3 * Fills the rev_sha1 and ar_sha1 columns of revision
4 * and archive tables for revisions created before MW 1.19.
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.
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.
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
22 * @ingroup Maintenance
25 require_once __DIR__
. '/Maintenance.php';
27 use MediaWiki\MediaWikiServices
;
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.
33 * @ingroup Maintenance
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 );
42 protected function getUpdateKey() {
43 return 'populate rev_sha1';
46 protected function doDBUpdates() {
47 $db = $this->getDB( DB_MASTER
);
49 if ( !$db->tableExists( 'revision' ) ) {
50 $this->fatalError( "revision table does not exist" );
51 } elseif ( !$db->tableExists( 'archive' ) ) {
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 );
58 $revStore = MediaWikiServices
::getInstance()->getRevisionStore();
60 $this->output( "Populating rev_sha1 column\n" );
61 $rc = $this->doSha1Updates( $revStore, 'revision', 'rev_id',
62 $revStore->getQueryInfo(), 'rev'
65 $this->output( "Populating ar_sha1 column\n" );
66 $ac = $this->doSha1Updates( $revStore, 'archive', 'ar_rev_id',
67 $revStore->getArchiveQueryInfo(), 'ar'
69 $this->output( "Populating ar_sha1 column legacy rows\n" );
70 $ac +
= $this->doSha1LegacyUpdates( $revStore );
72 $this->output( "rev_sha1 and ar_sha1 population complete "
73 . "[$rc revision rows, $ac archive rows].\n" );
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
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" );
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" );
105 $cond = "$idCol BETWEEN " . (int)$blockStart . " AND " . (int)$blockEnd .
106 " AND $idCol IS NOT NULL AND {$prefix}_sha1 = ''";
108 $queryInfo['tables'], $queryInfo['fields'], $cond, __METHOD__
, [], $queryInfo['joins']
111 $this->beginTransaction( $db, __METHOD__
);
112 foreach ( $res as $row ) {
113 if ( $this->upgradeRow( $revStore, $row, $table, $idCol, $prefix ) ) {
117 $this->commitTransaction( $db, __METHOD__
);
119 $blockStart +
= $batchSize;
120 $blockEnd +
= $batchSize;
127 * @param MediaWiki\Revision\RevisionStore $revStore
128 * @param string $emptySha1
131 protected function doSha1LegacyUpdates( $revStore ) {
133 $db = $this->getDB( DB_MASTER
);
134 $arQuery = $revStore->getArchiveQueryInfo();
135 $res = $db->select( $arQuery['tables'], $arQuery['fields'],
136 [ 'ar_rev_id IS NULL', 'ar_sha1' => '' ], __METHOD__
, [], $arQuery['joins'] );
139 $this->beginTransaction( $db, __METHOD__
);
140 foreach ( $res as $row ) {
141 if ( $this->upgradeLegacyArchiveRow( $revStore, $row ) ) {
144 if ( ++
$updateSize >= 100 ) {
146 $this->commitTransaction( $db, __METHOD__
);
147 $this->output( "Commited row with ar_timestamp={$row->ar_timestamp}\n" );
148 $this->beginTransaction( $db, __METHOD__
);
151 $this->commitTransaction( $db, __METHOD__
);
157 * @param MediaWiki\Revision\RevisionStore $revStore
158 * @param stdClass $row
159 * @param string $table
160 * @param string $idCol
161 * @param string $prefix
164 protected function upgradeRow( $revStore, $row, $table, $idCol, $prefix ) {
165 $db = $this->getDB( DB_MASTER
);
167 // Create a revision and use it to get the sha1 from the content table, if possible.
169 $rev = ( $table === 'archive' )
170 ?
$revStore->newRevisionFromArchiveRow( $row )
171 : $revStore->newRevisionFromRow( $row );
172 $sha1 = $rev->getSha1();
173 } catch ( Exception
$e ) {
174 $this->output( "Data of revision with {$idCol}={$row->$idCol} unavailable!\n" );
175 return false; // T24624? T22757?
179 [ "{$prefix}_sha1" => $sha1 ],
180 [ $idCol => $row->$idCol ],
188 * @param MediaWiki\Revision\RevisionStore $revStore
189 * @param stdClass $row
192 protected function upgradeLegacyArchiveRow( $revStore, $row ) {
193 $db = $this->getDB( DB_MASTER
);
195 // Create a revision and use it to get the sha1 from the content table, if possible.
197 $rev = $revStore->newRevisionFromArchiveRow( $row );
198 $sha1 = $rev->getSha1();
199 } catch ( Exception
$e ) {
200 $this->output( "Text of revision with timestamp {$row->ar_timestamp} unavailable!\n" );
201 return false; // T24624? T22757?
204 # Archive table has no PK, but (NS,title,time) should be near unique.
205 # Any duplicates on those should also have duplicated text anyway.
206 $db->update( 'archive',
207 [ 'ar_sha1' => $sha1 ],
209 'ar_namespace' => $row->ar_namespace
,
210 'ar_title' => $row->ar_title
,
211 'ar_timestamp' => $row->ar_timestamp
,
212 'ar_len' => $row->ar_len
// extra sanity
221 $maintClass = PopulateRevisionSha1
::class;
222 require_once RUN_MAINTENANCE_IF_MAIN
;