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
  28         /**
  29          * @var DatabaseBase
  30          */
  31         var $dbw;
  32         var $batchSize = 1000;
  33         var $minTs = false;
  34
  35         function execute() {
  36                 $this->dbw = wfGetDB( DB_MASTER );
  37                 $logging = $this->dbw->tableName( 'logging' );
  38                 $logging_1_10 = $this->dbw->tableName( 'logging_1_10' );
  39                 $logging_pre_1_10 = $this->dbw->tableName( 'logging_pre_1_10' );
  40
  41                 if ( $this->dbw->tableExists( 'logging_pre_1_10' ) && !$this->dbw->tableExists( 'logging' ) ) {
  42                         # Fix previous aborted run
  43                         echo "Cleaning up from previous aborted run\n";
  44                         $this->dbw->query( "RENAME TABLE $logging_pre_1_10 TO $logging", __METHOD__ );
  45                 }
  46
  47                 if ( $this->dbw->tableExists( 'logging_pre_1_10' ) ) {
  48                         echo "This script has already been run to completion\n";
  49                         return;
  50                 }
  51
  52                 # Create the target table
  53                 if ( !$this->dbw->tableExists( 'logging_1_10' ) ) {
  54                         global $wgDBTableOptions;
  55
  56                         $sql = <<<EOT
  57 CREATE TABLE $logging_1_10 (
  58   -- Log ID, for referring to this specific log entry, probably for deletion and such.
  59   log_id int unsigned NOT NULL auto_increment,
  60
  61   -- Symbolic keys for the general log type and the action type
  62   -- within the log. The output format will be controlled by the
  63   -- action field, but only the type controls categorization.
  64   log_type varbinary(10) NOT NULL default '',
  65   log_action varbinary(10) NOT NULL default '',
  66
  67   -- Timestamp. Duh.
  68   log_timestamp binary(14) NOT NULL default '19700101000000',
  69
  70   -- The user who performed this action; key to user_id
  71   log_user int unsigned NOT NULL default 0,
  72
  73   -- Key to the page affected. Where a user is the target,
  74   -- this will point to the user page.
  75   log_namespace int NOT NULL default 0,
  76   log_title varchar(255) binary NOT NULL default '',
  77
  78   -- Freeform text. Interpreted as edit history comments.
  79   log_comment varchar(255) NOT NULL default '',
  80
  81   -- LF separated list of miscellaneous parameters
  82   log_params blob NOT NULL,
  83
  84   -- rev_deleted for logs
  85   log_deleted tinyint unsigned NOT NULL default '0',
  86
  87   PRIMARY KEY log_id (log_id),
  88   KEY type_time (log_type, log_timestamp),
  89   KEY user_time (log_user, log_timestamp),
  90   KEY page_time (log_namespace, log_title, log_timestamp),
  91   KEY times (log_timestamp)
  92
  93 ) $wgDBTableOptions
  94 EOT;
  95                         echo "Creating table logging_1_10\n";
  96                         $this->dbw->query( $sql, __METHOD__ );
  97                 }
  98
  99                 # Synchronise the tables
 100                 echo "Doing initial sync...\n";
 101                 $this->sync( 'logging', 'logging_1_10' );
 102                 echo "Sync done\n\n";
 103
 104                 # Rename the old table away
 105                 echo "Renaming the old table to $logging_pre_1_10\n";
 106                 $this->dbw->query( "RENAME TABLE $logging TO $logging_pre_1_10", __METHOD__ );
 107
 108                 # Copy remaining old rows
 109                 # Done before the new table is active so that $copyPos is accurate
 110                 echo "Doing final sync...\n";
 111                 $this->sync( 'logging_pre_1_10', 'logging_1_10' );
 112
 113                 # Move the new table in
 114                 echo "Moving the new table in...\n";
 115                 $this->dbw->query( "RENAME TABLE $logging_1_10 TO $logging", __METHOD__ );
 116                 echo "Finished.\n";
 117         }
 118
 119         /**
 120          * Copy all rows from $srcTable to $dstTable
 121          */
 122         function sync( $srcTable, $dstTable ) {
 123                 $batchSize = 1000;
 124                 $minTs = $this->dbw->selectField( $srcTable, 'MIN(log_timestamp)', false, __METHOD__ );
 125                 $minTsUnix = wfTimestamp( TS_UNIX, $minTs );
 126                 $numRowsCopied = 0;
 127
 128                 while ( true ) {
 129                         $maxTs = $this->dbw->selectField( $srcTable, 'MAX(log_timestamp)', false, __METHOD__ );
 130                         $copyPos = $this->dbw->selectField( $dstTable, 'MAX(log_timestamp)', false, __METHOD__ );
 131                         $maxTsUnix = wfTimestamp( TS_UNIX, $maxTs );
 132                         $copyPosUnix = wfTimestamp( TS_UNIX, $copyPos );
 133
 134                         if ( $copyPos === null ) {
 135                                 $percent = 0;
 136                         } else {
 137                                 $percent = ( $copyPosUnix - $minTsUnix ) / ( $maxTsUnix - $minTsUnix ) * 100;
 138                         }
 139                         printf( "%s  %.2f%%\n", $copyPos, $percent );
 140
 141                         # Handle all entries with timestamp equal to $copyPos
 142                         if ( $copyPos !== null ) {
 143                                 $numRowsCopied += $this->copyExactMatch( $srcTable, $dstTable, $copyPos );
 144                         }
 145
 146                         # Now copy a batch of rows
 147                         if ( $copyPos === null ) {
 148                                 $conds = false;
 149                         } else {
 150                                 $conds = array( 'log_timestamp > ' . $this->dbw->addQuotes( $copyPos ) );
 151                         }
 152                         $srcRes = $this->dbw->select( $srcTable, '*', $conds, __METHOD__,
 153                                 array( 'LIMIT' => $batchSize, 'ORDER BY' => 'log_timestamp' ) );
 154
 155                         if ( ! $srcRes->numRows() ) {
 156                                 # All done
 157                                 break;
 158                         }
 159
 160                         $batch = array();
 161                         foreach ( $srcRes as $srcRow ) {
 162                                 $batch[] = (array)$srcRow;
 163                         }
 164                         $this->dbw->insert( $dstTable, $batch, __METHOD__ );
 165                         $numRowsCopied += count( $batch );
 166
 167                         wfWaitForSlaves();
 168                 }
 169                 echo "Copied $numRowsCopied rows\n";
 170         }
 171
 172         function copyExactMatch( $srcTable, $dstTable, $copyPos ) {
 173                 $numRowsCopied = 0;
 174                 $srcRes = $this->dbw->select( $srcTable, '*', array( 'log_timestamp' => $copyPos ), __METHOD__ );
 175                 $dstRes = $this->dbw->select( $dstTable, '*', array( 'log_timestamp' => $copyPos ), __METHOD__ );
 176
 177                 if ( $srcRes->numRows() ) {
 178                         $srcRow = $srcRes->fetchObject();
 179                         $srcFields = array_keys( (array)$srcRow );
 180                         $srcRes->seek( 0 );
 181                         $dstRowsSeen = array();
 182
 183                         # Make a hashtable of rows that already exist in the destination
 184                         foreach ( $dstRes as $dstRow ) {
 185                                 $reducedDstRow = array();
 186                                 foreach ( $srcFields as $field ) {
 187                                         $reducedDstRow[$field] = $dstRow->$field;
 188                                 }
 189                                 $hash = md5( serialize( $reducedDstRow ) );
 190                                 $dstRowsSeen[$hash] = true;
 191                         }
 192
 193                         # Copy all the source rows that aren't already in the destination
 194                         foreach ( $srcRes as $srcRow ) {
 195                                 $hash = md5( serialize( (array)$srcRow ) );
 196                                 if ( !isset( $dstRowsSeen[$hash] ) ) {
 197                                         $this->dbw->insert( $dstTable, (array)$srcRow, __METHOD__ );
 198                                         $numRowsCopied++;
 199                                 }
 200                         }
 201                 }
 202                 return $numRowsCopied;
 203         }
 204 }
 205
 206 $ul = new UpdateLogging;
 207 $ul->execute();
 208