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 use MediaWiki\MediaWikiServices;
  25
  26 require_once __DIR__ . '/Maintenance.php';
  27
  28 /**
  29  * Copy all files in FileRepo to an originals container using SHA1 paths.
  30  *
  31  * This script should be run while the repo is still set to the old layout.
  32  *
  33  * @ingroup Maintenance
  34  */
  35 class MigrateFileRepoLayout extends Maintenance {
  36         public function __construct() {
  37                 parent::__construct();
  38                 $this->addDescription( 'Copy files in repo to a different layout.' );
  39                 $this->addOption( 'oldlayout', "Old layout; one of 'name' or 'sha1'", true, true );
  40                 $this->addOption( 'newlayout', "New layout; one of 'name' or 'sha1'", true, true );
  41                 $this->addOption( 'since', "Copy only files from after this timestamp", false, true );
  42                 $this->setBatchSize( 50 );
  43         }
  44
  45         public function execute() {
  46                 $oldLayout = $this->getOption( 'oldlayout' );
  47                 if ( !in_array( $oldLayout, [ 'name', 'sha1' ] ) ) {
  48                         $this->fatalError( "Invalid old layout." );
  49                 }
  50                 $newLayout = $this->getOption( 'newlayout' );
  51                 if ( !in_array( $newLayout, [ 'name', 'sha1' ] ) ) {
  52                         $this->fatalError( "Invalid new layout." );
  53                 }
  54                 $since = $this->getOption( 'since' );
  55
  56                 $repo = $this->getRepo();
  57
  58                 $be = $repo->getBackend();
  59                 if ( $be instanceof FileBackendDBRepoWrapper ) {
  60                         // avoid path translations for this script
  61                         $be = $be->getInternalBackend();
  62                 }
  63
  64                 $dbw = $repo->getMasterDB();
  65
  66                 $origBase = $be->getContainerStoragePath( "{$repo->getName()}-original" );
  67                 $startTime = wfTimestampNow();
  68
  69                 // Do current and archived versions...
  70                 $conds = [];
  71                 if ( $since ) {
  72                         $conds[] = 'img_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
  73                 }
  74
  75                 $batchSize = $this->getBatchSize();
  76                 $batch = [];
  77                 $lastName = '';
  78                 do {
  79                         $res = $dbw->select( 'image',
  80                                 [ 'img_name', 'img_sha1' ],
  81                                 array_merge( [ 'img_name > ' . $dbw->addQuotes( $lastName ) ], $conds ),
  82                                 __METHOD__,
  83                                 [ 'LIMIT' => $batchSize, 'ORDER BY' => 'img_name' ]
  84                         );
  85
  86                         foreach ( $res as $row ) {
  87                                 $lastName = $row->img_name;
  88                                 /** @var LocalFile $file */
  89                                 $file = $repo->newFile( $row->img_name );
  90                                 // Check in case SHA1 rows are not populated for some files
  91                                 $sha1 = strlen( $row->img_sha1 ) ? $row->img_sha1 : $file->getSha1();
  92
  93                                 if ( !strlen( $sha1 ) ) {
  94                                         $this->error( "Image SHA-1 not known for {$row->img_name}." );
  95                                 } else {
  96                                         if ( $oldLayout === 'sha1' ) {
  97                                                 $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
  98                                         } else {
  99                                                 $spath = $file->getPath();
 100                                         }
 101
 102                                         if ( $newLayout === 'sha1' ) {
 103                                                 $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 104                                         } else {
 105                                                 $dpath = $file->getPath();
 106                                         }
 107
 108                                         $status = $be->prepare( [
 109                                                 'dir' => dirname( $dpath ), 'bypassReadOnly' => 1 ] );
 110                                         if ( !$status->isOK() ) {
 111                                                 $this->error( print_r( $status->getErrors(), true ) );
 112                                         }
 113
 114                                         $batch[] = [ 'op' => 'copy', 'overwrite' => true,
 115                                                 'src' => $spath, 'dst' => $dpath, 'img' => $row->img_name ];
 116                                 }
 117
 118                                 foreach ( $file->getHistory() as $ofile ) {
 119                                         $sha1 = $ofile->getSha1();
 120                                         if ( !strlen( $sha1 ) ) {
 121                                                 $this->error( "Image SHA-1 not set for {$ofile->getArchiveName()}." );
 122                                                 continue;
 123                                         }
 124
 125                                         if ( $oldLayout === 'sha1' ) {
 126                                                 $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 127                                         } elseif ( $ofile->isDeleted( File::DELETED_FILE ) ) {
 128                                                 $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 129                                                         '/' . $repo->getDeletedHashPath( $sha1 ) .
 130                                                         $sha1 . '.' . $ofile->getExtension();
 131                                         } else {
 132                                                 $spath = $ofile->getPath();
 133                                         }
 134
 135                                         if ( $newLayout === 'sha1' ) {
 136                                                 $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 137                                         } else {
 138                                                 $dpath = $ofile->getPath();
 139                                         }
 140
 141                                         $status = $be->prepare( [
 142                                                 'dir' => dirname( $dpath ), 'bypassReadOnly' => 1 ] );
 143                                         if ( !$status->isOK() ) {
 144                                                 $this->error( print_r( $status->getErrors(), true ) );
 145                                         }
 146                                         $batch[] = [ 'op' => 'copy', 'overwrite' => true,
 147                                                 'src' => $spath, 'dst' => $dpath, 'img' => $ofile->getArchiveName() ];
 148                                 }
 149
 150                                 if ( count( $batch ) >= $batchSize ) {
 151                                         $this->runBatch( $batch, $be );
 152                                         $batch = [];
 153                                 }
 154                         }
 155                 } while ( $res->numRows() );
 156
 157                 if ( count( $batch ) ) {
 158                         $this->runBatch( $batch, $be );
 159                 }
 160
 161                 // Do deleted versions...
 162                 $conds = [];
 163                 if ( $since ) {
 164                         $conds[] = 'fa_deleted_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
 165                 }
 166
 167                 $batch = [];
 168                 $lastId = 0;
 169                 do {
 170                         $res = $dbw->select( 'filearchive', [ 'fa_storage_key', 'fa_id', 'fa_name' ],
 171                                 array_merge( [ 'fa_id > ' . $dbw->addQuotes( $lastId ) ], $conds ),
 172                                 __METHOD__,
 173                                 [ 'LIMIT' => $batchSize, 'ORDER BY' => 'fa_id' ]
 174                         );
 175
 176                         foreach ( $res as $row ) {
 177                                 $lastId = $row->fa_id;
 178                                 $sha1Key = $row->fa_storage_key;
 179                                 if ( !strlen( $sha1Key ) ) {
 180                                         $this->error( "Image SHA-1 not set for file #{$row->fa_id} (deleted)." );
 181                                         continue;
 182                                 }
 183                                 $sha1 = substr( $sha1Key, 0, strpos( $sha1Key, '.' ) );
 184
 185                                 if ( $oldLayout === 'sha1' ) {
 186                                         $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 187                                 } else {
 188                                         $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 189                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 190                                 }
 191
 192                                 if ( $newLayout === 'sha1' ) {
 193                                         $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 194                                 } else {
 195                                         $dpath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 196                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 197                                 }
 198
 199                                 $status = $be->prepare( [
 200                                         'dir' => dirname( $dpath ), 'bypassReadOnly' => 1 ] );
 201                                 if ( !$status->isOK() ) {
 202                                         $this->error( print_r( $status->getErrors(), true ) );
 203                                 }
 204
 205                                 $batch[] = [ 'op' => 'copy', 'src' => $spath, 'dst' => $dpath,
 206                                         'overwriteSame' => true, 'img' => "(ID {$row->fa_id}) {$row->fa_name}" ];
 207
 208                                 if ( count( $batch ) >= $batchSize ) {
 209                                         $this->runBatch( $batch, $be );
 210                                         $batch = [];
 211                                 }
 212                         }
 213                 } while ( $res->numRows() );
 214
 215                 if ( count( $batch ) ) {
 216                         $this->runBatch( $batch, $be );
 217                 }
 218
 219                 $this->output( "Done (started $startTime)\n" );
 220         }
 221
 222         protected function getRepo() {
 223                 return MediaWikiServices::getInstance()->getRepoGroup()->getLocalRepo();
 224         }
 225
 226         /**
 227          * @param array[] $ops
 228          * @param FileBackend $be
 229          */
 230         protected function runBatch( array $ops, FileBackend $be ) {
 231                 $this->output( "Migrating file batch:\n" );
 232                 foreach ( $ops as $op ) {
 233                         $this->output( "\"{$op['img']}\" (dest: {$op['dst']})\n" );
 234                 }
 235
 236                 $status = $be->doOperations( $ops, [ 'bypassReadOnly' => 1 ] );
 237                 if ( !$status->isOK() ) {
 238                         $this->output( print_r( $status->getErrors(), true ) );
 239                 }
 240
 241                 $this->output( "Batch done\n\n" );
 242         }
 243 }
 244
 245 $maintClass = MigrateFileRepoLayout::class;
 246 require_once RUN_MAINTENANCE_IF_MAIN;