maintenance/convertLinks.php

   1 <?php
   2 /**
   3  * Convert from the old links schema (string->ID) to the new schema (ID->ID)
   4  * The wiki should be put into read-only mode while this script executes
   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  * @ingroup Maintenance
  22  */
  23
  24 require_once( dirname( __FILE__ ) . '/Maintenance.php' );
  25
  26 class ConvertLinks extends Maintenance {
  27
  28         public function __construct() {
  29                 parent::__construct();
  30                 $this->mDescription = "Convert from the old links schema (string->ID) to the new schema (ID->ID)
  31 The wiki should be put into read-only mode while this script executes";
  32         }
  33
  34         public function getDbType() {
  35                 return self::DB_ADMIN;
  36         }
  37
  38         public function execute() {
  39                 global $wgDBtype;
  40
  41                 $dbw = wfGetDB( DB_MASTER );
  42
  43                 $type = $dbw->getType();
  44                 if ( $type != 'mysql' ) {
  45                         $this->output( "Link table conversion not necessary for $type\n" );
  46                         return;
  47                 }
  48
  49                 global $wgLang, $noKeys, $logPerformance, $fh;
  50
  51                 $tuplesAdded = $numBadLinks = $curRowsRead = 0; # counters etc
  52                 $totalTuplesInserted = 0; # total tuples INSERTed into links_temp
  53
  54                 $reportCurReadProgress = true; # whether or not to give progress reports while reading IDs from cur table
  55                 $curReadReportInterval = 1000; # number of rows between progress reports
  56
  57                 $reportLinksConvProgress = true; # whether or not to give progress reports during conversion
  58                 $linksConvInsertInterval = 1000; # number of rows per INSERT
  59
  60                 $initialRowOffset = 0;
  61                 # $finalRowOffset = 0; # not used yet; highest row number from links table to process
  62
  63                 # Overwrite the old links table with the new one.  If this is set to false,
  64                 # the new table will be left at links_temp.
  65                 $overwriteLinksTable = true;
  66
  67                 # Don't create keys, and so allow duplicates in the new links table.
  68                 # This gives a huge speed improvement for very large links tables which are MyISAM. (What about InnoDB?)
  69                 $noKeys = false;
  70
  71
  72                 $logPerformance = false; # output performance data to a file
  73                 $perfLogFilename = "convLinksPerf.txt";
  74                 # --------------------------------------------------------------------
  75
  76                 list ( $cur, $links, $links_temp, $links_backup ) = $dbw->tableNamesN( 'cur', 'links', 'links_temp', 'links_backup' );
  77
  78                 if( $dbw->tableExists( 'pagelinks' ) ) {
  79                         $this->output( "...have pagelinks; skipping old links table updates\n" );
  80                         return;
  81                 }
  82
  83                 $res = $dbw->query( "SELECT l_from FROM $links LIMIT 1" );
  84                 if ( $dbw->fieldType( $res, 0 ) == "int" ) {
  85                         $this->output( "Schema already converted\n" );
  86                         return;
  87                 }
  88
  89                 $res = $dbw->query( "SELECT COUNT(*) AS count FROM $links" );
  90                 $row = $dbw->fetchObject( $res );
  91                 $numRows = $row->count;
  92                 $dbw->freeResult( $res );
  93
  94                 if ( $numRows == 0 ) {
  95                         $this->output( "Updating schema (no rows to convert)...\n" );
  96                         $this->createTempTable();
  97                 } else {
  98                         if ( $logPerformance ) { $fh = fopen ( $perfLogFilename, "w" ); }
  99                         $baseTime = $startTime = $this->getMicroTime();
 100                         # Create a title -> cur_id map
 101                         $this->output( "Loading IDs from $cur table...\n" );
 102                         $this->performanceLog ( "Reading $numRows rows from cur table...\n" );
 103                         $this->performanceLog ( "rows read vs seconds elapsed:\n" );
 104
 105                         $dbw->bufferResults( false );
 106                         $res = $dbw->query( "SELECT cur_namespace,cur_title,cur_id FROM $cur" );
 107                         $ids = array();
 108
 109                         while ( $row = $dbw->fetchObject( $res ) ) {
 110                                 $title = $row->cur_title;
 111                                 if ( $row->cur_namespace ) {
 112                                         $title = $wgLang->getNsText( $row->cur_namespace ) . ":$title";
 113                                 }
 114                                 $ids[$title] = $row->cur_id;
 115                                 $curRowsRead++;
 116                                 if ( $reportCurReadProgress ) {
 117                                         if ( ( $curRowsRead % $curReadReportInterval ) == 0 ) {
 118                                                 $this->performanceLog( $curRowsRead . " " . ( $this->getMicroTime() - $baseTime ) . "\n" );
 119                                                 $this->output( "\t$curRowsRead rows of $cur table read.\n" );
 120                                         }
 121                                 }
 122                         }
 123                         $dbw->freeResult( $res );
 124                         $dbw->bufferResults( true );
 125                         $this->output( "Finished loading IDs.\n\n" );
 126                         $this->performanceLog( "Took " . ( $this->getMicroTime() - $baseTime ) . " seconds to load IDs.\n\n" );
 127                 # --------------------------------------------------------------------
 128
 129                         # Now, step through the links table (in chunks of $linksConvInsertInterval rows),
 130                         # convert, and write to the new table.
 131                         $this->createTempTable();
 132                         $this->performanceLog( "Resetting timer.\n\n" );
 133                         $baseTime = $this->getMicroTime();
 134                         $this->output( "Processing $numRows rows from $links table...\n" );
 135                         $this->performanceLog( "Processing $numRows rows from $links table...\n" );
 136                         $this->performanceLog( "rows inserted vs seconds elapsed:\n" );
 137
 138                         for ( $rowOffset = $initialRowOffset; $rowOffset < $numRows; $rowOffset += $linksConvInsertInterval ) {
 139                                 $sqlRead = "SELECT * FROM $links ";
 140                                 $sqlRead = $dbw->limitResult( $sqlRead, $linksConvInsertInterval, $rowOffset );
 141                                 $res = $dbw->query( $sqlRead );
 142                                 if ( $noKeys ) {
 143                                         $sqlWrite = array( "INSERT INTO $links_temp (l_from,l_to) VALUES " );
 144                                 } else {
 145                                         $sqlWrite = array( "INSERT IGNORE INTO $links_temp (l_from,l_to) VALUES " );
 146                                 }
 147
 148                                 $tuplesAdded = 0; # no tuples added to INSERT yet
 149                                 while ( $row = $dbw->fetchObject( $res ) ) {
 150                                         $fromTitle = $row->l_from;
 151                                         if ( array_key_exists( $fromTitle, $ids ) ) { # valid title
 152                                                 $from = $ids[$fromTitle];
 153                                                 $to = $row->l_to;
 154                                                 if ( $tuplesAdded != 0 ) {
 155                                                         $sqlWrite[] = ",";
 156                                                 }
 157                                                 $sqlWrite[] = "($from,$to)";
 158                                                 $tuplesAdded++;
 159                                         } else { # invalid title
 160                                                 $numBadLinks++;
 161                                         }
 162                                 }
 163                                 $dbw->freeResult( $res );
 164                                 # $this->output( "rowOffset: $rowOffset\ttuplesAdded: $tuplesAdded\tnumBadLinks: $numBadLinks\n" );
 165                                 if ( $tuplesAdded != 0  ) {
 166                                         if ( $reportLinksConvProgress ) {
 167                                                 $this->output( "Inserting $tuplesAdded tuples into $links_temp..." );
 168                                         }
 169                                         $dbw->query( implode( "", $sqlWrite ) );
 170                                         $totalTuplesInserted += $tuplesAdded;
 171                                         if ( $reportLinksConvProgress )
 172                                                 $this->output( " done. Total $totalTuplesInserted tuples inserted.\n" );
 173                                                 $this->performanceLog( $totalTuplesInserted . " " . ( $this->getMicroTime() - $baseTime ) . "\n"  );
 174                                 }
 175                         }
 176                         $this->output( "$totalTuplesInserted valid titles and $numBadLinks invalid titles were processed.\n\n" );
 177                         $this->performanceLog( "$totalTuplesInserted valid titles and $numBadLinks invalid titles were processed.\n" );
 178                         $this->performanceLog( "Total execution time: " . ( $this->getMicroTime() - $startTime ) . " seconds.\n" );
 179                         if ( $logPerformance ) { fclose ( $fh ); }
 180                 }
 181                 # --------------------------------------------------------------------
 182
 183                 if ( $overwriteLinksTable ) {
 184                         # Check for existing links_backup, and delete it if it exists.
 185                         $this->output( "Dropping backup links table if it exists..." );
 186                         $dbw->query( "DROP TABLE IF EXISTS $links_backup", DB_MASTER );
 187                         $this->output( " done.\n" );
 188
 189                         # Swap in the new table, and move old links table to links_backup
 190                         $this->output( "Swapping tables '$links' to '$links_backup'; '$links_temp' to '$links'..." );
 191                         $dbw->query( "RENAME TABLE links TO $links_backup, $links_temp TO $links", DB_MASTER );
 192                         $this->output( " done.\n\n" );
 193
 194                         $dbw->close();
 195                         $this->output( "Conversion complete. The old table remains at $links_backup;\n" );
 196                         $this->output( "delete at your leisure.\n" );
 197                 } else {
 198                         $this->output( "Conversion complete.  The converted table is at $links_temp;\n" );
 199                         $this->output( "the original links table is unchanged.\n" );
 200                 }
 201         }
 202
 203         private function createTempTable() {
 204                 global $noKeys;
 205                 $dbConn = wfGetDB( DB_MASTER );
 206
 207                 if ( !( $dbConn->isOpen() ) ) {
 208                         $this->output( "Opening connection to database failed.\n" );
 209                         return;
 210                 }
 211                 $links_temp = $dbConn->tableName( 'links_temp' );
 212
 213                 $this->output( "Dropping temporary links table if it exists..." );
 214                 $dbConn->query( "DROP TABLE IF EXISTS $links_temp" );
 215                 $this->output( " done.\n" );
 216
 217                 $this->output( "Creating temporary links table..." );
 218                 if ( $noKeys ) {
 219                         $dbConn->query( "CREATE TABLE $links_temp ( " .
 220                         "l_from int(8) unsigned NOT NULL default '0', " .
 221                         "l_to int(8) unsigned NOT NULL default '0')" );
 222                 } else {
 223                         $dbConn->query( "CREATE TABLE $links_temp ( " .
 224                         "l_from int(8) unsigned NOT NULL default '0', " .
 225                         "l_to int(8) unsigned NOT NULL default '0', " .
 226                         "UNIQUE KEY l_from(l_from,l_to), " .
 227                         "KEY (l_to))" );
 228                 }
 229                 $this->output( " done.\n\n" );
 230         }
 231
 232         private function performanceLog( $text ) {
 233                 global $logPerformance, $fh;
 234                 if ( $logPerformance ) {
 235                         fwrite( $fh, $text );
 236                 }
 237         }
 238
 239         private function getMicroTime() { # return time in seconds, with microsecond accuracy
 240                 list( $usec, $sec ) = explode( " ", microtime() );
 241                 return ( (float)$usec + (float)$sec );
 242         }
 243 }
 244
 245 $maintClass = "ConvertLinks";
 246 require_once( DO_MAINTENANCE );