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