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