maintenance/archives/upgradeLogging.php

   1 <?php
   2 /**
   3  * Replication-safe online upgrade for log_id/log_deleted fields.
   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 __DIR__ . '/../commandLine.inc';
  25
  26 use Wikimedia\Rdbms\IMaintainableDatabase;
  27
  28 /**
  29  * Maintenance script that upgrade for log_id/log_deleted fields in a
  30  * replication-safe way.
  31  *
  32  * @ingroup Maintenance
  33  */
  34 class UpdateLogging {
  35
  36         /**
  37          * @var IMaintainableDatabase
  38          */
  39         public $dbw;
  40         public $batchSize = 1000;
  41         public $minTs = false;
  42
  43         function execute() {
  44                 $this->dbw = $this->getDB( DB_MASTER );
  45                 $logging = $this->dbw->tableName( 'logging' );
  46                 $logging_1_10 = $this->dbw->tableName( 'logging_1_10' );
  47                 $logging_pre_1_10 = $this->dbw->tableName( 'logging_pre_1_10' );
  48
  49                 if ( $this->dbw->tableExists( 'logging_pre_1_10' ) && !$this->dbw->tableExists( 'logging' ) ) {
  50                         # Fix previous aborted run
  51                         echo "Cleaning up from previous aborted run\n";
  52                         $this->dbw->query( "RENAME TABLE $logging_pre_1_10 TO $logging", __METHOD__ );
  53                 }
  54
  55                 if ( $this->dbw->tableExists( 'logging_pre_1_10' ) ) {
  56                         echo "This script has already been run to completion\n";
  57
  58                         return;
  59                 }
  60
  61                 # Create the target table
  62                 if ( !$this->dbw->tableExists( 'logging_1_10' ) ) {
  63                         global $wgDBTableOptions;
  64
  65                         $sql = <<<EOT
  66 CREATE TABLE $logging_1_10 (
  67   -- Log ID, for referring to this specific log entry, probably for deletion and such.
  68   log_id int unsigned NOT NULL auto_increment,
  69
  70   -- Symbolic keys for the general log type and the action type
  71   -- within the log. The output format will be controlled by the
  72   -- action field, but only the type controls categorization.
  73   log_type varbinary(10) NOT NULL default '',
  74   log_action varbinary(10) NOT NULL default '',
  75
  76   -- Timestamp. Duh.
  77   log_timestamp binary(14) NOT NULL default '19700101000000',
  78
  79   -- The user who performed this action; key to user_id
  80   log_user int unsigned NOT NULL default 0,
  81
  82   -- Key to the page affected. Where a user is the target,
  83   -- this will point to the user page.
  84   log_namespace int NOT NULL default 0,
  85   log_title varchar(255) binary NOT NULL default '',
  86
  87   -- Freeform text. Interpreted as edit history comments.
  88   log_comment varchar(255) NOT NULL default '',
  89
  90   -- LF separated list of miscellaneous parameters
  91   log_params blob NOT NULL,
  92
  93   -- rev_deleted for logs
  94   log_deleted tinyint unsigned NOT NULL default '0',
  95
  96   PRIMARY KEY log_id (log_id),
  97   KEY type_time (log_type, log_timestamp),
  98   KEY user_time (log_user, log_timestamp),
  99   KEY page_time (log_namespace, log_title, log_timestamp),
 100   KEY times (log_timestamp)
 101
 102 ) $wgDBTableOptions
 103 EOT;
 104                         echo "Creating table logging_1_10\n";
 105                         $this->dbw->query( $sql, __METHOD__ );
 106                 }
 107
 108                 # Synchronise the tables
 109                 echo "Doing initial sync...\n";
 110                 $this->sync( 'logging', 'logging_1_10' );
 111                 echo "Sync done\n\n";
 112
 113                 # Rename the old table away
 114                 echo "Renaming the old table to $logging_pre_1_10\n";
 115                 $this->dbw->query( "RENAME TABLE $logging TO $logging_pre_1_10", __METHOD__ );
 116
 117                 # Copy remaining old rows
 118                 # Done before the new table is active so that $copyPos is accurate
 119                 echo "Doing final sync...\n";
 120                 $this->sync( 'logging_pre_1_10', 'logging_1_10' );
 121
 122                 # Move the new table in
 123                 echo "Moving the new table in...\n";
 124                 $this->dbw->query( "RENAME TABLE $logging_1_10 TO $logging", __METHOD__ );
 125                 echo "Finished.\n";
 126         }
 127
 128         /**
 129          * Copy all rows from $srcTable to $dstTable
 130          * @param string $srcTable
 131          * @param string $dstTable
 132          */
 133         function sync( $srcTable, $dstTable ) {
 134                 $batchSize = 1000;
 135                 $minTs = $this->dbw->selectField( $srcTable, 'MIN(log_timestamp)', '', __METHOD__ );
 136                 $minTsUnix = wfTimestamp( TS_UNIX, $minTs );
 137                 $numRowsCopied = 0;
 138
 139                 while ( true ) {
 140                         $maxTs = $this->dbw->selectField( $srcTable, 'MAX(log_timestamp)', '', __METHOD__ );
 141                         $copyPos = $this->dbw->selectField( $dstTable, 'MAX(log_timestamp)', '', __METHOD__ );
 142                         $maxTsUnix = wfTimestamp( TS_UNIX, $maxTs );
 143                         $copyPosUnix = wfTimestamp( TS_UNIX, $copyPos );
 144
 145                         if ( $copyPos === null ) {
 146                                 $percent = 0;
 147                         } else {
 148                                 $percent = ( $copyPosUnix - $minTsUnix ) / ( $maxTsUnix - $minTsUnix ) * 100;
 149                         }
 150                         printf( "%s  %.2f%%\n", $copyPos, $percent );
 151
 152                         # Handle all entries with timestamp equal to $copyPos
 153                         if ( $copyPos !== null ) {
 154                                 $numRowsCopied += $this->copyExactMatch( $srcTable, $dstTable, $copyPos );
 155                         }
 156
 157                         # Now copy a batch of rows
 158                         if ( $copyPos === null ) {
 159                                 $conds = false;
 160                         } else {
 161                                 $conds = [ 'log_timestamp > ' . $this->dbw->addQuotes( $copyPos ) ];
 162                         }
 163                         $srcRes = $this->dbw->select( $srcTable, '*', $conds, __METHOD__,
 164                                 [ 'LIMIT' => $batchSize, 'ORDER BY' => 'log_timestamp' ] );
 165
 166                         if ( !$srcRes->numRows() ) {
 167                                 # All done
 168                                 break;
 169                         }
 170
 171                         $batch = [];
 172                         foreach ( $srcRes as $srcRow ) {
 173                                 $batch[] = (array)$srcRow;
 174                         }
 175                         $this->dbw->insert( $dstTable, $batch, __METHOD__ );
 176                         $numRowsCopied += count( $batch );
 177
 178                         wfWaitForSlaves();
 179                 }
 180                 echo "Copied $numRowsCopied rows\n";
 181         }
 182
 183         function copyExactMatch( $srcTable, $dstTable, $copyPos ) {
 184                 $numRowsCopied = 0;
 185                 $srcRes = $this->dbw->select( $srcTable, '*', [ 'log_timestamp' => $copyPos ], __METHOD__ );
 186                 $dstRes = $this->dbw->select( $dstTable, '*', [ 'log_timestamp' => $copyPos ], __METHOD__ );
 187
 188                 if ( $srcRes->numRows() ) {
 189                         $srcRow = $srcRes->fetchObject();
 190                         $srcFields = array_keys( (array)$srcRow );
 191                         $srcRes->seek( 0 );
 192                         $dstRowsSeen = [];
 193
 194                         # Make a hashtable of rows that already exist in the destination
 195                         foreach ( $dstRes as $dstRow ) {
 196                                 $reducedDstRow = [];
 197                                 foreach ( $srcFields as $field ) {
 198                                         $reducedDstRow[$field] = $dstRow->$field;
 199                                 }
 200                                 $hash = md5( serialize( $reducedDstRow ) );
 201                                 $dstRowsSeen[$hash] = true;
 202                         }
 203
 204                         # Copy all the source rows that aren't already in the destination
 205                         foreach ( $srcRes as $srcRow ) {
 206                                 $hash = md5( serialize( (array)$srcRow ) );
 207                                 if ( !isset( $dstRowsSeen[$hash] ) ) {
 208                                         $this->dbw->insert( $dstTable, (array)$srcRow, __METHOD__ );
 209                                         $numRowsCopied++;
 210                                 }
 211                         }
 212                 }
 213
 214                 return $numRowsCopied;
 215         }
 216 }
 217
 218 $ul = new UpdateLogging;
 219 $ul->execute();