maintenance/populateIpChanges.php

   1 <?php
   2 /**
   3  * Find all revisions by logged out users and copy the rev_id,
   4  * rev_timestamp, and a hex representation of rev_user_text to the
   5  * new ip_changes table. This table is used to efficiently query for
   6  * contributions within an IP range.
   7  *
   8  * This program is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with this program; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  21  * http://www.gnu.org/copyleft/gpl.html
  22  *
  23  * @file
  24  * @ingroup Maintenance
  25  */
  26
  27 require_once __DIR__ . '/Maintenance.php';
  28
  29 use MediaWiki\User\ActorMigration;
  30 use Wikimedia\IPUtils;
  31
  32 /**
  33  * Maintenance script that will find all rows in the revision table where
  34  * rev_user = 0 (user is an IP), and copy relevant fields to ip_changes so
  35  * that historical data will be available when querying for IP ranges.
  36  *
  37  * @ingroup Maintenance
  38  */
  39 class PopulateIpChanges extends LoggedUpdateMaintenance {
  40         public function __construct() {
  41                 parent::__construct();
  42
  43                 $this->addDescription( <<<TEXT
  44 This script will find all rows in the revision table where the user is an IP,
  45 and copy relevant fields to the ip_changes table. This backfilled data will
  46 then be available when querying for IP ranges at Special:Contributions.
  47 TEXT
  48                 );
  49                 $this->addOption( 'rev-id', 'The rev_id to start copying from. Default: 0', false, true );
  50                 $this->addOption(
  51                         'max-rev-id',
  52                         'The rev_id to stop at. Default: result of MAX(rev_id)',
  53                         false,
  54                         true
  55                 );
  56                 $this->addOption(
  57                         'throttle',
  58                         'Wait this many milliseconds after copying each batch of revisions. Default: 0',
  59                         false,
  60                         true
  61                 );
  62                 $this->addOption( 'force', 'Run regardless of whether the database says it\'s been run already' );
  63         }
  64
  65         public function doDBUpdates() {
  66                 $dbw = $this->getDB( DB_PRIMARY );
  67
  68                 if ( !$dbw->tableExists( 'ip_changes', __METHOD__ ) ) {
  69                         $this->fatalError( 'ip_changes table does not exist' );
  70                 }
  71
  72                 $dbr = $this->getDB( DB_REPLICA, [ 'vslow' ] );
  73                 $throttle = intval( $this->getOption( 'throttle', 0 ) );
  74                 $maxRevId = intval( $this->getOption( 'max-rev-id', 0 ) );
  75                 $start = $this->getOption( 'rev-id', 0 );
  76                 $end = $maxRevId > 0
  77                         ? $maxRevId
  78                         : $dbw->newSelectQueryBuilder()
  79                                 ->select( 'MAX(rev_id)' )
  80                                 ->from( 'revision' )
  81                                 ->caller( __METHOD__ )->fetchField();
  82
  83                 if ( !$end ) {
  84                         $this->output( "No revisions found, aborting.\n" );
  85                         return true;
  86                 }
  87
  88                 $blockStart = $start;
  89                 $attempted = 0;
  90                 $inserted = 0;
  91
  92                 $this->output( "Copying IP revisions to ip_changes, from rev_id $start to rev_id $end\n" );
  93
  94                 $actorMigration = ActorMigration::newMigration();
  95                 $actorQuery = $actorMigration->getJoin( 'rev_user' );
  96                 $revUserIsAnon = $actorMigration->isAnon( $actorQuery['fields']['rev_user'] );
  97
  98                 while ( $blockStart <= $end ) {
  99                         $blockEnd = min( $blockStart + $this->getBatchSize(), $end );
 100                         $rows = $dbr->select(
 101                                 [ 'revision' ] + $actorQuery['tables'],
 102                                 [ 'rev_id', 'rev_timestamp', 'rev_user_text' => $actorQuery['fields']['rev_user_text'] ],
 103                                 [ "rev_id BETWEEN " . (int)$blockStart . " AND " . (int)$blockEnd, $revUserIsAnon ],
 104                                 __METHOD__,
 105                                 [],
 106                                 $actorQuery['joins']
 107                         );
 108
 109                         $numRows = $rows->numRows();
 110
 111                         if ( !$rows || $numRows === 0 ) {
 112                                 $blockStart = $blockEnd + 1;
 113                                 continue;
 114                         }
 115
 116                         $this->output( "...checking $numRows revisions for IP edits that need copying, " .
 117                                 "between rev_ids $blockStart and $blockEnd\n" );
 118
 119                         $insertRows = [];
 120                         foreach ( $rows as $row ) {
 121                                 // Make sure this is really an IP, e.g. not maintenance user or imported revision.
 122                                 if ( IPUtils::isValid( $row->rev_user_text ) ) {
 123                                         $insertRows[] = [
 124                                                 'ipc_rev_id' => $row->rev_id,
 125                                                 'ipc_rev_timestamp' => $row->rev_timestamp,
 126                                                 'ipc_hex' => IPUtils::toHex( $row->rev_user_text ),
 127                                         ];
 128
 129                                         $attempted++;
 130                                 }
 131                         }
 132
 133                         if ( $insertRows ) {
 134                                 $dbw->insert( 'ip_changes', $insertRows, __METHOD__, [ 'IGNORE' ] );
 135
 136                                 $inserted += $dbw->affectedRows();
 137                         }
 138
 139                         $this->waitForReplication();
 140                         usleep( $throttle * 1000 );
 141
 142                         $blockStart = $blockEnd + 1;
 143                 }
 144
 145                 $this->output( "Attempted to insert $attempted IP revisions, $inserted actually done.\n" );
 146
 147                 return true;
 148         }
 149
 150         protected function getUpdateKey() {
 151                 return 'populate ip_changes';
 152         }
 153 }
 154
 155 $maintClass = PopulateIpChanges::class;
 156 require_once RUN_MAINTENANCE_IF_MAIN;