maintenance/backup.inc

   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  * @package MediaWiki
  22  * @subpackage SpecialPage
  23  */
  24
  25
  26 class BackupDumper {
  27         var $reportingInterval = 100;
  28         var $reporting = true;
  29         var $pageCount = 0;
  30         var $revCount  = 0;
  31         var $server    = null; // use default
  32         var $pages     = null; // all pages
  33         var $skipHeader = false; // don't output <mediawiki> and <siteinfo>
  34         var $skipFooter = false; // don't output </mediawiki>
  35         var $startId    = 0;
  36         var $endId      = 0;
  37         var $sink       = null; // Output filters
  38         var $stubText   = false; // include rev_text_id instead of text; for 2-pass dump
  39
  40         function BackupDumper( $args ) {
  41                 $this->stderr = fopen( "php://stderr", "wt" );
  42
  43                 // Built-in output and filter plugins
  44                 $this->registerOutput( 'file', 'DumpFileOutput' );
  45                 $this->registerOutput( 'gzip', 'DumpGZipOutput' );
  46                 $this->registerOutput( 'bzip2', 'DumpBZip2Output' );
  47                 $this->registerOutput( '7zip', 'Dump7ZipOutput' );
  48
  49                 $this->registerFilter( 'latest', 'DumpLatestFilter' );
  50                 $this->registerFilter( 'notalk', 'DumpNotalkFilter' );
  51                 $this->registerFilter( 'namespace', 'DumpNamespaceFilter' );
  52
  53                 $this->sink = $this->processArgs( $args );
  54         }
  55
  56         /**
  57          * @param string $name
  58          * @param string $class name of output filter plugin class
  59          */
  60         function registerOutput( $name, $class ) {
  61                 $this->outputTypes[$name] = $class;
  62         }
  63
  64         /**
  65          * @param string $name
  66          * @param string $class name of filter plugin class
  67          */
  68         function registerFilter( $name, $class ) {
  69                 $this->filterTypes[$name] = $class;
  70         }
  71
  72         /**
  73          * Load a plugin and register it
  74          * @param string $class Name of plugin class; must have a static 'register'
  75          *                      method that takes a BackupDumper as a parameter.
  76          * @param string $file Full or relative path to the PHP file to load, or empty
  77          */
  78         function loadPlugin( $class, $file ) {
  79                 if( $file != '' ) {
  80                         require_once( $file );
  81                 }
  82                 $register = array( $class, 'register' );
  83                 call_user_func_array( $register, array( &$this ) );
  84         }
  85
  86         /**
  87          * @param array $args
  88          * @return array
  89          * @static
  90          */
  91         function processArgs( $args ) {
  92                 $sink = null;
  93                 $sinks = array();
  94                 foreach( $args as $arg ) {
  95                         if( preg_match( '/^--(.+?)(?:=(.+?)(?::(.+?))?)?$/', $arg, $matches ) ) {
  96                                 @list( $full, $opt, $val, $param ) = $matches;
  97                                 switch( $opt ) {
  98                                 case "plugin":
  99                                         $this->loadPlugin( $val, $param );
 100                                         break;
 101                                 case "output":
 102                                         if( !is_null( $sink ) ) {
 103                                                 $sinks[] = $sink;
 104                                         }
 105                                         if( !isset( $this->outputTypes[$val] ) ) {
 106                                                 wfDie( "Unrecognized output sink type '$val'\n" );
 107                                         }
 108                                         $type = $this->outputTypes[$val];
 109                                         $sink = new $type( $param );
 110                                         break;
 111                                 case "filter":
 112                                         if( is_null( $sink ) ) {
 113                                                 $this->progress( "Warning: assuming stdout for filter output\n" );
 114                                                 $sink = new DumpOutput();
 115                                         }
 116                                         if( !isset( $this->filterTypes[$val] ) ) {
 117                                                 wfDie( "Unrecognized filter type '$val'\n" );
 118                                         }
 119                                         $type = $this->filterTypes[$val];
 120                                         $filter = new $type( $sink, $param );
 121
 122                                         // references are lame in php...
 123                                         unset( $sink );
 124                                         $sink = $filter;
 125
 126                                         break;
 127                                 case "report":
 128                                         $this->reportingInterval = intval( $val );
 129                                         break;
 130                                 case "server":
 131                                         $this->server = $val;
 132                                         break;
 133                                 case "force-normal":
 134                                         if( !function_exists( 'utf8_normalize' ) ) {
 135                                                 dl( "php_utfnormal.so" );
 136                                                 if( !function_exists( 'utf8_normalize' ) ) {
 137                                                         wfDie( "Failed to load UTF-8 normalization extension. " .
 138                                                                 "Install or remove --force-normal parameter to use slower code.\n" );
 139                                                 }
 140                                         }
 141                                         break;
 142                                 default:
 143                                         $this->processOption( $opt, $val, $param );
 144                                 }
 145                         }
 146                 }
 147
 148                 if( is_null( $sink ) ) {
 149                         $sink = new DumpOutput();
 150                 }
 151                 $sinks[] = $sink;
 152
 153                 if( count( $sinks ) > 1 ) {
 154                         return new DumpMultiWriter( $sinks );
 155                 } else {
 156                         return $sink;
 157                 }
 158         }
 159
 160         function processOption( $opt, $val, $param ) {
 161                 // extension point for subclasses to add options
 162         }
 163
 164         function dump( $history, $text = MW_EXPORT_TEXT ) {
 165                 # This shouldn't happen if on console... ;)
 166                 header( 'Content-type: text/html; charset=UTF-8' );
 167
 168                 # Notice messages will foul up your XML output even if they're
 169                 # relatively harmless.
 170                 ini_set( 'display_errors', false );
 171
 172                 $this->initProgress( $history );
 173
 174                 $db =& $this->backupDb();
 175                 $exporter = new WikiExporter( $db, $history, MW_EXPORT_STREAM, $text );
 176
 177                 $wrapper = new ExportProgressFilter( $this->sink, $this );
 178                 $exporter->setOutputSink( $wrapper );
 179
 180                 if( !$this->skipHeader )
 181                         $exporter->openStream();
 182
 183                 if( is_null( $this->pages ) ) {
 184                         if( $this->startId || $this->endId ) {
 185                                 $exporter->pagesByRange( $this->startId, $this->endId );
 186                         } else {
 187                                 $exporter->allPages();
 188                         }
 189                 } else {
 190                         $exporter->pagesByName( $this->pages );
 191                 }
 192
 193                 if( !$this->skipFooter )
 194                         $exporter->closeStream();
 195
 196                 $this->report( true );
 197         }
 198
 199         /**
 200          * Initialise starting time and maximum revision count.
 201          * We'll make ETA calculations based an progress, assuming relatively
 202          * constant per-revision rate.
 203          * @param int $history MW_EXPORT_CURRENT or MW_EXPORT_FULL
 204          */
 205         function initProgress( $history = MW_EXPORT_FULL ) {
 206                 $table = ($history == MW_EXPORT_CURRENT) ? 'page' : 'revision';
 207                 $field = ($history == MW_EXPORT_CURRENT) ? 'page_id' : 'rev_id';
 208
 209                 $dbr =& wfGetDB( DB_SLAVE );
 210                 $this->maxCount = $dbr->selectField( $table, "MAX($field)", '', 'BackupDumper::dump' );
 211                 $this->startTime = wfTime();
 212         }
 213
 214         function &backupDb() {
 215                 global $wgDBadminuser, $wgDBadminpassword;
 216                 global $wgDBname, $wgDebugDumpSql;
 217                 $flags = ($wgDebugDumpSql ? DBO_DEBUG : 0) | DBO_DEFAULT; // god-damn hack
 218                 $db =& new Database( $this->backupServer(), $wgDBadminuser, $wgDBadminpassword, $wgDBname, false, $flags );
 219                 $timeout = 3600 * 24;
 220                 $db->query( "SET net_read_timeout=$timeout" );
 221                 $db->query( "SET net_write_timeout=$timeout" );
 222                 return $db;
 223         }
 224
 225         function backupServer() {
 226                 global $wgDBserver;
 227                 return $this->server
 228                         ? $this->server
 229                         : $wgDBserver;
 230         }
 231
 232         function reportPage() {
 233                 $this->pageCount++;
 234         }
 235
 236         function revCount() {
 237                 $this->revCount++;
 238                 $this->report();
 239         }
 240
 241         function report( $final = false ) {
 242                 if( $final xor ( $this->revCount % $this->reportingInterval == 0 ) ) {
 243                         $this->showReport();
 244                 }
 245         }
 246
 247         function showReport() {
 248                 if( $this->reporting ) {
 249                         $delta = wfTime() - $this->startTime;
 250                         $now = wfTimestamp( TS_DB );
 251                         if( $delta ) {
 252                                 $rate = $this->pageCount / $delta;
 253                                 $revrate = $this->revCount / $delta;
 254                                 $portion = $this->revCount / $this->maxCount;
 255                                 $eta = $this->startTime + $delta / $portion;
 256                                 $etats = wfTimestamp( TS_DB, intval( $eta ) );
 257                         } else {
 258                                 $rate = '-';
 259                                 $revrate = '-';
 260                                 $etats = '-';
 261                         }
 262                         global $wgDBname;
 263                         $this->progress( sprintf( "%s: %s %d pages (%0.3f/sec), %d revs (%0.3f/sec), ETA %s [max %d]",
 264                                 $now, $wgDBname, $this->pageCount, $rate, $this->revCount, $revrate, $etats, $this->maxCount ) );
 265                 }
 266         }
 267
 268         function progress( $string ) {
 269                 fwrite( $this->stderr, $string . "\n" );
 270         }
 271 }
 272
 273 class ExportProgressFilter extends DumpFilter {
 274         function ExportProgressFilter( &$sink, &$progress ) {
 275                 parent::DumpFilter( $sink );
 276                 $this->progress = $progress;
 277         }
 278
 279         function writeClosePage( $string ) {
 280                 parent::writeClosePage( $string );
 281                 $this->progress->reportPage();
 282         }
 283
 284         function writeRevision( $rev, $string ) {
 285                 parent::writeRevision( $rev, $string );
 286                 $this->progress->revCount();
 287         }
 288 }
 289
 290 ?>