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