includes/filebackend/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)
  41  * notation, though any particular backend may not actually be using a local
  42  * filesystem. 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 of subclasses 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                         ? ( ( $config['fileJournal'] instanceof FileJournal )
 106                                 ? $config['fileJournal']
 107                                 : FileJournal::factory( $config['fileJournal'], $this->name ) )
 108                         : FileJournal::factory( array( 'class' => 'NullFileJournal' ), $this->name );
 109                 $this->readOnly = isset( $config['readOnly'] )
 110                         ? (string)$config['readOnly']
 111                         : '';
 112                 $this->parallelize = isset( $config['parallelize'] )
 113                         ? (string)$config['parallelize']
 114                         : 'off';
 115                 $this->concurrency = isset( $config['concurrency'] )
 116                         ? (int)$config['concurrency']
 117                         : 50;
 118         }
 119
 120         /**
 121          * Get the unique backend name.
 122          * We may have multiple different backends of the same type.
 123          * For example, we can have two Swift backends using different proxies.
 124          *
 125          * @return string
 126          */
 127         final public function getName() {
 128                 return $this->name;
 129         }
 130
 131         /**
 132          * Get the wiki identifier used for this backend (possibly empty).
 133          * Note that this might *not* be in the same format as wfWikiID().
 134          *
 135          * @return string
 136          * @since 1.20
 137          */
 138         final public function getWikiId() {
 139                 return $this->wikiId;
 140         }
 141
 142         /**
 143          * Check if this backend is read-only
 144          *
 145          * @return bool
 146          */
 147         final public function isReadOnly() {
 148                 return ( $this->readOnly != '' );
 149         }
 150
 151         /**
 152          * Get an explanatory message if this backend is read-only
 153          *
 154          * @return string|bool Returns false if the backend is not read-only
 155          */
 156         final public function getReadOnlyReason() {
 157                 return ( $this->readOnly != '' ) ? $this->readOnly : false;
 158         }
 159
 160         /**
 161          * This is the main entry point into the backend for write operations.
 162          * Callers supply an ordered list of operations to perform as a transaction.
 163          * Files will be locked, the stat cache cleared, and then the operations attempted.
 164          * If any serious errors occur, all attempted operations will be rolled back.
 165          *
 166          * $ops is an array of arrays. The outer array holds a list of operations.
 167          * Each inner array is a set of key value pairs that specify an operation.
 168          *
 169          * Supported operations and their parameters. The supported actions are:
 170          *  - create
 171          *  - store
 172          *  - copy
 173          *  - move
 174          *  - delete
 175          *  - describe (since 1.21)
 176          *  - null
 177          *
 178          * a) Create a new file in storage with the contents of a string
 179          * @code
 180          *     array(
 181          *         'op'                  => 'create',
 182          *         'dst'                 => <storage path>,
 183          *         'content'             => <string of new file contents>,
 184          *         'overwrite'           => <boolean>,
 185          *         'overwriteSame'       => <boolean>,
 186          *         'disposition'         => <Content-Disposition header value>,
 187          *         'headers'             => <HTTP header name/value map> # since 1.21
 188          *     );
 189          * @endcode
 190          *
 191          * b) Copy a file system file into storage
 192          * @code
 193          *     array(
 194          *         'op'                  => 'store',
 195          *         'src'                 => <file system path>,
 196          *         'dst'                 => <storage path>,
 197          *         'overwrite'           => <boolean>,
 198          *         'overwriteSame'       => <boolean>,
 199          *         'disposition'         => <Content-Disposition header value>,
 200          *         'headers'             => <HTTP header name/value map> # since 1.21
 201          *     )
 202          * @endcode
 203          *
 204          * c) Copy a file within storage
 205          * @code
 206          *     array(
 207          *         'op'                  => 'copy',
 208          *         'src'                 => <storage path>,
 209          *         'dst'                 => <storage path>,
 210          *         'overwrite'           => <boolean>,
 211          *         'overwriteSame'       => <boolean>,
 212          *         'ignoreMissingSource' => <boolean>, # since 1.21
 213          *         'disposition'         => <Content-Disposition header value>
 214          *     )
 215          * @endcode
 216          *
 217          * d) Move a file within storage
 218          * @code
 219          *     array(
 220          *         'op'                  => 'move',
 221          *         'src'                 => <storage path>,
 222          *         'dst'                 => <storage path>,
 223          *         'overwrite'           => <boolean>,
 224          *         'overwriteSame'       => <boolean>,
 225          *         'ignoreMissingSource' => <boolean>, # since 1.21
 226          *         'disposition'         => <Content-Disposition header value>
 227          *     )
 228          * @endcode
 229          *
 230          * e) Delete a file within storage
 231          * @code
 232          *     array(
 233          *         'op'                  => 'delete',
 234          *         'src'                 => <storage path>,
 235          *         'ignoreMissingSource' => <boolean>
 236          *     )
 237          * @endcode
 238          *
 239          * f) Update metadata for a file within storage
 240          * @code
 241          *     array(
 242          *         'op'                  => 'describe',
 243          *         'src'                 => <storage path>,
 244          *         'disposition'         => <Content-Disposition header value>,
 245          *         'headers'             => <HTTP header name/value map>
 246          *     )
 247          * @endcode
 248          *
 249          * g) Do nothing (no-op)
 250          * @code
 251          *     array(
 252          *         'op'                  => 'null',
 253          *     )
 254          * @endcode
 255          *
 256          * Boolean flags for operations (operation-specific):
 257          *   - ignoreMissingSource : The operation will simply succeed and do
 258          *                           nothing if the source file does not exist.
 259          *   - overwrite           : Any destination file will be overwritten.
 260          *   - overwriteSame       : An error will not be given if a file already
 261          *                           exists at the destination that has the same
 262          *                           contents as the new contents to be written there.
 263          *   - disposition         : If supplied, the backend will return a Content-Disposition
 264          *                           header when GETs/HEADs of the destination file are made.
 265          *                           Backends that don't support metadata ignore this.
 266          *                           See http://tools.ietf.org/html/rfc6266. (since 1.20)
 267          *   - headers             : If supplied, the backend will return these headers when
 268          *                           GETs/HEADs of the destination file are made. Header values
 269          *                           should be smaller than 256 bytes, often options or numbers.
 270          *                           Existing headers will remain, but these will replace any
 271          *                           conflicting previous headers, and headers will be removed
 272          *                           if they are set to an empty string.
 273          *                           Backends that don't support metadata ignore this. (since 1.21)
 274          *
 275          * $opts is an associative of boolean flags, including:
 276          *   - force               : Operation precondition errors no longer trigger an abort.
 277          *                           Any remaining operations are still attempted. Unexpected
 278          *                           failures may still cause remaning operations to be aborted.
 279          *   - nonLocking          : No locks are acquired for the operations.
 280          *                           This can increase performance for non-critical writes.
 281          *                           This has no effect unless the 'force' flag is set.
 282          *   - nonJournaled        : Don't log this operation batch in the file journal.
 283          *                           This limits the ability of recovery scripts.
 284          *   - parallelize         : Try to do operations in parallel when possible.
 285          *   - bypassReadOnly      : Allow writes in read-only mode. (since 1.20)
 286          *   - preserveCache       : Don't clear the process cache before checking files.
 287          *                           This should only be used if all entries in the process
 288          *                           cache were added after the files were already locked. (since 1.20)
 289          *
 290          * @remarks Remarks on locking:
 291          * File system paths given to operations should refer to files that are
 292          * already locked or otherwise safe from modification from other processes.
 293          * Normally these files will be new temp files, which should be adequate.
 294          *
 295          * @par Return value:
 296          *
 297          * This returns a Status, which contains all warnings and fatals that occurred
 298          * during the operation. The 'failCount', 'successCount', and 'success' members
 299          * will reflect each operation attempted.
 300          *
 301          * The status will be "OK" unless:
 302          *   - a) unexpected operation errors occurred (network partitions, disk full...)
 303          *   - b) significant operation errors occurred and 'force' was not set
 304          *
 305          * @param $ops Array List of operations to execute in order
 306          * @param $opts Array Batch operation options
 307          * @return Status
 308          */
 309         final public function doOperations( array $ops, array $opts = array() ) {
 310                 if ( empty( $opts['bypassReadOnly'] ) && $this->isReadOnly() ) {
 311                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 312                 }
 313                 if ( empty( $opts['force'] ) ) { // sanity
 314                         unset( $opts['nonLocking'] );
 315                 }
 316                 return $this->doOperationsInternal( $ops, $opts );
 317         }
 318
 319         /**
 320          * @see FileBackend::doOperations()
 321          */
 322         abstract protected function doOperationsInternal( array $ops, array $opts );
 323
 324         /**
 325          * Same as doOperations() except it takes a single operation.
 326          * If you are doing a batch of operations that should either
 327          * all succeed or all fail, then use that function instead.
 328          *
 329          * @see FileBackend::doOperations()
 330          *
 331          * @param $op Array Operation
 332          * @param $opts Array Operation options
 333          * @return Status
 334          */
 335         final public function doOperation( array $op, array $opts = array() ) {
 336                 return $this->doOperations( array( $op ), $opts );
 337         }
 338
 339         /**
 340          * Performs a single create operation.
 341          * This sets $params['op'] to 'create' and passes it to doOperation().
 342          *
 343          * @see FileBackend::doOperation()
 344          *
 345          * @param $params Array Operation parameters
 346          * @param $opts Array Operation options
 347          * @return Status
 348          */
 349         final public function create( array $params, array $opts = array() ) {
 350                 return $this->doOperation( array( 'op' => 'create' ) + $params, $opts );
 351         }
 352
 353         /**
 354          * Performs a single store operation.
 355          * This sets $params['op'] to 'store' and passes it to doOperation().
 356          *
 357          * @see FileBackend::doOperation()
 358          *
 359          * @param $params Array Operation parameters
 360          * @param $opts Array Operation options
 361          * @return Status
 362          */
 363         final public function store( array $params, array $opts = array() ) {
 364                 return $this->doOperation( array( 'op' => 'store' ) + $params, $opts );
 365         }
 366
 367         /**
 368          * Performs a single copy operation.
 369          * This sets $params['op'] to 'copy' and passes it to doOperation().
 370          *
 371          * @see FileBackend::doOperation()
 372          *
 373          * @param $params Array Operation parameters
 374          * @param $opts Array Operation options
 375          * @return Status
 376          */
 377         final public function copy( array $params, array $opts = array() ) {
 378                 return $this->doOperation( array( 'op' => 'copy' ) + $params, $opts );
 379         }
 380
 381         /**
 382          * Performs a single move operation.
 383          * This sets $params['op'] to 'move' and passes it to doOperation().
 384          *
 385          * @see FileBackend::doOperation()
 386          *
 387          * @param $params Array Operation parameters
 388          * @param $opts Array Operation options
 389          * @return Status
 390          */
 391         final public function move( array $params, array $opts = array() ) {
 392                 return $this->doOperation( array( 'op' => 'move' ) + $params, $opts );
 393         }
 394
 395         /**
 396          * Performs a single delete operation.
 397          * This sets $params['op'] to 'delete' and passes it to doOperation().
 398          *
 399          * @see FileBackend::doOperation()
 400          *
 401          * @param $params Array Operation parameters
 402          * @param $opts Array Operation options
 403          * @return Status
 404          */
 405         final public function delete( array $params, array $opts = array() ) {
 406                 return $this->doOperation( array( 'op' => 'delete' ) + $params, $opts );
 407         }
 408
 409         /**
 410          * Performs a single describe operation.
 411          * This sets $params['op'] to 'describe' and passes it to doOperation().
 412          *
 413          * @see FileBackend::doOperation()
 414          *
 415          * @param $params Array Operation parameters
 416          * @param $opts Array Operation options
 417          * @return Status
 418          * @since 1.21
 419          */
 420         final public function describe( array $params, array $opts = array() ) {
 421                 return $this->doOperation( array( 'op' => 'describe' ) + $params, $opts );
 422         }
 423
 424         /**
 425          * Perform a set of independent file operations on some files.
 426          *
 427          * This does no locking, nor journaling, and possibly no stat calls.
 428          * Any destination files that already exist will be overwritten.
 429          * This should *only* be used on non-original files, like cache files.
 430          *
 431          * Supported operations and their parameters:
 432          *  - create
 433          *  - store
 434          *  - copy
 435          *  - move
 436          *  - delete
 437          *  - describe (since 1.21)
 438          *  - null
 439          *
 440          * a) Create a new file in storage with the contents of a string
 441          * @code
 442          *     array(
 443          *         'op'                  => 'create',
 444          *         'dst'                 => <storage path>,
 445          *         'content'             => <string of new file contents>,
 446          *         'disposition'         => <Content-Disposition header value>,
 447          *         'headers'             => <HTTP header name/value map> # since 1.21
 448          *     )
 449          * @endcode
 450          *
 451          * b) Copy a file system file into storage
 452          * @code
 453          *     array(
 454          *         'op'                  => 'store',
 455          *         'src'                 => <file system path>,
 456          *         'dst'                 => <storage path>,
 457          *         'disposition'         => <Content-Disposition header value>,
 458          *         'headers'             => <HTTP header name/value map> # since 1.21
 459          *     )
 460          * @endcode
 461          *
 462          * c) Copy a file within storage
 463          * @code
 464          *     array(
 465          *         'op'                  => 'copy',
 466          *         'src'                 => <storage path>,
 467          *         'dst'                 => <storage path>,
 468          *         'ignoreMissingSource' => <boolean>, # since 1.21
 469          *         'disposition'         => <Content-Disposition header value>
 470          *     )
 471          * @endcode
 472          *
 473          * d) Move a file within storage
 474          * @code
 475          *     array(
 476          *         'op'                  => 'move',
 477          *         'src'                 => <storage path>,
 478          *         'dst'                 => <storage path>,
 479          *         'ignoreMissingSource' => <boolean>, # since 1.21
 480          *         'disposition'         => <Content-Disposition header value>
 481          *     )
 482          * @endcode
 483          *
 484          * e) Delete a file within storage
 485          * @code
 486          *     array(
 487          *         'op'                  => 'delete',
 488          *         'src'                 => <storage path>,
 489          *         'ignoreMissingSource' => <boolean>
 490          *     )
 491          * @endcode
 492          *
 493          * f) Update metadata for a file within storage
 494          * @code
 495          *     array(
 496          *         'op'                  => 'describe',
 497          *         'src'                 => <storage path>,
 498          *         'disposition'         => <Content-Disposition header value>,
 499          *         'headers'             => <HTTP header name/value map>
 500          *     )
 501          * @endcode
 502          *
 503          * g) Do nothing (no-op)
 504          * @code
 505          *     array(
 506          *         'op'                  => 'null',
 507          *     )
 508          * @endcode
 509          *
 510          * @par Boolean flags for operations (operation-specific):
 511          *   - ignoreMissingSource : The operation will simply succeed and do
 512          *                           nothing if the source file does not exist.
 513          *   - disposition         : When supplied, the backend will add a Content-Disposition
 514          *                           header when GETs/HEADs of the destination file are made.
 515          *                           Backends that don't support file metadata will ignore this.
 516          *                           See http://tools.ietf.org/html/rfc6266 (since 1.20).
 517          *   - headers             : If supplied with a header name/value map, the backend will
 518          *                           reply with these headers when GETs/HEADs of the destination
 519          *                           file are made. Header values should be smaller than 256 bytes.
 520          *                           Existing headers will remain, but these will replace any
 521          *                           conflicting previous headers, and headers will be removed
 522          *                           if they are set to an empty string.
 523          *                           Backends that don't support metadata ignore this. (since 1.21)
 524          *
 525          * $opts is an associative of boolean flags, including:
 526          *   - bypassReadOnly      : Allow writes in read-only mode (since 1.20)
 527          *
 528          * @par Return value:
 529          * This returns a Status, which contains all warnings and fatals that occurred
 530          * during the operation. The 'failCount', 'successCount', and 'success' members
 531          * will reflect each operation attempted for the given files. The status will be
 532          * considered "OK" as long as no fatal errors occurred.
 533          *
 534          * @param $ops Array Set of operations to execute
 535          * @param $opts Array Batch operation options
 536          * @return Status
 537          * @since 1.20
 538          */
 539         final public function doQuickOperations( array $ops, array $opts = array() ) {
 540                 if ( empty( $opts['bypassReadOnly'] ) && $this->isReadOnly() ) {
 541                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 542                 }
 543                 foreach ( $ops as &$op ) {
 544                         $op['overwrite'] = true; // avoids RTTs in key/value stores
 545                 }
 546                 return $this->doQuickOperationsInternal( $ops );
 547         }
 548
 549         /**
 550          * @see FileBackend::doQuickOperations()
 551          * @since 1.20
 552          */
 553         abstract protected function doQuickOperationsInternal( array $ops );
 554
 555         /**
 556          * Same as doQuickOperations() except it takes a single operation.
 557          * If you are doing a batch of operations, then use that function instead.
 558          *
 559          * @see FileBackend::doQuickOperations()
 560          *
 561          * @param $op Array Operation
 562          * @return Status
 563          * @since 1.20
 564          */
 565         final public function doQuickOperation( array $op ) {
 566                 return $this->doQuickOperations( array( $op ) );
 567         }
 568
 569         /**
 570          * Performs a single quick create operation.
 571          * This sets $params['op'] to 'create' and passes it to doQuickOperation().
 572          *
 573          * @see FileBackend::doQuickOperation()
 574          *
 575          * @param $params Array Operation parameters
 576          * @return Status
 577          * @since 1.20
 578          */
 579         final public function quickCreate( array $params ) {
 580                 return $this->doQuickOperation( array( 'op' => 'create' ) + $params );
 581         }
 582
 583         /**
 584          * Performs a single quick store operation.
 585          * This sets $params['op'] to 'store' and passes it to doQuickOperation().
 586          *
 587          * @see FileBackend::doQuickOperation()
 588          *
 589          * @param $params Array Operation parameters
 590          * @return Status
 591          * @since 1.20
 592          */
 593         final public function quickStore( array $params ) {
 594                 return $this->doQuickOperation( array( 'op' => 'store' ) + $params );
 595         }
 596
 597         /**
 598          * Performs a single quick copy operation.
 599          * This sets $params['op'] to 'copy' and passes it to doQuickOperation().
 600          *
 601          * @see FileBackend::doQuickOperation()
 602          *
 603          * @param $params Array Operation parameters
 604          * @return Status
 605          * @since 1.20
 606          */
 607         final public function quickCopy( array $params ) {
 608                 return $this->doQuickOperation( array( 'op' => 'copy' ) + $params );
 609         }
 610
 611         /**
 612          * Performs a single quick move operation.
 613          * This sets $params['op'] to 'move' and passes it to doQuickOperation().
 614          *
 615          * @see FileBackend::doQuickOperation()
 616          *
 617          * @param $params Array Operation parameters
 618          * @return Status
 619          * @since 1.20
 620          */
 621         final public function quickMove( array $params ) {
 622                 return $this->doQuickOperation( array( 'op' => 'move' ) + $params );
 623         }
 624
 625         /**
 626          * Performs a single quick delete operation.
 627          * This sets $params['op'] to 'delete' and passes it to doQuickOperation().
 628          *
 629          * @see FileBackend::doQuickOperation()
 630          *
 631          * @param $params Array Operation parameters
 632          * @return Status
 633          * @since 1.20
 634          */
 635         final public function quickDelete( array $params ) {
 636                 return $this->doQuickOperation( array( 'op' => 'delete' ) + $params );
 637         }
 638
 639         /**
 640          * Performs a single quick describe operation.
 641          * This sets $params['op'] to 'describe' and passes it to doQuickOperation().
 642          *
 643          * @see FileBackend::doQuickOperation()
 644          *
 645          * @param $params Array Operation parameters
 646          * @return Status
 647          * @since 1.21
 648          */
 649         final public function quickDescribe( array $params ) {
 650                 return $this->doQuickOperation( array( 'op' => 'describe' ) + $params );
 651         }
 652
 653         /**
 654          * Concatenate a list of storage files into a single file system file.
 655          * The target path should refer to a file that is already locked or
 656          * otherwise safe from modification from other processes. Normally,
 657          * the file will be a new temp file, which should be adequate.
 658          *
 659          * @param $params Array Operation parameters
 660          * $params include:
 661          *   - srcs        : ordered source storage paths (e.g. chunk1, chunk2, ...)
 662          *   - dst         : file system path to 0-byte temp file
 663          *   - parallelize : try to do operations in parallel when possible
 664          * @return Status
 665          */
 666         abstract public function concatenate( array $params );
 667
 668         /**
 669          * Prepare a storage directory for usage.
 670          * This will create any required containers and parent directories.
 671          * Backends using key/value stores only need to create the container.
 672          *
 673          * The 'noAccess' and 'noListing' parameters works the same as in secure(),
 674          * except they are only applied *if* the directory/container had to be created.
 675          * These flags should always be set for directories that have private files.
 676          *
 677          * @param $params Array
 678          * $params include:
 679          *   - dir            : storage directory
 680          *   - noAccess       : try to deny file access (since 1.20)
 681          *   - noListing      : try to deny file listing (since 1.20)
 682          *   - bypassReadOnly : allow writes in read-only mode (since 1.20)
 683          * @return Status
 684          */
 685         final public function prepare( array $params ) {
 686                 if ( empty( $params['bypassReadOnly'] ) && $this->isReadOnly() ) {
 687                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 688                 }
 689                 return $this->doPrepare( $params );
 690         }
 691
 692         /**
 693          * @see FileBackend::prepare()
 694          */
 695         abstract protected function doPrepare( array $params );
 696
 697         /**
 698          * Take measures to block web access to a storage directory and
 699          * the container it belongs to. FS backends might add .htaccess
 700          * files whereas key/value store backends might revoke container
 701          * access to the storage user representing end-users in web requests.
 702          * This is not guaranteed to actually do anything.
 703          *
 704          * @param $params Array
 705          * $params include:
 706          *   - dir            : storage directory
 707          *   - noAccess       : try to deny file access
 708          *   - noListing      : try to deny file listing
 709          *   - bypassReadOnly : allow writes in read-only mode (since 1.20)
 710          * @return Status
 711          */
 712         final public function secure( array $params ) {
 713                 if ( empty( $params['bypassReadOnly'] ) && $this->isReadOnly() ) {
 714                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 715                 }
 716                 return $this->doSecure( $params );
 717         }
 718
 719         /**
 720          * @see FileBackend::secure()
 721          */
 722         abstract protected function doSecure( array $params );
 723
 724         /**
 725          * Remove measures to block web access to a storage directory and
 726          * the container it belongs to. FS backends might remove .htaccess
 727          * files whereas key/value store backends might grant container
 728          * access to the storage user representing end-users in web requests.
 729          * This essentially can undo the result of secure() calls.
 730          *
 731          * @param $params Array
 732          * $params include:
 733          *   - dir            : storage directory
 734          *   - access         : try to allow file access
 735          *   - listing        : try to allow file listing
 736          *   - bypassReadOnly : allow writes in read-only mode (since 1.20)
 737          * @return Status
 738          * @since 1.20
 739          */
 740         final public function publish( array $params ) {
 741                 if ( empty( $params['bypassReadOnly'] ) && $this->isReadOnly() ) {
 742                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 743                 }
 744                 return $this->doPublish( $params );
 745         }
 746
 747         /**
 748          * @see FileBackend::publish()
 749          */
 750         abstract protected function doPublish( array $params );
 751
 752         /**
 753          * Delete a storage directory if it is empty.
 754          * Backends using key/value stores may do nothing unless the directory
 755          * is that of an empty container, in which case it will be deleted.
 756          *
 757          * @param $params Array
 758          * $params include:
 759          *   - dir            : storage directory
 760          *   - recursive      : recursively delete empty subdirectories first (since 1.20)
 761          *   - bypassReadOnly : allow writes in read-only mode (since 1.20)
 762          * @return Status
 763          */
 764         final public function clean( array $params ) {
 765                 if ( empty( $params['bypassReadOnly'] ) && $this->isReadOnly() ) {
 766                         return Status::newFatal( 'backend-fail-readonly', $this->name, $this->readOnly );
 767                 }
 768                 return $this->doClean( $params );
 769         }
 770
 771         /**
 772          * @see FileBackend::clean()
 773          */
 774         abstract protected function doClean( array $params );
 775
 776         /**
 777          * Check if a file exists at a storage path in the backend.
 778          * This returns false if only a directory exists at the path.
 779          *
 780          * @param $params Array
 781          * $params include:
 782          *   - src    : source storage path
 783          *   - latest : use the latest available data
 784          * @return bool|null Returns null on failure
 785          */
 786         abstract public function fileExists( array $params );
 787
 788         /**
 789          * Get the last-modified timestamp of the file at a storage path.
 790          *
 791          * @param $params Array
 792          * $params include:
 793          *   - src    : source storage path
 794          *   - latest : use the latest available data
 795          * @return string|bool TS_MW timestamp or false on failure
 796          */
 797         abstract public function getFileTimestamp( array $params );
 798
 799         /**
 800          * Get the contents of a file at a storage path in the backend.
 801          * This should be avoided for potentially large files.
 802          *
 803          * @param $params Array
 804          * $params include:
 805          *   - src    : source storage path
 806          *   - latest : use the latest available data
 807          * @return string|bool Returns false on failure
 808          */
 809         final public function getFileContents( array $params ) {
 810                 $contents = $this->getFileContentsMulti(
 811                         array( 'srcs' => array( $params['src'] ) ) + $params );
 812
 813                 return $contents[$params['src']];
 814         }
 815
 816         /**
 817          * Like getFileContents() except it takes an array of storage paths
 818          * and returns a map of storage paths to strings (or null on failure).
 819          * The map keys (paths) are in the same order as the provided list of paths.
 820          *
 821          * @see FileBackend::getFileContents()
 822          *
 823          * @param $params Array
 824          * $params include:
 825          *   - srcs        : list of source storage paths
 826          *   - latest      : use the latest available data
 827          *   - parallelize : try to do operations in parallel when possible
 828          * @return Array Map of (path name => string or false on failure)
 829          * @since 1.20
 830          */
 831         abstract public function getFileContentsMulti( array $params );
 832
 833         /**
 834          * Get the size (bytes) of a file at a storage path in the backend.
 835          *
 836          * @param $params Array
 837          * $params include:
 838          *   - src    : source storage path
 839          *   - latest : use the latest available data
 840          * @return integer|bool Returns false on failure
 841          */
 842         abstract public function getFileSize( array $params );
 843
 844         /**
 845          * Get quick information about a file at a storage path in the backend.
 846          * If the file does not exist, then this returns false.
 847          * Otherwise, the result is an associative array that includes:
 848          *   - mtime  : the last-modified timestamp (TS_MW)
 849          *   - size   : the file size (bytes)
 850          * Additional values may be included for internal use only.
 851          *
 852          * @param $params Array
 853          * $params include:
 854          *   - src    : source storage path
 855          *   - latest : use the latest available data
 856          * @return Array|bool|null Returns null on failure
 857          */
 858         abstract public function getFileStat( array $params );
 859
 860         /**
 861          * Get a SHA-1 hash of the file at a storage path in the backend.
 862          *
 863          * @param $params Array
 864          * $params include:
 865          *   - src    : source storage path
 866          *   - latest : use the latest available data
 867          * @return string|bool Hash string or false on failure
 868          */
 869         abstract public function getFileSha1Base36( array $params );
 870
 871         /**
 872          * Get the properties of the file at a storage path in the backend.
 873          * This gives the result of FSFile::getProps() on a local copy of the file.
 874          *
 875          * @param $params Array
 876          * $params include:
 877          *   - src    : source storage path
 878          *   - latest : use the latest available data
 879          * @return Array Returns FSFile::placeholderProps() on failure
 880          */
 881         abstract public function getFileProps( array $params );
 882
 883         /**
 884          * Stream the file at a storage path in the backend.
 885          * If the file does not exists, an HTTP 404 error will be given.
 886          * Appropriate HTTP headers (Status, Content-Type, Content-Length)
 887          * will be sent if streaming began, while none will be sent otherwise.
 888          * Implementations should flush the output buffer before sending data.
 889          *
 890          * @param $params Array
 891          * $params include:
 892          *   - src     : source storage path
 893          *   - headers : list of additional HTTP headers to send on success
 894          *   - latest  : use the latest available data
 895          * @return Status
 896          */
 897         abstract public function streamFile( array $params );
 898
 899         /**
 900          * Returns a file system file, identical to the file at a storage path.
 901          * The file returned is either:
 902          *   - a) A local copy of the file at a storage path in the backend.
 903          *        The temporary copy will have the same extension as the source.
 904          *   - b) An original of the file at a storage path in the backend.
 905          * Temporary files may be purged when the file object falls out of scope.
 906          *
 907          * Write operations should *never* be done on this file as some backends
 908          * may do internal tracking or may be instances of FileBackendMultiWrite.
 909          * In that later case, there are copies of the file that must stay in sync.
 910          * Additionally, further calls to this function may return the same file.
 911          *
 912          * @param $params Array
 913          * $params include:
 914          *   - src    : source storage path
 915          *   - latest : use the latest available data
 916          * @return FSFile|null Returns null on failure
 917          */
 918         final public function getLocalReference( array $params ) {
 919                 $fsFiles = $this->getLocalReferenceMulti(
 920                         array( 'srcs' => array( $params['src'] ) ) + $params );
 921
 922                 return $fsFiles[$params['src']];
 923         }
 924
 925         /**
 926          * Like getLocalReference() except it takes an array of storage paths
 927          * and returns a map of storage paths to FSFile objects (or null on failure).
 928          * The map keys (paths) are in the same order as the provided list of paths.
 929          *
 930          * @see FileBackend::getLocalReference()
 931          *
 932          * @param $params Array
 933          * $params include:
 934          *   - srcs        : list of source storage paths
 935          *   - latest      : use the latest available data
 936          *   - parallelize : try to do operations in parallel when possible
 937          * @return Array Map of (path name => FSFile or null on failure)
 938          * @since 1.20
 939          */
 940         abstract public function getLocalReferenceMulti( array $params );
 941
 942         /**
 943          * Get a local copy on disk of the file at a storage path in the backend.
 944          * The temporary copy will have the same file extension as the source.
 945          * Temporary files may be purged when the file object falls out of scope.
 946          *
 947          * @param $params Array
 948          * $params include:
 949          *   - src    : source storage path
 950          *   - latest : use the latest available data
 951          * @return TempFSFile|null Returns null on failure
 952          */
 953         final public function getLocalCopy( array $params ) {
 954                 $tmpFiles = $this->getLocalCopyMulti(
 955                         array( 'srcs' => array( $params['src'] ) ) + $params );
 956
 957                 return $tmpFiles[$params['src']];
 958         }
 959
 960         /**
 961          * Like getLocalCopy() except it takes an array of storage paths and
 962          * returns a map of storage paths to TempFSFile objects (or null on failure).
 963          * The map keys (paths) are in the same order as the provided list of paths.
 964          *
 965          * @see FileBackend::getLocalCopy()
 966          *
 967          * @param $params Array
 968          * $params include:
 969          *   - srcs        : list of source storage paths
 970          *   - latest      : use the latest available data
 971          *   - parallelize : try to do operations in parallel when possible
 972          * @return Array Map of (path name => TempFSFile or null on failure)
 973          * @since 1.20
 974          */
 975         abstract public function getLocalCopyMulti( array $params );
 976
 977         /**
 978          * Return an HTTP URL to a given file that requires no authentication to use.
 979          * The URL may be pre-authenticated (via some token in the URL) and temporary.
 980          * This will return null if the backend cannot make an HTTP URL for the file.
 981          *
 982          * This is useful for key/value stores when using scripts that seek around
 983          * large files and those scripts (and the backend) support HTTP Range headers.
 984          * Otherwise, one would need to use getLocalReference(), which involves loading
 985          * the entire file on to local disk.
 986          *
 987          * @param $params Array
 988          * $params include:
 989          *   - src : source storage path
 990          * @return string|null
 991          * @since 1.21
 992          */
 993         abstract public function getFileHttpUrl( array $params );
 994
 995         /**
 996          * Check if a directory exists at a given storage path.
 997          * Backends using key/value stores will check if the path is a
 998          * virtual directory, meaning there are files under the given directory.
 999          *
1000          * Storage backends with eventual consistency might return stale data.
1001          *
1002          * @param $params array
1003          * $params include:
1004          *   - dir : storage directory
1005          * @return bool|null Returns null on failure
1006          * @since 1.20
1007          */
1008         abstract public function directoryExists( array $params );
1009
1010         /**
1011          * Get an iterator to list *all* directories under a storage directory.
1012          * If the directory is of the form "mwstore://backend/container",
1013          * then all directories in the container will be listed.
1014          * If the directory is of form "mwstore://backend/container/dir",
1015          * then all directories directly under that directory will be listed.
1016          * Results will be storage directories relative to the given directory.
1017          *
1018          * Storage backends with eventual consistency might return stale data.
1019          *
1020          * @param $params array
1021          * $params include:
1022          *   - dir     : storage directory
1023          *   - topOnly : only return direct child dirs of the directory
1024          * @return Traversable|Array|null Returns null on failure
1025          * @since 1.20
1026          */
1027         abstract public function getDirectoryList( array $params );
1028
1029         /**
1030          * Same as FileBackend::getDirectoryList() except only lists
1031          * directories that are immediately under the given directory.
1032          *
1033          * Storage backends with eventual consistency might return stale data.
1034          *
1035          * @param $params array
1036          * $params include:
1037          *   - dir : storage directory
1038          * @return Traversable|Array|null Returns null on failure
1039          * @since 1.20
1040          */
1041         final public function getTopDirectoryList( array $params ) {
1042                 return $this->getDirectoryList( array( 'topOnly' => true ) + $params );
1043         }
1044
1045         /**
1046          * Get an iterator to list *all* stored files under a storage directory.
1047          * If the directory is of the form "mwstore://backend/container",
1048          * then all files in the container will be listed.
1049          * If the directory is of form "mwstore://backend/container/dir",
1050          * then all files under that directory will be listed.
1051          * Results will be storage paths relative to the given directory.
1052          *
1053          * Storage backends with eventual consistency might return stale data.
1054          *
1055          * @param $params array
1056          * $params include:
1057          *   - dir     : storage directory
1058          *   - topOnly : only return direct child files of the directory (since 1.20)
1059          * @return Traversable|Array|null Returns null on failure
1060          */
1061         abstract public function getFileList( array $params );
1062
1063         /**
1064          * Same as FileBackend::getFileList() except only lists
1065          * files that are immediately under the given directory.
1066          *
1067          * Storage backends with eventual consistency might return stale data.
1068          *
1069          * @param $params array
1070          * $params include:
1071          *   - dir : storage directory
1072          * @return Traversable|Array|null Returns null on failure
1073          * @since 1.20
1074          */
1075         final public function getTopFileList( array $params ) {
1076                 return $this->getFileList( array( 'topOnly' => true ) + $params );
1077         }
1078
1079         /**
1080          * Preload persistent file stat and property cache into in-process cache.
1081          * This should be used when stat calls will be made on a known list of a many files.
1082          *
1083          * @param $paths Array Storage paths
1084          * @return void
1085          */
1086         public function preloadCache( array $paths ) {}
1087
1088         /**
1089          * Invalidate any in-process file stat and property cache.
1090          * If $paths is given, then only the cache for those files will be cleared.
1091          *
1092          * @param $paths Array Storage paths (optional)
1093          * @return void
1094          */
1095         public function clearCache( array $paths = null ) {}
1096
1097         /**
1098          * Lock the files at the given storage paths in the backend.
1099          * This will either lock all the files or none (on failure).
1100          *
1101          * Callers should consider using getScopedFileLocks() instead.
1102          *
1103          * @param $paths Array Storage paths
1104          * @param $type integer LockManager::LOCK_* constant
1105          * @return Status
1106          */
1107         final public function lockFiles( array $paths, $type ) {
1108                 return $this->lockManager->lock( $paths, $type );
1109         }
1110
1111         /**
1112          * Unlock the files at the given storage paths in the backend.
1113          *
1114          * @param $paths Array Storage paths
1115          * @param $type integer LockManager::LOCK_* constant
1116          * @return Status
1117          */
1118         final public function unlockFiles( array $paths, $type ) {
1119                 return $this->lockManager->unlock( $paths, $type );
1120         }
1121
1122         /**
1123          * Lock the files at the given storage paths in the backend.
1124          * This will either lock all the files or none (on failure).
1125          * On failure, the status object will be updated with errors.
1126          *
1127          * Once the return value goes out scope, the locks will be released and
1128          * the status updated. Unlock fatals will not change the status "OK" value.
1129          *
1130          * @param $paths Array Storage paths
1131          * @param $type integer LockManager::LOCK_* constant
1132          * @param $status Status Status to update on lock/unlock
1133          * @return ScopedLock|null Returns null on failure
1134          */
1135         final public function getScopedFileLocks( array $paths, $type, Status $status ) {
1136                 return ScopedLock::factory( $this->lockManager, $paths, $type, $status );
1137         }
1138
1139         /**
1140          * Get an array of scoped locks needed for a batch of file operations.
1141          *
1142          * Normally, FileBackend::doOperations() handles locking, unless
1143          * the 'nonLocking' param is passed in. This function is useful if you
1144          * want the files to be locked for a broader scope than just when the
1145          * files are changing. For example, if you need to update DB metadata,
1146          * you may want to keep the files locked until finished.
1147          *
1148          * @see FileBackend::doOperations()
1149          *
1150          * @param $ops Array List of file operations to FileBackend::doOperations()
1151          * @param $status Status Status to update on lock/unlock
1152          * @return Array List of ScopedFileLocks or null values
1153          * @since 1.20
1154          */
1155         abstract public function getScopedLocksForOps( array $ops, Status $status );
1156
1157         /**
1158          * Get the root storage path of this backend.
1159          * All container paths are "subdirectories" of this path.
1160          *
1161          * @return string Storage path
1162          * @since 1.20
1163          */
1164         final public function getRootStoragePath() {
1165                 return "mwstore://{$this->name}";
1166         }
1167
1168         /**
1169          * Get the storage path for the given container for this backend
1170          *
1171          * @param $container string Container name
1172          * @return string Storage path
1173          * @since 1.21
1174          */
1175         final public function getContainerStoragePath( $container ) {
1176                 return $this->getRootStoragePath() . "/{$container}";
1177         }
1178
1179         /**
1180          * Get the file journal object for this backend
1181          *
1182          * @return FileJournal
1183          */
1184         final public function getJournal() {
1185                 return $this->fileJournal;
1186         }
1187
1188         /**
1189          * Check if a given path is a "mwstore://" path.
1190          * This does not do any further validation or any existence checks.
1191          *
1192          * @param $path string
1193          * @return bool
1194          */
1195         final public static function isStoragePath( $path ) {
1196                 return ( strpos( $path, 'mwstore://' ) === 0 );
1197         }
1198
1199         /**
1200          * Split a storage path into a backend name, a container name,
1201          * and a relative file path. The relative path may be the empty string.
1202          * This does not do any path normalization or traversal checks.
1203          *
1204          * @param $storagePath string
1205          * @return Array (backend, container, rel object) or (null, null, null)
1206          */
1207         final public static function splitStoragePath( $storagePath ) {
1208                 if ( self::isStoragePath( $storagePath ) ) {
1209                         // Remove the "mwstore://" prefix and split the path
1210                         $parts = explode( '/', substr( $storagePath, 10 ), 3 );
1211                         if ( count( $parts ) >= 2 && $parts[0] != '' && $parts[1] != '' ) {
1212                                 if ( count( $parts ) == 3 ) {
1213                                         return $parts; // e.g. "backend/container/path"
1214                                 } else {
1215                                         return array( $parts[0], $parts[1], '' ); // e.g. "backend/container"
1216                                 }
1217                         }
1218                 }
1219                 return array( null, null, null );
1220         }
1221
1222         /**
1223          * Normalize a storage path by cleaning up directory separators.
1224          * Returns null if the path is not of the format of a valid storage path.
1225          *
1226          * @param $storagePath string
1227          * @return string|null
1228          */
1229         final public static function normalizeStoragePath( $storagePath ) {
1230                 list( $backend, $container, $relPath ) = self::splitStoragePath( $storagePath );
1231                 if ( $relPath !== null ) { // must be for this backend
1232                         $relPath = self::normalizeContainerPath( $relPath );
1233                         if ( $relPath !== null ) {
1234                                 return ( $relPath != '' )
1235                                         ? "mwstore://{$backend}/{$container}/{$relPath}"
1236                                         : "mwstore://{$backend}/{$container}";
1237                         }
1238                 }
1239                 return null;
1240         }
1241
1242         /**
1243          * Get the parent storage directory of a storage path.
1244          * This returns a path like "mwstore://backend/container",
1245          * "mwstore://backend/container/...", or null if there is no parent.
1246          *
1247          * @param $storagePath string
1248          * @return string|null
1249          */
1250         final public static function parentStoragePath( $storagePath ) {
1251                 $storagePath = dirname( $storagePath );
1252                 list( $b, $cont, $rel ) = self::splitStoragePath( $storagePath );
1253                 return ( $rel === null ) ? null : $storagePath;
1254         }
1255
1256         /**
1257          * Get the final extension from a storage or FS path
1258          *
1259          * @param $path string
1260          * @return string
1261          */
1262         final public static function extensionFromPath( $path ) {
1263                 $i = strrpos( $path, '.' );
1264                 return strtolower( $i ? substr( $path, $i + 1 ) : '' );
1265         }
1266
1267         /**
1268          * Check if a relative path has no directory traversals
1269          *
1270          * @param $path string
1271          * @return bool
1272          * @since 1.20
1273          */
1274         final public static function isPathTraversalFree( $path ) {
1275                 return ( self::normalizeContainerPath( $path ) !== null );
1276         }
1277
1278         /**
1279          * Build a Content-Disposition header value per RFC 6266.
1280          *
1281          * @param $type string One of (attachment, inline)
1282          * @param $filename string Suggested file name (should not contain slashes)
1283          * @throws MWException
1284          * @return string
1285          * @since 1.20
1286          */
1287         final public static function makeContentDisposition( $type, $filename = '' ) {
1288                 $parts = array();
1289
1290                 $type = strtolower( $type );
1291                 if ( !in_array( $type, array( 'inline', 'attachment' ) ) ) {
1292                         throw new MWException( "Invalid Content-Disposition type '$type'." );
1293                 }
1294                 $parts[] = $type;
1295
1296                 if ( strlen( $filename ) ) {
1297                         $parts[] = "filename*=UTF-8''" . rawurlencode( basename( $filename ) );
1298                 }
1299
1300                 return implode( ';', $parts );
1301         }
1302
1303         /**
1304          * Validate and normalize a relative storage path.
1305          * Null is returned if the path involves directory traversal.
1306          * Traversal is insecure for FS backends and broken for others.
1307          *
1308          * This uses the same traversal protection as Title::secureAndSplit().
1309          *
1310          * @param $path string Storage path relative to a container
1311          * @return string|null
1312          */
1313         final protected static function normalizeContainerPath( $path ) {
1314                 // Normalize directory separators
1315                 $path = strtr( $path, '\\', '/' );
1316                 // Collapse any consecutive directory separators
1317                 $path = preg_replace( '![/]{2,}!', '/', $path );
1318                 // Remove any leading directory separator
1319                 $path = ltrim( $path, '/' );
1320                 // Use the same traversal protection as Title::secureAndSplit()
1321                 if ( strpos( $path, '.' ) !== false ) {
1322                         if (
1323                                 $path === '.' ||
1324                                 $path === '..' ||
1325                                 strpos( $path, './' ) === 0 ||
1326                                 strpos( $path, '../' ) === 0 ||
1327                                 strpos( $path, '/./' ) !== false ||
1328                                 strpos( $path, '/../' ) !== false
1329                         ) {
1330                                 return null;
1331                         }
1332                 }
1333                 return $path;
1334         }
1335 }