maintenance/populateRevisionSha1.php

   1 <?php
   2 /**
   3  * Fills the rev_sha1 and ar_sha1 columns of revision
   4  * and archive tables for revisions created before MW 1.19.
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License along
  17  * with this program; if not, write to the Free Software Foundation, Inc.,
  18  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  19  * http://www.gnu.org/copyleft/gpl.html
  20  *
  21  * @file
  22  * @ingroup Maintenance
  23  */
  24
  25 require_once __DIR__ . '/Maintenance.php';
  26
  27 use MediaWiki\MediaWikiServices;
  28
  29 /**
  30  * Maintenance script that fills the rev_sha1 and ar_sha1 columns of revision
  31  * and archive tables for revisions created before MW 1.19.
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class PopulateRevisionSha1 extends LoggedUpdateMaintenance {
  36         public function __construct() {
  37                 parent::__construct();
  38                 $this->addDescription( 'Populates the rev_sha1 and ar_sha1 fields' );
  39                 $this->setBatchSize( 200 );
  40         }
  41
  42         protected function getUpdateKey() {
  43                 return 'populate rev_sha1';
  44         }
  45
  46         protected function doDBUpdates() {
  47                 $db = $this->getDB( DB_MASTER );
  48
  49                 if ( !$db->tableExists( 'revision', __METHOD__ ) ) {
  50                         $this->fatalError( "revision table does not exist" );
  51                 } elseif ( !$db->tableExists( 'archive', __METHOD__ ) ) {
  52                         $this->fatalError( "archive table does not exist" );
  53                 } elseif ( !$db->fieldExists( 'revision', 'rev_sha1', __METHOD__ ) ) {
  54                         $this->output( "rev_sha1 column does not exist\n\n", true );
  55                         return false;
  56                 }
  57
  58                 $revStore = MediaWikiServices::getInstance()->getRevisionStore();
  59
  60                 $this->output( "Populating rev_sha1 column\n" );
  61                 $rc = $this->doSha1Updates( $revStore, 'revision', 'rev_id',
  62                         $revStore->getQueryInfo(), 'rev'
  63                 );
  64
  65                 $this->output( "Populating ar_sha1 column\n" );
  66                 $ac = $this->doSha1Updates( $revStore, 'archive', 'ar_rev_id',
  67                         $revStore->getArchiveQueryInfo(), 'ar'
  68                 );
  69                 $this->output( "Populating ar_sha1 column legacy rows\n" );
  70                 $ac += $this->doSha1LegacyUpdates( $revStore );
  71
  72                 $this->output( "rev_sha1 and ar_sha1 population complete "
  73                         . "[$rc revision rows, $ac archive rows].\n" );
  74
  75                 return true;
  76         }
  77
  78         /**
  79          * @param MediaWiki\Revision\RevisionStore $revStore
  80          * @param string $table
  81          * @param string $idCol
  82          * @param array $queryInfo
  83          * @param string $prefix
  84          * @return int Rows changed
  85          */
  86         protected function doSha1Updates( $revStore, $table, $idCol, $queryInfo, $prefix ) {
  87                 $db = $this->getDB( DB_MASTER );
  88                 $batchSize = $this->getBatchSize();
  89                 $start = $db->selectField( $table, "MIN($idCol)", '', __METHOD__ );
  90                 $end = $db->selectField( $table, "MAX($idCol)", '', __METHOD__ );
  91                 if ( !$start || !$end ) {
  92                         $this->output( "...$table table seems to be empty.\n" );
  93
  94                         return 0;
  95                 }
  96
  97                 $count = 0;
  98                 # Do remaining chunk
  99                 $end += $batchSize - 1;
 100                 $blockStart = $start;
 101                 $blockEnd = $start + $batchSize - 1;
 102                 while ( $blockEnd <= $end ) {
 103                         $this->output( "...doing $idCol from $blockStart to $blockEnd\n" );
 104
 105                         $cond = "$idCol BETWEEN " . (int)$blockStart . " AND " . (int)$blockEnd .
 106                                 " AND $idCol IS NOT NULL AND {$prefix}_sha1 = ''";
 107                         $res = $db->select(
 108                                 $queryInfo['tables'], $queryInfo['fields'], $cond, __METHOD__, [], $queryInfo['joins']
 109                         );
 110
 111                         $this->beginTransaction( $db, __METHOD__ );
 112                         foreach ( $res as $row ) {
 113                                 if ( $this->upgradeRow( $revStore, $row, $table, $idCol, $prefix ) ) {
 114                                         $count++;
 115                                 }
 116                         }
 117                         $this->commitTransaction( $db, __METHOD__ );
 118
 119                         $blockStart += $batchSize;
 120                         $blockEnd += $batchSize;
 121                 }
 122
 123                 return $count;
 124         }
 125
 126         /**
 127          * @param MediaWiki\Revision\RevisionStore $revStore
 128          * @return int
 129          */
 130         protected function doSha1LegacyUpdates( $revStore ) {
 131                 $count = 0;
 132                 $db = $this->getDB( DB_MASTER );
 133                 $arQuery = $revStore->getArchiveQueryInfo();
 134                 $res = $db->select( $arQuery['tables'], $arQuery['fields'],
 135                         [ 'ar_rev_id IS NULL', 'ar_sha1' => '' ], __METHOD__, [], $arQuery['joins'] );
 136
 137                 $updateSize = 0;
 138                 $this->beginTransaction( $db, __METHOD__ );
 139                 foreach ( $res as $row ) {
 140                         if ( $this->upgradeLegacyArchiveRow( $revStore, $row ) ) {
 141                                 ++$count;
 142                         }
 143                         if ( ++$updateSize >= 100 ) {
 144                                 $updateSize = 0;
 145                                 $this->commitTransaction( $db, __METHOD__ );
 146                                 $this->output( "Commited row with ar_timestamp={$row->ar_timestamp}\n" );
 147                                 $this->beginTransaction( $db, __METHOD__ );
 148                         }
 149                 }
 150                 $this->commitTransaction( $db, __METHOD__ );
 151
 152                 return $count;
 153         }
 154
 155         /**
 156          * @param MediaWiki\Revision\RevisionStore $revStore
 157          * @param stdClass $row
 158          * @param string $table
 159          * @param string $idCol
 160          * @param string $prefix
 161          * @return bool
 162          */
 163         protected function upgradeRow( $revStore, $row, $table, $idCol, $prefix ) {
 164                 $db = $this->getDB( DB_MASTER );
 165
 166                 // Create a revision and use it to get the sha1 from the content table, if possible.
 167                 try {
 168                         $rev = ( $table === 'archive' )
 169                                 ? $revStore->newRevisionFromArchiveRow( $row )
 170                                 : $revStore->newRevisionFromRow( $row );
 171                         $sha1 = $rev->getSha1();
 172                 } catch ( Exception $e ) {
 173                         $this->output( "Data of revision with {$idCol}={$row->$idCol} unavailable!\n" );
 174                         return false; // T24624? T22757?
 175                 }
 176
 177                 $db->update( $table,
 178                         [ "{$prefix}_sha1" => $sha1 ],
 179                         [ $idCol => $row->$idCol ],
 180                         __METHOD__
 181                 );
 182
 183                 return true;
 184         }
 185
 186         /**
 187          * @param MediaWiki\Revision\RevisionStore $revStore
 188          * @param stdClass $row
 189          * @return bool
 190          */
 191         protected function upgradeLegacyArchiveRow( $revStore, $row ) {
 192                 $db = $this->getDB( DB_MASTER );
 193
 194                 // Create a revision and use it to get the sha1 from the content table, if possible.
 195                 try {
 196                         $rev = $revStore->newRevisionFromArchiveRow( $row );
 197                         $sha1 = $rev->getSha1();
 198                 } catch ( Exception $e ) {
 199                         $this->output( "Text of revision with timestamp {$row->ar_timestamp} unavailable!\n" );
 200                         return false; // T24624? T22757?
 201                 }
 202
 203                 # Archive table has no PK, but (NS,title,time) should be near unique.
 204                 # Any duplicates on those should also have duplicated text anyway.
 205                 $db->update( 'archive',
 206                         [ 'ar_sha1' => $sha1 ],
 207                         [
 208                                 'ar_namespace' => $row->ar_namespace,
 209                                 'ar_title' => $row->ar_title,
 210                                 'ar_timestamp' => $row->ar_timestamp,
 211                                 'ar_len' => $row->ar_len // extra sanity
 212                         ],
 213                         __METHOD__
 214                 );
 215
 216                 return true;
 217         }
 218 }
 219
 220 $maintClass = PopulateRevisionSha1::class;
 221 require_once RUN_MAINTENANCE_IF_MAIN;