includes/filerepo/backend/FileBackend.php

   1 <?php
   2 /**
   3  * @defgroup FileBackend File backend
   4  * @ingroup  FileRepo
   5  *
   6  * File backend is used to interact with file storage systems,
   7  * such as the local file system, NFS, or cloud storage systems.
   8  */
   9
  10 /**
  11  * Base class for all file backends.
  12  *
  13  * This program is free software; you can redistribute it and/or modify
  14  * it under the terms of the GNU General Public License as published by
  15  * the Free Software Foundation; either version 2 of the License, or
  16  * (at your option) any later version.
  17  *
  18  * This program is distributed in the hope that it will be useful,
  19  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  20  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  21  * GNU General Public License for more details.
  22  *
  23  * You should have received a copy of the GNU General Public License along
  24  * with this program; if not, write to the Free Software Foundation, Inc.,
  25  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  26  * http://www.gnu.org/copyleft/gpl.html
  27  *
  28  * @file
  29  * @ingroup FileBackend
  30  * @author Aaron Schulz
  31  */
  32
  33 /**
  34  * @brief Base class for all file backend classes (including multi-write backends).
  35  *
  36  * This class defines the methods as abstract that subclasses must implement.
  37  * Outside callers can assume that all backends will have these functions.
  38  *
  39  * All "storage paths" are of the format "mwstore://<backend>/<container>/<path>".
  40  * The <path> portion is a relative path that uses UNIX file system (FS) notation,
  41  * though any particular backend may not actually be using a local filesystem.
  42  * Therefore, the relative paths are only virtual.
  43  *
  44  * Backend contents are stored under wiki-specific container names by default.
  45  * For legacy reasons, this has no effect for the FS backend class, and per-wiki
  46  * segregation must be done by setting the container paths appropriately.
  47  *
  48  * FS-based backends are somewhat more restrictive due to the existence of real
  49  * directory files; a regular file cannot have the same name as a directory. Other
  50  * backends with virtual directories may not have this limitation. Callers should
  51  * store files in such a way that no files and directories are under the same path.
  52  *
  53  * Methods should avoid throwing exceptions at all costs.
  54  * As a corollary, external dependencies should be kept to a minimum.
  55  *
  56  * @ingroup FileBackend
  57  * @since 1.19
  58  */
  59 abstract class FileBackend {
  60         protected $name; // string; unique backend name
  61         protected $wikiId; // string; unique wiki name
  62         protected $readOnly; // string; read-only explanation message
  63         protected $parallelize; // string; when to do operations in parallel
  64         protected $concurrency; // integer; how many operations can be done in parallel
  65
  66         /** @var LockManager */
  67         protected $lockManager;
  68         /** @var FileJournal */
  69         protected $fileJournal;
  70
  71         /**
  72          * Create a new backend instance from configuration.
  73          * This should only be called from within FileBackendGroup.
  74          *
  75          * $config includes:
  76          *     'name'        : The unique name of this backend.
  77          *                     This should consist of alphanumberic, '-', and '_' characters.
  78          *                     This name should not be changed after use.
  79          *     'wikiId'      : Prefix to container names that is unique to this wiki.
  80          *                     It should only consist of alphanumberic, '-', and '_' characters.
  81          *     'lockManager' : Registered name of a file lock manager to use.
  82          *     'fileJournal' : File journal configuration; see FileJournal::factory().
  83          *                     Journals simply log changes to files stored in the backend.
  84          *     'readOnly'    : Write operations are disallowed if this is a non-empty string.
  85          *                     It should be an explanation for the backend being read-only.
  86          *     'parallelize' : When to do file operations in parallel (when possible).
  87          *                     Allowed values are "implicit", "explicit" and "off".
  88          *     'concurrency' : How many file operations can be done in parallel.
  89          *
  90          * @param $config Array
  91          * @throws MWException
  92          */
  93         public function __construct( array $config ) {
  94                 $this->name = $config['name'];
  95                 if ( !preg_match( '!^[a-zA-Z0-9-_]{1,255}$!', $this->name ) ) {
  96                         throw new MWException( "Backend name `{$this->name}` is invalid." );
  97                 }
  98                 $this->wikiId = isset( $config['wikiId'] )
  99                         ? $config['wikiId']
 100                         : wfWikiID(); // e.g. "my_wiki-en_"
 101                 $this->lockManager = ( $config['lockManager'] instanceof LockManager )
 102                         ? $config['lockManager']
 103                         : LockManagerGroup::singleton()->get( $config['lockManager'] );
 104                 $this->fileJournal = isset( $config['fileJournal'] )
 105                         ? FileJournal::factory( $config['fileJournal'], $this->name )
 106                         : FileJournal::factory( array( 'class' => 'NullFileJournal' ), $this->name );
 107                 $this->readOnly = isset( $config['readOnly'] )
 108                         ? (string)$config['readOnly']
 109                         : '';
 110                 $this->parallelize = isset( $config['parallelize'] )
 111                         ? (string)$config['parallelize']
 112                         : 'off';
 113                 $this->concurrency = isset( $config['concurrency'] )
 114                         ? (int)$config['concurrency']
 115                         : 50;
 116         }
 117
 118         /**
 119          * Get the unique backend name.
 120          * We may have multiple different backends of the same type.
 121          * For example, we can have two Swift backends using different proxies.
 122          *
 123          * @return string
 124          */
 125         final public function getName() {
 126                 return $this->name;
 127         }
 128
 129         /**
 130          * Check if this backend is read-only
 131          *
 132          * @return bool
 133          */
 134         final public function isReadOnly() {
 135                 return ( $this->readOnly != '' );
 136         }
 137
 138         /**
 139          * Get an explanatory message if this backend is read-only
 140          *
 141          * @return string|bool Returns false if the backend is not read-only
 142          */
 143         final public function getReadOnlyReason() {
 144                 return ( $this->readOnly != '' ) ? $this->readOnly : false;
 145         }
 146
 147         /**
 148          * This is the main entry point into the backend for write operations.
 149          * Callers supply an ordered list of operations to perform as a transaction.
 150          * Files will be locked, the stat cache cleared, and then the operations attempted.
 151          * If any serious errors occur, all attempted operations will be rolled back.
 152          *
 153          * $ops is an array of arrays. The outer array holds a list of operations.
 154          * Each inner array is a set of key value pairs that specify an operation.
 155          *
 156          * Supported operations and their parameters:
 157          * a) Create a new file in storage with the contents of a string
 158          *     array(
 159          *         'op'                  => 'create',
 160          *         'dst'                 => <storage path>,
 161          *         'content'             => <string of new file contents>,
 162          *         'overwrite'           => <boolean>,
 163          *         'overwriteSame'       => <boolean>
 164          *     )
 165          * b) Copy a file system file into storage
 166          *     array(
 167          *         'op'                  => 'store',
 168          *         'src'                 => <file system path>,
 169          *         'dst'                 => <storage path>,
 170          *         'overwrite'           => <boolean>,
 171          *         'overwriteSame'       => <boolean>
 172          *     )
 173          * c) Copy a file within storage
 174          *     array(
 175          *         'op'                  => 'copy',
 176          *         'src'                 => <storage path>,
 177          *         'dst'                 => <storage path>,
 178          *         'overwrite'           => <boolean>,
 179          *         'overwriteSame'       => <boolean>
 180          *     )
 181          * d) Move a file within storage
 182          *     array(
 183          *         'op'                  => 'move',
 184          *         'src'                 => <storage path>,
 185          *         'dst'                 => <storage path>,
 186          *         'overwrite'           => <boolean>,
 187          *         'overwriteSame'       => <boolean>
 188          *     )
 189          * e) Delete a file within storage
 190          *     array(
 191          *         'op'                  => 'delete',
 192          *         'src'                 => <storage path>,
 193          *         'ignoreMissingSource' => <boolean>
 194          *     )
 195          * f) Do nothing (no-op)
 196          *     array(
 197          *         'op'                  => 'null',
 198          *     )
 199          *
 200          * Boolean flags for operations (operation-specific):
 201          * 'ignoreMissingSource' : The operation will simply succeed and do
 202          *                         nothing if the source file does not exist.
 203          * 'overwrite'           : Any destination file will be overwritten.
 204          * 'overwriteSame'       : An error will not be given if a file already
 205          *                         exists at the destination that has the same
 206          *                         contents as the new contents to be written there.
 207          *
 208          * $opts is an associative of boolean flags, including:
 209          * 'force'               : Operation precondition errors no longer trigger an abort.
 210          *                         Any remaining operations are still attempted. Unexpected
 211          *                         failures may still cause remaning operations to be aborted.
 212          * 'nonLocking'          : No locks are acquired for the operations.
 213          *                         This can increase performance for non-critical writes.
 214          *                         This has no effect unless the 'force' flag is set.
 215          * 'allowStale'          : Don't require the latest available data.
 216          *                         This can increase performance for non-critical writes.
 217          *                         This has no effect unless the 'force' flag is set.
 218          * 'nonJournaled'        : Don't log this operation batch in the file journal.
 219          *                         This limits the ability of recovery scripts.
 220          * 'parallelize'         : Try to do operations in parallel when possible.
 221          *
 222          * Remarks on locking:
 223          * File system paths given to operations should refer to files that are
 224          * already locked or otherwise safe from modification from other processes.
 225          * Normally these files will be new temp files, which should be adequate.
 226          *
 227          * Return value:
 228          * This returns a Status, which contains all warnings and fatals that occured
 229          * during the operation. The 'failCount', 'successCount', and 'success' members
 230          * will reflect each operation attempted. The status will be "OK" unless:
 231          *     a) unexpected operation errors occurred (network partitions, disk full...)
 232          *     b) significant operation errors occured and 'force' was not set
 233          *
 234          * @param $ops Array List of operations to execute in order
 235          * @param $opts Array Batch operation options
 236          * @return Status
 237          */
 238         final public function doOperations( array $ops, array $opts = array() ) {
 239                 if ( $this->isReadOnly() ) {
 240                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 241                 }
 242                 if ( empty( $opts['force'] ) ) { // sanity
 243                         unset( $opts['nonLocking'] );
 244                         unset( $opts['allowStale'] );
 245                 }
 246                 $opts['concurrency'] = 1; // off
 247                 if ( $this->parallelize === 'implicit' ) {
 248                         if ( !isset( $opts['parallelize'] ) || $opts['parallelize'] ) {
 249                                 $opts['concurrency'] = $this->concurrency;
 250                         }
 251                 } elseif ( $this->parallelize === 'explicit' ) {
 252                         if ( !empty( $opts['parallelize'] ) ) {
 253                                 $opts['concurrency'] = $this->concurrency;
 254                         }
 255                 }
 256                 return $this->doOperationsInternal( $ops, $opts );
 257         }
 258
 259         /**
 260          * @see FileBackend::doOperations()
 261          */
 262         abstract protected function doOperationsInternal( array $ops, array $opts );
 263
 264         /**
 265          * Same as doOperations() except it takes a single operation.
 266          * If you are doing a batch of operations that should either
 267          * all succeed or all fail, then use that function instead.
 268          *
 269          * @see FileBackend::doOperations()
 270          *
 271          * @param $op Array Operation
 272          * @param $opts Array Operation options
 273          * @return Status
 274          */
 275         final public function doOperation( array $op, array $opts = array() ) {
 276                 return $this->doOperations( array( $op ), $opts );
 277         }
 278
 279         /**
 280          * Performs a single create operation.
 281          * This sets $params['op'] to 'create' and passes it to doOperation().
 282          *
 283          * @see FileBackend::doOperation()
 284          *
 285          * @param $params Array Operation parameters
 286          * @param $opts Array Operation options
 287          * @return Status
 288          */
 289         final public function create( array $params, array $opts = array() ) {
 290                 return $this->doOperation( array( 'op' => 'create' ) + $params, $opts );
 291         }
 292
 293         /**
 294          * Performs a single store operation.
 295          * This sets $params['op'] to 'store' and passes it to doOperation().
 296          *
 297          * @see FileBackend::doOperation()
 298          *
 299          * @param $params Array Operation parameters
 300          * @param $opts Array Operation options
 301          * @return Status
 302          */
 303         final public function store( array $params, array $opts = array() ) {
 304                 return $this->doOperation( array( 'op' => 'store' ) + $params, $opts );
 305         }
 306
 307         /**
 308          * Performs a single copy operation.
 309          * This sets $params['op'] to 'copy' and passes it to doOperation().
 310          *
 311          * @see FileBackend::doOperation()
 312          *
 313          * @param $params Array Operation parameters
 314          * @param $opts Array Operation options
 315          * @return Status
 316          */
 317         final public function copy( array $params, array $opts = array() ) {
 318                 return $this->doOperation( array( 'op' => 'copy' ) + $params, $opts );
 319         }
 320
 321         /**
 322          * Performs a single move operation.
 323          * This sets $params['op'] to 'move' and passes it to doOperation().
 324          *
 325          * @see FileBackend::doOperation()
 326          *
 327          * @param $params Array Operation parameters
 328          * @param $opts Array Operation options
 329          * @return Status
 330          */
 331         final public function move( array $params, array $opts = array() ) {
 332                 return $this->doOperation( array( 'op' => 'move' ) + $params, $opts );
 333         }
 334
 335         /**
 336          * Performs a single delete operation.
 337          * This sets $params['op'] to 'delete' and passes it to doOperation().
 338          *
 339          * @see FileBackend::doOperation()
 340          *
 341          * @param $params Array Operation parameters
 342          * @param $opts Array Operation options
 343          * @return Status
 344          */
 345         final public function delete( array $params, array $opts = array() ) {
 346                 return $this->doOperation( array( 'op' => 'delete' ) + $params, $opts );
 347         }
 348
 349         /**
 350          * Perform a set of independent file operations on some files.
 351          *
 352          * This does no locking, nor journaling, and possibly no stat calls.
 353          * Any destination files that already exist will be overwritten.
 354          * This should *only* be used on non-original files, like cache files.
 355          *
 356          * Supported operations and their parameters:
 357          * a) Create a new file in storage with the contents of a string
 358          *     array(
 359          *         'op'                  => 'create',
 360          *         'dst'                 => <storage path>,
 361          *         'content'             => <string of new file contents>
 362          *     )
 363          * b) Copy a file system file into storage
 364          *     array(
 365          *         'op'                  => 'store',
 366          *         'src'                 => <file system path>,
 367          *         'dst'                 => <storage path>
 368          *     )
 369          * c) Copy a file within storage
 370          *     array(
 371          *         'op'                  => 'copy',
 372          *         'src'                 => <storage path>,
 373          *         'dst'                 => <storage path>
 374          *     )
 375          * d) Move a file within storage
 376          *     array(
 377          *         'op'                  => 'move',
 378          *         'src'                 => <storage path>,
 379          *         'dst'                 => <storage path>
 380          *     )
 381          * e) Delete a file within storage
 382          *     array(
 383          *         'op'                  => 'delete',
 384          *         'src'                 => <storage path>,
 385          *         'ignoreMissingSource' => <boolean>
 386          *     )
 387          * f) Do nothing (no-op)
 388          *     array(
 389          *         'op'                  => 'null',
 390          *     )
 391          *
 392          * Boolean flags for operations (operation-specific):
 393          * 'ignoreMissingSource' : The operation will simply succeed and do
 394          *                         nothing if the source file does not exist.
 395          *
 396          * Return value:
 397          * This returns a Status, which contains all warnings and fatals that occured
 398          * during the operation. The 'failCount', 'successCount', and 'success' members
 399          * will reflect each operation attempted for the given files. The status will be
 400          * considered "OK" as long as no fatal errors occured.
 401          *
 402          * @param $ops Array Set of operations to execute
 403          * @return Status
 404          * @since 1.20
 405          */
 406         final public function doQuickOperations( array $ops ) {
 407                 if ( $this->isReadOnly() ) {
 408                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 409                 }
 410                 foreach ( $ops as &$op ) {
 411                         $op['overwrite'] = true; // avoids RTTs in key/value stores
 412                 }
 413                 return $this->doQuickOperationsInternal( $ops );
 414         }
 415
 416         /**
 417          * @see FileBackend::doQuickOperations()
 418          * @since 1.20
 419          */
 420         abstract protected function doQuickOperationsInternal( array $ops );
 421
 422         /**
 423          * Same as doQuickOperations() except it takes a single operation.
 424          * If you are doing a batch of operations, then use that function instead.
 425          *
 426          * @see FileBackend::doQuickOperations()
 427          *
 428          * @param $op Array Operation
 429          * @return Status
 430          * @since 1.20
 431          */
 432         final public function doQuickOperation( array $op ) {
 433                 return $this->doQuickOperations( array( $op ) );
 434         }
 435
 436         /**
 437          * Performs a single quick create operation.
 438          * This sets $params['op'] to 'create' and passes it to doQuickOperation().
 439          *
 440          * @see FileBackend::doQuickOperation()
 441          *
 442          * @param $params Array Operation parameters
 443          * @return Status
 444          * @since 1.20
 445          */
 446         final public function quickCreate( array $params ) {
 447                 return $this->doQuickOperation( array( 'op' => 'create' ) + $params );
 448         }
 449
 450         /**
 451          * Performs a single quick store operation.
 452          * This sets $params['op'] to 'store' and passes it to doQuickOperation().
 453          *
 454          * @see FileBackend::doQuickOperation()
 455          *
 456          * @param $params Array Operation parameters
 457          * @return Status
 458          * @since 1.20
 459          */
 460         final public function quickStore( array $params ) {
 461                 return $this->doQuickOperation( array( 'op' => 'store' ) + $params );
 462         }
 463
 464         /**
 465          * Performs a single quick copy operation.
 466          * This sets $params['op'] to 'copy' and passes it to doQuickOperation().
 467          *
 468          * @see FileBackend::doQuickOperation()
 469          *
 470          * @param $params Array Operation parameters
 471          * @return Status
 472          * @since 1.20
 473          */
 474         final public function quickCopy( array $params ) {
 475                 return $this->doQuickOperation( array( 'op' => 'copy' ) + $params );
 476         }
 477
 478         /**
 479          * Performs a single quick move operation.
 480          * This sets $params['op'] to 'move' and passes it to doQuickOperation().
 481          *
 482          * @see FileBackend::doQuickOperation()
 483          *
 484          * @param $params Array Operation parameters
 485          * @return Status
 486          * @since 1.20
 487          */
 488         final public function quickMove( array $params ) {
 489                 return $this->doQuickOperation( array( 'op' => 'move' ) + $params );
 490         }
 491
 492         /**
 493          * Performs a single quick delete operation.
 494          * This sets $params['op'] to 'delete' and passes it to doQuickOperation().
 495          *
 496          * @see FileBackend::doQuickOperation()
 497          *
 498          * @param $params Array Operation parameters
 499          * @return Status
 500          * @since 1.20
 501          */
 502         final public function quickDelete( array $params ) {
 503                 return $this->doQuickOperation( array( 'op' => 'delete' ) + $params );
 504         }
 505
 506         /**
 507          * Concatenate a list of storage files into a single file system file.
 508          * The target path should refer to a file that is already locked or
 509          * otherwise safe from modification from other processes. Normally,
 510          * the file will be a new temp file, which should be adequate.
 511          * $params include:
 512          *     srcs : ordered source storage paths (e.g. chunk1, chunk2, ...)
 513          *     dst  : file system path to 0-byte temp file
 514          *
 515          * @param $params Array Operation parameters
 516          * @return Status
 517          */
 518         abstract public function concatenate( array $params );
 519
 520         /**
 521          * Prepare a storage directory for usage.
 522          * This will create any required containers and parent directories.
 523          * Backends using key/value stores only need to create the container.
 524          *
 525          * $params include:
 526          *     dir : storage directory
 527          *
 528          * @param $params Array
 529          * @return Status
 530          */
 531         final public function prepare( array $params ) {
 532                 if ( $this->isReadOnly() ) {
 533                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 534                 }
 535                 return $this->doPrepare( $params );
 536         }
 537
 538         /**
 539          * @see FileBackend::prepare()
 540          */
 541         abstract protected function doPrepare( array $params );
 542
 543         /**
 544          * Take measures to block web access to a storage directory and
 545          * the container it belongs to. FS backends might add .htaccess
 546          * files whereas key/value store backends might restrict container
 547          * access to the auth user that represents end-users in web request.
 548          * This is not guaranteed to actually do anything.
 549          *
 550          * $params include:
 551          *     dir       : storage directory
 552          *     noAccess  : try to deny file access
 553          *     noListing : try to deny file listing
 554          *
 555          * @param $params Array
 556          * @return Status
 557          */
 558         final public function secure( array $params ) {
 559                 if ( $this->isReadOnly() ) {
 560                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 561                 }
 562                 $status = $this->doPrepare( $params ); // dir must exist to restrict it
 563                 if ( $status->isOK() ) {
 564                         $status->merge( $this->doSecure( $params ) );
 565                 }
 566                 return $status;
 567         }
 568
 569         /**
 570          * @see FileBackend::secure()
 571          */
 572         abstract protected function doSecure( array $params );
 573
 574         /**
 575          * Delete a storage directory if it is empty.
 576          * Backends using key/value stores may do nothing unless the directory
 577          * is that of an empty container, in which case it should be deleted.
 578          *
 579          * $params include:
 580          *     dir       : storage directory
 581          *     recursive : recursively delete empty subdirectories first (@since 1.20)
 582          *
 583          * @param $params Array
 584          * @return Status
 585          */
 586         final public function clean( array $params ) {
 587                 if ( $this->isReadOnly() ) {
 588                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 589                 }
 590                 return $this->doClean( $params );
 591         }
 592
 593         /**
 594          * @see FileBackend::clean()
 595          */
 596         abstract protected function doClean( array $params );
 597
 598         /**
 599          * Check if a file exists at a storage path in the backend.
 600          * This returns false if only a directory exists at the path.
 601          *
 602          * $params include:
 603          *     src    : source storage path
 604          *     latest : use the latest available data
 605          *
 606          * @param $params Array
 607          * @return bool|null Returns null on failure
 608          */
 609         abstract public function fileExists( array $params );
 610
 611         /**
 612          * Get the last-modified timestamp of the file at a storage path.
 613          *
 614          * $params include:
 615          *     src    : source storage path
 616          *     latest : use the latest available data
 617          *
 618          * @param $params Array
 619          * @return string|bool TS_MW timestamp or false on failure
 620          */
 621         abstract public function getFileTimestamp( array $params );
 622
 623         /**
 624          * Get the contents of a file at a storage path in the backend.
 625          * This should be avoided for potentially large files.
 626          *
 627          * $params include:
 628          *     src    : source storage path
 629          *     latest : use the latest available data
 630          *
 631          * @param $params Array
 632          * @return string|bool Returns false on failure
 633          */
 634         abstract public function getFileContents( array $params );
 635
 636         /**
 637          * Get the size (bytes) of a file at a storage path in the backend.
 638          *
 639          * $params include:
 640          *     src    : source storage path
 641          *     latest : use the latest available data
 642          *
 643          * @param $params Array
 644          * @return integer|bool Returns false on failure
 645          */
 646         abstract public function getFileSize( array $params );
 647
 648         /**
 649          * Get quick information about a file at a storage path in the backend.
 650          * If the file does not exist, then this returns false.
 651          * Otherwise, the result is an associative array that includes:
 652          *     mtime  : the last-modified timestamp (TS_MW)
 653          *     size   : the file size (bytes)
 654          * Additional values may be included for internal use only.
 655          *
 656          * $params include:
 657          *     src    : source storage path
 658          *     latest : use the latest available data
 659          *
 660          * @param $params Array
 661          * @return Array|bool|null Returns null on failure
 662          */
 663         abstract public function getFileStat( array $params );
 664
 665         /**
 666          * Get a SHA-1 hash of the file at a storage path in the backend.
 667          *
 668          * $params include:
 669          *     src    : source storage path
 670          *     latest : use the latest available data
 671          *
 672          * @param $params Array
 673          * @return string|bool Hash string or false on failure
 674          */
 675         abstract public function getFileSha1Base36( array $params );
 676
 677         /**
 678          * Get the properties of the file at a storage path in the backend.
 679          * Returns FSFile::placeholderProps() on failure.
 680          *
 681          * $params include:
 682          *     src    : source storage path
 683          *     latest : use the latest available data
 684          *
 685          * @param $params Array
 686          * @return Array
 687          */
 688         abstract public function getFileProps( array $params );
 689
 690         /**
 691          * Stream the file at a storage path in the backend.
 692          * If the file does not exists, a 404 error will be given.
 693          * Appropriate HTTP headers (Status, Content-Type, Content-Length)
 694          * must be sent if streaming began, while none should be sent otherwise.
 695          * Implementations should flush the output buffer before sending data.
 696          *
 697          * $params include:
 698          *     src     : source storage path
 699          *     headers : additional HTTP headers to send on success
 700          *     latest  : use the latest available data
 701          *
 702          * @param $params Array
 703          * @return Status
 704          */
 705         abstract public function streamFile( array $params );
 706
 707         /**
 708          * Returns a file system file, identical to the file at a storage path.
 709          * The file returned is either:
 710          * a) A local copy of the file at a storage path in the backend.
 711          *    The temporary copy will have the same extension as the source.
 712          * b) An original of the file at a storage path in the backend.
 713          * Temporary files may be purged when the file object falls out of scope.
 714          *
 715          * Write operations should *never* be done on this file as some backends
 716          * may do internal tracking or may be instances of FileBackendMultiWrite.
 717          * In that later case, there are copies of the file that must stay in sync.
 718          * Additionally, further calls to this function may return the same file.
 719          *
 720          * $params include:
 721          *     src    : source storage path
 722          *     latest : use the latest available data
 723          *
 724          * @param $params Array
 725          * @return FSFile|null Returns null on failure
 726          */
 727         abstract public function getLocalReference( array $params );
 728
 729         /**
 730          * Get a local copy on disk of the file at a storage path in the backend.
 731          * The temporary copy will have the same file extension as the source.
 732          * Temporary files may be purged when the file object falls out of scope.
 733          *
 734          * $params include:
 735          *     src    : source storage path
 736          *     latest : use the latest available data
 737          *
 738          * @param $params Array
 739          * @return TempFSFile|null Returns null on failure
 740          */
 741         abstract public function getLocalCopy( array $params );
 742
 743         /**
 744          * Check if a directory exists at a given storage path.
 745          * Backends using key/value stores will check if the path is a
 746          * virtual directory, meaning there are files under the given directory.
 747          *
 748          * Storage backends with eventual consistency might return stale data.
 749          *
 750          * $params include:
 751          *     dir : storage directory
 752          *
 753          * @param $params array
 754          * @return bool|null Returns null on failure
 755          * @since 1.20
 756          */
 757         abstract public function directoryExists( array $params );
 758
 759         /**
 760          * Get an iterator to list *all* directories under a storage directory.
 761          * If the directory is of the form "mwstore://backend/container",
 762          * then all directories in the container should be listed.
 763          * If the directory is of form "mwstore://backend/container/dir",
 764          * then all directories directly under that directory should be listed.
 765          * Results should be storage directories relative to the given directory.
 766          *
 767          * Storage backends with eventual consistency might return stale data.
 768          *
 769          * $params include:
 770          *     dir     : storage directory
 771          *     topOnly : only return direct child dirs of the directory
 772          *
 773          * @param $params array
 774          * @return Traversable|Array|null Returns null on failure
 775          * @since 1.20
 776          */
 777         abstract public function getDirectoryList( array $params );
 778
 779         /**
 780          * Same as FileBackend::getDirectoryList() except only lists
 781          * directories that are immediately under the given directory.
 782          *
 783          * Storage backends with eventual consistency might return stale data.
 784          *
 785          * $params include:
 786          *     dir : storage directory
 787          *
 788          * @param $params array
 789          * @return Traversable|Array|null Returns null on failure
 790          * @since 1.20
 791          */
 792         final public function getTopDirectoryList( array $params ) {
 793                 return $this->getDirectoryList( array( 'topOnly' => true ) + $params );
 794         }
 795
 796         /**
 797          * Get an iterator to list *all* stored files under a storage directory.
 798          * If the directory is of the form "mwstore://backend/container",
 799          * then all files in the container should be listed.
 800          * If the directory is of form "mwstore://backend/container/dir",
 801          * then all files under that directory should be listed.
 802          * Results should be storage paths relative to the given directory.
 803          *
 804          * Storage backends with eventual consistency might return stale data.
 805          *
 806          * $params include:
 807          *     dir     : storage directory
 808          *     topOnly : only return direct child files of the directory (@since 1.20)
 809          *
 810          * @param $params array
 811          * @return Traversable|Array|null Returns null on failure
 812          */
 813         abstract public function getFileList( array $params );
 814
 815         /**
 816          * Same as FileBackend::getFileList() except only lists
 817          * files that are immediately under the given directory.
 818          *
 819          * Storage backends with eventual consistency might return stale data.
 820          *
 821          * $params include:
 822          *     dir : storage directory
 823          *
 824          * @param $params array
 825          * @return Traversable|Array|null Returns null on failure
 826          * @since 1.20
 827          */
 828         final public function getTopFileList( array $params ) {
 829                 return $this->getFileList( array( 'topOnly' => true ) + $params );
 830         }
 831
 832         /**
 833          * Invalidate any in-process file existence and property cache.
 834          * If $paths is given, then only the cache for those files will be cleared.
 835          *
 836          * @param $paths Array Storage paths (optional)
 837          * @return void
 838          */
 839         public function clearCache( array $paths = null ) {}
 840
 841         /**
 842          * Lock the files at the given storage paths in the backend.
 843          * This will either lock all the files or none (on failure).
 844          *
 845          * Callers should consider using getScopedFileLocks() instead.
 846          *
 847          * @param $paths Array Storage paths
 848          * @param $type integer LockManager::LOCK_* constant
 849          * @return Status
 850          */
 851         final public function lockFiles( array $paths, $type ) {
 852                 return $this->lockManager->lock( $paths, $type );
 853         }
 854
 855         /**
 856          * Unlock the files at the given storage paths in the backend.
 857          *
 858          * @param $paths Array Storage paths
 859          * @param $type integer LockManager::LOCK_* constant
 860          * @return Status
 861          */
 862         final public function unlockFiles( array $paths, $type ) {
 863                 return $this->lockManager->unlock( $paths, $type );
 864         }
 865
 866         /**
 867          * Lock the files at the given storage paths in the backend.
 868          * This will either lock all the files or none (on failure).
 869          * On failure, the status object will be updated with errors.
 870          *
 871          * Once the return value goes out scope, the locks will be released and
 872          * the status updated. Unlock fatals will not change the status "OK" value.
 873          *
 874          * @param $paths Array Storage paths
 875          * @param $type integer LockManager::LOCK_* constant
 876          * @param $status Status Status to update on lock/unlock
 877          * @return ScopedLock|null Returns null on failure
 878          */
 879         final public function getScopedFileLocks( array $paths, $type, Status $status ) {
 880                 return ScopedLock::factory( $this->lockManager, $paths, $type, $status );
 881         }
 882
 883         /**
 884          * Get an array of scoped locks needed for a batch of file operations.
 885          *
 886          * Normally, FileBackend::doOperations() handles locking, unless
 887          * the 'nonLocking' param is passed in. This function is useful if you
 888          * want the files to be locked for a broader scope than just when the
 889          * files are changing. For example, if you need to update DB metadata,
 890          * you may want to keep the files locked until finished.
 891          *
 892          * @see FileBackend::doOperations()
 893          *
 894          * @param $ops Array List of file operations to FileBackend::doOperations()
 895          * @param $status Status Status to update on lock/unlock
 896          * @return Array List of ScopedFileLocks or null values
 897          * @since 1.20
 898          */
 899         abstract public function getScopedLocksForOps( array $ops, Status $status );
 900
 901         /**
 902          * Get the root storage path of this backend.
 903          * All container paths are "subdirectories" of this path.
 904          *
 905          * @return string Storage path
 906          * @since 1.20
 907          */
 908         final public function getRootStoragePath() {
 909                 return "mwstore://{$this->name}";
 910         }
 911
 912         /**
 913          * Get the file journal object for this backend
 914          *
 915          * @return FileJournal
 916          */
 917         final public function getJournal() {
 918                 return $this->fileJournal;
 919         }
 920
 921         /**
 922          * Check if a given path is a "mwstore://" path.
 923          * This does not do any further validation or any existence checks.
 924          *
 925          * @param $path string
 926          * @return bool
 927          */
 928         final public static function isStoragePath( $path ) {
 929                 return ( strpos( $path, 'mwstore://' ) === 0 );
 930         }
 931
 932         /**
 933          * Split a storage path into a backend name, a container name,
 934          * and a relative file path. The relative path may be the empty string.
 935          * This does not do any path normalization or traversal checks.
 936          *
 937          * @param $storagePath string
 938          * @return Array (backend, container, rel object) or (null, null, null)
 939          */
 940         final public static function splitStoragePath( $storagePath ) {
 941                 if ( self::isStoragePath( $storagePath ) ) {
 942                         // Remove the "mwstore://" prefix and split the path
 943                         $parts = explode( '/', substr( $storagePath, 10 ), 3 );
 944                         if ( count( $parts ) >= 2 && $parts[0] != '' && $parts[1] != '' ) {
 945                                 if ( count( $parts ) == 3 ) {
 946                                         return $parts; // e.g. "backend/container/path"
 947                                 } else {
 948                                         return array( $parts[0], $parts[1], '' ); // e.g. "backend/container"
 949                                 }
 950                         }
 951                 }
 952                 return array( null, null, null );
 953         }
 954
 955         /**
 956          * Normalize a storage path by cleaning up directory separators.
 957          * Returns null if the path is not of the format of a valid storage path.
 958          *
 959          * @param $storagePath string
 960          * @return string|null
 961          */
 962         final public static function normalizeStoragePath( $storagePath ) {
 963                 list( $backend, $container, $relPath ) = self::splitStoragePath( $storagePath );
 964                 if ( $relPath !== null ) { // must be for this backend
 965                         $relPath = self::normalizeContainerPath( $relPath );
 966                         if ( $relPath !== null ) {
 967                                 return ( $relPath != '' )
 968                                         ? "mwstore://{$backend}/{$container}/{$relPath}"
 969                                         : "mwstore://{$backend}/{$container}";
 970                         }
 971                 }
 972                 return null;
 973         }
 974
 975         /**
 976          * Get the parent storage directory of a storage path.
 977          * This returns a path like "mwstore://backend/container",
 978          * "mwstore://backend/container/...", or null if there is no parent.
 979          *
 980          * @param $storagePath string
 981          * @return string|null
 982          */
 983         final public static function parentStoragePath( $storagePath ) {
 984                 $storagePath = dirname( $storagePath );
 985                 list( $b, $cont, $rel ) = self::splitStoragePath( $storagePath );
 986                 return ( $rel === null ) ? null : $storagePath;
 987         }
 988
 989         /**
 990          * Get the final extension from a storage or FS path
 991          *
 992          * @param $path string
 993          * @return string
 994          */
 995         final public static function extensionFromPath( $path ) {
 996                 $i = strrpos( $path, '.' );
 997                 return strtolower( $i ? substr( $path, $i + 1 ) : '' );
 998         }
 999
1000         /**
1001          * Check if a relative path has no directory traversals
1002          *
1003          * @param $path string
1004          * @return bool
1005          * @since 1.20
1006          */
1007         final public static function isPathTraversalFree( $path ) {
1008                 return ( self::normalizeContainerPath( $path ) !== null );
1009         }
1010
1011         /**
1012          * Validate and normalize a relative storage path.
1013          * Null is returned if the path involves directory traversal.
1014          * Traversal is insecure for FS backends and broken for others.
1015          *
1016          * This uses the same traversal protection as Title::secureAndSplit().
1017          *
1018          * @param $path string Storage path relative to a container
1019          * @return string|null
1020          */
1021         final protected static function normalizeContainerPath( $path ) {
1022                 // Normalize directory separators
1023                 $path = strtr( $path, '\\', '/' );
1024                 // Collapse any consecutive directory separators
1025                 $path = preg_replace( '![/]{2,}!', '/', $path );
1026                 // Remove any leading directory separator
1027                 $path = ltrim( $path, '/' );
1028                 // Use the same traversal protection as Title::secureAndSplit()
1029                 if ( strpos( $path, '.' ) !== false ) {
1030                         if (
1031                                 $path === '.' ||
1032                                 $path === '..' ||
1033                                 strpos( $path, './' ) === 0 ||
1034                                 strpos( $path, '../' ) === 0 ||
1035                                 strpos( $path, '/./' ) !== false ||
1036                                 strpos( $path, '/../' ) !== false
1037                         ) {
1038                                 return null;
1039                         }
1040                 }
1041                 return $path;
1042         }
1043 }