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->addDescription( '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, [ 'name', 'sha1' ] ) ) {
  46                         $this->fatalError( "Invalid old layout." );
  47                 }
  48                 $newLayout = $this->getOption( 'newlayout' );
  49                 if ( !in_array( $newLayout, [ 'name', 'sha1' ] ) ) {
  50                         $this->fatalError( "Invalid new layout." );
  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 = [];
  68                 if ( $since ) {
  69                         $conds[] = 'img_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
  70                 }
  71
  72                 $batchSize = $this->getBatchSize();
  73                 $batch = [];
  74                 $lastName = '';
  75                 do {
  76                         $res = $dbw->select( 'image',
  77                                 [ 'img_name', 'img_sha1' ],
  78                                 array_merge( [ 'img_name > ' . $dbw->addQuotes( $lastName ) ], $conds ),
  79                                 __METHOD__,
  80                                 [ 'LIMIT' => $batchSize, 'ORDER BY' => 'img_name' ]
  81                         );
  82
  83                         foreach ( $res as $row ) {
  84                                 $lastName = $row->img_name;
  85                                 /** @var LocalFile $file */
  86                                 $file = $repo->newFile( $row->img_name );
  87                                 // Check in case SHA1 rows are not populated for some files
  88                                 $sha1 = strlen( $row->img_sha1 ) ? $row->img_sha1 : $file->getSha1();
  89
  90                                 if ( !strlen( $sha1 ) ) {
  91                                         $this->error( "Image SHA-1 not known for {$row->img_name}." );
  92                                 } else {
  93                                         if ( $oldLayout === 'sha1' ) {
  94                                                 $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
  95                                         } else {
  96                                                 $spath = $file->getPath();
  97                                         }
  98
  99                                         if ( $newLayout === 'sha1' ) {
 100                                                 $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 101                                         } else {
 102                                                 $dpath = $file->getPath();
 103                                         }
 104
 105                                         $status = $be->prepare( [
 106                                                 'dir' => dirname( $dpath ), 'bypassReadOnly' => 1 ] );
 107                                         if ( !$status->isOK() ) {
 108                                                 $this->error( print_r( $status->getErrors(), true ) );
 109                                         }
 110
 111                                         $batch[] = [ 'op' => 'copy', 'overwrite' => true,
 112                                                 'src' => $spath, 'dst' => $dpath, 'img' => $row->img_name ];
 113                                 }
 114
 115                                 foreach ( $file->getHistory() as $ofile ) {
 116                                         $sha1 = $ofile->getSha1();
 117                                         if ( !strlen( $sha1 ) ) {
 118                                                 $this->error( "Image SHA-1 not set for {$ofile->getArchiveName()}." );
 119                                                 continue;
 120                                         }
 121
 122                                         if ( $oldLayout === 'sha1' ) {
 123                                                 $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 124                                         } elseif ( $ofile->isDeleted( File::DELETED_FILE ) ) {
 125                                                 $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 126                                                         '/' . $repo->getDeletedHashPath( $sha1 ) .
 127                                                         $sha1 . '.' . $ofile->getExtension();
 128                                         } else {
 129                                                 $spath = $ofile->getPath();
 130                                         }
 131
 132                                         if ( $newLayout === 'sha1' ) {
 133                                                 $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 134                                         } else {
 135                                                 $dpath = $ofile->getPath();
 136                                         }
 137
 138                                         $status = $be->prepare( [
 139                                                 'dir' => dirname( $dpath ), 'bypassReadOnly' => 1 ] );
 140                                         if ( !$status->isOK() ) {
 141                                                 $this->error( print_r( $status->getErrors(), true ) );
 142                                         }
 143                                         $batch[] = [ 'op' => 'copy', 'overwrite' => true,
 144                                                 'src' => $spath, 'dst' => $dpath, 'img' => $ofile->getArchiveName() ];
 145                                 }
 146
 147                                 if ( count( $batch ) >= $batchSize ) {
 148                                         $this->runBatch( $batch, $be );
 149                                         $batch = [];
 150                                 }
 151                         }
 152                 } while ( $res->numRows() );
 153
 154                 if ( count( $batch ) ) {
 155                         $this->runBatch( $batch, $be );
 156                 }
 157
 158                 // Do deleted versions...
 159                 $conds = [];
 160                 if ( $since ) {
 161                         $conds[] = 'fa_deleted_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
 162                 }
 163
 164                 $batch = [];
 165                 $lastId = 0;
 166                 do {
 167                         $res = $dbw->select( 'filearchive', [ 'fa_storage_key', 'fa_id', 'fa_name' ],
 168                                 array_merge( [ 'fa_id > ' . $dbw->addQuotes( $lastId ) ], $conds ),
 169                                 __METHOD__,
 170                                 [ 'LIMIT' => $batchSize, 'ORDER BY' => 'fa_id' ]
 171                         );
 172
 173                         foreach ( $res as $row ) {
 174                                 $lastId = $row->fa_id;
 175                                 $sha1Key = $row->fa_storage_key;
 176                                 if ( !strlen( $sha1Key ) ) {
 177                                         $this->error( "Image SHA-1 not set for file #{$row->fa_id} (deleted)." );
 178                                         continue;
 179                                 }
 180                                 $sha1 = substr( $sha1Key, 0, strpos( $sha1Key, '.' ) );
 181
 182                                 if ( $oldLayout === 'sha1' ) {
 183                                         $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 184                                 } else {
 185                                         $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 186                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 187                                 }
 188
 189                                 if ( $newLayout === 'sha1' ) {
 190                                         $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 191                                 } else {
 192                                         $dpath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 193                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 194                                 }
 195
 196                                 $status = $be->prepare( [
 197                                         'dir' => dirname( $dpath ), 'bypassReadOnly' => 1 ] );
 198                                 if ( !$status->isOK() ) {
 199                                         $this->error( print_r( $status->getErrors(), true ) );
 200                                 }
 201
 202                                 $batch[] = [ 'op' => 'copy', 'src' => $spath, 'dst' => $dpath,
 203                                         'overwriteSame' => true, 'img' => "(ID {$row->fa_id}) {$row->fa_name}" ];
 204
 205                                 if ( count( $batch ) >= $batchSize ) {
 206                                         $this->runBatch( $batch, $be );
 207                                         $batch = [];
 208                                 }
 209                         }
 210                 } while ( $res->numRows() );
 211
 212                 if ( count( $batch ) ) {
 213                         $this->runBatch( $batch, $be );
 214                 }
 215
 216                 $this->output( "Done (started $startTime)\n" );
 217         }
 218
 219         protected function getRepo() {
 220                 return RepoGroup::singleton()->getLocalRepo();
 221         }
 222
 223         protected function runBatch( array $ops, FileBackend $be ) {
 224                 $this->output( "Migrating file batch:\n" );
 225                 foreach ( $ops as $op ) {
 226                         $this->output( "\"{$op['img']}\" (dest: {$op['dst']})\n" );
 227                 }
 228
 229                 $status = $be->doOperations( $ops, [ 'bypassReadOnly' => 1 ] );
 230                 if ( !$status->isOK() ) {
 231                         $this->output( print_r( $status->getErrors(), true ) );
 232                 }
 233
 234                 $this->output( "Batch done\n\n" );
 235         }
 236 }
 237
 238 $maintClass = MigrateFileRepoLayout::class;
 239 require_once RUN_MAINTENANCE_IF_MAIN;