2 Read and write ZIP files.
4 import struct
, os
, time
, sys
5 import binascii
, cStringIO
8 import zlib
# We may need its compression method
12 __all__
= ["BadZipfile", "error", "ZIP_STORED", "ZIP_DEFLATED", "is_zipfile",
13 "ZipInfo", "ZipFile", "PyZipFile", "LargeZipFile" ]
15 class BadZipfile(Exception):
19 class LargeZipFile(Exception):
21 Raised when writing a zipfile, the zipfile requires ZIP64 extensions
22 and those extensions are disabled.
25 error
= BadZipfile
# The exception raised by this module
27 ZIP64_LIMIT
= (1 << 31) - 1
29 # constants for Zip file compression methods
32 # Other ZIP compression methods not supported
34 # Here are some struct module formats for reading headers
35 structEndArchive
= "<4s4H2lH" # 9 items, end of archive, 22 bytes
36 stringEndArchive
= "PK\005\006" # magic number for end of archive record
37 structCentralDir
= "<4s4B4HlLL5HLl"# 19 items, central directory, 46 bytes
38 stringCentralDir
= "PK\001\002" # magic number for central directory
39 structFileHeader
= "<4s2B4HlLL2H" # 12 items, file header record, 30 bytes
40 stringFileHeader
= "PK\003\004" # magic number for file header
41 structEndArchive64Locator
= "<4slql" # 4 items, locate Zip64 header, 20 bytes
42 stringEndArchive64Locator
= "PK\x06\x07" # magic token for locator header
43 structEndArchive64
= "<4sqhhllqqqq" # 10 items, end of archive (Zip64), 56 bytes
44 stringEndArchive64
= "PK\x06\x06" # magic token for Zip64 header
47 # indexes of entries in the central directory structure
49 _CD_CREATE_VERSION
= 1
51 _CD_EXTRACT_VERSION
= 3
52 _CD_EXTRACT_SYSTEM
= 4 # is this meaningful?
58 _CD_COMPRESSED_SIZE
= 10
59 _CD_UNCOMPRESSED_SIZE
= 11
60 _CD_FILENAME_LENGTH
= 12
61 _CD_EXTRA_FIELD_LENGTH
= 13
62 _CD_COMMENT_LENGTH
= 14
63 _CD_DISK_NUMBER_START
= 15
64 _CD_INTERNAL_FILE_ATTRIBUTES
= 16
65 _CD_EXTERNAL_FILE_ATTRIBUTES
= 17
66 _CD_LOCAL_HEADER_OFFSET
= 18
68 # indexes of entries in the local file header structure
70 _FH_EXTRACT_VERSION
= 1
71 _FH_EXTRACT_SYSTEM
= 2 # is this meaningful?
72 _FH_GENERAL_PURPOSE_FLAG_BITS
= 3
73 _FH_COMPRESSION_METHOD
= 4
77 _FH_COMPRESSED_SIZE
= 8
78 _FH_UNCOMPRESSED_SIZE
= 9
79 _FH_FILENAME_LENGTH
= 10
80 _FH_EXTRA_FIELD_LENGTH
= 11
82 def is_zipfile(filename
):
83 """Quickly see if file is a ZIP file by checking the magic number."""
85 fpin
= open(filename
, "rb")
86 endrec
= _EndRecData(fpin
)
89 return True # file has correct magic number
94 def _EndRecData64(fpin
, offset
, endrec
):
96 Read the ZIP64 end-of-archive records and use that to update endrec
98 locatorSize
= struct
.calcsize(structEndArchive64Locator
)
99 fpin
.seek(offset
- locatorSize
, 2)
100 data
= fpin
.read(locatorSize
)
101 sig
, diskno
, reloff
, disks
= struct
.unpack(structEndArchive64Locator
, data
)
102 if sig
!= stringEndArchive64Locator
:
105 if diskno
!= 0 or disks
!= 1:
106 raise BadZipfile("zipfiles that span multiple disks are not supported")
108 # Assume no 'zip64 extensible data'
109 endArchiveSize
= struct
.calcsize(structEndArchive64
)
110 fpin
.seek(offset
- locatorSize
- endArchiveSize
, 2)
111 data
= fpin
.read(endArchiveSize
)
112 sig
, sz
, create_version
, read_version
, disk_num
, disk_dir
, \
113 dircount
, dircount2
, dirsize
, diroffset
= \
114 struct
.unpack(structEndArchive64
, data
)
115 if sig
!= stringEndArchive64
:
118 # Update the original endrec using data from the ZIP64 record
122 endrec
[4] = dircount2
124 endrec
[6] = diroffset
128 def _EndRecData(fpin
):
129 """Return data from the "End of Central Directory" record, or None.
131 The data is a list of the nine items in the ZIP "End of central dir"
132 record followed by a tenth item, the file seek offset of this record."""
133 fpin
.seek(-22, 2) # Assume no archive comment.
134 filesize
= fpin
.tell() + 22 # Get file size
136 if data
[0:4] == stringEndArchive
and data
[-2:] == "\000\000":
137 endrec
= struct
.unpack(structEndArchive
, data
)
138 endrec
= list(endrec
)
139 endrec
.append("") # Append the archive comment
140 endrec
.append(filesize
- 22) # Append the record start offset
141 if endrec
[-4] == -1 or endrec
[-4] == 0xffffffff:
142 return _EndRecData64(fpin
, -22, endrec
)
144 # Search the last END_BLOCK bytes of the file for the record signature.
145 # The comment is appended to the ZIP file and has a 16 bit length.
146 # So the comment may be up to 64K long. We limit the search for the
147 # signature to a few Kbytes at the end of the file for efficiency.
148 # also, the signature must not appear in the comment.
149 END_BLOCK
= min(filesize
, 1024 * 4)
150 fpin
.seek(filesize
- END_BLOCK
, 0)
152 start
= data
.rfind(stringEndArchive
)
153 if start
>= 0: # Correct signature string was found
154 endrec
= struct
.unpack(structEndArchive
, data
[start
:start
+22])
155 endrec
= list(endrec
)
156 comment
= data
[start
+22:]
157 if endrec
[7] == len(comment
): # Comment length checks out
158 # Append the archive comment and start offset
159 endrec
.append(comment
)
160 endrec
.append(filesize
- END_BLOCK
+ start
)
161 if endrec
[-4] == -1 or endrec
[-4] == 0xffffffff:
162 return _EndRecData64(fpin
, - END_BLOCK
+ start
, endrec
)
164 return # Error, return None
167 class ZipInfo (object):
168 """Class with attributes describing each file in the ZIP archive."""
191 def __init__(self
, filename
="NoName", date_time
=(1980,1,1,0,0,0)):
192 self
.orig_filename
= filename
# Original file name in archive
194 # Terminate the file name at the first null byte. Null bytes in file
195 # names are used as tricks by viruses in archives.
196 null_byte
= filename
.find(chr(0))
198 filename
= filename
[0:null_byte
]
199 # This is used to ensure paths in generated ZIP files always use
200 # forward slashes as the directory separator, as required by the
201 # ZIP format specification.
202 if os
.sep
!= "/" and os
.sep
in filename
:
203 filename
= filename
.replace(os
.sep
, "/")
205 self
.filename
= filename
# Normalized file name
206 self
.date_time
= date_time
# year, month, day, hour, min, sec
208 self
.compress_type
= ZIP_STORED
# Type of compression for the file
209 self
.comment
= "" # Comment for each file
210 self
.extra
= "" # ZIP extra data
211 if sys
.platform
== 'win32':
212 self
.create_system
= 0 # System which created ZIP archive
214 # Assume everything else is unix-y
215 self
.create_system
= 3 # System which created ZIP archive
216 self
.create_version
= 20 # Version which created ZIP archive
217 self
.extract_version
= 20 # Version needed to extract archive
218 self
.reserved
= 0 # Must be zero
219 self
.flag_bits
= 0 # ZIP flag bits
220 self
.volume
= 0 # Volume number of file header
221 self
.internal_attr
= 0 # Internal attributes
222 self
.external_attr
= 0 # External file attributes
223 # Other attributes are set by class ZipFile:
224 # header_offset Byte offset to the file header
225 # CRC CRC-32 of the uncompressed file
226 # compress_size Size of the compressed file
227 # file_size Size of the uncompressed file
229 def FileHeader(self
):
230 """Return the per-file header as a string."""
232 dosdate
= (dt
[0] - 1980) << 9 | dt
[1] << 5 | dt
[2]
233 dostime
= dt
[3] << 11 | dt
[4] << 5 |
(dt
[5] // 2)
234 if self
.flag_bits
& 0x08:
235 # Set these to zero because we write them after the file data
236 CRC
= compress_size
= file_size
= 0
239 compress_size
= self
.compress_size
240 file_size
= self
.file_size
244 if file_size
> ZIP64_LIMIT
or compress_size
> ZIP64_LIMIT
:
245 # File is larger than what fits into a 4 byte integer,
246 # fall back to the ZIP64 extension
248 extra
= extra
+ struct
.pack(fmt
,
249 1, struct
.calcsize(fmt
)-4, file_size
, compress_size
)
250 file_size
= 0xffffffff # -1
251 compress_size
= 0xffffffff # -1
252 self
.extract_version
= max(45, self
.extract_version
)
253 self
.create_version
= max(45, self
.extract_version
)
255 header
= struct
.pack(structFileHeader
, stringFileHeader
,
256 self
.extract_version
, self
.reserved
, self
.flag_bits
,
257 self
.compress_type
, dostime
, dosdate
, CRC
,
258 compress_size
, file_size
,
259 len(self
.filename
), len(extra
))
260 return header
+ self
.filename
+ extra
262 def _decodeExtra(self
):
263 # Try to decode the extra field.
265 unpack
= struct
.unpack
267 tp
, ln
= unpack('<hh', extra
[:4])
270 counts
= unpack('<qqq', extra
[4:28])
272 counts
= unpack('<qq', extra
[4:20])
274 counts
= unpack('<q', extra
[4:12])
278 raise RuntimeError, "Corrupt extra field %s"%(ln
,)
282 # ZIP64 extension (large files and/or large archives)
283 if self
.file_size
== -1 or self
.file_size
== 0xFFFFFFFFL
:
284 self
.file_size
= counts
[idx
]
287 if self
.compress_size
== -1 or self
.compress_size
== 0xFFFFFFFFL
:
288 self
.compress_size
= counts
[idx
]
291 if self
.header_offset
== -1 or self
.header_offset
== 0xffffffffL
:
292 old
= self
.header_offset
293 self
.header_offset
= counts
[idx
]
300 """ Class with methods to open, read, write, close, list zip files.
302 z = ZipFile(file, mode="r", compression=ZIP_STORED, allowZip64=True)
304 file: Either the path to the file, or a file-like object.
305 If it is a path, the file will be opened and closed by ZipFile.
306 mode: The mode can be either read "r", write "w" or append "a".
307 compression: ZIP_STORED (no compression) or ZIP_DEFLATED (requires zlib).
308 allowZip64: if True ZipFile will create files with ZIP64 extensions when
309 needed, otherwise it will raise an exception when this would
314 fp
= None # Set here since __del__ checks it
316 def __init__(self
, file, mode
="r", compression
=ZIP_STORED
, allowZip64
=False):
317 """Open the ZIP file with mode read "r", write "w" or append "a"."""
318 self
._allowZip
64 = allowZip64
319 self
._didModify
= False
320 if compression
== ZIP_STORED
:
322 elif compression
== ZIP_DEFLATED
:
325 "Compression requires the (missing) zlib module"
327 raise RuntimeError, "That compression method is not supported"
328 self
.debug
= 0 # Level of printing: 0 through 3
329 self
.NameToInfo
= {} # Find file info given name
330 self
.filelist
= [] # List of ZipInfo instances for archive
331 self
.compression
= compression
# Method of compression
332 self
.mode
= key
= mode
.replace('b', '')[0]
334 # Check if we were passed a file-like object
335 if isinstance(file, basestring
):
338 modeDict
= {'r' : 'rb', 'w': 'wb', 'a' : 'r+b'}
339 self
.fp
= open(file, modeDict
[mode
])
343 self
.filename
= getattr(file, 'name', None)
350 try: # See if file is a zip file
351 self
._RealGetContents
()
352 # seek to start of directory and overwrite
353 self
.fp
.seek(self
.start_dir
, 0)
354 except BadZipfile
: # file is not a zip file, just append
357 if not self
._filePassed
:
360 raise RuntimeError, 'Mode must be "r", "w" or "a"'
362 def _GetContents(self
):
363 """Read the directory, making sure we close the file if the format
366 self
._RealGetContents
()
368 if not self
._filePassed
:
373 def _RealGetContents(self
):
374 """Read in the table of contents for the ZIP file."""
376 endrec
= _EndRecData(fp
)
378 raise BadZipfile
, "File is not a zip file"
381 size_cd
= endrec
[5] # bytes in central directory
382 offset_cd
= endrec
[6] # offset of central directory
383 self
.comment
= endrec
[8] # archive comment
384 # endrec[9] is the offset of the "End of Central Dir" record
385 if endrec
[9] > ZIP64_LIMIT
:
386 x
= endrec
[9] - size_cd
- 56 - 20
388 x
= endrec
[9] - size_cd
389 # "concat" is zero, unless zip was concatenated to another file
390 concat
= x
- offset_cd
392 print "given, inferred, offset", offset_cd
, x
, concat
393 # self.start_dir: Position of start of central directory
394 self
.start_dir
= offset_cd
+ concat
395 fp
.seek(self
.start_dir
, 0)
396 data
= fp
.read(size_cd
)
397 fp
= cStringIO
.StringIO(data
)
399 while total
< size_cd
:
400 centdir
= fp
.read(46)
402 if centdir
[0:4] != stringCentralDir
:
403 raise BadZipfile
, "Bad magic number for central directory"
404 centdir
= struct
.unpack(structCentralDir
, centdir
)
407 filename
= fp
.read(centdir
[_CD_FILENAME_LENGTH
])
408 # Create ZipInfo instance to store file information
409 x
= ZipInfo(filename
)
410 x
.extra
= fp
.read(centdir
[_CD_EXTRA_FIELD_LENGTH
])
411 x
.comment
= fp
.read(centdir
[_CD_COMMENT_LENGTH
])
412 total
= (total
+ centdir
[_CD_FILENAME_LENGTH
]
413 + centdir
[_CD_EXTRA_FIELD_LENGTH
]
414 + centdir
[_CD_COMMENT_LENGTH
])
415 x
.header_offset
= centdir
[_CD_LOCAL_HEADER_OFFSET
]
416 (x
.create_version
, x
.create_system
, x
.extract_version
, x
.reserved
,
417 x
.flag_bits
, x
.compress_type
, t
, d
,
418 x
.CRC
, x
.compress_size
, x
.file_size
) = centdir
[1:12]
419 x
.volume
, x
.internal_attr
, x
.external_attr
= centdir
[15:18]
420 # Convert date/time code to (year, month, day, hour, min, sec)
421 x
.date_time
= ( (d
>>9)+1980, (d
>>5)&0xF, d
&0x1F,
422 t
>>11, (t
>>5)&0x3F, (t
&0x1F) * 2 )
425 x
.header_offset
= x
.header_offset
+ concat
426 self
.filelist
.append(x
)
427 self
.NameToInfo
[x
.filename
] = x
433 """Return a list of file names in the archive."""
435 for data
in self
.filelist
:
436 l
.append(data
.filename
)
440 """Return a list of class ZipInfo instances for files in the
445 """Print a table of contents for the zip file."""
446 print "%-46s %19s %12s" % ("File Name", "Modified ", "Size")
447 for zinfo
in self
.filelist
:
448 date
= "%d-%02d-%02d %02d:%02d:%02d" % zinfo
.date_time
449 print "%-46s %s %12d" % (zinfo
.filename
, date
, zinfo
.file_size
)
452 """Read all the files and check the CRC."""
453 for zinfo
in self
.filelist
:
455 self
.read(zinfo
.filename
) # Check CRC-32
457 return zinfo
.filename
460 def getinfo(self
, name
):
461 """Return the instance of ZipInfo given 'name'."""
462 return self
.NameToInfo
[name
]
464 def read(self
, name
):
465 """Return file bytes (as a string) for name."""
466 if self
.mode
not in ("r", "a"):
467 raise RuntimeError, 'read() requires mode "r" or "a"'
469 raise RuntimeError, \
470 "Attempt to read ZIP archive that was already closed"
471 zinfo
= self
.getinfo(name
)
472 filepos
= self
.fp
.tell()
474 self
.fp
.seek(zinfo
.header_offset
, 0)
476 # Skip the file header:
477 fheader
= self
.fp
.read(30)
478 if fheader
[0:4] != stringFileHeader
:
479 raise BadZipfile
, "Bad magic number for file header"
481 fheader
= struct
.unpack(structFileHeader
, fheader
)
482 fname
= self
.fp
.read(fheader
[_FH_FILENAME_LENGTH
])
483 if fheader
[_FH_EXTRA_FIELD_LENGTH
]:
484 self
.fp
.read(fheader
[_FH_EXTRA_FIELD_LENGTH
])
486 if fname
!= zinfo
.orig_filename
:
488 'File name in directory "%s" and header "%s" differ.' % (
489 zinfo
.orig_filename
, fname
)
491 bytes
= self
.fp
.read(zinfo
.compress_size
)
492 self
.fp
.seek(filepos
, 0)
493 if zinfo
.compress_type
== ZIP_STORED
:
495 elif zinfo
.compress_type
== ZIP_DEFLATED
:
497 raise RuntimeError, \
498 "De-compression requires the (missing) zlib module"
499 # zlib compress/decompress code by Jeremy Hylton of CNRI
500 dc
= zlib
.decompressobj(-15)
501 bytes
= dc
.decompress(bytes
)
502 # need to feed in unused pad byte so that zlib won't choke
503 ex
= dc
.decompress('Z') + dc
.flush()
508 "Unsupported compression method %d for file %s" % \
509 (zinfo
.compress_type
, name
)
510 crc
= binascii
.crc32(bytes
)
512 raise BadZipfile
, "Bad CRC-32 for file %s" % name
515 def _writecheck(self
, zinfo
):
516 """Check for errors before writing a file to the archive."""
517 if zinfo
.filename
in self
.NameToInfo
:
518 if self
.debug
: # Warning for duplicate names
519 print "Duplicate name:", zinfo
.filename
520 if self
.mode
not in ("w", "a"):
521 raise RuntimeError, 'write() requires mode "w" or "a"'
523 raise RuntimeError, \
524 "Attempt to write ZIP archive that was already closed"
525 if zinfo
.compress_type
== ZIP_DEFLATED
and not zlib
:
526 raise RuntimeError, \
527 "Compression requires the (missing) zlib module"
528 if zinfo
.compress_type
not in (ZIP_STORED
, ZIP_DEFLATED
):
529 raise RuntimeError, \
530 "That compression method is not supported"
531 if zinfo
.file_size
> ZIP64_LIMIT
:
532 if not self
._allowZip
64:
533 raise LargeZipFile("Filesize would require ZIP64 extensions")
534 if zinfo
.header_offset
> ZIP64_LIMIT
:
535 if not self
._allowZip
64:
536 raise LargeZipFile("Zipfile size would require ZIP64 extensions")
538 def write(self
, filename
, arcname
=None, compress_type
=None):
539 """Put the bytes from filename into the archive under the name
541 st
= os
.stat(filename
)
542 mtime
= time
.localtime(st
.st_mtime
)
543 date_time
= mtime
[0:6]
544 # Create ZipInfo instance to store file information
547 arcname
= os
.path
.normpath(os
.path
.splitdrive(arcname
)[1])
548 while arcname
[0] in (os
.sep
, os
.altsep
):
549 arcname
= arcname
[1:]
550 zinfo
= ZipInfo(arcname
, date_time
)
551 zinfo
.external_attr
= (st
[0] & 0xFFFF) << 16L # Unix attributes
552 if compress_type
is None:
553 zinfo
.compress_type
= self
.compression
555 zinfo
.compress_type
= compress_type
557 zinfo
.file_size
= st
.st_size
558 zinfo
.flag_bits
= 0x00
559 zinfo
.header_offset
= self
.fp
.tell() # Start of header bytes
561 self
._writecheck
(zinfo
)
562 self
._didModify
= True
563 fp
= open(filename
, "rb")
564 # Must overwrite CRC and sizes with correct data later
566 zinfo
.compress_size
= compress_size
= 0
567 zinfo
.file_size
= file_size
= 0
568 self
.fp
.write(zinfo
.FileHeader())
569 if zinfo
.compress_type
== ZIP_DEFLATED
:
570 cmpr
= zlib
.compressobj(zlib
.Z_DEFAULT_COMPRESSION
,
575 buf
= fp
.read(1024 * 8)
578 file_size
= file_size
+ len(buf
)
579 CRC
= binascii
.crc32(buf
, CRC
)
581 buf
= cmpr
.compress(buf
)
582 compress_size
= compress_size
+ len(buf
)
587 compress_size
= compress_size
+ len(buf
)
589 zinfo
.compress_size
= compress_size
591 zinfo
.compress_size
= file_size
593 zinfo
.file_size
= file_size
594 # Seek backwards and write CRC and file sizes
595 position
= self
.fp
.tell() # Preserve current position in file
596 self
.fp
.seek(zinfo
.header_offset
+ 14, 0)
597 self
.fp
.write(struct
.pack("<lLL", zinfo
.CRC
, zinfo
.compress_size
,
599 self
.fp
.seek(position
, 0)
600 self
.filelist
.append(zinfo
)
601 self
.NameToInfo
[zinfo
.filename
] = zinfo
603 def writestr(self
, zinfo_or_arcname
, bytes
):
604 """Write a file into the archive. The contents is the string
605 'bytes'. 'zinfo_or_arcname' is either a ZipInfo instance or
606 the name of the file in the archive."""
607 if not isinstance(zinfo_or_arcname
, ZipInfo
):
608 zinfo
= ZipInfo(filename
=zinfo_or_arcname
,
609 date_time
=time
.localtime(time
.time()))
610 zinfo
.compress_type
= self
.compression
612 zinfo
= zinfo_or_arcname
613 zinfo
.file_size
= len(bytes
) # Uncompressed size
614 zinfo
.header_offset
= self
.fp
.tell() # Start of header bytes
615 self
._writecheck
(zinfo
)
616 self
._didModify
= True
617 zinfo
.CRC
= binascii
.crc32(bytes
) # CRC-32 checksum
618 if zinfo
.compress_type
== ZIP_DEFLATED
:
619 co
= zlib
.compressobj(zlib
.Z_DEFAULT_COMPRESSION
,
621 bytes
= co
.compress(bytes
) + co
.flush()
622 zinfo
.compress_size
= len(bytes
) # Compressed size
624 zinfo
.compress_size
= zinfo
.file_size
625 zinfo
.header_offset
= self
.fp
.tell() # Start of header bytes
626 self
.fp
.write(zinfo
.FileHeader())
629 if zinfo
.flag_bits
& 0x08:
630 # Write CRC and file sizes after the file data
631 self
.fp
.write(struct
.pack("<lLL", zinfo
.CRC
, zinfo
.compress_size
,
633 self
.filelist
.append(zinfo
)
634 self
.NameToInfo
[zinfo
.filename
] = zinfo
637 """Call the "close()" method in case the user forgot."""
641 """Close the file, and for mode "w" and "a" write the ending
646 if self
.mode
in ("w", "a") and self
._didModify
: # write ending records
648 pos1
= self
.fp
.tell()
649 for zinfo
in self
.filelist
: # write central directory
652 dosdate
= (dt
[0] - 1980) << 9 | dt
[1] << 5 | dt
[2]
653 dostime
= dt
[3] << 11 | dt
[4] << 5 |
(dt
[5] // 2)
655 if zinfo
.file_size
> ZIP64_LIMIT \
656 or zinfo
.compress_size
> ZIP64_LIMIT
:
657 extra
.append(zinfo
.file_size
)
658 extra
.append(zinfo
.compress_size
)
659 file_size
= 0xffffffff #-1
660 compress_size
= 0xffffffff #-1
662 file_size
= zinfo
.file_size
663 compress_size
= zinfo
.compress_size
665 if zinfo
.header_offset
> ZIP64_LIMIT
:
666 extra
.append(zinfo
.header_offset
)
667 header_offset
= 0xffffffff #-1
669 header_offset
= zinfo
.header_offset
671 extra_data
= zinfo
.extra
673 # Append a ZIP64 field to the extra's
674 extra_data
= struct
.pack(
675 '<hh' + 'q'*len(extra
),
676 1, 8*len(extra
), *extra
) + extra_data
678 extract_version
= max(45, zinfo
.extract_version
)
679 create_version
= max(45, zinfo
.create_version
)
681 extract_version
= zinfo
.extract_version
682 create_version
= zinfo
.create_version
684 centdir
= struct
.pack(structCentralDir
,
685 stringCentralDir
, create_version
,
686 zinfo
.create_system
, extract_version
, zinfo
.reserved
,
687 zinfo
.flag_bits
, zinfo
.compress_type
, dostime
, dosdate
,
688 zinfo
.CRC
, compress_size
, file_size
,
689 len(zinfo
.filename
), len(extra_data
), len(zinfo
.comment
),
690 0, zinfo
.internal_attr
, zinfo
.external_attr
,
692 self
.fp
.write(centdir
)
693 self
.fp
.write(zinfo
.filename
)
694 self
.fp
.write(extra_data
)
695 self
.fp
.write(zinfo
.comment
)
697 pos2
= self
.fp
.tell()
698 # Write end-of-zip-archive record
699 if pos1
> ZIP64_LIMIT
:
700 # Need to write the ZIP64 end-of-archive records
701 zip64endrec
= struct
.pack(
702 structEndArchive64
, stringEndArchive64
,
703 44, 45, 45, 0, 0, count
, count
, pos2
- pos1
, pos1
)
704 self
.fp
.write(zip64endrec
)
706 zip64locrec
= struct
.pack(
707 structEndArchive64Locator
,
708 stringEndArchive64Locator
, 0, pos2
, 1)
709 self
.fp
.write(zip64locrec
)
711 pos3
= self
.fp
.tell()
712 endrec
= struct
.pack(structEndArchive
, stringEndArchive
,
713 0, 0, count
, count
, pos2
- pos1
, 0xffffffff, 0) # -1, 0)
714 self
.fp
.write(endrec
)
717 endrec
= struct
.pack(structEndArchive
, stringEndArchive
,
718 0, 0, count
, count
, pos2
- pos1
, pos1
, 0)
719 self
.fp
.write(endrec
)
721 if not self
._filePassed
:
726 class PyZipFile(ZipFile
):
727 """Class to create ZIP archives with Python library files and packages."""
729 def writepy(self
, pathname
, basename
= ""):
730 """Add all files from "pathname" to the ZIP archive.
732 If pathname is a package directory, search the directory and
733 all package subdirectories recursively for all *.py and enter
734 the modules into the archive. If pathname is a plain
735 directory, listdir *.py and enter all modules. Else, pathname
736 must be a Python *.py file and the module will be put into the
737 archive. Added modules are always module.pyo or module.pyc.
738 This method will compile the module.py into module.pyc if
741 dir, name
= os
.path
.split(pathname
)
742 if os
.path
.isdir(pathname
):
743 initname
= os
.path
.join(pathname
, "__init__.py")
744 if os
.path
.isfile(initname
):
745 # This is a package directory, add it
747 basename
= "%s/%s" % (basename
, name
)
751 print "Adding package in", pathname
, "as", basename
752 fname
, arcname
= self
._get
_codename
(initname
[0:-3], basename
)
754 print "Adding", arcname
755 self
.write(fname
, arcname
)
756 dirlist
= os
.listdir(pathname
)
757 dirlist
.remove("__init__.py")
758 # Add all *.py files and package subdirectories
759 for filename
in dirlist
:
760 path
= os
.path
.join(pathname
, filename
)
761 root
, ext
= os
.path
.splitext(filename
)
762 if os
.path
.isdir(path
):
763 if os
.path
.isfile(os
.path
.join(path
, "__init__.py")):
764 # This is a package directory, add it
765 self
.writepy(path
, basename
) # Recursive call
767 fname
, arcname
= self
._get
_codename
(path
[0:-3],
770 print "Adding", arcname
771 self
.write(fname
, arcname
)
773 # This is NOT a package directory, add its files at top level
775 print "Adding files from directory", pathname
776 for filename
in os
.listdir(pathname
):
777 path
= os
.path
.join(pathname
, filename
)
778 root
, ext
= os
.path
.splitext(filename
)
780 fname
, arcname
= self
._get
_codename
(path
[0:-3],
783 print "Adding", arcname
784 self
.write(fname
, arcname
)
786 if pathname
[-3:] != ".py":
787 raise RuntimeError, \
788 'Files added with writepy() must end with ".py"'
789 fname
, arcname
= self
._get
_codename
(pathname
[0:-3], basename
)
791 print "Adding file", arcname
792 self
.write(fname
, arcname
)
794 def _get_codename(self
, pathname
, basename
):
795 """Return (filename, archivename) for the path.
797 Given a module name path, return the correct file path and
798 archive name, compiling if necessary. For example, given
799 /python/lib/string, return (/python/lib/string.pyc, string).
801 file_py
= pathname
+ ".py"
802 file_pyc
= pathname
+ ".pyc"
803 file_pyo
= pathname
+ ".pyo"
804 if os
.path
.isfile(file_pyo
) and \
805 os
.stat(file_pyo
).st_mtime
>= os
.stat(file_py
).st_mtime
:
806 fname
= file_pyo
# Use .pyo file
807 elif not os
.path
.isfile(file_pyc
) or \
808 os
.stat(file_pyc
).st_mtime
< os
.stat(file_py
).st_mtime
:
811 print "Compiling", file_py
813 py_compile
.compile(file_py
, file_pyc
, None, True)
814 except py_compile
.PyCompileError
,err
:
819 archivename
= os
.path
.split(fname
)[1]
821 archivename
= "%s/%s" % (basename
, archivename
)
822 return (fname
, archivename
)
825 def main(args
= None):
827 USAGE
=textwrap
.dedent("""\
829 zipfile.py -l zipfile.zip # Show listing of a zipfile
830 zipfile.py -t zipfile.zip # Test if a zipfile is valid
831 zipfile.py -e zipfile.zip target # Extract zipfile into target dir
832 zipfile.py -c zipfile.zip src ... # Create zipfile from sources
837 if not args
or args
[0] not in ('-l', '-c', '-e', '-t'):
845 zf
= ZipFile(args
[1], 'r')
849 elif args
[0] == '-t':
853 zf
= ZipFile(args
[1], 'r')
857 elif args
[0] == '-e':
862 zf
= ZipFile(args
[1], 'r')
864 for path
in zf
.namelist():
865 if path
.startswith('./'):
866 tgt
= os
.path
.join(out
, path
[2:])
868 tgt
= os
.path
.join(out
, path
)
870 tgtdir
= os
.path
.dirname(tgt
)
871 if not os
.path
.exists(tgtdir
):
874 fp
.write(zf
.read(path
))
878 elif args
[0] == '-c':
883 def addToZip(zf
, path
, zippath
):
884 if os
.path
.isfile(path
):
885 zf
.write(path
, zippath
, ZIP_DEFLATED
)
886 elif os
.path
.isdir(path
):
887 for nm
in os
.listdir(path
):
889 os
.path
.join(path
, nm
), os
.path
.join(zippath
, nm
))
892 zf
= ZipFile(args
[1], 'w', allowZip64
=True)
894 addToZip(zf
, src
, os
.path
.basename(src
))
898 if __name__
== "__main__":