maintenance/migrateFileRepoLayout.php

   1 <?php
   2 /**
   3  * Copy all files in FileRepo to an originals container using SHA1 paths.
   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  * Copy all files in FileRepo to an originals container using SHA1 paths.
  28  *
  29  * This script should be run while the repo is still set to the old layout.
  30  *
  31  * @ingroup Maintenance
  32  */
  33 class MigrateFileRepoLayout extends Maintenance {
  34         public function __construct() {
  35                 parent::__construct();
  36                 $this->mDescription = "Copy files in repo to a different layout.";
  37                 $this->addOption( 'oldlayout', "Old layout; one of 'name' or 'sha1'", true, true );
  38                 $this->addOption( 'newlayout', "New layout; one of 'name' or 'sha1'", true, true );
  39                 $this->addOption( 'since', "Copy only files from after this timestamp", false, true );
  40                 $this->setBatchSize( 50 );
  41         }
  42
  43         public function execute() {
  44                 $oldLayout = $this->getOption( 'oldlayout' );
  45                 if ( !in_array( $oldLayout, array( 'name', 'sha1' ) ) ) {
  46                         $this->error( "Invalid old layout.", 1 );
  47                 }
  48                 $newLayout = $this->getOption( 'newlayout' );
  49                 if ( !in_array( $newLayout, array( 'name', 'sha1' ) ) ) {
  50                         $this->error( "Invalid new layout.", 1 );
  51                 }
  52                 $since = $this->getOption( 'since' );
  53
  54                 $repo = $this->getRepo();
  55
  56                 $be = $repo->getBackend();
  57                 if ( $be instanceof FileBackendDBRepoWrapper ) {
  58                         $be = $be->getInternalBackend(); // avoid path translations for this script
  59                 }
  60
  61                 $dbw = $repo->getMasterDB();
  62
  63                 $origBase = $be->getContainerStoragePath( "{$repo->getName()}-original" );
  64                 $startTime = wfTimestampNow();
  65
  66                 // Do current and archived versions...
  67                 $conds = array();
  68                 if ( $since ) {
  69                         $conds[] = 'img_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
  70                 }
  71
  72                 $batch = array();
  73                 $lastName = '';
  74                 do {
  75                         $res = $dbw->select( 'image', array( 'img_name', 'img_sha1' ),
  76                                 array_merge( array( 'img_name > ' . $dbw->addQuotes( $lastName ) ), $conds ),
  77                                 __METHOD__,
  78                                 array( 'LIMIT' => $this->mBatchSize, 'ORDER BY' => 'img_name' )
  79                         );
  80
  81                         foreach ( $res as $row ) {
  82                                 $lastName = $row->img_name;
  83                                 $sha1 = $row->img_sha1;
  84                                 if ( !strlen( $sha1 ) ) {
  85                                         $this->error( "Image SHA-1 not set for {$row->img_name}." );
  86                                 } else {
  87                                         $file = $repo->newFile( $row->img_name );
  88
  89                                         if ( $oldLayout === 'sha1' ) {
  90                                                 $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
  91                                         } else {
  92                                                 $spath = $file->getPath();
  93                                         }
  94
  95                                         if ( $newLayout === 'sha1' ) {
  96                                                 $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
  97                                         } else {
  98                                                 $dpath = $file->getPath();
  99                                         }
 100
 101                                         $status = $be->prepare( array( 'dir' => dirname( $dpath ) ) );
 102                                         if ( !$status->isOK() ) {
 103                                                 $this->error( print_r( $status->getErrorsArray(), true ) );
 104                                         }
 105
 106                                         $batch[] = array( 'op' => 'copy', 'overwrite' => true,
 107                                                 'src' => $spath, 'dst' => $dpath, 'img' => $row->img_name );
 108                                 }
 109
 110                                 foreach ( $file->getHistory() as $ofile ) {
 111                                         $sha1 = $ofile->getSha1();
 112                                         if ( !strlen( $sha1 ) ) {
 113                                                 $this->error( "Image SHA-1 not set for {$ofile->getArchiveName()}." );
 114                                                 continue;
 115                                         }
 116
 117                                         if ( $oldLayout === 'sha1' ) {
 118                                                 $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 119                                         } elseif ( $ofile->isDeleted( File::DELETED_FILE ) ) {
 120                                                 $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 121                                                         '/' . $repo->getDeletedHashPath( $sha1 ) .
 122                                                         $sha1 . '.' . $ofile->getExtension();
 123                                         } else {
 124                                                 $spath = $ofile->getPath();
 125                                         }
 126
 127                                         if ( $newLayout === 'sha1' ) {
 128                                                 $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 129                                         } else {
 130                                                 $dpath = $ofile->getPath();
 131                                         }
 132
 133                                         $status = $be->prepare( array( 'dir' => dirname( $dpath ) ) );
 134                                         if ( !$status->isOK() ) {
 135                                                 $this->error( print_r( $status->getErrorsArray(), true ) );
 136                                         }
 137                                         $batch[] = array( 'op' => 'copy', 'overwrite' => true,
 138                                                 'src' => $spath, 'dst' => $dpath, 'img' => $ofile->getArchiveName() );
 139                                 }
 140
 141                                 if ( count( $batch ) >= $this->mBatchSize ) {
 142                                         $this->runBatch( $batch, $be );
 143                                         $batch = array();
 144                                 }
 145                         }
 146                 } while ( $res->numRows() );
 147
 148                 if ( count( $batch ) ) {
 149                         $this->runBatch( $batch, $be );
 150                 }
 151
 152                 // Do deleted versions...
 153                 $conds = array();
 154                 if ( $since ) {
 155                         $conds[] = 'fa_deleted_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
 156                 }
 157
 158                 $batch = array();
 159                 $lastId = 0;
 160                 do {
 161                         $res = $dbw->select( 'filearchive', array( 'fa_storage_key', 'fa_id', 'fa_name' ),
 162                                 array_merge( array( 'fa_id > ' . $dbw->addQuotes( $lastId ) ), $conds ),
 163                                 __METHOD__,
 164                                 array( 'LIMIT' => $this->mBatchSize, 'ORDER BY' => 'fa_id' )
 165                         );
 166
 167                         foreach ( $res as $row ) {
 168                                 $lastId = $row->fa_id;
 169                                 $sha1Key = $row->fa_storage_key;
 170                                 if ( !strlen( $sha1Key ) ) {
 171                                         $this->error( "Image SHA-1 not set for file #{$row->fa_id} (deleted)." );
 172                                         continue;
 173                                 }
 174                                 $sha1 = substr( $sha1Key, 0, strpos( $sha1Key, '.' ) );
 175
 176                                 if ( $oldLayout === 'sha1' ) {
 177                                         $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 178                                 } else {
 179                                         $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 180                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 181                                 }
 182
 183                                 if ( $newLayout === 'sha1' ) {
 184                                         $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 185                                 } else {
 186                                         $dpath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 187                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 188                                 }
 189
 190                                 $status = $be->prepare( array( 'dir' => dirname( $dpath ) ) );
 191                                 if ( !$status->isOK() ) {
 192                                         $this->error( print_r( $status->getErrorsArray(), true ) );
 193                                 }
 194
 195                                 $batch[] = array( 'op' => 'copy', 'src' => $spath, 'dst' => $dpath,
 196                                         'overwriteSame' => true, 'img' => "(ID {$row->fa_id}) {$row->fa_name}" );
 197
 198                                 if ( count( $batch ) >= $this->mBatchSize ) {
 199                                         $this->runBatch( $batch, $be );
 200                                         $batch = array();
 201                                 }
 202                         }
 203                 } while ( $res->numRows() );
 204
 205                 if ( count( $batch ) ) {
 206                         $this->runBatch( $batch, $be );
 207                 }
 208
 209                 $this->output( "Done (started $startTime)\n" );
 210         }
 211
 212         protected function getRepo() {
 213                 return RepoGroup::singleton()->getLocalRepo();
 214         }
 215
 216         protected function runBatch( array $ops, FileBackend $be ) {
 217                 $this->output( "Migrating file batch:\n" );
 218                 foreach ( $ops as $op ) {
 219                         $this->output( "\"{$op['img']}\" (dest: {$op['dst']})\n" );
 220                 }
 221
 222                 $status = $be->doOperations( $ops );
 223                 if ( !$status->isOK() ) {
 224                         $this->output( print_r( $status->getErrorsArray(), true ) );
 225                 }
 226
 227                 $this->output( "Batch done\n\n" );
 228         }
 229 }
 230
 231 $maintClass = 'MigrateFileRepoLayout';
 232 require_once RUN_MAINTENANCE_IF_MAIN;