1
# pack.py -- For dealing wih packed git objects.
2
# Copyright (C) 2007 James Westby <jw+debian@jameswestby.net>
3
# Copryight (C) 2008 Jelmer Vernooij <jelmer@samba.org>
4
# The code is loosely based on that in the sha1_file.c file from git itself,
5
# which is Copyright (C) Linus Torvalds, 2005 and distributed under the
8
# This program is free software; you can redistribute it and/or
9
# modify it under the terms of the GNU General Public License
10
# as published by the Free Software Foundation; version 2
13
# This program is distributed in the hope that it will be useful,
14
# but WITHOUT ANY WARRANTY; without even the implied warranty of
15
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16
# GNU General Public License for more details.
18
# You should have received a copy of the GNU General Public License
19
# along with this program; if not, write to the Free Software
20
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
23
"""Classes for dealing with packed git objects.
25
A pack is a compact representation of a bunch of objects, stored
26
using deltas where possible.
28
They have two parts, the pack file, which stores the data, and an index
29
that tells you where the data is.
31
To find an object you look in all of the index files 'til you find a
32
match for the object name. You then use the pointer got from this as
33
a pointer in to the corresponding packfile.
36
from collections import defaultdict
43
supports_mmap_offset = (sys.version_info[0] >= 3 or
44
(sys.version_info[0] == 2 and sys.version_info[1] >= 6))
46
from objects import (ShaFile,
52
for i in range(0, len(hex), 2):
53
ret += chr(int(hex[i:i+2], 16))
59
ret += "%02x" % ord(i)
62
MAX_MMAP_SIZE = 256 * 1024 * 1024
64
def simple_mmap(f, offset, size, access=mmap.ACCESS_READ):
65
if offset+size > MAX_MMAP_SIZE and not supports_mmap_offset:
66
raise AssertionError("%s is larger than 256 meg, and this version "
67
"of Python does not support the offset argument to mmap().")
68
if supports_mmap_offset:
69
return mmap.mmap(f.fileno(), size, access=access, offset=offset)
71
class ArraySkipper(object):
73
def __init__(self, array, offset):
77
def __getslice__(self, i, j):
78
return self.array[i+self.offset:j+self.offset]
80
def __getitem__(self, i):
81
return self.array[i+self.offset]
84
return len(self.array) - self.offset
87
return str(self.array[self.offset:])
89
mem = mmap.mmap(f.fileno(), size+offset, access=access)
92
return ArraySkipper(mem, offset)
95
def multi_ord(map, start, count):
97
for i in range(count):
98
value = value * 0x100 + ord(map[start+i])
102
class PackIndex(object):
103
"""An index in to a packfile.
105
Given a sha id of an object a pack index can tell you the location in the
106
packfile of that object if it has it.
108
To do the loop it opens the file, and indexes first 256 4 byte groups
109
with the first byte of the sha id. The value in the four byte group indexed
110
is the end of the group that shares the same starting byte. Subtract one
111
from the starting byte and index again to find the start of the group.
112
The values are sorted by sha id within the group, so do the math to find
113
the start and end offset and then bisect in to find if the value is present.
116
PACK_INDEX_HEADER_SIZE = 0x100 * 4
118
record_size = sha_bytes + 4
120
def __init__(self, filename):
121
"""Create a pack index object.
123
Provide it with the name of the index file to consider, and it will map
124
it whenever required.
126
self._filename = filename
127
assert os.path.exists(filename), "%s is not a pack index" % filename
128
# Take the size now, so it can be checked each time we map the file to
129
# ensure that it hasn't changed.
130
self._size = os.path.getsize(filename)
131
self._file = open(filename, 'r')
132
self._contents = simple_mmap(self._file, 0, self._size)
133
if self._contents[:4] != '\377tOc':
135
self._fan_out_table = self._read_fan_out_table(0)
137
(self.version, ) = struct.unpack_from(">L", self._contents, 4)
138
assert self.version in (2,), "Version was %d" % self.version
139
self._fan_out_table = self._read_fan_out_table(8)
140
self._name_table_offset = 8 + 0x100 * 4
141
self._crc32_table_offset = self._name_table_offset + 20 * len(self)
142
self._pack_offset_table_offset = self._crc32_table_offset + 4 * len(self)
148
"""Return the number of entries in this pack index."""
149
return self._fan_out_table[-1]
151
def _unpack_entry(self, i):
152
"""Unpack the i-th entry in the index file.
154
:return: Tuple with object name (SHA), offset in pack file and
155
CRC32 checksum (if known)."""
156
if self.version == 1:
157
(offset, name) = struct.unpack_from(">L20s", self._contents,
158
self.PACK_INDEX_HEADER_SIZE + (i * self.record_size))
159
return (name, offset, None)
161
return (self._unpack_name(i), self._unpack_offset(i),
162
self._unpack_crc32_checksum(i))
164
def _unpack_name(self, i):
165
if self.version == 1:
166
return self._unpack_entry(i)[0]
168
return struct.unpack_from("20s", self._contents,
169
self._name_table_offset + i * 20)[0]
171
def _unpack_offset(self, i):
172
if self.version == 1:
173
return self._unpack_entry(i)[1]
175
return struct.unpack_from(">L", self._contents,
176
self._pack_offset_table_offset + i * 4)[0]
178
def _unpack_crc32_checksum(self, i):
179
if self.version == 1:
182
return struct.unpack_from(">L", self._contents,
183
self._crc32_table_offset + i * 4)[0]
185
def iterentries(self):
186
"""Iterate over the entries in this pack index.
188
Will yield tuples with object name, offset in packfile and crc32 checksum.
190
for i in range(len(self)):
191
yield self._unpack_entry(i)
193
def _read_fan_out_table(self, start_offset):
195
for i in range(0x100):
196
ret.append(struct.unpack(">L", self._contents[start_offset+i*4:start_offset+(i+1)*4])[0])
200
"""Check that the stored checksum matches the actual checksum."""
201
return self.calculate_checksum() == self.get_stored_checksums()[1]
203
def calculate_checksum(self):
204
f = open(self._filename, 'r')
206
return hashlib.sha1(self._contents[:-20]).digest()
210
def get_stored_checksums(self):
211
"""Return the SHA1 checksums stored for the corresponding packfile and
212
this header file itself."""
213
return str(self._contents[-40:-20]), str(self._contents[-20:])
215
def object_index(self, sha):
216
"""Return the index in to the corresponding packfile for the object.
218
Given the name of an object it will return the offset that object lives
219
at within the corresponding pack file. If the pack file doesn't have the
220
object then None will be returned.
222
size = os.path.getsize(self._filename)
223
assert size == self._size, "Pack index %s has changed size, I don't " \
224
"like that" % self._filename
225
return self._object_index(sha)
227
def _object_index(self, hexsha):
228
"""See object_index"""
229
sha = hex_to_sha(hexsha)
230
start = self._fan_out_table[ord(sha[0])-1]
231
end = self._fan_out_table[ord(sha[0])]
234
file_sha = self._unpack_name(i)
236
return self._unpack_offset(i)
244
class PackData(object):
245
"""The data contained in a packfile.
247
Pack files can be accessed both sequentially for exploding a pack, and
248
directly with the help of an index to retrieve a specific object.
250
The objects within are either complete or a delta aginst another.
252
The header is variable length. If the MSB of each byte is set then it
253
indicates that the subsequent byte is still part of the header.
254
For the first byte the next MS bits are the type, which tells you the type
255
of object, and whether it is a delta. The LS byte is the lowest bits of the
256
size. For each subsequent byte the LS 7 bits are the next MS bits of the
257
size, i.e. the last byte of the header contains the MS bits of the size.
259
For the complete objects the data is stored as zlib deflated data.
260
The size in the header is the uncompressed object size, so to uncompress
261
you need to just keep feeding data to zlib until you get an object back,
262
or it errors on bad data. This is done here by just giving the complete
263
buffer from the start of the deflated object on. This is bad, but until I
264
get mmap sorted out it will have to do.
266
Currently there are no integrity checks done. Also no attempt is made to try
267
and detect the delta case, or a request for an object at the wrong position.
268
It will all just throw a zlib or KeyError.
271
def __init__(self, filename):
272
"""Create a PackData object that represents the pack in the given filename.
274
The file must exist and stay readable until the object is disposed of. It
275
must also stay the same size. It will be mapped whenever needed.
277
Currently there is a restriction on the size of the pack as the python
278
mmap implementation is flawed.
280
self._filename = filename
281
assert os.path.exists(filename), "%s is not a packfile" % filename
282
self._size = os.path.getsize(filename)
285
def _read_header(self):
286
f = open(self._filename, 'rb')
289
f.seek(self._size-20)
290
self._stored_checksum = f.read(20)
293
assert header[:4] == "PACK"
294
(version,) = struct.unpack_from(">L", header, 4)
295
assert version in (2, 3), "Version was %d" % version
296
(self._num_objects,) = struct.unpack_from(">L", header, 8)
299
"""Returns the number of objects in this pack."""
300
return self._num_objects
302
def calculate_checksum(self):
303
f = open(self._filename, 'rb')
305
map = simple_mmap(f, 0, self._size)
306
return hashlib.sha1(map[:-20]).digest()
311
return (self.calculate_checksum() == self._stored_checksum)
313
def get_object_at(self, offset):
314
"""Given an offset in to the packfile return the object that is there.
316
Using the associated index the location of an object can be looked up, and
317
then the packfile can be asked directly for that object using this
320
Currently only non-delta objects are supported.
322
assert isinstance(offset, long) or isinstance(offset, int)
323
size = os.path.getsize(self._filename)
324
assert size == self._size, "Pack data %s has changed size, I don't " \
325
"like that" % self._filename
326
f = open(self._filename, 'rb')
328
map = simple_mmap(f, offset, size-offset)
329
return self._get_object_at(map)
333
def _get_object_at(self, map):
334
first_byte = ord(map[0])
335
sign_extend = first_byte & 0x80
336
type = (first_byte >> 4) & 0x07
337
size = first_byte & 0x0f
339
while sign_extend > 0:
340
byte = ord(map[cur_offset+1])
341
sign_extend = byte & 0x80
342
size_part = byte & 0x7f
343
size += size_part << ((cur_offset * 7) + 4)
345
raw_base = cur_offset+1
346
# The size is the inflated size, so we have no idea what the deflated size
347
# is, so for now give it as much as we have. It should really iterate
348
# feeding it more data if it doesn't decompress, but as we have the whole
349
# thing then just use it.
351
uncomp = _decompress(raw)
352
obj = ShaFile.from_raw_string(type, uncomp)
356
class SHA1Writer(object):
358
def __init__(self, f):
360
self.sha1 = hashlib.sha1("")
362
def write(self, data):
363
self.sha1.update(data)
367
sha = self.sha1.digest()
368
assert len(sha) == 20
374
def write_pack(filename, objects):
375
"""Write a new pack file.
377
:param filename: The filename of the new pack file.
378
:param objects: List of objects to write.
379
:return: List with (name, offset, crc32 checksum) entries, pack checksum
381
f = open(filename, 'w')
384
f.write("PACK") # Pack header
385
f.write(struct.pack(">L", 2)) # Pack version
386
f.write(struct.pack(">L", len(objects))) # Number of objects in pack
388
pass # FIXME: Write object
389
return entries, f.close()
392
def write_pack_index_v1(filename, entries, pack_checksum):
393
"""Write a new pack index file.
395
:param filename: The filename of the new pack index file.
396
:param entries: List of tuples with object name (sha), offset_in_pack, and
398
:param pack_checksum: Checksum of the pack file.
402
entries = sorted(entries)
403
f = open(filename, 'w')
405
fan_out_table = defaultdict(lambda: 0)
406
for (name, offset, entry_checksum) in entries:
407
fan_out_table[ord(name[0])] += 1
409
for i in range(0x100):
410
f.write(struct.pack(">L", fan_out_table[i]))
411
fan_out_table[i+1] += fan_out_table[i]
412
for (name, offset, entry_checksum) in entries:
413
f.write(struct.pack(">L20s", offset, name))
414
assert len(pack_checksum) == 20
415
f.write(pack_checksum)
419
def write_pack_index_v2(filename, entries, pack_checksum):
420
"""Write a new pack index file.
422
:param filename: The filename of the new pack index file.
423
:param entries: List of tuples with object name (sha), offset_in_pack, and
425
:param pack_checksum: Checksum of the pack file.
428
entries = sorted(entries)
429
f = open(filename, 'w')
432
f.write(struct.pack(">L", 2))
433
fan_out_table = defaultdict(lambda: 0)
434
for (name, offset, entry_checksum) in entries:
435
fan_out_table[ord(name[0])] += 1
437
for i in range(0x100):
438
f.write(struct.pack(">L", fan_out_table[i]))
439
fan_out_table[i+1] += fan_out_table[i]
440
for (name, offset, entry_checksum) in entries:
442
for (name, offset, entry_checksum) in entries:
443
f.write(struct.pack(">L", entry_checksum))
444
for (name, offset, entry_checksum) in entries:
445
# FIXME: handle if MSBit is set in offset
446
f.write(struct.pack(">L", offset))
447
# FIXME: handle table for pack files > 8 Gb
448
assert len(pack_checksum) == 20
449
f.write(pack_checksum)
added
removed
dulwich/pack.py
