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