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 $nsFilter  = false;
  42
  43         function __construct() {
  44                 $this->stderr = fopen( "php://stderr", "wt" );
  45         }
  46
  47         function setNsfilter( array $namespaces ) {
  48                 if ( count( $namespaces ) == 0 ) {
  49                         $this->nsFilter = false;
  50                         return;
  51                 }
  52                 $this->nsFilter = array_unique( array_map( array( $this, 'getNsIndex' ), $namespaces ) );
  53         }
  54
  55         private function getNsIndex( $namespace ) {
  56                 global $wgContLang;
  57                 if ( ( $result = $wgContLang->getNsIndex( $namespace ) ) !== false ) {
  58                         return $result;
  59                 }
  60                 $ns = intval( $namespace );
  61                 if ( strval( $ns ) === $namespace && $wgContLang->getNsText( $ns ) !== false ) {
  62                         return $ns;
  63                 }
  64                 wfDie( "Unknown namespace text / index specified: $namespace\n" );
  65         }
  66
  67         private function skippedNamespace( $obj ) {
  68                 if ( $obj instanceof Title ) {
  69                         $ns = $obj->getNamespace();
  70                 } elseif ( $obj instanceof Revision ) {
  71                         $ns = $obj->getTitle()->getNamespace();
  72                 } elseif ( $obj instanceof WikiRevision ) {
  73                         $ns = $obj->title->getNamespace();
  74                 } else {
  75                         echo wfBacktrace();
  76                         wfDie( "Cannot get namespace of object in " . __METHOD__ . "\n" );
  77                 }
  78                 return is_array( $this->nsFilter ) && !in_array( $ns, $this->nsFilter );
  79         }
  80
  81         function reportPage( $page ) {
  82                 $this->pageCount++;
  83         }
  84
  85         function handleRevision( $rev ) {
  86                 $title = $rev->getTitle();
  87                 if ( !$title ) {
  88                         $this->progress( "Got bogus revision with null title!" );
  89                         return;
  90                 }
  91
  92                 if ( $this->skippedNamespace( $title ) ) {
  93                         return;
  94                 }
  95
  96                 $this->revCount++;
  97                 $this->report();
  98
  99                 if ( !$this->dryRun ) {
 100                         call_user_func( $this->importCallback, $rev );
 101                 }
 102         }
 103
 104         function handleUpload( $revision ) {
 105                 if ( $this->uploads ) {
 106                         if ( $this->skippedNamespace( $revision ) ) {
 107                                 return;
 108                         }
 109                         $this->uploadCount++;
 110                         // $this->report();
 111                         $this->progress( "upload: " . $revision->getFilename() );
 112
 113                         if ( !$this->dryRun ) {
 114                                 // bluuuh hack
 115                                 // call_user_func( $this->uploadCallback, $revision );
 116                                 $dbw = wfGetDB( DB_MASTER );
 117                                 return $dbw->deadlockLoop( array( $revision, 'importUpload' ) );
 118                         }
 119                 }
 120         }
 121
 122         function handleLogItem( $rev ) {
 123                 if ( $this->skippedNamespace( $rev ) ) {
 124                         return;
 125                 }
 126                 $this->revCount++;
 127                 $this->report();
 128
 129                 if ( !$this->dryRun ) {
 130                         call_user_func( $this->logItemCallback, $rev );
 131                 }
 132         }
 133
 134         function report( $final = false ) {
 135                 if ( $final xor ( $this->pageCount % $this->reportingInterval == 0 ) ) {
 136                         $this->showReport();
 137                 }
 138         }
 139
 140         function showReport() {
 141                 if ( $this->reporting ) {
 142                         $delta = wfTime() - $this->startTime;
 143                         if ( $delta ) {
 144                                 $rate = sprintf( "%.2f", $this->pageCount / $delta );
 145                                 $revrate = sprintf( "%.2f", $this->revCount / $delta );
 146                         } else {
 147                                 $rate = '-';
 148                                 $revrate = '-';
 149                         }
 150                         # Logs dumps don't have page tallies
 151                         if ( $this->pageCount ) {
 152                                 $this->progress( "$this->pageCount ($rate pages/sec $revrate revs/sec)" );
 153                         } else {
 154                                 $this->progress( "$this->revCount ($revrate revs/sec)" );
 155                         }
 156                 }
 157                 wfWaitForSlaves( 5 );
 158                 // XXX: Don't let deferred jobs array get absurdly large (bug 24375)
 159                 wfDoUpdates( 'commit' );
 160         }
 161
 162         function progress( $string ) {
 163                 fwrite( $this->stderr, $string . "\n" );
 164         }
 165
 166         function importFromFile( $filename ) {
 167                 if ( preg_match( '/\.gz$/', $filename ) ) {
 168                         $filename = 'compress.zlib://' . $filename;
 169                 }
 170                 elseif ( preg_match( '/\.bz2$/', $filename ) ) {
 171                         $filename = 'compress.bzip2://' . $filename;
 172                 }
 173                 elseif ( preg_match( '/\.7z$/', $filename ) ) {
 174                         $filename = 'mediawiki.compress.7z://' . $filename;
 175                 }
 176
 177                 $file = fopen( $filename, 'rt' );
 178                 return $this->importFromHandle( $file );
 179         }
 180
 181         function importFromStdin() {
 182                 $file = fopen( 'php://stdin', 'rt' );
 183                 if( posix_isatty( $file ) ) {
 184                         $this->showHelp();
 185                         exit();
 186                 }
 187                 return $this->importFromHandle( $file );
 188         }
 189
 190         function importFromHandle( $handle ) {
 191                 $this->startTime = wfTime();
 192
 193                 $source = new ImportStreamSource( $handle );
 194                 $importer = new WikiImporter( $source );
 195
 196                 $importer->setDebug( $this->debug );
 197                 $importer->setPageCallback( array( &$this, 'reportPage' ) );
 198                 $this->importCallback =  $importer->setRevisionCallback(
 199                         array( &$this, 'handleRevision' ) );
 200                 $this->uploadCallback = $importer->setUploadCallback(
 201                         array( &$this, 'handleUpload' ) );
 202                 $this->logItemCallback = $importer->setLogItemCallback(
 203                         array( &$this, 'handleLogItem' ) );
 204
 205                 if ( $this->dryRun ) {
 206                         $importer->setPageOutCallback( null );
 207                 }
 208
 209                 return $importer->doImport();
 210         }
 211
 212         function showHelp() {
 213                 $gz = in_array('compress.zlib', stream_get_wrappers()) ? 'ok' : '(disabled; requires PHP zlib module)';
 214                 $bz2 = in_array('compress.bzip2', stream_get_wrappers()) ? 'ok' : '(disabled; requires PHP bzip2 module)';
 215                 echo "This script reads pages from an XML file as produced from Special:Export\n";
 216                 echo "or dumpBackup.php, and saves them into the current wiki.\n";
 217                 echo "\n";
 218                 echo "Note that for very large data sets, importDump.php may be slow; there are\n";
 219                 echo "alternate methods which can be much faster for full site restoration:\n";
 220                 echo "http://www.mediawiki.org/wiki/Manual:Importing_XML_dumps\n";
 221                 echo "\n";
 222                 echo "Usage: php importDump.php [<options>] [<file>]\n";
 223                 echo "If no file is listed, input may be piped from stdin.\n";
 224                 echo "\n";
 225                 echo "Options:\n";
 226                 echo "  --quiet    Don't dump status reports to stderr.\n";
 227                 echo "  --report=n Report position and speed after every n pages processed.\n";
 228                 echo "  --namespaces=a|b|..|z Import only the pages from namespaces belonging to\n";
 229                 echo "    the list of pipe-separated namespace names or namespace indexes\n";
 230                 echo "  --dry-run  Parse dump without actually importing pages.\n";
 231                 echo "  --debug    Output extra verbose debug information\n";
 232                 echo "  --uploads  Process file upload data if included (experimental)\n";
 233                 echo "\n";
 234                 echo "Compressed XML files may be read directly:\n";
 235                 echo "  .gz $gz\n";
 236                 echo "  .bz2 $bz2\n";
 237                 echo "  .7z (if 7za executable is in PATH)\n";
 238                 echo "\n";
 239         }
 240 }
 241
 242 if ( wfReadOnly() ) {
 243         wfDie( "Wiki is in read-only mode; you'll need to disable it for import to work.\n" );
 244 }
 245
 246 $reader = new BackupReader();
 247 if ( isset( $options['quiet'] ) ) {
 248         $reader->reporting = false;
 249 }
 250 if ( isset( $options['report'] ) ) {
 251         $reader->reportingInterval = intval( $options['report'] );
 252 }
 253 if ( isset( $options['dry-run'] ) ) {
 254         $reader->dryRun = true;
 255 }
 256 if ( isset( $options['debug'] ) ) {
 257         $reader->debug = true;
 258 }
 259 if ( isset( $options['uploads'] ) ) {
 260         $reader->uploads = true; // experimental!
 261 }
 262 if ( isset( $options['namespaces'] ) ) {
 263         $reader->setNsfilter( explode( '|', $options['namespaces'] ) );
 264 }
 265
 266 if ( isset( $options['help'] ) ) {
 267         $reader->showHelp();
 268         exit();
 269 } elseif ( isset( $args[0] ) ) {
 270         $result = $reader->importFromFile( $args[0] );
 271 } else {
 272         $result = $reader->importFromStdin();
 273 }
 274
 275 echo "Done!\n";
 276 echo "You might want to run rebuildrecentchanges.php to regenerate\n";
 277 echo "the recentchanges page.\n";