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 $optionsWithArgs = array( 'report', 'namespaces' );
  26
  27 require_once( dirname( __FILE__ ) . '/commandLine.inc' );
  28
  29 /**
  30  * @ingroup Maintenance
  31  * @todo port to Maintenance class
  32  */
  33 class BackupReader {
  34         var $reportingInterval = 100;
  35         var $reporting = true;
  36         var $pageCount = 0;
  37         var $revCount  = 0;
  38         var $dryRun    = false;
  39         var $debug     = false;
  40         var $uploads   = false;
  41         var $imageBasePath = false;
  42         var $nsFilter  = false;
  43
  44         function __construct() {
  45                 $this->stderr = fopen( "php://stderr", "wt" );
  46         }
  47
  48         function setNsfilter( array $namespaces ) {
  49                 if ( count( $namespaces ) == 0 ) {
  50                         $this->nsFilter = false;
  51                         return;
  52                 }
  53                 $this->nsFilter = array_unique( array_map( array( $this, 'getNsIndex' ), $namespaces ) );
  54         }
  55
  56         private function getNsIndex( $namespace ) {
  57                 global $wgContLang;
  58                 if ( ( $result = $wgContLang->getNsIndex( $namespace ) ) !== false ) {
  59                         return $result;
  60                 }
  61                 $ns = intval( $namespace );
  62                 if ( strval( $ns ) === $namespace && $wgContLang->getNsText( $ns ) !== false ) {
  63                         return $ns;
  64                 }
  65                 wfDie( "Unknown namespace text / index specified: $namespace\n" );
  66         }
  67
  68         private function skippedNamespace( $obj ) {
  69                 if ( $obj instanceof Title ) {
  70                         $ns = $obj->getNamespace();
  71                 } elseif ( $obj instanceof Revision ) {
  72                         $ns = $obj->getTitle()->getNamespace();
  73                 } elseif ( $obj instanceof WikiRevision ) {
  74                         $ns = $obj->title->getNamespace();
  75                 } else {
  76                         echo wfBacktrace();
  77                         wfDie( "Cannot get namespace of object in " . __METHOD__ . "\n" );
  78                 }
  79                 return is_array( $this->nsFilter ) && !in_array( $ns, $this->nsFilter );
  80         }
  81
  82         function reportPage( $page ) {
  83                 $this->pageCount++;
  84         }
  85
  86         function handleRevision( $rev ) {
  87                 $title = $rev->getTitle();
  88                 if ( !$title ) {
  89                         $this->progress( "Got bogus revision with null title!" );
  90                         return;
  91                 }
  92
  93                 if ( $this->skippedNamespace( $title ) ) {
  94                         return;
  95                 }
  96
  97                 $this->revCount++;
  98                 $this->report();
  99
 100                 if ( !$this->dryRun ) {
 101                         call_user_func( $this->importCallback, $rev );
 102                 }
 103         }
 104
 105         function handleUpload( $revision ) {
 106                 if ( $this->uploads ) {
 107                         if ( $this->skippedNamespace( $revision ) ) {
 108                                 return;
 109                         }
 110                         $this->uploadCount++;
 111                         // $this->report();
 112                         $this->progress( "upload: " . $revision->getFilename() );
 113
 114                         if ( !$this->dryRun ) {
 115                                 // bluuuh hack
 116                                 // call_user_func( $this->uploadCallback, $revision );
 117                                 $dbw = wfGetDB( DB_MASTER );
 118                                 return $dbw->deadlockLoop( array( $revision, 'importUpload' ) );
 119                         }
 120                 }
 121         }
 122
 123         function handleLogItem( $rev ) {
 124                 if ( $this->skippedNamespace( $rev ) ) {
 125                         return;
 126                 }
 127                 $this->revCount++;
 128                 $this->report();
 129
 130                 if ( !$this->dryRun ) {
 131                         call_user_func( $this->logItemCallback, $rev );
 132                 }
 133         }
 134
 135         function report( $final = false ) {
 136                 if ( $final xor ( $this->pageCount % $this->reportingInterval == 0 ) ) {
 137                         $this->showReport();
 138                 }
 139         }
 140
 141         function showReport() {
 142                 if ( $this->reporting ) {
 143                         $delta = wfTime() - $this->startTime;
 144                         if ( $delta ) {
 145                                 $rate = sprintf( "%.2f", $this->pageCount / $delta );
 146                                 $revrate = sprintf( "%.2f", $this->revCount / $delta );
 147                         } else {
 148                                 $rate = '-';
 149                                 $revrate = '-';
 150                         }
 151                         # Logs dumps don't have page tallies
 152                         if ( $this->pageCount ) {
 153                                 $this->progress( "$this->pageCount ($rate pages/sec $revrate revs/sec)" );
 154                         } else {
 155                                 $this->progress( "$this->revCount ($revrate revs/sec)" );
 156                         }
 157                 }
 158                 wfWaitForSlaves();
 159                 // XXX: Don't let deferred jobs array get absurdly large (bug 24375)
 160                 wfDoUpdates( 'commit' );
 161         }
 162
 163         function progress( $string ) {
 164                 fwrite( $this->stderr, $string . "\n" );
 165         }
 166
 167         function importFromFile( $filename ) {
 168                 if ( preg_match( '/\.gz$/', $filename ) ) {
 169                         $filename = 'compress.zlib://' . $filename;
 170                 }
 171                 elseif ( preg_match( '/\.bz2$/', $filename ) ) {
 172                         $filename = 'compress.bzip2://' . $filename;
 173                 }
 174                 elseif ( preg_match( '/\.7z$/', $filename ) ) {
 175                         $filename = 'mediawiki.compress.7z://' . $filename;
 176                 }
 177
 178                 $file = fopen( $filename, 'rt' );
 179                 return $this->importFromHandle( $file );
 180         }
 181
 182         function importFromStdin() {
 183                 $file = fopen( 'php://stdin', 'rt' );
 184                 if( posix_isatty( $file ) ) {
 185                         $this->showHelp();
 186                         exit();
 187                 }
 188                 return $this->importFromHandle( $file );
 189         }
 190
 191         function importFromHandle( $handle ) {
 192                 $this->startTime = wfTime();
 193
 194                 $source = new ImportStreamSource( $handle );
 195                 $importer = new WikiImporter( $source );
 196
 197                 $importer->setDebug( $this->debug );
 198                 $importer->setPageCallback( array( &$this, 'reportPage' ) );
 199                 $this->importCallback =  $importer->setRevisionCallback(
 200                         array( &$this, 'handleRevision' ) );
 201                 $this->uploadCallback = $importer->setUploadCallback(
 202                         array( &$this, 'handleUpload' ) );
 203                 $this->logItemCallback = $importer->setLogItemCallback(
 204                         array( &$this, 'handleLogItem' ) );
 205                 if ( $this->uploads ) {
 206                         $importer->setImportUploads( true );
 207                 }
 208                 if ( $this->imageBasePath ) {
 209                         $importer->setImageBasePath( $this->imageBasePath );
 210                 }
 211
 212                 if ( $this->dryRun ) {
 213                         $importer->setPageOutCallback( null );
 214                 }
 215
 216                 return $importer->doImport();
 217         }
 218
 219         function showHelp() {
 220                 $gz = in_array('compress.zlib', stream_get_wrappers()) ? 'ok' : '(disabled; requires PHP zlib module)';
 221                 $bz2 = in_array('compress.bzip2', stream_get_wrappers()) ? 'ok' : '(disabled; requires PHP bzip2 module)';
 222                 echo "This script reads pages from an XML file as produced from Special:Export\n";
 223                 echo "or dumpBackup.php, and saves them into the current wiki.\n";
 224                 echo "\n";
 225                 echo "Note that for very large data sets, importDump.php may be slow; there are\n";
 226                 echo "alternate methods which can be much faster for full site restoration:\n";
 227                 echo "http://www.mediawiki.org/wiki/Manual:Importing_XML_dumps\n";
 228                 echo "\n";
 229                 echo "Usage: php importDump.php [<options>] [<file>]\n";
 230                 echo "If no file is listed, input may be piped from stdin.\n";
 231                 echo "\n";
 232                 echo "Options:\n";
 233                 echo "  --quiet    Don't dump status reports to stderr.\n";
 234                 echo "  --report=n Report position and speed after every n pages processed.\n";
 235                 echo "  --namespaces=a|b|..|z Import only the pages from namespaces belonging to\n";
 236                 echo "    the list of pipe-separated namespace names or namespace indexes\n";
 237                 echo "  --dry-run  Parse dump without actually importing pages.\n";
 238                 echo "  --debug    Output extra verbose debug information\n";
 239                 echo "  --uploads  Process file upload data if included (experimental)\n";
 240                 echo "  --image-base-path=path  Import files from a specified path\n";
 241                 echo "\n";
 242                 echo "Compressed XML files may be read directly:\n";
 243                 echo "  .gz $gz\n";
 244                 echo "  .bz2 $bz2\n";
 245                 echo "  .7z (if 7za executable is in PATH)\n";
 246                 echo "\n";
 247         }
 248 }
 249
 250 if ( wfReadOnly() ) {
 251         wfDie( "Wiki is in read-only mode; you'll need to disable it for import to work.\n" );
 252 }
 253
 254 $reader = new BackupReader();
 255 if ( isset( $options['quiet'] ) ) {
 256         $reader->reporting = false;
 257 }
 258 if ( isset( $options['report'] ) ) {
 259         $reader->reportingInterval = intval( $options['report'] );
 260 }
 261 if ( isset( $options['dry-run'] ) ) {
 262         $reader->dryRun = true;
 263 }
 264 if ( isset( $options['debug'] ) ) {
 265         $reader->debug = true;
 266 }
 267 if ( isset( $options['uploads'] ) ) {
 268         $reader->uploads = true; // experimental!
 269 }
 270 if ( isset( $options['image-base-path'] ) ) {
 271         $reader->imageBasePath = $options['image-base-path'];
 272 }
 273 if ( isset( $options['namespaces'] ) ) {
 274         $reader->setNsfilter( explode( '|', $options['namespaces'] ) );
 275 }
 276
 277 if ( isset( $options['help'] ) ) {
 278         $reader->showHelp();
 279         exit();
 280 } elseif ( isset( $args[0] ) ) {
 281         $result = $reader->importFromFile( $args[0] );
 282 } else {
 283         $result = $reader->importFromStdin();
 284 }
 285
 286 echo "Done!\n";
 287 echo "You might want to run rebuildrecentchanges.php to regenerate\n";
 288 echo "the recentchanges page.\n";