includes/utils/ZipDirectoryReader.php

   1 <?php
   2 /**
   3  * ZIP file directories reader, for the purposes of upload verification.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  */
  22
  23 /**
  24  * A class for reading ZIP file directories, for the purposes of upload
  25  * verification.
  26  *
  27  * Only a functional interface is provided: ZipFileReader::read(). No access is
  28  * given to object instances.
  29  */
  30 class ZipDirectoryReader {
  31         /**
  32          * Read a ZIP file and call a function for each file discovered in it.
  33          *
  34          * Because this class is aimed at verification, an error is raised on
  35          * suspicious or ambiguous input, instead of emulating some standard
  36          * behavior.
  37          *
  38          * @param string $fileName The archive file name
  39          * @param array $callback The callback function. It will be called for each file
  40          *   with a single associative array each time, with members:
  41          *
  42          *      - name: The file name. Directories conventionally have a trailing
  43          *        slash.
  44          *
  45          *      - mtime: The file modification time, in MediaWiki 14-char format
  46          *
  47          *      - size: The uncompressed file size
  48          *
  49          * @param array $options An associative array of read options, with the option
  50          *    name in the key. This may currently contain:
  51          *
  52          *      - zip64: If this is set to true, then we will emulate a
  53          *        library with ZIP64 support, like OpenJDK 7. If it is set to
  54          *        false, then we will emulate a library with no knowledge of
  55          *        ZIP64.
  56          *
  57          *        NOTE: The ZIP64 code is untested and probably doesn't work. It
  58          *        turned out to be easier to just reject ZIP64 archive uploads,
  59          *        since they are likely to be very rare. Confirming safety of a
  60          *        ZIP64 file is fairly complex. What do you do with a file that is
  61          *        ambiguous and broken when read with a non-ZIP64 reader, but valid
  62          *        when read with a ZIP64 reader? This situation is normal for a
  63          *        valid ZIP64 file, and working out what non-ZIP64 readers will make
  64          *        of such a file is not trivial.
  65          *
  66          * @return Status A Status object. The following fatal errors are defined:
  67          *
  68          *      - zip-file-open-error: The file could not be opened.
  69          *
  70          *      - zip-wrong-format: The file does not appear to be a ZIP file.
  71          *
  72          *      - zip-bad: There was something wrong or ambiguous about the file
  73          *        data.
  74          *
  75          *      - zip-unsupported: The ZIP file uses features which
  76          *        ZipDirectoryReader does not support.
  77          *
  78          * The default messages for those fatal errors are written in a way that
  79          * makes sense for upload verification.
  80          *
  81          * If a fatal error is returned, more information about the error will be
  82          * available in the debug log.
  83          *
  84          * Note that the callback function may be called any number of times before
  85          * a fatal error is returned. If this occurs, the data sent to the callback
  86          * function should be discarded.
  87          */
  88         public static function read( $fileName, $callback, $options = [] ) {
  89                 $zdr = new self( $fileName, $callback, $options );
  90
  91                 return $zdr->execute();
  92         }
  93
  94         /** The file name */
  95         protected $fileName;
  96
  97         /** The opened file resource */
  98         protected $file;
  99
 100         /** The cached length of the file, or null if it has not been loaded yet. */
 101         protected $fileLength;
 102
 103         /** A segmented cache of the file contents */
 104         protected $buffer;
 105
 106         /** The file data callback */
 107         protected $callback;
 108
 109         /** The ZIP64 mode */
 110         protected $zip64 = false;
 111
 112         /** Stored headers */
 113         protected $eocdr, $eocdr64, $eocdr64Locator;
 114
 115         protected $data;
 116
 117         /** The "extra field" ID for ZIP64 central directory entries */
 118         const ZIP64_EXTRA_HEADER = 0x0001;
 119
 120         /** The segment size for the file contents cache */
 121         const SEGSIZE = 16384;
 122
 123         /** The index of the "general field" bit for UTF-8 file names */
 124         const GENERAL_UTF8 = 11;
 125
 126         /** The index of the "general field" bit for central directory encryption */
 127         const GENERAL_CD_ENCRYPTED = 13;
 128
 129         /**
 130          * Private constructor
 131          * @param string $fileName
 132          * @param callable $callback
 133          * @param array $options
 134          */
 135         protected function __construct( $fileName, $callback, $options ) {
 136                 $this->fileName = $fileName;
 137                 $this->callback = $callback;
 138
 139                 if ( isset( $options['zip64'] ) ) {
 140                         $this->zip64 = $options['zip64'];
 141                 }
 142         }
 143
 144         /**
 145          * Read the directory according to settings in $this.
 146          *
 147          * @return Status
 148          */
 149         function execute() {
 150                 $this->file = fopen( $this->fileName, 'r' );
 151                 $this->data = [];
 152                 if ( !$this->file ) {
 153                         return Status::newFatal( 'zip-file-open-error' );
 154                 }
 155
 156                 $status = Status::newGood();
 157                 try {
 158                         $this->readEndOfCentralDirectoryRecord();
 159                         if ( $this->zip64 ) {
 160                                 list( $offset, $size ) = $this->findZip64CentralDirectory();
 161                                 $this->readCentralDirectory( $offset, $size );
 162                         } else {
 163                                 if ( $this->eocdr['CD size'] == 0xffffffff
 164                                         || $this->eocdr['CD offset'] == 0xffffffff
 165                                         || $this->eocdr['CD entries total'] == 0xffff
 166                                 ) {
 167                                         $this->error( 'zip-unsupported', 'Central directory header indicates ZIP64, ' .
 168                                                 'but we are in legacy mode. Rejecting this upload is necessary to avoid ' .
 169                                                 'opening vulnerabilities on clients using OpenJDK 7 or later.' );
 170                                 }
 171
 172                                 list( $offset, $size ) = $this->findOldCentralDirectory();
 173                                 $this->readCentralDirectory( $offset, $size );
 174                         }
 175                 } catch ( ZipDirectoryReaderError $e ) {
 176                         $status->fatal( $e->getErrorCode() );
 177                 }
 178
 179                 fclose( $this->file );
 180
 181                 return $status;
 182         }
 183
 184         /**
 185          * Throw an error, and log a debug message
 186          * @param mixed $code
 187          * @param string $debugMessage
 188          * @throws ZipDirectoryReaderError
 189          */
 190         function error( $code, $debugMessage ) {
 191                 wfDebug( __CLASS__ . ": Fatal error: $debugMessage\n" );
 192                 throw new ZipDirectoryReaderError( $code );
 193         }
 194
 195         /**
 196          * Read the header which is at the end of the central directory,
 197          * unimaginatively called the "end of central directory record" by the ZIP
 198          * spec.
 199          */
 200         function readEndOfCentralDirectoryRecord() {
 201                 $info = [
 202                         'signature' => 4,
 203                         'disk' => 2,
 204                         'CD start disk' => 2,
 205                         'CD entries this disk' => 2,
 206                         'CD entries total' => 2,
 207                         'CD size' => 4,
 208                         'CD offset' => 4,
 209                         'file comment length' => 2,
 210                 ];
 211                 $structSize = $this->getStructSize( $info );
 212                 $startPos = $this->getFileLength() - 65536 - $structSize;
 213                 if ( $startPos < 0 ) {
 214                         $startPos = 0;
 215                 }
 216
 217                 if ( $this->getFileLength() === 0 ) {
 218                         $this->error( 'zip-wrong-format', "The file is empty." );
 219                 }
 220
 221                 $block = $this->getBlock( $startPos );
 222                 $sigPos = strrpos( $block, "PK\x05\x06" );
 223                 if ( $sigPos === false ) {
 224                         $this->error( 'zip-wrong-format',
 225                                 "zip file lacks EOCDR signature. It probably isn't a zip file." );
 226                 }
 227
 228                 $this->eocdr = $this->unpack( substr( $block, $sigPos ), $info );
 229                 $this->eocdr['EOCDR size'] = $structSize + $this->eocdr['file comment length'];
 230
 231                 if ( $structSize + $this->eocdr['file comment length'] != strlen( $block ) - $sigPos ) {
 232                         $this->error( 'zip-bad', 'trailing bytes after the end of the file comment' );
 233                 }
 234                 if ( $this->eocdr['disk'] !== 0
 235                         || $this->eocdr['CD start disk'] !== 0
 236                 ) {
 237                         $this->error( 'zip-unsupported', 'more than one disk (in EOCDR)' );
 238                 }
 239                 $this->eocdr += $this->unpack(
 240                         $block,
 241                         [ 'file comment' => [ 'string', $this->eocdr['file comment length'] ] ],
 242                         $sigPos + $structSize );
 243                 $this->eocdr['position'] = $startPos + $sigPos;
 244         }
 245
 246         /**
 247          * Read the header called the "ZIP64 end of central directory locator". An
 248          * error will be raised if it does not exist.
 249          */
 250         function readZip64EndOfCentralDirectoryLocator() {
 251                 $info = [
 252                         'signature' => [ 'string', 4 ],
 253                         'eocdr64 start disk' => 4,
 254                         'eocdr64 offset' => 8,
 255                         'number of disks' => 4,
 256                 ];
 257                 $structSize = $this->getStructSize( $info );
 258
 259                 $start = $this->getFileLength() - $this->eocdr['EOCDR size'] - $structSize;
 260                 $block = $this->getBlock( $start, $structSize );
 261                 $this->eocdr64Locator = $data = $this->unpack( $block, $info );
 262
 263                 if ( $data['signature'] !== "PK\x06\x07" ) {
 264                         // Note: Java will allow this and continue to read the
 265                         // EOCDR64, so we have to reject the upload, we can't
 266                         // just use the EOCDR header instead.
 267                         $this->error( 'zip-bad', 'wrong signature on Zip64 end of central directory locator' );
 268                 }
 269         }
 270
 271         /**
 272          * Read the header called the "ZIP64 end of central directory record". It
 273          * may replace the regular "end of central directory record" in ZIP64 files.
 274          */
 275         function readZip64EndOfCentralDirectoryRecord() {
 276                 if ( $this->eocdr64Locator['eocdr64 start disk'] != 0
 277                         || $this->eocdr64Locator['number of disks'] != 0
 278                 ) {
 279                         $this->error( 'zip-unsupported', 'more than one disk (in EOCDR64 locator)' );
 280                 }
 281
 282                 $info = [
 283                         'signature' => [ 'string', 4 ],
 284                         'EOCDR64 size' => 8,
 285                         'version made by' => 2,
 286                         'version needed' => 2,
 287                         'disk' => 4,
 288                         'CD start disk' => 4,
 289                         'CD entries this disk' => 8,
 290                         'CD entries total' => 8,
 291                         'CD size' => 8,
 292                         'CD offset' => 8
 293                 ];
 294                 $structSize = $this->getStructSize( $info );
 295                 $block = $this->getBlock( $this->eocdr64Locator['eocdr64 offset'], $structSize );
 296                 $this->eocdr64 = $data = $this->unpack( $block, $info );
 297                 if ( $data['signature'] !== "PK\x06\x06" ) {
 298                         $this->error( 'zip-bad', 'wrong signature on Zip64 end of central directory record' );
 299                 }
 300                 if ( $data['disk'] !== 0
 301                         || $data['CD start disk'] !== 0
 302                 ) {
 303                         $this->error( 'zip-unsupported', 'more than one disk (in EOCDR64)' );
 304                 }
 305         }
 306
 307         /**
 308          * Find the location of the central directory, as would be seen by a
 309          * non-ZIP64 reader.
 310          *
 311          * @return array List containing offset, size and end position.
 312          */
 313         function findOldCentralDirectory() {
 314                 $size = $this->eocdr['CD size'];
 315                 $offset = $this->eocdr['CD offset'];
 316                 $endPos = $this->eocdr['position'];
 317
 318                 // Some readers use the EOCDR position instead of the offset field
 319                 // to find the directory, so to be safe, we check if they both agree.
 320                 if ( $offset + $size != $endPos ) {
 321                         $this->error( 'zip-bad', 'the central directory does not immediately precede the end ' .
 322                                 'of central directory record' );
 323                 }
 324
 325                 return [ $offset, $size ];
 326         }
 327
 328         /**
 329          * Find the location of the central directory, as would be seen by a
 330          * ZIP64-compliant reader.
 331          *
 332          * @return array List containing offset, size and end position.
 333          */
 334         function findZip64CentralDirectory() {
 335                 // The spec is ambiguous about the exact rules of precedence between the
 336                 // ZIP64 headers and the original headers. Here we follow zip_util.c
 337                 // from OpenJDK 7.
 338                 $size = $this->eocdr['CD size'];
 339                 $offset = $this->eocdr['CD offset'];
 340                 $numEntries = $this->eocdr['CD entries total'];
 341                 $endPos = $this->eocdr['position'];
 342                 if ( $size == 0xffffffff
 343                         || $offset == 0xffffffff
 344                         || $numEntries == 0xffff
 345                 ) {
 346                         $this->readZip64EndOfCentralDirectoryLocator();
 347
 348                         if ( isset( $this->eocdr64Locator['eocdr64 offset'] ) ) {
 349                                 $this->readZip64EndOfCentralDirectoryRecord();
 350                                 if ( isset( $this->eocdr64['CD offset'] ) ) {
 351                                         $size = $this->eocdr64['CD size'];
 352                                         $offset = $this->eocdr64['CD offset'];
 353                                         $endPos = $this->eocdr64Locator['eocdr64 offset'];
 354                                 }
 355                         }
 356                 }
 357                 // Some readers use the EOCDR position instead of the offset field
 358                 // to find the directory, so to be safe, we check if they both agree.
 359                 if ( $offset + $size != $endPos ) {
 360                         $this->error( 'zip-bad', 'the central directory does not immediately precede the end ' .
 361                                 'of central directory record' );
 362                 }
 363
 364                 return [ $offset, $size ];
 365         }
 366
 367         /**
 368          * Read the central directory at the given location
 369          * @param int $offset
 370          * @param int $size
 371          */
 372         function readCentralDirectory( $offset, $size ) {
 373                 $block = $this->getBlock( $offset, $size );
 374
 375                 $fixedInfo = [
 376                         'signature' => [ 'string', 4 ],
 377                         'version made by' => 2,
 378                         'version needed' => 2,
 379                         'general bits' => 2,
 380                         'compression method' => 2,
 381                         'mod time' => 2,
 382                         'mod date' => 2,
 383                         'crc-32' => 4,
 384                         'compressed size' => 4,
 385                         'uncompressed size' => 4,
 386                         'name length' => 2,
 387                         'extra field length' => 2,
 388                         'comment length' => 2,
 389                         'disk number start' => 2,
 390                         'internal attrs' => 2,
 391                         'external attrs' => 4,
 392                         'local header offset' => 4,
 393                 ];
 394                 $fixedSize = $this->getStructSize( $fixedInfo );
 395
 396                 $pos = 0;
 397                 while ( $pos < $size ) {
 398                         $data = $this->unpack( $block, $fixedInfo, $pos );
 399                         $pos += $fixedSize;
 400
 401                         if ( $data['signature'] !== "PK\x01\x02" ) {
 402                                 $this->error( 'zip-bad', 'Invalid signature found in directory entry' );
 403                         }
 404
 405                         $variableInfo = [
 406                                 'name' => [ 'string', $data['name length'] ],
 407                                 'extra field' => [ 'string', $data['extra field length'] ],
 408                                 'comment' => [ 'string', $data['comment length'] ],
 409                         ];
 410                         $data += $this->unpack( $block, $variableInfo, $pos );
 411                         $pos += $this->getStructSize( $variableInfo );
 412
 413                         if ( $this->zip64 && (
 414                                         $data['compressed size'] == 0xffffffff
 415                                         || $data['uncompressed size'] == 0xffffffff
 416                                         || $data['local header offset'] == 0xffffffff )
 417                         ) {
 418                                 $zip64Data = $this->unpackZip64Extra( $data['extra field'] );
 419                                 if ( $zip64Data ) {
 420                                         $data = $zip64Data + $data;
 421                                 }
 422                         }
 423
 424                         if ( $this->testBit( $data['general bits'], self::GENERAL_CD_ENCRYPTED ) ) {
 425                                 $this->error( 'zip-unsupported', 'central directory encryption is not supported' );
 426                         }
 427
 428                         // Convert the timestamp into MediaWiki format
 429                         // For the format, please see the MS-DOS 2.0 Programmer's Reference,
 430                         // pages 3-5 and 3-6.
 431                         $time = $data['mod time'];
 432                         $date = $data['mod date'];
 433
 434                         $year = 1980 + ( $date >> 9 );
 435                         $month = ( $date >> 5 ) & 15;
 436                         $day = $date & 31;
 437                         $hour = ( $time >> 11 ) & 31;
 438                         $minute = ( $time >> 5 ) & 63;
 439                         $second = ( $time & 31 ) * 2;
 440                         $timestamp = sprintf( "%04d%02d%02d%02d%02d%02d",
 441                                 $year, $month, $day, $hour, $minute, $second );
 442
 443                         // Convert the character set in the file name
 444                         if ( $this->testBit( $data['general bits'], self::GENERAL_UTF8 ) ) {
 445                                 $name = $data['name'];
 446                         } else {
 447                                 $name = iconv( 'CP437', 'UTF-8', $data['name'] );
 448                         }
 449
 450                         // Compile a data array for the user, with a sensible format
 451                         $userData = [
 452                                 'name' => $name,
 453                                 'mtime' => $timestamp,
 454                                 'size' => $data['uncompressed size'],
 455                         ];
 456                         call_user_func( $this->callback, $userData );
 457                 }
 458         }
 459
 460         /**
 461          * Interpret ZIP64 "extra field" data and return an associative array.
 462          * @param string $extraField
 463          * @return array|bool
 464          */
 465         function unpackZip64Extra( $extraField ) {
 466                 $extraHeaderInfo = [
 467                         'id' => 2,
 468                         'size' => 2,
 469                 ];
 470                 $extraHeaderSize = $this->getStructSize( $extraHeaderInfo );
 471
 472                 $zip64ExtraInfo = [
 473                         'uncompressed size' => 8,
 474                         'compressed size' => 8,
 475                         'local header offset' => 8,
 476                         'disk number start' => 4,
 477                 ];
 478
 479                 $extraPos = 0;
 480                 while ( $extraPos < strlen( $extraField ) ) {
 481                         $extra = $this->unpack( $extraField, $extraHeaderInfo, $extraPos );
 482                         $extraPos += $extraHeaderSize;
 483                         $extra += $this->unpack( $extraField,
 484                                 [ 'data' => [ 'string', $extra['size'] ] ],
 485                                 $extraPos );
 486                         $extraPos += $extra['size'];
 487
 488                         if ( $extra['id'] == self::ZIP64_EXTRA_HEADER ) {
 489                                 return $this->unpack( $extra['data'], $zip64ExtraInfo );
 490                         }
 491                 }
 492
 493                 return false;
 494         }
 495
 496         /**
 497          * Get the length of the file.
 498          * @return int
 499          */
 500         function getFileLength() {
 501                 if ( $this->fileLength === null ) {
 502                         $stat = fstat( $this->file );
 503                         $this->fileLength = $stat['size'];
 504                 }
 505
 506                 return $this->fileLength;
 507         }
 508
 509         /**
 510          * Get the file contents from a given offset. If there are not enough bytes
 511          * in the file to satisfy the request, an exception will be thrown.
 512          *
 513          * @param int $start The byte offset of the start of the block.
 514          * @param int $length The number of bytes to return. If omitted, the remainder
 515          *    of the file will be returned.
 516          *
 517          * @return string
 518          */
 519         function getBlock( $start, $length = null ) {
 520                 $fileLength = $this->getFileLength();
 521                 if ( $start >= $fileLength ) {
 522                         $this->error( 'zip-bad', "getBlock() requested position $start, " .
 523                                 "file length is $fileLength" );
 524                 }
 525                 if ( $length === null ) {
 526                         $length = $fileLength - $start;
 527                 }
 528                 $end = $start + $length;
 529                 if ( $end > $fileLength ) {
 530                         $this->error( 'zip-bad', "getBlock() requested end position $end, " .
 531                                 "file length is $fileLength" );
 532                 }
 533                 $startSeg = floor( $start / self::SEGSIZE );
 534                 $endSeg = ceil( $end / self::SEGSIZE );
 535
 536                 $block = '';
 537                 for ( $segIndex = $startSeg; $segIndex <= $endSeg; $segIndex++ ) {
 538                         $block .= $this->getSegment( $segIndex );
 539                 }
 540
 541                 $block = substr( $block,
 542                         $start - $startSeg * self::SEGSIZE,
 543                         $length );
 544
 545                 if ( strlen( $block ) < $length ) {
 546                         $this->error( 'zip-bad', 'getBlock() returned an unexpectedly small amount of data' );
 547                 }
 548
 549                 return $block;
 550         }
 551
 552         /**
 553          * Get a section of the file starting at position $segIndex * self::SEGSIZE,
 554          * of length self::SEGSIZE. The result is cached. This is a helper function
 555          * for getBlock().
 556          *
 557          * If there are not enough bytes in the file to satisfy the request, the
 558          * return value will be truncated. If a request is made for a segment beyond
 559          * the end of the file, an empty string will be returned.
 560          *
 561          * @param int $segIndex
 562          *
 563          * @return string
 564          */
 565         function getSegment( $segIndex ) {
 566                 if ( !isset( $this->buffer[$segIndex] ) ) {
 567                         $bytePos = $segIndex * self::SEGSIZE;
 568                         if ( $bytePos >= $this->getFileLength() ) {
 569                                 $this->buffer[$segIndex] = '';
 570
 571                                 return '';
 572                         }
 573                         if ( fseek( $this->file, $bytePos ) ) {
 574                                 $this->error( 'zip-bad', "seek to $bytePos failed" );
 575                         }
 576                         $seg = fread( $this->file, self::SEGSIZE );
 577                         if ( $seg === false ) {
 578                                 $this->error( 'zip-bad', "read from $bytePos failed" );
 579                         }
 580                         $this->buffer[$segIndex] = $seg;
 581                 }
 582
 583                 return $this->buffer[$segIndex];
 584         }
 585
 586         /**
 587          * Get the size of a structure in bytes. See unpack() for the format of $struct.
 588          * @param array $struct
 589          * @return int
 590          */
 591         function getStructSize( $struct ) {
 592                 $size = 0;
 593                 foreach ( $struct as $type ) {
 594                         if ( is_array( $type ) ) {
 595                                 list( , $fieldSize ) = $type;
 596                                 $size += $fieldSize;
 597                         } else {
 598                                 $size += $type;
 599                         }
 600                 }
 601
 602                 return $size;
 603         }
 604
 605         /**
 606          * Unpack a binary structure. This is like the built-in unpack() function
 607          * except nicer.
 608          *
 609          * @param string $string The binary data input
 610          *
 611          * @param array $struct An associative array giving structure members and their
 612          *    types. In the key is the field name. The value may be either an
 613          *    integer, in which case the field is a little-endian unsigned integer
 614          *    encoded in the given number of bytes, or an array, in which case the
 615          *    first element of the array is the type name, and the subsequent
 616          *    elements are type-dependent parameters. Only one such type is defined:
 617          *       - "string": The second array element gives the length of string.
 618          *          Not null terminated.
 619          *
 620          * @param int $offset The offset into the string at which to start unpacking.
 621          *
 622          * @throws MWException
 623          * @return array Unpacked associative array. Note that large integers in the input
 624          *    may be represented as floating point numbers in the return value, so
 625          *    the use of weak comparison is advised.
 626          */
 627         function unpack( $string, $struct, $offset = 0 ) {
 628                 $size = $this->getStructSize( $struct );
 629                 if ( $offset + $size > strlen( $string ) ) {
 630                         $this->error( 'zip-bad', 'unpack() would run past the end of the supplied string' );
 631                 }
 632
 633                 $data = [];
 634                 $pos = $offset;
 635                 foreach ( $struct as $key => $type ) {
 636                         if ( is_array( $type ) ) {
 637                                 list( $typeName, $fieldSize ) = $type;
 638                                 switch ( $typeName ) {
 639                                         case 'string':
 640                                                 $data[$key] = substr( $string, $pos, $fieldSize );
 641                                                 $pos += $fieldSize;
 642                                                 break;
 643                                         default:
 644                                                 throw new MWException( __METHOD__ . ": invalid type \"$typeName\"" );
 645                                 }
 646                         } else {
 647                                 // Unsigned little-endian integer
 648                                 $length = intval( $type );
 649
 650                                 // Calculate the value. Use an algorithm which automatically
 651                                 // upgrades the value to floating point if necessary.
 652                                 $value = 0;
 653                                 for ( $i = $length - 1; $i >= 0; $i-- ) {
 654                                         $value *= 256;
 655                                         $value += ord( $string[$pos + $i] );
 656                                 }
 657
 658                                 // Throw an exception if there was loss of precision
 659                                 if ( $value > pow( 2, 52 ) ) {
 660                                         $this->error( 'zip-unsupported', 'number too large to be stored in a double. ' .
 661                                                 'This could happen if we tried to unpack a 64-bit structure ' .
 662                                                 'at an invalid location.' );
 663                                 }
 664                                 $data[$key] = $value;
 665                                 $pos += $length;
 666                         }
 667                 }
 668
 669                 return $data;
 670         }
 671
 672         /**
 673          * Returns a bit from a given position in an integer value, converted to
 674          * boolean.
 675          *
 676          * @param int $value
 677          * @param int $bitIndex The index of the bit, where 0 is the LSB.
 678          * @return bool
 679          */
 680         function testBit( $value, $bitIndex ) {
 681                 return (bool)( ( $value >> $bitIndex ) & 1 );
 682         }
 683
 684         /**
 685          * Debugging helper function which dumps a string in hexdump -C format.
 686          * @param string $s
 687          */
 688         function hexDump( $s ) {
 689                 $n = strlen( $s );
 690                 for ( $i = 0; $i < $n; $i += 16 ) {
 691                         printf( "%08X ", $i );
 692                         for ( $j = 0; $j < 16; $j++ ) {
 693                                 print " ";
 694                                 if ( $j == 8 ) {
 695                                         print " ";
 696                                 }
 697                                 if ( $i + $j >= $n ) {
 698                                         print "  ";
 699                                 } else {
 700                                         printf( "%02X", ord( $s[$i + $j] ) );
 701                                 }
 702                         }
 703
 704                         print "  |";
 705                         for ( $j = 0; $j < 16; $j++ ) {
 706                                 if ( $i + $j >= $n ) {
 707                                         print " ";
 708                                 } elseif ( ctype_print( $s[$i + $j] ) ) {
 709                                         print $s[$i + $j];
 710                                 } else {
 711                                         print '.';
 712                                 }
 713                         }
 714                         print "|\n";
 715                 }
 716         }
 717 }
 718
 719 /**
 720  * Internal exception class. Will be caught by private code.
 721  */
 722 class ZipDirectoryReaderError extends Exception {
 723         protected $errorCode;
 724
 725         function __construct( $code ) {
 726                 $this->errorCode = $code;
 727                 parent::__construct( "ZipDirectoryReader error: $code" );
 728         }
 729
 730         /**
 731          * @return mixed
 732          */
 733         function getErrorCode() {
 734                 return $this->errorCode;
 735         }
 736 }