Black Lives Matter. Support the Equal Justice Initiative.

Source file src/archive/zip/struct.go

Documentation: archive/zip

     1  // Copyright 2010 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  /*
     6  Package zip provides support for reading and writing ZIP archives.
     7  
     8  See: https://www.pkware.com/appnote
     9  
    10  This package does not support disk spanning.
    11  
    12  A note about ZIP64:
    13  
    14  To be backwards compatible the FileHeader has both 32 and 64 bit Size
    15  fields. The 64 bit fields will always contain the correct value and
    16  for normal archives both fields will be the same. For files requiring
    17  the ZIP64 format the 32 bit fields will be 0xffffffff and the 64 bit
    18  fields must be used instead.
    19  */
    20  package zip
    21  
    22  import (
    23  	"io/fs"
    24  	"path"
    25  	"time"
    26  )
    27  
    28  // Compression methods.
    29  const (
    30  	Store   uint16 = 0 // no compression
    31  	Deflate uint16 = 8 // DEFLATE compressed
    32  )
    33  
    34  const (
    35  	fileHeaderSignature      = 0x04034b50
    36  	directoryHeaderSignature = 0x02014b50
    37  	directoryEndSignature    = 0x06054b50
    38  	directory64LocSignature  = 0x07064b50
    39  	directory64EndSignature  = 0x06064b50
    40  	dataDescriptorSignature  = 0x08074b50 // de-facto standard; required by OS X Finder
    41  	fileHeaderLen            = 30         // + filename + extra
    42  	directoryHeaderLen       = 46         // + filename + extra + comment
    43  	directoryEndLen          = 22         // + comment
    44  	dataDescriptorLen        = 16         // four uint32: descriptor signature, crc32, compressed size, size
    45  	dataDescriptor64Len      = 24         // two uint32: signature, crc32 | two uint64: compressed size, size
    46  	directory64LocLen        = 20         //
    47  	directory64EndLen        = 56         // + extra
    48  
    49  	// Constants for the first byte in CreatorVersion.
    50  	creatorFAT    = 0
    51  	creatorUnix   = 3
    52  	creatorNTFS   = 11
    53  	creatorVFAT   = 14
    54  	creatorMacOSX = 19
    55  
    56  	// Version numbers.
    57  	zipVersion20 = 20 // 2.0
    58  	zipVersion45 = 45 // 4.5 (reads and writes zip64 archives)
    59  
    60  	// Limits for non zip64 files.
    61  	uint16max = (1 << 16) - 1
    62  	uint32max = (1 << 32) - 1
    63  
    64  	// Extra header IDs.
    65  	//
    66  	// IDs 0..31 are reserved for official use by PKWARE.
    67  	// IDs above that range are defined by third-party vendors.
    68  	// Since ZIP lacked high precision timestamps (nor a official specification
    69  	// of the timezone used for the date fields), many competing extra fields
    70  	// have been invented. Pervasive use effectively makes them "official".
    71  	//
    72  	// See http://mdfs.net/Docs/Comp/Archiving/Zip/ExtraField
    73  	zip64ExtraID       = 0x0001 // Zip64 extended information
    74  	ntfsExtraID        = 0x000a // NTFS
    75  	unixExtraID        = 0x000d // UNIX
    76  	extTimeExtraID     = 0x5455 // Extended timestamp
    77  	infoZipUnixExtraID = 0x5855 // Info-ZIP Unix extension
    78  )
    79  
    80  // FileHeader describes a file within a zip file.
    81  // See the zip spec for details.
    82  type FileHeader struct {
    83  	// Name is the name of the file.
    84  	//
    85  	// It must be a relative path, not start with a drive letter (such as "C:"),
    86  	// and must use forward slashes instead of back slashes. A trailing slash
    87  	// indicates that this file is a directory and should have no data.
    88  	//
    89  	// When reading zip files, the Name field is populated from
    90  	// the zip file directly and is not validated for correctness.
    91  	// It is the caller's responsibility to sanitize it as
    92  	// appropriate, including canonicalizing slash directions,
    93  	// validating that paths are relative, and preventing path
    94  	// traversal through filenames ("../../../").
    95  	Name string
    96  
    97  	// Comment is any arbitrary user-defined string shorter than 64KiB.
    98  	Comment string
    99  
   100  	// NonUTF8 indicates that Name and Comment are not encoded in UTF-8.
   101  	//
   102  	// By specification, the only other encoding permitted should be CP-437,
   103  	// but historically many ZIP readers interpret Name and Comment as whatever
   104  	// the system's local character encoding happens to be.
   105  	//
   106  	// This flag should only be set if the user intends to encode a non-portable
   107  	// ZIP file for a specific localized region. Otherwise, the Writer
   108  	// automatically sets the ZIP format's UTF-8 flag for valid UTF-8 strings.
   109  	NonUTF8 bool
   110  
   111  	CreatorVersion uint16
   112  	ReaderVersion  uint16
   113  	Flags          uint16
   114  
   115  	// Method is the compression method. If zero, Store is used.
   116  	Method uint16
   117  
   118  	// Modified is the modified time of the file.
   119  	//
   120  	// When reading, an extended timestamp is preferred over the legacy MS-DOS
   121  	// date field, and the offset between the times is used as the timezone.
   122  	// If only the MS-DOS date is present, the timezone is assumed to be UTC.
   123  	//
   124  	// When writing, an extended timestamp (which is timezone-agnostic) is
   125  	// always emitted. The legacy MS-DOS date field is encoded according to the
   126  	// location of the Modified time.
   127  	Modified     time.Time
   128  	ModifiedTime uint16 // Deprecated: Legacy MS-DOS date; use Modified instead.
   129  	ModifiedDate uint16 // Deprecated: Legacy MS-DOS time; use Modified instead.
   130  
   131  	CRC32              uint32
   132  	CompressedSize     uint32 // Deprecated: Use CompressedSize64 instead.
   133  	UncompressedSize   uint32 // Deprecated: Use UncompressedSize64 instead.
   134  	CompressedSize64   uint64
   135  	UncompressedSize64 uint64
   136  	Extra              []byte
   137  	ExternalAttrs      uint32 // Meaning depends on CreatorVersion
   138  }
   139  
   140  // FileInfo returns an fs.FileInfo for the FileHeader.
   141  func (h *FileHeader) FileInfo() fs.FileInfo {
   142  	return headerFileInfo{h}
   143  }
   144  
   145  // headerFileInfo implements fs.FileInfo.
   146  type headerFileInfo struct {
   147  	fh *FileHeader
   148  }
   149  
   150  func (fi headerFileInfo) Name() string { return path.Base(fi.fh.Name) }
   151  func (fi headerFileInfo) Size() int64 {
   152  	if fi.fh.UncompressedSize64 > 0 {
   153  		return int64(fi.fh.UncompressedSize64)
   154  	}
   155  	return int64(fi.fh.UncompressedSize)
   156  }
   157  func (fi headerFileInfo) IsDir() bool { return fi.Mode().IsDir() }
   158  func (fi headerFileInfo) ModTime() time.Time {
   159  	if fi.fh.Modified.IsZero() {
   160  		return fi.fh.ModTime()
   161  	}
   162  	return fi.fh.Modified.UTC()
   163  }
   164  func (fi headerFileInfo) Mode() fs.FileMode { return fi.fh.Mode() }
   165  func (fi headerFileInfo) Type() fs.FileMode { return fi.fh.Mode().Type() }
   166  func (fi headerFileInfo) Sys() interface{}  { return fi.fh }
   167  
   168  func (fi headerFileInfo) Info() (fs.FileInfo, error) { return fi, nil }
   169  
   170  // FileInfoHeader creates a partially-populated FileHeader from an
   171  // fs.FileInfo.
   172  // Because fs.FileInfo's Name method returns only the base name of
   173  // the file it describes, it may be necessary to modify the Name field
   174  // of the returned header to provide the full path name of the file.
   175  // If compression is desired, callers should set the FileHeader.Method
   176  // field; it is unset by default.
   177  func FileInfoHeader(fi fs.FileInfo) (*FileHeader, error) {
   178  	size := fi.Size()
   179  	fh := &FileHeader{
   180  		Name:               fi.Name(),
   181  		UncompressedSize64: uint64(size),
   182  	}
   183  	fh.SetModTime(fi.ModTime())
   184  	fh.SetMode(fi.Mode())
   185  	if fh.UncompressedSize64 > uint32max {
   186  		fh.UncompressedSize = uint32max
   187  	} else {
   188  		fh.UncompressedSize = uint32(fh.UncompressedSize64)
   189  	}
   190  	return fh, nil
   191  }
   192  
   193  type directoryEnd struct {
   194  	diskNbr            uint32 // unused
   195  	dirDiskNbr         uint32 // unused
   196  	dirRecordsThisDisk uint64 // unused
   197  	directoryRecords   uint64
   198  	directorySize      uint64
   199  	directoryOffset    uint64 // relative to file
   200  	commentLen         uint16
   201  	comment            string
   202  }
   203  
   204  // timeZone returns a *time.Location based on the provided offset.
   205  // If the offset is non-sensible, then this uses an offset of zero.
   206  func timeZone(offset time.Duration) *time.Location {
   207  	const (
   208  		minOffset   = -12 * time.Hour  // E.g., Baker island at -12:00
   209  		maxOffset   = +14 * time.Hour  // E.g., Line island at +14:00
   210  		offsetAlias = 15 * time.Minute // E.g., Nepal at +5:45
   211  	)
   212  	offset = offset.Round(offsetAlias)
   213  	if offset < minOffset || maxOffset < offset {
   214  		offset = 0
   215  	}
   216  	return time.FixedZone("", int(offset/time.Second))
   217  }
   218  
   219  // msDosTimeToTime converts an MS-DOS date and time into a time.Time.
   220  // The resolution is 2s.
   221  // See: https://msdn.microsoft.com/en-us/library/ms724247(v=VS.85).aspx
   222  func msDosTimeToTime(dosDate, dosTime uint16) time.Time {
   223  	return time.Date(
   224  		// date bits 0-4: day of month; 5-8: month; 9-15: years since 1980
   225  		int(dosDate>>9+1980),
   226  		time.Month(dosDate>>5&0xf),
   227  		int(dosDate&0x1f),
   228  
   229  		// time bits 0-4: second/2; 5-10: minute; 11-15: hour
   230  		int(dosTime>>11),
   231  		int(dosTime>>5&0x3f),
   232  		int(dosTime&0x1f*2),
   233  		0, // nanoseconds
   234  
   235  		time.UTC,
   236  	)
   237  }
   238  
   239  // timeToMsDosTime converts a time.Time to an MS-DOS date and time.
   240  // The resolution is 2s.
   241  // See: https://msdn.microsoft.com/en-us/library/ms724274(v=VS.85).aspx
   242  func timeToMsDosTime(t time.Time) (fDate uint16, fTime uint16) {
   243  	fDate = uint16(t.Day() + int(t.Month())<<5 + (t.Year()-1980)<<9)
   244  	fTime = uint16(t.Second()/2 + t.Minute()<<5 + t.Hour()<<11)
   245  	return
   246  }
   247  
   248  // ModTime returns the modification time in UTC using the legacy
   249  // ModifiedDate and ModifiedTime fields.
   250  //
   251  // Deprecated: Use Modified instead.
   252  func (h *FileHeader) ModTime() time.Time {
   253  	return msDosTimeToTime(h.ModifiedDate, h.ModifiedTime)
   254  }
   255  
   256  // SetModTime sets the Modified, ModifiedTime, and ModifiedDate fields
   257  // to the given time in UTC.
   258  //
   259  // Deprecated: Use Modified instead.
   260  func (h *FileHeader) SetModTime(t time.Time) {
   261  	t = t.UTC() // Convert to UTC for compatibility
   262  	h.Modified = t
   263  	h.ModifiedDate, h.ModifiedTime = timeToMsDosTime(t)
   264  }
   265  
   266  const (
   267  	// Unix constants. The specification doesn't mention them,
   268  	// but these seem to be the values agreed on by tools.
   269  	s_IFMT   = 0xf000
   270  	s_IFSOCK = 0xc000
   271  	s_IFLNK  = 0xa000
   272  	s_IFREG  = 0x8000
   273  	s_IFBLK  = 0x6000
   274  	s_IFDIR  = 0x4000
   275  	s_IFCHR  = 0x2000
   276  	s_IFIFO  = 0x1000
   277  	s_ISUID  = 0x800
   278  	s_ISGID  = 0x400
   279  	s_ISVTX  = 0x200
   280  
   281  	msdosDir      = 0x10
   282  	msdosReadOnly = 0x01
   283  )
   284  
   285  // Mode returns the permission and mode bits for the FileHeader.
   286  func (h *FileHeader) Mode() (mode fs.FileMode) {
   287  	switch h.CreatorVersion >> 8 {
   288  	case creatorUnix, creatorMacOSX:
   289  		mode = unixModeToFileMode(h.ExternalAttrs >> 16)
   290  	case creatorNTFS, creatorVFAT, creatorFAT:
   291  		mode = msdosModeToFileMode(h.ExternalAttrs)
   292  	}
   293  	if len(h.Name) > 0 && h.Name[len(h.Name)-1] == '/' {
   294  		mode |= fs.ModeDir
   295  	}
   296  	return mode
   297  }
   298  
   299  // SetMode changes the permission and mode bits for the FileHeader.
   300  func (h *FileHeader) SetMode(mode fs.FileMode) {
   301  	h.CreatorVersion = h.CreatorVersion&0xff | creatorUnix<<8
   302  	h.ExternalAttrs = fileModeToUnixMode(mode) << 16
   303  
   304  	// set MSDOS attributes too, as the original zip does.
   305  	if mode&fs.ModeDir != 0 {
   306  		h.ExternalAttrs |= msdosDir
   307  	}
   308  	if mode&0200 == 0 {
   309  		h.ExternalAttrs |= msdosReadOnly
   310  	}
   311  }
   312  
   313  // isZip64 reports whether the file size exceeds the 32 bit limit
   314  func (h *FileHeader) isZip64() bool {
   315  	return h.CompressedSize64 >= uint32max || h.UncompressedSize64 >= uint32max
   316  }
   317  
   318  func (f *FileHeader) hasDataDescriptor() bool {
   319  	return f.Flags&0x8 != 0
   320  }
   321  
   322  func msdosModeToFileMode(m uint32) (mode fs.FileMode) {
   323  	if m&msdosDir != 0 {
   324  		mode = fs.ModeDir | 0777
   325  	} else {
   326  		mode = 0666
   327  	}
   328  	if m&msdosReadOnly != 0 {
   329  		mode &^= 0222
   330  	}
   331  	return mode
   332  }
   333  
   334  func fileModeToUnixMode(mode fs.FileMode) uint32 {
   335  	var m uint32
   336  	switch mode & fs.ModeType {
   337  	default:
   338  		m = s_IFREG
   339  	case fs.ModeDir:
   340  		m = s_IFDIR
   341  	case fs.ModeSymlink:
   342  		m = s_IFLNK
   343  	case fs.ModeNamedPipe:
   344  		m = s_IFIFO
   345  	case fs.ModeSocket:
   346  		m = s_IFSOCK
   347  	case fs.ModeDevice:
   348  		m = s_IFBLK
   349  	case fs.ModeDevice | fs.ModeCharDevice:
   350  		m = s_IFCHR
   351  	}
   352  	if mode&fs.ModeSetuid != 0 {
   353  		m |= s_ISUID
   354  	}
   355  	if mode&fs.ModeSetgid != 0 {
   356  		m |= s_ISGID
   357  	}
   358  	if mode&fs.ModeSticky != 0 {
   359  		m |= s_ISVTX
   360  	}
   361  	return m | uint32(mode&0777)
   362  }
   363  
   364  func unixModeToFileMode(m uint32) fs.FileMode {
   365  	mode := fs.FileMode(m & 0777)
   366  	switch m & s_IFMT {
   367  	case s_IFBLK:
   368  		mode |= fs.ModeDevice
   369  	case s_IFCHR:
   370  		mode |= fs.ModeDevice | fs.ModeCharDevice
   371  	case s_IFDIR:
   372  		mode |= fs.ModeDir
   373  	case s_IFIFO:
   374  		mode |= fs.ModeNamedPipe
   375  	case s_IFLNK:
   376  		mode |= fs.ModeSymlink
   377  	case s_IFREG:
   378  		// nothing to do
   379  	case s_IFSOCK:
   380  		mode |= fs.ModeSocket
   381  	}
   382  	if m&s_ISGID != 0 {
   383  		mode |= fs.ModeSetgid
   384  	}
   385  	if m&s_ISUID != 0 {
   386  		mode |= fs.ModeSetuid
   387  	}
   388  	if m&s_ISVTX != 0 {
   389  		mode |= fs.ModeSticky
   390  	}
   391  	return mode
   392  }
   393  
   394  // dataDescriptor holds the data descriptor that optionally follows the file
   395  // contents in the zip file.
   396  type dataDescriptor struct {
   397  	crc32            uint32
   398  	compressedSize   uint64
   399  	uncompressedSize uint64
   400  }
   401  

View as plain text