdates( 'commit' );
 208         }
 209
 210         function progress( $string ) {
 211                 fwrite( $this->stderr, $string . "\n" );
 212         }
 213
 214         function importFromFile( $filename ) {
 215                 if ( preg_match( '/\.gz$/', $filename ) ) {
 216                         $filename = 'compress.zlib://' . $filename;
 217                 }
 218                 elseif ( preg_match( '/\.bz2$/', $filename ) ) {
 219                         $filename = 'compress.bzip2://' . $filename;
 220                 }
 221                 elseif ( preg_match( '/\.7z$/', $filename ) ) {
 222                         $filename = 'mediawiki.compress.7z://' . $filename;
 223                 }
 224
 225                 $file = fopen( $filename, 'rt' );
 226                 return $this->importFromHandle( $file );
 227         }
 228
 229         function importFromStdin() {
 230                 $file = fopen( 'php://stdin', 'rt' );
 231                 if( posix_isatty( $file ) ) {
 232                         $this->maybeHelp( true );
 233                 }
 234                 return $this->importFromHandle( $file );
 235         }
 236
 237         function importFromHandle( $handle ) {
 238                 $this->startTime = wfTime();
 239
 240                 $source = new ImportStreamSource( $handle );
 241                 $importer = new WikiImporter( $source );
 242
 243                 if( $this->hasOption( 'debug' ) ) {
 244                         $importer->setDebug( true );
 245                 }
 246                 $importer->setPageCallback( array( &$this, 'reportPage' ) );
 247                 $this->importCallback =  $importer->setRevisionCallback(
 248                         array( &$this, 'handleRevision' ) );
 249                 $this->uploadCallback = $importer->setUploadCallback(
 250                         array( &$this, 'handleUpload' ) );
 251                 $this->logItemCallback = $importer->setLogItemCallback(
 252                         array( &$this, 'handleLogItem' ) );
 253                 if ( $this->uploads ) {
 254                         $importer->setImportUploads( true );
 255                 }
 256                 if ( $this->imageBasePath ) {
 257                         $importer->setImageBasePath( $this->imageBasePath );
 258                 }
 259
 260                 if ( $this->dryRun ) {
 261                         $importer->setPageOutCallback( null );
 262                 }
 263
 264                 return $importer->doImport();
 265         }
 266 }
 267
 268 $maintClass = 'BackupReader';
 269 require_once( RUN_MAINTENANCE_IF_MAIN );