maintenance/populateImageSha1.php

   1 <?php
   2 /**
   3  * Optional upgrade script to populate the img_sha1 field
   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 Maintenance
  22  */
  23
  24 use MediaWiki\MediaWikiServices;
  25 use MediaWiki\Shell\Shell;
  26
  27 require_once __DIR__ . '/Maintenance.php';
  28
  29 /**
  30  * Maintenance script to populate the img_sha1 field.
  31  *
  32  * @ingroup Maintenance
  33  */
  34 class PopulateImageSha1 extends LoggedUpdateMaintenance {
  35         public function __construct() {
  36                 parent::__construct();
  37                 $this->addDescription( 'Populate the img_sha1 field' );
  38                 $this->addOption( 'force', "Recalculate sha1 for rows that already have a value" );
  39                 $this->addOption( 'multiversiononly', "Calculate only for files with several versions" );
  40                 $this->addOption( 'method', "Use 'pipe' to pipe to mysql command line,\n" .
  41                         "\t\tdefault uses Database class", false, true );
  42                 $this->addOption(
  43                         'file',
  44                         'Fix for a specific file, without File: namespace prefixed',
  45                         false,
  46                         true
  47                 );
  48         }
  49
  50         protected function getUpdateKey() {
  51                 return 'populate img_sha1';
  52         }
  53
  54         protected function updateSkippedMessage() {
  55                 return 'img_sha1 column of image table already populated.';
  56         }
  57
  58         public function execute() {
  59                 if ( $this->getOption( 'file' ) || $this->hasOption( 'multiversiononly' ) ) {
  60                         $this->doDBUpdates(); // skip update log checks/saves
  61                 } else {
  62                         parent::execute();
  63                 }
  64         }
  65
  66         public function doDBUpdates() {
  67                 $method = $this->getOption( 'method', 'normal' );
  68                 $file = $this->getOption( 'file', '' );
  69                 $force = $this->getOption( 'force' );
  70                 $isRegen = ( $force || $file != '' ); // forced recalculation?
  71
  72                 $t = -microtime( true );
  73                 $dbw = $this->getDB( DB_MASTER );
  74                 if ( $file != '' ) {
  75                         $res = $dbw->select(
  76                                 'image',
  77                                 [ 'img_name' ],
  78                                 [ 'img_name' => $file ],
  79                                 __METHOD__
  80                         );
  81                         if ( !$res ) {
  82                                 $this->fatalError( "No such file: $file" );
  83                         }
  84                         $this->output( "Populating img_sha1 field for specified files\n" );
  85                 } else {
  86                         if ( $this->hasOption( 'multiversiononly' ) ) {
  87                                 $conds = [];
  88                                 $this->output( "Populating and recalculating img_sha1 field for versioned files\n" );
  89                         } elseif ( $force ) {
  90                                 $conds = [];
  91                                 $this->output( "Populating and recalculating img_sha1 field\n" );
  92                         } else {
  93                                 $conds = [ 'img_sha1' => '' ];
  94                                 $this->output( "Populating img_sha1 field\n" );
  95                         }
  96                         if ( $this->hasOption( 'multiversiononly' ) ) {
  97                                 $res = $dbw->select( 'oldimage',
  98                                         [ 'img_name' => 'DISTINCT(oi_name)' ], $conds, __METHOD__ );
  99                         } else {
 100                                 $res = $dbw->select( 'image', [ 'img_name' ], $conds, __METHOD__ );
 101                         }
 102                 }
 103
 104                 $imageTable = $dbw->tableName( 'image' );
 105                 $oldImageTable = $dbw->tableName( 'oldimage' );
 106
 107                 if ( $method == 'pipe' ) {
 108                         // Opening a pipe allows the SHA-1 operation to be done in parallel
 109                         // with the database write operation, because the writes are queued
 110                         // in the pipe buffer. This can improve performance by up to a
 111                         // factor of 2.
 112                         $config = $this->getConfig();
 113                         $cmd = 'mysql -u' . Shell::escape( $config->get( 'DBuser' ) ) .
 114                                 ' -h' . Shell::escape( $config->get( 'DBserver' ) ) .
 115                                 ' -p' . Shell::escape( $config->get( 'DBpassword' ), $config->get( 'DBname' ) );
 116                         $this->output( "Using pipe method\n" );
 117                         $pipe = popen( $cmd, 'w' );
 118                 }
 119
 120                 $numRows = $res->numRows();
 121                 $i = 0;
 122                 $lbFactory = MediaWikiServices::getInstance()->getDBLoadBalancerFactory();
 123                 foreach ( $res as $row ) {
 124                         if ( $i % $this->getBatchSize() == 0 ) {
 125                                 $this->output( sprintf(
 126                                         "Done %d of %d, %5.3f%%  \r", $i, $numRows, $i / $numRows * 100 ) );
 127                                 $lbFactory->waitForReplication();
 128                         }
 129
 130                         $file = MediaWikiServices::getInstance()->getRepoGroup()->getLocalRepo()
 131                                 ->newFile( $row->img_name );
 132                         if ( !$file ) {
 133                                 continue;
 134                         }
 135
 136                         // Upgrade the current file version...
 137                         $sha1 = $file->getRepo()->getFileSha1( $file->getPath() );
 138                         if ( strval( $sha1 ) !== '' ) { // file on disk and hashed properly
 139                                 if ( $isRegen && $file->getSha1() !== $sha1 ) {
 140                                         // The population was probably done already. If the old SHA1
 141                                         // does not match, then both fix the SHA1 and the metadata.
 142                                         $file->upgradeRow();
 143                                 } else {
 144                                         $sql = "UPDATE $imageTable SET img_sha1=" . $dbw->addQuotes( $sha1 ) .
 145                                                 " WHERE img_name=" . $dbw->addQuotes( $file->getName() );
 146                                         if ( $method == 'pipe' ) {
 147                                                 fwrite( $pipe, "$sql;\n" );
 148                                         } else {
 149                                                 $dbw->query( $sql, __METHOD__ );
 150                                         }
 151                                 }
 152                         }
 153                         // Upgrade the old file versions...
 154                         foreach ( $file->getHistory() as $oldFile ) {
 155                                 /** @var OldLocalFile $oldFile */
 156                                 '@phan-var OldLocalFile $oldFile';
 157                                 $sha1 = $oldFile->getRepo()->getFileSha1( $oldFile->getPath() );
 158                                 if ( strval( $sha1 ) !== '' ) { // file on disk and hashed properly
 159                                         if ( $isRegen && $oldFile->getSha1() !== $sha1 ) {
 160                                                 // The population was probably done already. If the old SHA1
 161                                                 // does not match, then both fix the SHA1 and the metadata.
 162                                                 $oldFile->upgradeRow();
 163                                         } else {
 164                                                 $sql = "UPDATE $oldImageTable SET oi_sha1=" . $dbw->addQuotes( $sha1 ) .
 165                                                         " WHERE (oi_name=" . $dbw->addQuotes( $oldFile->getName() ) . " AND" .
 166                                                         " oi_archive_name=" . $dbw->addQuotes( $oldFile->getArchiveName() ) . ")";
 167                                                 if ( $method == 'pipe' ) {
 168                                                         fwrite( $pipe, "$sql;\n" );
 169                                                 } else {
 170                                                         $dbw->query( $sql, __METHOD__ );
 171                                                 }
 172                                         }
 173                                 }
 174                         }
 175                         $i++;
 176                 }
 177                 if ( $method == 'pipe' ) {
 178                         fflush( $pipe );
 179                         pclose( $pipe );
 180                 }
 181                 $t += microtime( true );
 182                 $this->output( sprintf( "\nDone %d files in %.1f seconds\n", $numRows, $t ) );
 183
 184                 return !$file; // we only updated *some* files, don't log
 185         }
 186 }
 187
 188 $maintClass = PopulateImageSha1::class;
 189 require_once RUN_MAINTENANCE_IF_MAIN;