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\MediaWikiServices;
  30
  31 /**
  32  * Maintenance script that will find all rows in the revision table where
  33  * rev_user = 0 (user is an IP), and copy relevant fields to ip_changes so
  34  * that historical data will be available when querying for IP ranges.
  35  *
  36  * @ingroup Maintenance
  37  */
  38 class PopulateIpChanges extends LoggedUpdateMaintenance {
  39         public function __construct() {
  40                 parent::__construct();
  41
  42                 $this->addDescription( <<<TEXT
  43 This script will find all rows in the revision table where the user is an IP,
  44 and copy relevant fields to the ip_changes table. This backfilled data will
  45 then be available when querying for IP ranges at Special:Contributions.
  46 TEXT
  47                 );
  48                 $this->addOption( 'rev-id', 'The rev_id to start copying from. Default: 0', false, true );
  49                 $this->addOption(
  50                         'max-rev-id',
  51                         'The rev_id to stop at. Default: result of MAX(rev_id)',
  52                         false,
  53                         true
  54                 );
  55                 $this->addOption(
  56                         'throttle',
  57                         'Wait this many milliseconds after copying each batch of revisions. Default: 0',
  58                         false,
  59                         true
  60                 );
  61                 $this->addOption( 'force', 'Run regardless of whether the database says it\'s been run already' );
  62         }
  63
  64         public function doDBUpdates() {
  65                 $dbw = $this->getDB( DB_MASTER );
  66
  67                 if ( !$dbw->tableExists( 'ip_changes' ) ) {
  68                         $this->fatalError( 'ip_changes table does not exist' );
  69                 }
  70
  71                 $lbFactory = MediaWikiServices::getInstance()->getDBLoadBalancerFactory();
  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->selectField( 'revision', 'MAX(rev_id)', false, __METHOD__ );
  79
  80                 if ( empty( $end ) ) {
  81                         $this->output( "No revisions found, aborting.\n" );
  82                         return true;
  83                 }
  84
  85                 $blockStart = $start;
  86                 $attempted = 0;
  87                 $inserted = 0;
  88
  89                 $this->output( "Copying IP revisions to ip_changes, from rev_id $start to rev_id $end\n" );
  90
  91                 $actorMigration = ActorMigration::newMigration();
  92                 $actorQuery = $actorMigration->getJoin( 'rev_user' );
  93                 $revUserIsAnon = $actorMigration->isAnon( $actorQuery['fields']['rev_user'] );
  94
  95                 while ( $blockStart <= $end ) {
  96                         $blockEnd = min( $blockStart + $this->getBatchSize(), $end );
  97                         $rows = $dbr->select(
  98                                 [ 'revision' ] + $actorQuery['tables'],
  99                                 [ 'rev_id', 'rev_timestamp', 'rev_user_text' => $actorQuery['fields']['rev_user_text'] ],
 100                                 [ "rev_id BETWEEN " . (int)$blockStart . " AND " . (int)$blockEnd, $revUserIsAnon ],
 101                                 __METHOD__,
 102                                 [],
 103                                 $actorQuery['joins']
 104                         );
 105
 106                         $numRows = $rows->numRows();
 107
 108                         if ( !$rows || $numRows === 0 ) {
 109                                 $blockStart = $blockEnd + 1;
 110                                 continue;
 111                         }
 112
 113                         $this->output( "...checking $numRows revisions for IP edits that need copying, " .
 114                                 "between rev_ids $blockStart and $blockEnd\n" );
 115
 116                         $insertRows = [];
 117                         foreach ( $rows as $row ) {
 118                                 // Make sure this is really an IP, e.g. not maintenance user or imported revision.
 119                                 if ( IP::isValid( $row->rev_user_text ) ) {
 120                                         $insertRows[] = [
 121                                                 'ipc_rev_id' => $row->rev_id,
 122                                                 'ipc_rev_timestamp' => $row->rev_timestamp,
 123                                                 'ipc_hex' => IP::toHex( $row->rev_user_text ),
 124                                         ];
 125
 126                                         $attempted++;
 127                                 }
 128                         }
 129
 130                         if ( $insertRows ) {
 131                                 $dbw->insert( 'ip_changes', $insertRows, __METHOD__, 'IGNORE' );
 132
 133                                 $inserted += $dbw->affectedRows();
 134                         }
 135
 136                         $lbFactory->waitForReplication();
 137                         usleep( $throttle * 1000 );
 138
 139                         $blockStart = $blockEnd + 1;
 140                 }
 141
 142                 $this->output( "Attempted to insert $attempted IP revisions, $inserted actually done.\n" );
 143
 144                 return true;
 145         }
 146
 147         protected function getUpdateKey() {
 148                 return 'populate ip_changes';
 149         }
 150 }
 151
 152 $maintClass = PopulateIpChanges::class;
 153 require_once RUN_MAINTENANCE_IF_MAIN;