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                         // avoid path translations for this script
  59                         $be = $be->getInternalBackend();
  60                 }
  61
  62                 $dbw = $repo->getPrimaryDB();
  63
  64                 $origBase = $be->getContainerStoragePath( "{$repo->getName()}-original" );
  65                 $startTime = wfTimestampNow();
  66
  67                 // Do current and archived versions...
  68                 $conds = [];
  69                 if ( $since ) {
  70                         $conds[] = 'img_timestamp >= ' . $dbw->addQuotes( $dbw->timestamp( $since ) );
  71                 }
  72
  73                 $batchSize = $this->getBatchSize();
  74                 $batch = [];
  75                 $lastName = '';
  76                 do {
  77                         $res = $dbw->newSelectQueryBuilder()
  78                                 ->select( [ 'img_name', 'img_sha1' ] )
  79                                 ->from( 'image' )
  80                                 ->where( 'img_name > ' . $dbw->addQuotes( $lastName ) )
  81                                 ->andWhere( $conds )
  82                                 ->orderBy( 'img_name' )
  83                                 ->limit( $batchSize )
  84                                 ->caller( __METHOD__ )->fetchResultSet();
  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( $dpath ), 'bypassReadOnly' => true ] );
 110                                         if ( !$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( $dpath ), 'bypassReadOnly' => true ] );
 143                                         if ( !$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->newSelectQueryBuilder()
 171                                 ->select( [ 'fa_storage_key', 'fa_id', 'fa_name' ] )
 172                                 ->from( 'filearchive' )
 173                                 ->where( 'fa_id > ' . $dbw->addQuotes( $lastId ) )
 174                                 ->andWhere( $conds )
 175                                 ->orderBy( 'fa_id' )
 176                                 ->limit( $batchSize )
 177                                 ->caller( __METHOD__ )->fetchResultSet();
 178
 179                         foreach ( $res as $row ) {
 180                                 $lastId = $row->fa_id;
 181                                 $sha1Key = $row->fa_storage_key;
 182                                 if ( !strlen( $sha1Key ) ) {
 183                                         $this->error( "Image SHA-1 not set for file #{$row->fa_id} (deleted)." );
 184                                         continue;
 185                                 }
 186                                 $sha1 = substr( $sha1Key, 0, strpos( $sha1Key, '.' ) );
 187
 188                                 if ( $oldLayout === 'sha1' ) {
 189                                         $spath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 190                                 } else {
 191                                         $spath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 192                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 193                                 }
 194
 195                                 if ( $newLayout === 'sha1' ) {
 196                                         $dpath = "{$origBase}/{$sha1[0]}/{$sha1[1]}/{$sha1[2]}/{$sha1}";
 197                                 } else {
 198                                         $dpath = $be->getContainerStoragePath( "{$repo->getName()}-deleted" ) .
 199                                                 '/' . $repo->getDeletedHashPath( $sha1Key ) . $sha1Key;
 200                                 }
 201
 202                                 $status = $be->prepare( [
 203                                         'dir' => dirname( $dpath ), 'bypassReadOnly' => true ] );
 204                                 if ( !$status->isOK() ) {
 205                                         $this->error( print_r( $status->getErrors(), true ) );
 206                                 }
 207
 208                                 $batch[] = [ 'op' => 'copy', 'src' => $spath, 'dst' => $dpath,
 209                                         'overwriteSame' => true, 'img' => "(ID {$row->fa_id}) {$row->fa_name}" ];
 210
 211                                 if ( count( $batch ) >= $batchSize ) {
 212                                         $this->runBatch( $batch, $be );
 213                                         $batch = [];
 214                                 }
 215                         }
 216                 } while ( $res->numRows() );
 217
 218                 if ( count( $batch ) ) {
 219                         $this->runBatch( $batch, $be );
 220                 }
 221
 222                 $this->output( "Done (started $startTime)\n" );
 223         }
 224
 225         protected function getRepo() {
 226                 return $this->getServiceContainer()->getRepoGroup()->getLocalRepo();
 227         }
 228
 229         /**
 230          * @param array[] $ops
 231          * @param FileBackend $be
 232          */
 233         protected function runBatch( array $ops, FileBackend $be ) {
 234                 $this->output( "Migrating file batch:\n" );
 235                 foreach ( $ops as $op ) {
 236                         $this->output( "\"{$op['img']}\" (dest: {$op['dst']})\n" );
 237                 }
 238
 239                 $status = $be->doOperations( $ops, [ 'bypassReadOnly' => true ] );
 240                 if ( !$status->isOK() ) {
 241                         $this->output( print_r( $status->getErrors(), true ) );
 242                 }
 243
 244                 $this->output( "Batch done\n\n" );
 245         }
 246 }
 247
 248 $maintClass = MigrateFileRepoLayout::class;
 249 require_once RUN_MAINTENANCE_IF_MAIN;