1 # pack.py -- For dealing with packed git objects.
2 # Copyright (C) 2007 James Westby <jw+debian@jameswestby.net>
3 # Copyright (C) 2008-2009 Jelmer Vernooij <jelmer@samba.org>
5 # This program is free software; you can redistribute it and/or
6 # modify it under the terms of the GNU General Public License
7 # as published by the Free Software Foundation; version 2
8 # of the License or (at your option) a later version.
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.
15 # You should have received a copy of the GNU General Public License
16 # along with this program; if not, write to the Free Software
17 # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
20 """Classes for dealing with packed git objects.
22 A pack is a compact representation of a bunch of objects, stored
23 using deltas where possible.
25 They have two parts, the pack file, which stores the data, and an index
26 that tells you where the data is.
28 To find an object you look in all of the index files 'til you find a
29 match for the object name. You then use the pointer got from this as
30 a pointer in to the corresponding packfile.
34 from collections import defaultdict
36 from misc import defaultdict
38 from cStringIO import (
41 from collections import (
45 from itertools import (
54 from struct import unpack_from
56 from dulwich.misc import unpack_from
60 from dulwich.errors import (
64 from dulwich.file import GitFile
65 from dulwich.lru_cache import (
68 from dulwich.misc import (
72 from dulwich.objects import (
79 supports_mmap_offset = (sys.version_info[0] >= 3 or
80 (sys.version_info[0] == 2 and sys.version_info[1] >= 6))
86 DELTA_TYPES = (OFS_DELTA, REF_DELTA)
89 def take_msb_bytes(read):
90 """Read bytes marked with most significant bit.
92 :param read: Read function
95 while len(ret) == 0 or ret[-1] & 0x80:
96 ret.append(ord(read(1)))
100 def read_zlib_chunks(read_some, dec_size, buffer_size=4096):
101 """Read zlib data from a buffer.
103 This function requires that the buffer have additional data following the
104 compressed data, which is guaranteed to be the case for git pack files.
106 :param read_some: Read function that returns at least one byte, but may
107 return less than the requested size
108 :param dec_size: Expected size of the decompressed buffer
109 :param buffer_size: Size of the read buffer
110 :return: Tuple with list of chunks, length of compressed data length and
111 and unused read data.
112 :raise zlib.error: if a decompression error occurred.
115 raise ValueError("non-negative zlib data stream size expected")
116 obj = zlib.decompressobj()
120 while obj.unused_data == "":
121 add = read_some(buffer_size)
123 raise zlib.error("EOF before end of zlib stream")
125 decomp = obj.decompress(add)
129 raise zlib.error("decompressed data does not match expected size")
130 comp_len = fed - len(obj.unused_data)
131 return ret, comp_len, obj.unused_data
135 """Return the hexdigest of the SHA1 over a set of names.
137 :param iter: Iterator over string objects
138 :return: 40-byte hex sha1 digest
143 return sha1.hexdigest()
146 def load_pack_index(path):
147 """Load an index file by path.
149 :param filename: Path to the index file
150 :return: A PackIndex loaded from the given path
152 f = GitFile(path, 'rb')
154 return load_pack_index_file(path, f)
159 def _load_file_contents(f, size=None):
160 fileno = getattr(f, 'fileno', None)
161 # Attempt to use mmap if possible
162 if fileno is not None:
165 size = os.fstat(fd).st_size
167 contents = mmap.mmap(fd, size, access=mmap.ACCESS_READ)
172 return contents, size
175 return contents, size
178 def load_pack_index_file(path, f):
179 """Load an index file from a file-like object.
181 :param path: Path for the index file
182 :param f: File-like object
183 :return: A PackIndex loaded from the given file
185 contents, size = _load_file_contents(f)
186 if contents[:4] == '\377tOc':
187 version = struct.unpack(">L", contents[4:8])[0]
189 return PackIndex2(path, file=f, contents=contents,
192 raise KeyError("Unknown pack index format %d" % version)
194 return PackIndex1(path, file=f, contents=contents, size=size)
197 def bisect_find_sha(start, end, sha, unpack_name):
198 """Find a SHA in a data blob with sorted SHAs.
200 :param start: Start index of range to search
201 :param end: End index of range to search
202 :param sha: Sha to find
203 :param unpack_name: Callback to retrieve SHA by index
204 :return: Index of the SHA, or None if it wasn't found
209 file_sha = unpack_name(i)
210 x = cmp(file_sha, sha)
220 class PackIndex(object):
221 """An index in to a packfile.
223 Given a sha id of an object a pack index can tell you the location in the
224 packfile of that object if it has it.
227 def __eq__(self, other):
228 if not isinstance(other, PackIndex):
231 for (name1, _, _), (name2, _, _) in izip(self.iterentries(),
232 other.iterentries()):
237 def __ne__(self, other):
238 return not self.__eq__(other)
241 """Return the number of entries in this pack index."""
242 raise NotImplementedError(self.__len__)
245 """Iterate over the SHAs in this pack."""
246 raise NotImplementedError(self.__iter__)
248 def iterentries(self):
249 """Iterate over the entries in this pack index.
251 :return: iterator over tuples with object name, offset in packfile and
254 raise NotImplementedError(self.iterentries)
256 def get_pack_checksum(self):
257 """Return the SHA1 checksum stored for the corresponding packfile.
259 :return: 20-byte binary digest
261 raise NotImplementedError(self.get_pack_checksum)
263 def object_index(self, sha):
264 """Return the index in to the corresponding packfile for the object.
266 Given the name of an object it will return the offset that object
267 lives at within the corresponding pack file. If the pack file doesn't
268 have the object then None will be returned.
271 sha = hex_to_sha(sha)
272 return self._object_index(sha)
274 def _object_index(self, sha):
277 :param sha: A *binary* SHA string. (20 characters long)_
279 raise NotImplementedError(self._object_index)
282 class FilePackIndex(PackIndex):
283 """Pack index that is based on a file.
285 To do the loop it opens the file, and indexes first 256 4 byte groups
286 with the first byte of the sha id. The value in the four byte group indexed
287 is the end of the group that shares the same starting byte. Subtract one
288 from the starting byte and index again to find the start of the group.
289 The values are sorted by sha id within the group, so do the math to find
290 the start and end offset and then bisect in to find if the value is present.
293 def __init__(self, filename, file=None, contents=None, size=None):
294 """Create a pack index object.
296 Provide it with the name of the index file to consider, and it will map
297 it whenever required.
299 self._filename = filename
300 # Take the size now, so it can be checked each time we map the file to
301 # ensure that it hasn't changed.
303 self._file = GitFile(filename, 'rb')
307 self._contents, self._size = _load_file_contents(file, size)
309 self._contents, self._size = (contents, size)
311 def __eq__(self, other):
312 # Quick optimization:
313 if isinstance(other, FilePackIndex) and self._fan_out_table != other._fan_out_table:
316 return super(FilePackIndex, self).__eq__(other)
322 """Return the number of entries in this pack index."""
323 return self._fan_out_table[-1]
325 def _unpack_entry(self, i):
326 """Unpack the i-th entry in the index file.
328 :return: Tuple with object name (SHA), offset in pack file and CRC32
331 raise NotImplementedError(self._unpack_entry)
333 def _unpack_name(self, i):
334 """Unpack the i-th name from the index file."""
335 raise NotImplementedError(self._unpack_name)
337 def _unpack_offset(self, i):
338 """Unpack the i-th object offset from the index file."""
339 raise NotImplementedError(self._unpack_offset)
341 def _unpack_crc32_checksum(self, i):
342 """Unpack the crc32 checksum for the i-th object from the index file."""
343 raise NotImplementedError(self._unpack_crc32_checksum)
346 """Iterate over the SHAs in this pack."""
347 return imap(sha_to_hex, self._itersha())
350 for i in range(len(self)):
351 yield self._unpack_name(i)
353 def objects_sha1(self):
354 """Return the hex SHA1 over all the shas of all objects in this pack.
356 :note: This is used for the filename of the pack.
358 return iter_sha1(self._itersha())
360 def iterentries(self):
361 """Iterate over the entries in this pack index.
363 :return: iterator over tuples with object name, offset in packfile and
366 for i in range(len(self)):
367 yield self._unpack_entry(i)
369 def _read_fan_out_table(self, start_offset):
371 for i in range(0x100):
372 fanout_entry = self._contents[start_offset+i*4:start_offset+(i+1)*4]
373 ret.append(struct.unpack(">L", fanout_entry)[0])
377 """Check that the stored checksum matches the actual checksum."""
378 actual = self.calculate_checksum()
379 stored = self.get_stored_checksum()
381 raise ChecksumMismatch(stored, actual)
383 def calculate_checksum(self):
384 """Calculate the SHA1 checksum over this pack index.
386 :return: This is a 20-byte binary digest
388 return make_sha(self._contents[:-20]).digest()
390 def get_pack_checksum(self):
391 """Return the SHA1 checksum stored for the corresponding packfile.
393 :return: 20-byte binary digest
395 return str(self._contents[-40:-20])
397 def get_stored_checksum(self):
398 """Return the SHA1 checksum stored for this index.
400 :return: 20-byte binary digest
402 return str(self._contents[-20:])
404 def _object_index(self, sha):
407 :param sha: A *binary* SHA string. (20 characters long)_
409 assert len(sha) == 20
414 start = self._fan_out_table[idx-1]
415 end = self._fan_out_table[idx]
416 i = bisect_find_sha(start, end, sha, self._unpack_name)
419 return self._unpack_offset(i)
422 class PackIndex1(FilePackIndex):
423 """Version 1 Pack Index file."""
425 def __init__(self, filename, file=None, contents=None, size=None):
426 super(PackIndex1, self).__init__(filename, file, contents, size)
428 self._fan_out_table = self._read_fan_out_table(0)
430 def _unpack_entry(self, i):
431 (offset, name) = unpack_from(">L20s", self._contents,
432 (0x100 * 4) + (i * 24))
433 return (name, offset, None)
435 def _unpack_name(self, i):
436 offset = (0x100 * 4) + (i * 24) + 4
437 return self._contents[offset:offset+20]
439 def _unpack_offset(self, i):
440 offset = (0x100 * 4) + (i * 24)
441 return unpack_from(">L", self._contents, offset)[0]
443 def _unpack_crc32_checksum(self, i):
444 # Not stored in v1 index files
448 class PackIndex2(FilePackIndex):
449 """Version 2 Pack Index file."""
451 def __init__(self, filename, file=None, contents=None, size=None):
452 super(PackIndex2, self).__init__(filename, file, contents, size)
453 assert self._contents[:4] == '\377tOc', "Not a v2 pack index file"
454 (self.version, ) = unpack_from(">L", self._contents, 4)
455 assert self.version == 2, "Version was %d" % self.version
456 self._fan_out_table = self._read_fan_out_table(8)
457 self._name_table_offset = 8 + 0x100 * 4
458 self._crc32_table_offset = self._name_table_offset + 20 * len(self)
459 self._pack_offset_table_offset = (self._crc32_table_offset +
462 def _unpack_entry(self, i):
463 return (self._unpack_name(i), self._unpack_offset(i),
464 self._unpack_crc32_checksum(i))
466 def _unpack_name(self, i):
467 offset = self._name_table_offset + i * 20
468 return self._contents[offset:offset+20]
470 def _unpack_offset(self, i):
471 offset = self._pack_offset_table_offset + i * 4
472 return unpack_from(">L", self._contents, offset)[0]
474 def _unpack_crc32_checksum(self, i):
475 return unpack_from(">L", self._contents,
476 self._crc32_table_offset + i * 4)[0]
479 def read_pack_header(read):
480 """Read the header of a pack file.
482 :param read: Read function
483 :return: Tuple with pack version and number of objects
486 assert header[:4] == "PACK"
487 (version,) = unpack_from(">L", header, 4)
488 assert version in (2, 3), "Version was %d" % version
489 (num_objects,) = unpack_from(">L", header, 8)
490 return (version, num_objects)
493 def chunks_length(chunks):
494 return sum(imap(len, chunks))
497 def unpack_object(read_all, read_some=None):
498 """Unpack a Git object.
500 :param read_all: Read function that blocks until the number of requested
502 :param read_some: Read function that returns at least one byte, but may not
503 return the number of bytes requested.
504 :return: tuple with type, uncompressed data, compressed size and tail data.
506 if read_some is None:
508 bytes = take_msb_bytes(read_all)
509 type = (bytes[0] >> 4) & 0x07
510 size = bytes[0] & 0x0f
511 for i, byte in enumerate(bytes[1:]):
512 size += (byte & 0x7f) << ((i * 7) + 4)
513 raw_base = len(bytes)
514 if type == OFS_DELTA:
515 bytes = take_msb_bytes(read_all)
516 raw_base += len(bytes)
517 assert not (bytes[-1] & 0x80)
518 delta_base_offset = bytes[0] & 0x7f
519 for byte in bytes[1:]:
520 delta_base_offset += 1
521 delta_base_offset <<= 7
522 delta_base_offset += (byte & 0x7f)
523 uncomp, comp_len, unused = read_zlib_chunks(read_some, size)
524 assert size == chunks_length(uncomp)
525 return type, (delta_base_offset, uncomp), comp_len+raw_base, unused
526 elif type == REF_DELTA:
527 basename = read_all(20)
529 uncomp, comp_len, unused = read_zlib_chunks(read_some, size)
530 assert size == chunks_length(uncomp)
531 return type, (basename, uncomp), comp_len+raw_base, unused
533 uncomp, comp_len, unused = read_zlib_chunks(read_some, size)
534 assert chunks_length(uncomp) == size
535 return type, uncomp, comp_len+raw_base, unused
538 def _compute_object_size((num, obj)):
539 """Compute the size of a unresolved object for use with LRUSizeCache."""
540 if num in DELTA_TYPES:
541 return chunks_length(obj[1])
542 return chunks_length(obj)
545 class PackStreamReader(object):
546 """Class to read a pack stream.
548 The pack is read from a ReceivableProtocol using read() or recv() as
552 def __init__(self, read_all, read_some=None):
553 self.read_all = read_all
554 if read_some is None:
555 self.read_some = read_all
557 self.read_some = read_some
558 self.sha = make_sha()
560 self._rbuf = StringIO()
561 # trailer is a deque to avoid memory allocation on small reads
562 self._trailer = deque()
564 def _read(self, read, size):
565 """Read up to size bytes using the given callback.
567 As a side effect, update the verifier's hash (excluding the last 20
568 bytes read) and write through to the output file.
570 :param read: The read callback to read from.
571 :param size: The maximum number of bytes to read; the particular
572 behavior is callback-specific.
576 # maintain a trailer of the last 20 bytes we've read
579 tn = len(self._trailer)
584 to_pop = max(n + tn - 20, 0)
586 for _ in xrange(to_pop):
587 self.sha.update(self._trailer.popleft())
588 self._trailer.extend(data[-to_add:])
590 # hash everything but the trailer
591 self.sha.update(data[:-to_add])
597 buf.seek(0, SEEK_END)
604 return self._offset - self._buf_len()
606 def read(self, size):
607 """Read, blocking until size bytes are read."""
608 buf_len = self._buf_len()
610 return self._rbuf.read(size)
611 buf_data = self._rbuf.read()
612 self._rbuf = StringIO()
613 return buf_data + self._read(self.read_all, size - buf_len)
615 def recv(self, size):
616 """Read up to size bytes, blocking until one byte is read."""
617 buf_len = self._buf_len()
619 data = self._rbuf.read(size)
621 self._rbuf = StringIO()
623 return self._read(self.read_some, size)
626 return self._num_objects
628 def read_objects(self):
629 """Read the objects in this pack file.
631 :raise AssertionError: if there is an error in the pack format.
632 :raise ChecksumMismatch: if the checksum of the pack contents does not
633 match the checksum in the pack trailer.
634 :raise zlib.error: if an error occurred during zlib decompression.
635 :raise IOError: if an error occurred writing to the output file.
637 pack_version, self._num_objects = read_pack_header(self.read)
638 for i in xrange(self._num_objects):
639 type, uncomp, comp_len, unused = unpack_object(self.read, self.recv)
640 yield type, uncomp, comp_len
642 # prepend any unused data to current read buffer
645 buf.write(self._rbuf.read())
649 pack_sha = sha_to_hex(''.join([c for c in self._trailer]))
650 calculated_sha = self.sha.hexdigest()
651 if pack_sha != calculated_sha:
652 raise ChecksumMismatch(pack_sha, calculated_sha)
655 class PackObjectIterator(object):
657 def __init__(self, pack, progress=None):
659 self.offset = pack._header_size
661 self.map = pack._file
662 self._progress = progress
671 if self.i == self.num:
673 self.map.seek(self.offset)
674 (type, obj, total_size, unused) = unpack_object(self.map.read)
675 self.map.seek(self.offset)
676 crc32 = zlib.crc32(self.map.read(total_size)) & 0xffffffff
677 ret = (self.offset, type, obj, crc32)
678 self.offset += total_size
679 if self._progress is not None:
680 self._progress(self.i, self.num)
684 def obj_sha(type, chunks):
685 """Compute the SHA for a numeric type and object chunks."""
687 sha.update(object_header(type, chunks_length(chunks)))
693 class PackData(object):
694 """The data contained in a packfile.
696 Pack files can be accessed both sequentially for exploding a pack, and
697 directly with the help of an index to retrieve a specific object.
699 The objects within are either complete or a delta aginst another.
701 The header is variable length. If the MSB of each byte is set then it
702 indicates that the subsequent byte is still part of the header.
703 For the first byte the next MS bits are the type, which tells you the type
704 of object, and whether it is a delta. The LS byte is the lowest bits of the
705 size. For each subsequent byte the LS 7 bits are the next MS bits of the
706 size, i.e. the last byte of the header contains the MS bits of the size.
708 For the complete objects the data is stored as zlib deflated data.
709 The size in the header is the uncompressed object size, so to uncompress
710 you need to just keep feeding data to zlib until you get an object back,
711 or it errors on bad data. This is done here by just giving the complete
712 buffer from the start of the deflated object on. This is bad, but until I
713 get mmap sorted out it will have to do.
715 Currently there are no integrity checks done. Also no attempt is made to
716 try and detect the delta case, or a request for an object at the wrong
717 position. It will all just throw a zlib or KeyError.
720 def __init__(self, filename, file=None, size=None):
721 """Create a PackData object representing the pack in the given filename.
723 The file must exist and stay readable until the object is disposed of. It
724 must also stay the same size. It will be mapped whenever needed.
726 Currently there is a restriction on the size of the pack as the python
727 mmap implementation is flawed.
729 self._filename = filename
731 self._header_size = 12
733 self._file = GitFile(self._filename, 'rb')
736 (version, self._num_objects) = read_pack_header(self._file.read)
737 self._offset_cache = LRUSizeCache(1024*1024*20,
738 compute_size=_compute_object_size)
742 def from_file(cls, file, size):
743 return cls(str(file), file=file, size=size)
746 def from_path(cls, path):
747 return cls(filename=path)
753 if self._size is not None:
755 self._size = os.path.getsize(self._filename)
756 if self._size < self._header_size:
757 errmsg = ("%s is too small for a packfile (%d < %d)" %
758 (self._filename, self._size, self._header_size))
759 raise AssertionError(errmsg)
763 """Returns the number of objects in this pack."""
764 return self._num_objects
766 def calculate_checksum(self):
767 """Calculate the checksum for this pack.
769 :return: 20-byte binary SHA1 digest
773 todo = self._get_size() - 20
775 x = self._file.read(min(todo, 1<<16))
780 def get_ref(self, sha):
781 """Get the object for a ref SHA, only looking in this pack."""
782 # TODO: cache these results
783 if self.pack is None:
785 offset = self.pack.index.object_index(sha)
788 type, obj = self.get_object_at(offset)
789 return offset, type, obj
791 def resolve_object(self, offset, type, obj, get_ref=None):
792 """Resolve an object, possibly resolving deltas when necessary.
794 :return: Tuple with object type and contents.
796 if type not in DELTA_TYPES:
800 get_ref = self.get_ref
801 if type == OFS_DELTA:
802 (delta_offset, delta) = obj
803 # TODO: clean up asserts and replace with nicer error messages
804 assert isinstance(offset, int)
805 assert isinstance(delta_offset, int)
806 base_offset = offset-delta_offset
807 type, base_obj = self.get_object_at(base_offset)
808 assert isinstance(type, int)
809 elif type == REF_DELTA:
810 (basename, delta) = obj
811 assert isinstance(basename, str) and len(basename) == 20
812 base_offset, type, base_obj = get_ref(basename)
813 assert isinstance(type, int)
814 type, base_chunks = self.resolve_object(base_offset, type, base_obj)
815 chunks = apply_delta(base_chunks, delta)
816 # TODO(dborowitz): This can result in poor performance if large base
817 # objects are separated from deltas in the pack. We should reorganize
818 # so that we apply deltas to all objects in a chain one after the other
819 # to optimize cache performance.
820 if offset is not None:
821 self._offset_cache[offset] = type, chunks
824 def iterobjects(self, progress=None):
825 return PackObjectIterator(self, progress)
827 def iterentries(self, progress=None):
828 """Yield entries summarizing the contents of this pack.
830 :param progress: Progress function, called with current and total
832 :return: iterator of tuples with (sha, offset, crc32)
834 for offset, type, obj, crc32 in self.iterobjects(progress=progress):
835 assert isinstance(offset, int)
836 assert isinstance(type, int)
837 assert isinstance(obj, list) or isinstance(obj, tuple)
838 type, obj = self.resolve_object(offset, type, obj)
839 yield obj_sha(type, obj), offset, crc32
841 def sorted_entries(self, progress=None):
842 """Return entries in this pack, sorted by SHA.
844 :param progress: Progress function, called with current and total
846 :return: List of tuples with (sha, offset, crc32)
848 ret = list(self.iterentries(progress=progress))
852 def create_index_v1(self, filename, progress=None):
853 """Create a version 1 file for this data file.
855 :param filename: Index filename.
856 :param progress: Progress report function
857 :return: Checksum of index file
859 entries = self.sorted_entries(progress=progress)
860 f = GitFile(filename, 'wb')
862 return write_pack_index_v1(f, entries, self.calculate_checksum())
866 def create_index_v2(self, filename, progress=None):
867 """Create a version 2 index file for this data file.
869 :param filename: Index filename.
870 :param progress: Progress report function
871 :return: Checksum of index file
873 entries = self.sorted_entries(progress=progress)
874 f = GitFile(filename, 'wb')
876 return write_pack_index_v2(f, entries, self.calculate_checksum())
880 def create_index(self, filename, progress=None,
882 """Create an index file for this data file.
884 :param filename: Index filename.
885 :param progress: Progress report function
886 :return: Checksum of index file
889 return self.create_index_v1(filename, progress)
891 return self.create_index_v2(filename, progress)
893 raise ValueError("unknown index format %d" % version)
895 def get_stored_checksum(self):
896 """Return the expected checksum stored in this pack."""
897 self._file.seek(self._get_size()-20)
898 return self._file.read(20)
901 """Check the consistency of this pack."""
902 actual = self.calculate_checksum()
903 stored = self.get_stored_checksum()
905 raise ChecksumMismatch(stored, actual)
907 def get_object_at(self, offset):
908 """Given an offset in to the packfile return the object that is there.
910 Using the associated index the location of an object can be looked up,
911 and then the packfile can be asked directly for that object using this
914 if offset in self._offset_cache:
915 return self._offset_cache[offset]
916 assert isinstance(offset, long) or isinstance(offset, int),\
917 "offset was %r" % offset
918 assert offset >= self._header_size
919 self._file.seek(offset)
920 return unpack_object(self._file.read)[:2]
923 class ThinPackData(PackData):
924 """PackData for thin packs, which require an ObjectStore for resolving."""
926 def __init__(self, resolve_ext_ref, *args, **kwargs):
927 super(ThinPackData, self).__init__(*args, **kwargs)
928 self.resolve_ext_ref = resolve_ext_ref
931 def from_file(cls, resolve_ext_ref, file, size):
932 return cls(resolve_ext_ref, str(file), file=file, size=size)
934 def get_ref(self, sha):
935 """Resolve a reference looking in both this pack and the store."""
937 # As part of completing a pack we create a Pack object with a
938 # ThinPackData and a full PackIndex, so check in the index first if
940 # TODO(dborowitz): reevaluate this when the pack completion code is
942 return super(ThinPackData, self).get_ref(sha)
944 type, obj = self.resolve_ext_ref(sha)
945 return None, type, obj
947 def iterentries(self, progress=None):
948 """Yield entries summarizing the contents of this pack.
950 :param progress: Progress function, called with current and
953 This will yield tuples with (sha, offset, crc32)
956 postponed = defaultdict(list)
958 class Postpone(Exception):
959 """Raised to postpone delta resolving."""
961 def __init__(self, sha):
964 def get_ref_text(sha):
965 assert len(sha) == 20
968 type, obj = self.get_object_at(offset)
969 return offset, type, obj
971 return self.get_ref(sha)
976 todo = chain(self.iterobjects(progress=progress), extra)
977 for (offset, type, obj, crc32) in todo:
978 assert isinstance(offset, int)
980 # Inflate postponed delta
981 obj, type = self.get_object_at(offset)
982 assert isinstance(type, int)
983 assert isinstance(obj, list) or isinstance(obj, tuple)
985 type, obj = self.resolve_object(offset, type, obj, get_ref_text)
987 # Save memory by not storing the inflated obj in postponed
988 postponed[e.sha].append((offset, type, None, crc32))
990 sha = obj_sha(type, obj)
992 yield sha, offset, crc32
993 extra.extend(postponed.pop(sha, []))
995 raise KeyError([sha_to_hex(h) for h in postponed.keys()])
998 class SHA1Reader(object):
999 """Wrapper around a file-like object that remembers the SHA1 of its data."""
1001 def __init__(self, f):
1003 self.sha1 = make_sha("")
1005 def read(self, num=None):
1006 data = self.f.read(num)
1007 self.sha1.update(data)
1010 def check_sha(self):
1011 stored = self.f.read(20)
1012 if stored != self.sha1.digest():
1013 raise ChecksumMismatch(self.sha1.hexdigest(), sha_to_hex(stored))
1016 return self.f.close()
1019 return self.f.tell()
1022 class SHA1Writer(object):
1023 """Wrapper around a file-like object that remembers the SHA1 of its data."""
1025 def __init__(self, f):
1027 self.sha1 = make_sha("")
1029 def write(self, data):
1030 self.sha1.update(data)
1033 def write_sha(self):
1034 sha = self.sha1.digest()
1035 assert len(sha) == 20
1040 sha = self.write_sha()
1045 return self.f.tell()
1048 def write_pack_object(f, type, object):
1049 """Write pack object to a file.
1051 :param f: File to write to
1052 :param type: Numeric type of the object
1053 :param object: Object to write
1054 :return: Tuple with offset at which the object was written, and crc32
1057 packed_data_hdr = ""
1058 if type == OFS_DELTA:
1059 (delta_base_offset, object) = object
1060 elif type == REF_DELTA:
1061 (basename, object) = object
1063 c = (type << 4) | (size & 15)
1066 packed_data_hdr += (chr(c | 0x80))
1069 packed_data_hdr += chr(c)
1070 if type == OFS_DELTA:
1071 ret = [delta_base_offset & 0x7f]
1072 delta_base_offset >>= 7
1073 while delta_base_offset:
1074 delta_base_offset -= 1
1075 ret.insert(0, 0x80 | (delta_base_offset & 0x7f))
1076 delta_base_offset >>= 7
1077 packed_data_hdr += "".join([chr(x) for x in ret])
1078 elif type == REF_DELTA:
1079 assert len(basename) == 20
1080 packed_data_hdr += basename
1081 packed_data = packed_data_hdr + zlib.compress(object)
1082 f.write(packed_data)
1083 return (offset, (zlib.crc32(packed_data) & 0xffffffff))
1086 def write_pack(filename, objects, num_objects):
1087 """Write a new pack data file.
1089 :param filename: Path to the new pack file (without .pack extension)
1090 :param objects: Iterable over (object, path) tuples to write
1091 :param num_objects: Number of objects to write
1092 :return: Tuple with checksum of pack file and index file
1094 f = GitFile(filename + ".pack", 'wb')
1096 entries, data_sum = write_pack_data(f, objects, num_objects)
1100 f = GitFile(filename + ".idx", 'wb')
1102 return data_sum, write_pack_index_v2(f, entries, data_sum)
1107 def write_pack_header(f, num_objects):
1108 """Write a pack header for the given number of objects."""
1109 f.write('PACK') # Pack header
1110 f.write(struct.pack('>L', 2)) # Pack version
1111 f.write(struct.pack('>L', num_objects)) # Number of objects in pack
1114 def write_pack_data(f, objects, num_objects, window=10):
1115 """Write a new pack data file.
1117 :param f: File to write to
1118 :param objects: Iterable over (object, path) tuples to write
1119 :param num_objects: Number of objects to write
1120 :param window: Sliding window size for searching for deltas; currently
1122 :return: List with (name, offset, crc32 checksum) entries, pack checksum
1124 recency = list(objects)
1125 # FIXME: Somehow limit delta depth
1126 # FIXME: Make thin-pack optional (its not used when cloning a pack)
1127 # Build a list of objects ordered by the magic Linus heuristic
1128 # This helps us find good objects to diff against us
1130 for obj, path in recency:
1131 magic.append( (obj.type_num, path, 1, -obj.raw_length(), obj) )
1133 # Build a map of objects and their index in magic - so we can find
1134 # preceeding objects to diff against
1136 for i in range(len(magic)):
1137 offs[magic[i][4]] = i
1141 write_pack_header(f, num_objects)
1142 for o, path in recency:
1143 sha1 = o.sha().digest()
1145 raw = o.as_raw_string()
1148 #for i in range(offs[o]-window, window):
1149 # if i < 0 or i >= len(offs): continue
1151 # if b.type_num != orig_t: continue
1152 # base = b.as_raw_string()
1153 # delta = create_delta(base, raw)
1154 # if len(delta) < len(winner):
1156 # t = 6 if magic[i][2] == 1 else 7
1157 offset, crc32 = write_pack_object(f, t, winner)
1158 entries.append((sha1, offset, crc32))
1159 return entries, f.write_sha()
1162 def write_pack_index_v1(f, entries, pack_checksum):
1163 """Write a new pack index file.
1165 :param f: A file-like object to write to
1166 :param entries: List of tuples with object name (sha), offset_in_pack,
1168 :param pack_checksum: Checksum of the pack file.
1169 :return: The SHA of the written index file
1172 fan_out_table = defaultdict(lambda: 0)
1173 for (name, offset, entry_checksum) in entries:
1174 fan_out_table[ord(name[0])] += 1
1176 for i in range(0x100):
1177 f.write(struct.pack(">L", fan_out_table[i]))
1178 fan_out_table[i+1] += fan_out_table[i]
1179 for (name, offset, entry_checksum) in entries:
1180 f.write(struct.pack(">L20s", offset, name))
1181 assert len(pack_checksum) == 20
1182 f.write(pack_checksum)
1183 return f.write_sha()
1186 def create_delta(base_buf, target_buf):
1187 """Use python difflib to work out how to transform base_buf to target_buf.
1189 :param base_buf: Base buffer
1190 :param target_buf: Target buffer
1192 assert isinstance(base_buf, str)
1193 assert isinstance(target_buf, str)
1195 # write delta header
1196 def encode_size(size):
1201 ret += chr(c | 0x80)
1206 out_buf += encode_size(len(base_buf))
1207 out_buf += encode_size(len(target_buf))
1208 # write out delta opcodes
1209 seq = difflib.SequenceMatcher(a=base_buf, b=target_buf)
1210 for opcode, i1, i2, j1, j2 in seq.get_opcodes():
1211 # Git patch opcodes don't care about deletes!
1212 #if opcode == "replace" or opcode == "delete":
1214 if opcode == "equal":
1215 # If they are equal, unpacker will use data from base_buf
1216 # Write out an opcode that says what range to use
1222 scratch += chr((o >> i*8) & 0xff)
1227 scratch += chr((s >> i*8) & 0xff)
1231 if opcode == "replace" or opcode == "insert":
1232 # If we are replacing a range or adding one, then we just
1233 # output it to the stream (prefixed by its size)
1238 out_buf += target_buf[o:o+127]
1242 out_buf += target_buf[o:o+s]
1246 def apply_delta(src_buf, delta):
1247 """Based on the similar function in git's patch-delta.c.
1249 :param src_buf: Source buffer
1250 :param delta: Delta instructions
1252 if type(src_buf) != str:
1253 src_buf = "".join(src_buf)
1254 if type(delta) != str:
1255 delta = "".join(delta)
1258 delta_length = len(delta)
1259 def get_delta_header_size(delta, index):
1263 cmd = ord(delta[index])
1265 size |= (cmd & ~0x80) << i
1270 src_size, index = get_delta_header_size(delta, index)
1271 dest_size, index = get_delta_header_size(delta, index)
1272 assert src_size == len(src_buf), "%d vs %d" % (src_size, len(src_buf))
1273 while index < delta_length:
1274 cmd = ord(delta[index])
1280 x = ord(delta[index])
1282 cp_off |= x << (i * 8)
1285 if cmd & (1 << (4+i)):
1286 x = ord(delta[index])
1288 cp_size |= x << (i * 8)
1291 if (cp_off + cp_size < cp_size or
1292 cp_off + cp_size > src_size or
1293 cp_size > dest_size):
1295 out.append(src_buf[cp_off:cp_off+cp_size])
1297 out.append(delta[index:index+cmd])
1300 raise ApplyDeltaError("Invalid opcode 0")
1302 if index != delta_length:
1303 raise ApplyDeltaError("delta not empty: %r" % delta[index:])
1305 if dest_size != chunks_length(out):
1306 raise ApplyDeltaError("dest size incorrect")
1311 def write_pack_index_v2(f, entries, pack_checksum):
1312 """Write a new pack index file.
1314 :param f: File-like object to write to
1315 :param entries: List of tuples with object name (sha), offset_in_pack, and
1317 :param pack_checksum: Checksum of the pack file.
1318 :return: The SHA of the index file written
1321 f.write('\377tOc') # Magic!
1322 f.write(struct.pack(">L", 2))
1323 fan_out_table = defaultdict(lambda: 0)
1324 for (name, offset, entry_checksum) in entries:
1325 fan_out_table[ord(name[0])] += 1
1327 for i in range(0x100):
1328 f.write(struct.pack(">L", fan_out_table[i]))
1329 fan_out_table[i+1] += fan_out_table[i]
1330 for (name, offset, entry_checksum) in entries:
1332 for (name, offset, entry_checksum) in entries:
1333 f.write(struct.pack(">L", entry_checksum))
1334 for (name, offset, entry_checksum) in entries:
1335 # FIXME: handle if MSBit is set in offset
1336 f.write(struct.pack(">L", offset))
1337 # FIXME: handle table for pack files > 8 Gb
1338 assert len(pack_checksum) == 20
1339 f.write(pack_checksum)
1340 return f.write_sha()
1344 """A Git pack object."""
1346 def __init__(self, basename):
1347 self._basename = basename
1350 self._idx_path = self._basename + ".idx"
1351 self._data_path = self._basename + ".pack"
1352 self._data_load = lambda: PackData(self._data_path)
1353 self._idx_load = lambda: load_pack_index(self._idx_path)
1356 def from_lazy_objects(self, data_fn, idx_fn):
1357 """Create a new pack object from callables to load pack data and
1360 ret._data_load = data_fn
1361 ret._idx_load = idx_fn
1365 def from_objects(self, data, idx):
1366 """Create a new pack object from pack data and index objects."""
1368 ret._data_load = lambda: data
1369 ret._idx_load = lambda: idx
1373 """The SHA over the SHAs of the objects in this pack."""
1374 return self.index.objects_sha1()
1378 """The pack data object being used."""
1379 if self._data is None:
1380 self._data = self._data_load()
1381 self._data.pack = self
1382 assert len(self.index) == len(self._data)
1383 idx_stored_checksum = self.index.get_pack_checksum()
1384 data_stored_checksum = self._data.get_stored_checksum()
1385 if idx_stored_checksum != data_stored_checksum:
1386 raise ChecksumMismatch(sha_to_hex(idx_stored_checksum),
1387 sha_to_hex(data_stored_checksum))
1392 """The index being used.
1394 :note: This may be an in-memory index
1396 if self._idx is None:
1397 self._idx = self._idx_load()
1401 if self._data is not None:
1405 def __eq__(self, other):
1406 return type(self) == type(other) and self.index == other.index
1409 """Number of entries in this pack."""
1410 return len(self.index)
1413 return "%s(%r)" % (self.__class__.__name__, self._basename)
1416 """Iterate over all the sha1s of the objects in this pack."""
1417 return iter(self.index)
1420 """Check the integrity of this pack.
1422 :raise ChecksumMismatch: if a checksum for the index or data is wrong
1426 for obj in self.iterobjects():
1428 # TODO: object connectivity checks
1430 def get_stored_checksum(self):
1431 return self.data.get_stored_checksum()
1433 def __contains__(self, sha1):
1434 """Check whether this pack contains a particular SHA1."""
1436 self.index.object_index(sha1)
1441 def get_raw(self, sha1):
1442 offset = self.index.object_index(sha1)
1443 obj_type, obj = self.data.get_object_at(offset)
1444 if type(offset) is long:
1445 offset = int(offset)
1446 type_num, chunks = self.data.resolve_object(offset, obj_type, obj)
1447 return type_num, "".join(chunks)
1449 def __getitem__(self, sha1):
1450 """Retrieve the specified SHA1."""
1451 type, uncomp = self.get_raw(sha1)
1452 return ShaFile.from_raw_string(type, uncomp)
1454 def iterobjects(self):
1455 """Iterate over the objects in this pack."""
1456 for offset, type, obj, crc32 in self.data.iterobjects():
1457 assert isinstance(offset, int)
1458 yield ShaFile.from_raw_chunks(
1459 *self.data.resolve_object(offset, type, obj))
1463 from dulwich._pack import apply_delta, bisect_find_sha