maintenance/archives/upgradeLogging.php

   1 <?php
   2 /**
   3  * Replication-safe online upgrade script for log_id/log_deleted
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup MaintenanceArchive
  22  */
  23
  24 require( dirname( __FILE__ ) . '/../commandLine.inc' );
  25
  26 class UpdateLogging {
  27         var $dbw;
  28         var $batchSize = 1000;
  29         var $minTs = false;
  30
  31         function execute() {
  32                 $this->dbw = wfGetDB( DB_MASTER );
  33                 $logging = $this->dbw->tableName( 'logging' );
  34                 $logging_1_10 = $this->dbw->tableName( 'logging_1_10' );
  35                 $logging_pre_1_10 = $this->dbw->tableName( 'logging_pre_1_10' );
  36
  37                 if ( $this->dbw->tableExists( 'logging_pre_1_10' ) && !$this->dbw->tableExists( 'logging' ) ) {
  38                         # Fix previous aborted run
  39                         echo "Cleaning up from previous aborted run\n";
  40                         $this->dbw->query( "RENAME TABLE $logging_pre_1_10 TO $logging", __METHOD__ );
  41                 }
  42
  43                 if ( $this->dbw->tableExists( 'logging_pre_1_10' ) ) {
  44                         echo "This script has already been run to completion\n";
  45                         return;
  46                 }
  47
  48                 # Create the target table
  49                 if ( !$this->dbw->tableExists( 'logging_1_10' ) ) {
  50                         global $wgDBTableOptions;
  51
  52                         $sql = <<<EOT
  53 CREATE TABLE $logging_1_10 (
  54   -- Log ID, for referring to this specific log entry, probably for deletion and such.
  55   log_id int unsigned NOT NULL auto_increment,
  56
  57   -- Symbolic keys for the general log type and the action type
  58   -- within the log. The output format will be controlled by the
  59   -- action field, but only the type controls categorization.
  60   log_type varbinary(10) NOT NULL default '',
  61   log_action varbinary(10) NOT NULL default '',
  62
  63   -- Timestamp. Duh.
  64   log_timestamp binary(14) NOT NULL default '19700101000000',
  65
  66   -- The user who performed this action; key to user_id
  67   log_user int unsigned NOT NULL default 0,
  68
  69   -- Key to the page affected. Where a user is the target,
  70   -- this will point to the user page.
  71   log_namespace int NOT NULL default 0,
  72   log_title varchar(255) binary NOT NULL default '',
  73
  74   -- Freeform text. Interpreted as edit history comments.
  75   log_comment varchar(255) NOT NULL default '',
  76
  77   -- LF separated list of miscellaneous parameters
  78   log_params blob NOT NULL,
  79
  80   -- rev_deleted for logs
  81   log_deleted tinyint unsigned NOT NULL default '0',
  82
  83   PRIMARY KEY log_id (log_id),
  84   KEY type_time (log_type, log_timestamp),
  85   KEY user_time (log_user, log_timestamp),
  86   KEY page_time (log_namespace, log_title, log_timestamp),
  87   KEY times (log_timestamp)
  88
  89 ) $wgDBTableOptions
  90 EOT;
  91                         echo "Creating table logging_1_10\n";
  92                         $this->dbw->query( $sql, __METHOD__ );
  93                 }
  94
  95                 # Synchronise the tables
  96                 echo "Doing initial sync...\n";
  97                 $this->sync( 'logging', 'logging_1_10' );
  98                 echo "Sync done\n\n";
  99
 100                 # Rename the old table away
 101                 echo "Renaming the old table to $logging_pre_1_10\n";
 102                 $this->dbw->query( "RENAME TABLE $logging TO $logging_pre_1_10", __METHOD__ );
 103
 104                 # Copy remaining old rows
 105                 # Done before the new table is active so that $copyPos is accurate
 106                 echo "Doing final sync...\n";
 107                 $this->sync( 'logging_pre_1_10', 'logging_1_10' );
 108
 109                 # Move the new table in
 110                 echo "Moving the new table in...\n";
 111                 $this->dbw->query( "RENAME TABLE $logging_1_10 TO $logging", __METHOD__ );
 112                 echo "Finished.\n";
 113         }
 114
 115         /**
 116          * Copy all rows from $srcTable to $dstTable
 117          */
 118         function sync( $srcTable, $dstTable ) {
 119                 $batchSize = 1000;
 120                 $minTs = $this->dbw->selectField( $srcTable, 'MIN(log_timestamp)', false, __METHOD__ );
 121                 $minTsUnix = wfTimestamp( TS_UNIX, $minTs );
 122                 $numRowsCopied = 0;
 123
 124                 while ( true ) {
 125                         $maxTs = $this->dbw->selectField( $srcTable, 'MAX(log_timestamp)', false, __METHOD__ );
 126                         $copyPos = $this->dbw->selectField( $dstTable, 'MAX(log_timestamp)', false, __METHOD__ );
 127                         $maxTsUnix = wfTimestamp( TS_UNIX, $maxTs );
 128                         $copyPosUnix = wfTimestamp( TS_UNIX, $copyPos );
 129
 130                         if ( $copyPos === null ) {
 131                                 $percent = 0;
 132                         } else {
 133                                 $percent = ( $copyPosUnix - $minTsUnix ) / ( $maxTsUnix - $minTsUnix ) * 100;
 134                         }
 135                         printf( "%s  %.2f%%\n", $copyPos, $percent );
 136
 137                         # Handle all entries with timestamp equal to $copyPos
 138                         if ( $copyPos !== null ) {
 139                                 $numRowsCopied += $this->copyExactMatch( $srcTable, $dstTable, $copyPos );
 140                         }
 141
 142                         # Now copy a batch of rows
 143                         if ( $copyPos === null ) {
 144                                 $conds = false;
 145                         } else {
 146                                 $conds = array( 'log_timestamp > ' . $this->dbw->addQuotes( $copyPos ) );
 147                         }
 148                         $srcRes = $this->dbw->select( $srcTable, '*', $conds, __METHOD__,
 149                                 array( 'LIMIT' => $batchSize, 'ORDER BY' => 'log_timestamp' ) );
 150
 151                         if ( ! $srcRes->numRows() ) {
 152                                 # All done
 153                                 break;
 154                         }
 155
 156                         $batch = array();
 157                         foreach ( $srcRes as $srcRow ) {
 158                                 $batch[] = (array)$srcRow;
 159                         }
 160                         $this->dbw->insert( $dstTable, $batch, __METHOD__ );
 161                         $numRowsCopied += count( $batch );
 162
 163                         wfWaitForSlaves( 5 );
 164                 }
 165                 echo "Copied $numRowsCopied rows\n";
 166         }
 167
 168         function copyExactMatch( $srcTable, $dstTable, $copyPos ) {
 169                 $numRowsCopied = 0;
 170                 $srcRes = $this->dbw->select( $srcTable, '*', array( 'log_timestamp' => $copyPos ), __METHOD__ );
 171                 $dstRes = $this->dbw->select( $dstTable, '*', array( 'log_timestamp' => $copyPos ), __METHOD__ );
 172
 173                 if ( $srcRes->numRows() ) {
 174                         $srcRow = $srcRes->fetchObject();
 175                         $srcFields = array_keys( (array)$srcRow );
 176                         $srcRes->seek( 0 );
 177                         $dstRowsSeen = array();
 178
 179                         # Make a hashtable of rows that already exist in the destination
 180                         foreach ( $dstRes as $dstRow ) {
 181                                 $reducedDstRow = array();
 182                                 foreach ( $srcFields as $field ) {
 183                                         $reducedDstRow[$field] = $dstRow->$field;
 184                                 }
 185                                 $hash = md5( serialize( $reducedDstRow ) );
 186                                 $dstRowsSeen[$hash] = true;
 187                         }
 188
 189                         # Copy all the source rows that aren't already in the destination
 190                         foreach ( $srcRes as $srcRow ) {
 191                                 $hash = md5( serialize( (array)$srcRow ) );
 192                                 if ( !isset( $dstRowsSeen[$hash] ) ) {
 193                                         $this->dbw->insert( $dstTable, (array)$srcRow, __METHOD__ );
 194                                         $numRowsCopied++;
 195                                 }
 196                         }
 197                 }
 198                 return $numRowsCopied;
 199         }
 200 }
 201
 202 $ul = new UpdateLogging;
 203 $ul->execute();
 204