1
# Copyright (C) 2009 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19
cdef extern from "python-compat.h":
23
ctypedef unsigned int size_t
24
int memcmp(void *, void*, size_t)
25
void memcpy(void *, void*, size_t)
26
void *memchr(void *s, int c, size_t len)
27
long strtol(char *, char **, int)
28
void sprintf(char *, char *, ...)
30
cdef extern from "Python.h":
31
ctypedef int Py_ssize_t # Required for older pyrex versions
34
ctypedef _PyObject PyObject
35
int PyTuple_CheckExact(object p)
36
Py_ssize_t PyTuple_GET_SIZE(object t)
37
int PyString_CheckExact(object)
38
char *PyString_AS_STRING(object s)
39
Py_ssize_t PyString_GET_SIZE(object)
41
int PyDict_SetItem(object d, object k, object v) except -1
43
object PyTuple_New(Py_ssize_t count)
44
void PyTuple_SET_ITEM(object t, Py_ssize_t offset, object)
46
void Py_INCREF(object)
48
PyObject * PyTuple_GET_ITEM_ptr "PyTuple_GET_ITEM" (object t,
50
int PyString_CheckExact_ptr "PyString_CheckExact" (PyObject *p)
51
Py_ssize_t PyString_GET_SIZE_ptr "PyString_GET_SIZE" (PyObject *s)
52
char *PyString_AS_STRING_ptr "PyString_AS_STRING" (PyObject *s)
53
object PyString_FromStringAndSize(char*, Py_ssize_t)
55
cdef extern from "zlib.h":
56
ctypedef unsigned long uLong
57
ctypedef unsigned int uInt
58
ctypedef unsigned char Bytef
60
uLong crc32(uLong crc, Bytef *buf, uInt len)
62
# It seems we need to import the definitions so that the pyrex compiler has
63
# local names to access them.
64
from _static_tuple_c cimport StaticTuple,\
65
import_static_tuple_c, StaticTuple_New, \
66
StaticTuple_Intern, StaticTuple_SET_ITEM, StaticTuple_CheckExact
69
# This sets up the StaticTuple C_API functionality
70
if import_static_tuple_c() != 0:
71
raise ImportError('der borken')
75
cdef object _InternalNode
80
# We shouldn't just copy this from _dirstate_helpers_pyx
81
cdef void* _my_memrchr(void *s, int c, size_t n):
82
# memrchr seems to be a GNU extension, so we have to implement it ourselves
95
def _search_key_16(key):
96
"""See chk_map._search_key_16."""
97
cdef Py_ssize_t num_bits
99
cdef Py_ssize_t num_out_bytes
103
cdef Py_ssize_t out_off
107
if not PyTuple_CheckExact(key) and not StaticTuple_CheckExact(key):
108
raise TypeError('key %r is not a tuple' % (key,))
110
# 4 bytes per crc32, and another 1 byte between bits
111
num_out_bytes = (9 * num_bits) - 1
112
out = PyString_FromStringAndSize(NULL, num_out_bytes)
113
c_out = PyString_AS_STRING(out)
114
for i from 0 <= i < num_bits:
118
# We use the _ptr variant, because GET_ITEM returns a borrowed
119
# reference, and Pyrex assumes that returned 'object' are a new
121
# XXX: This needs to be updated for PySequence_GetItem since both
122
# PyTuple and StaticTuple support that api
123
bit = key[i]# PyTuple_GET_ITEM_ptr(key, i)
124
if not PyString_CheckExact(bit):
125
raise TypeError('Bit %d of %r is not a string' % (i, key))
126
c_bit = <Bytef *>PyString_AS_STRING(bit)
127
c_len = PyString_GET_SIZE(bit)
128
crc_val = crc32(0, c_bit, c_len)
130
sprintf(c_out, '%08X', crc_val)
135
def _search_key_255(key):
136
"""See chk_map._search_key_255."""
137
cdef Py_ssize_t num_bits
139
cdef Py_ssize_t num_out_bytes
143
cdef Py_ssize_t out_off
147
if not PyTuple_CheckExact(key) and not StaticTuple_CheckExact(key):
148
raise TypeError('key %r is not a tuple' % (key,))
150
# 4 bytes per crc32, and another 1 byte between bits
151
num_out_bytes = (5 * num_bits) - 1
152
out = PyString_FromStringAndSize(NULL, num_out_bytes)
153
c_out = PyString_AS_STRING(out)
154
for i from 0 <= i < num_bits:
158
bit = key[i] # PyTuple_GET_ITEM_ptr(key, i)
159
if not PyString_CheckExact(bit):
160
raise TypeError('Bit %d of %r is not a string: %r' % (i, key,
162
c_bit = <Bytef *>PyString_AS_STRING(bit)
163
c_len = PyString_GET_SIZE(bit)
164
crc_val = crc32(0, c_bit, c_len)
166
c_out[0] = (crc_val >> 24) & 0xFF
167
c_out[1] = (crc_val >> 16) & 0xFF
168
c_out[2] = (crc_val >> 8) & 0xFF
169
c_out[3] = (crc_val >> 0) & 0xFF
170
for j from 0 <= j < 4:
171
if c_out[j] == c'\n':
177
cdef int _get_int_from_line(char **cur, char *end, char *message) except -1:
178
"""Read a positive integer from the data stream.
180
:param cur: The start of the data, this will be moved to after the
181
trailing newline when done.
182
:param end: Do not parse any data past this byte.
183
:return: The integer stored in those bytes
186
cdef char *next_line, *next
188
next_line = <char *>memchr(cur[0], c'\n', end - cur[0])
189
if next_line == NULL:
190
raise ValueError("Missing %s line\n" % message)
192
value = strtol(cur[0], &next, 10)
193
if next != next_line:
194
raise ValueError("%s line not a proper int\n" % message)
195
cur[0] = next_line + 1
199
def _deserialise_leaf_node(bytes, key, search_key_func=None):
200
"""Deserialise bytes, with key key, into a LeafNode.
202
:param bytes: The bytes of the node.
203
:param key: The key that the serialised node has.
205
cdef char *c_bytes, *cur, *next, *end
207
cdef Py_ssize_t c_bytes_len, prefix_length, items_length
208
cdef int maximum_size, width, length, i, prefix_tail_len
209
cdef int num_value_lines, num_prefix_bits
210
cdef char *prefix, *value_start, *prefix_tail
211
cdef char *next_null, *last_null, *line_start
212
cdef char *c_entry, *entry_start
213
cdef StaticTuple entry_bits
215
if _LeafNode is None:
216
from bzrlib import chk_map
217
_LeafNode = chk_map.LeafNode
218
_InternalNode = chk_map.InternalNode
219
_unknown = chk_map._unknown
221
result = _LeafNode(search_key_func=search_key_func)
222
# Splitlines can split on '\r' so don't use it, split('\n') adds an
223
# extra '' if the bytes ends in a final newline.
224
if not PyString_CheckExact(bytes):
225
raise TypeError('bytes must be a plain string not %s' % (type(bytes),))
227
c_bytes = PyString_AS_STRING(bytes)
228
c_bytes_len = PyString_GET_SIZE(bytes)
230
if c_bytes_len < 9 or memcmp(c_bytes, "chkleaf:\n", 9) != 0:
231
raise ValueError("not a serialised leaf node: %r" % bytes)
232
if c_bytes[c_bytes_len - 1] != c'\n':
233
raise ValueError("bytes does not end in a newline")
235
end = c_bytes + c_bytes_len
237
maximum_size = _get_int_from_line(&cur, end, "maximum_size")
238
width = _get_int_from_line(&cur, end, "width")
239
length = _get_int_from_line(&cur, end, "length")
241
next_line = <char *>memchr(cur, c'\n', end - cur)
242
if next_line == NULL:
243
raise ValueError('Missing the prefix line\n')
245
prefix_length = next_line - cur
251
next_null = <char *>memchr(prefix, c'\0', prefix_length)
252
while next_null != NULL:
253
num_prefix_bits = num_prefix_bits + 1
255
PyString_FromStringAndSize(prefix_tail, next_null - prefix_tail))
256
prefix_tail = next_null + 1
257
next_null = <char *>memchr(prefix_tail, c'\0', next_line - prefix_tail)
258
prefix_tail_len = next_line - prefix_tail
260
if num_prefix_bits >= width:
261
raise ValueError('Prefix has too many nulls versus width')
263
items_length = end - cur
267
next_line = <char *>memchr(cur, c'\n', end - cur)
268
if next_line == NULL:
269
raise ValueError('null line\n')
270
last_null = <char *>_my_memrchr(cur, c'\0', next_line - cur)
271
if last_null == NULL:
272
raise ValueError('fail to find the num value lines null')
273
next_null = last_null + 1 # move past NULL
274
num_value_lines = _get_int_from_line(&next_null, next_line + 1,
278
# Walk num_value_lines forward
279
for i from 0 <= i < num_value_lines:
280
next_line = <char *>memchr(cur, c'\n', end - cur)
281
if next_line == NULL:
282
raise ValueError('missing trailing newline')
284
entry_bits = StaticTuple_New(width)
285
for i from 0 <= i < num_prefix_bits:
286
entry = prefix_bits[i]
287
# SET_ITEM 'steals' a reference
289
StaticTuple_SET_ITEM(entry_bits, i, entry)
290
value = PyString_FromStringAndSize(value_start, next_line - value_start)
291
# The next entry bit needs the 'tail' from the prefix, and first part
293
entry_start = line_start
294
next_null = <char *>memchr(entry_start, c'\0',
295
last_null - entry_start + 1)
296
if next_null == NULL:
297
raise ValueError('bad no null, bad')
298
entry = PyString_FromStringAndSize(NULL,
299
prefix_tail_len + next_null - line_start)
300
c_entry = PyString_AS_STRING(entry)
301
if prefix_tail_len > 0:
302
memcpy(c_entry, prefix_tail, prefix_tail_len)
303
if next_null - line_start > 0:
304
memcpy(c_entry + prefix_tail_len, line_start, next_null - line_start)
307
StaticTuple_SET_ITEM(entry_bits, i, entry)
308
while next_null != last_null: # We have remaining bits
311
raise ValueError("Too many bits for entry")
312
entry_start = next_null + 1
313
next_null = <char *>memchr(entry_start, c'\0',
314
last_null - entry_start + 1)
315
if next_null == NULL:
316
raise ValueError('bad no null')
317
entry = PyString_FromStringAndSize(entry_start,
318
next_null - entry_start)
320
StaticTuple_SET_ITEM(entry_bits, i, entry)
321
if len(entry_bits) != width:
322
raise AssertionError(
323
'Incorrect number of elements (%d vs %d)'
324
% (len(entry_bits)+1, width + 1))
325
entry_bits = StaticTuple_Intern(entry_bits)
326
PyDict_SetItem(items, entry_bits, value)
327
if len(items) != length:
328
raise ValueError("item count (%d) mismatch for key %s,"
329
" bytes %r" % (length, entry_bits, bytes))
330
result._items = items
332
result._maximum_size = maximum_size
334
result._key_width = width
335
result._raw_size = items_length + length * prefix_length
337
result._search_prefix = None
338
result._common_serialised_prefix = None
340
result._search_prefix = _unknown
341
result._common_serialised_prefix = PyString_FromStringAndSize(prefix,
343
if c_bytes_len != result._current_size():
344
raise AssertionError('_current_size computed incorrectly %d != %d',
345
c_bytes_len, result._current_size())
349
def _deserialise_internal_node(bytes, key, search_key_func=None):
350
cdef char *c_bytes, *cur, *next, *end
352
cdef Py_ssize_t c_bytes_len, prefix_length
353
cdef int maximum_size, width, length, i, prefix_tail_len
354
cdef char *prefix, *line_prefix, *next_null, *c_item_prefix
356
if _InternalNode is None:
357
from bzrlib import chk_map
358
_LeafNode = chk_map.LeafNode
359
_InternalNode = chk_map.InternalNode
360
_unknown = chk_map._unknown
361
result = _InternalNode(search_key_func=search_key_func)
363
if not PyString_CheckExact(bytes):
364
raise TypeError('bytes must be a plain string not %s' % (type(bytes),))
366
c_bytes = PyString_AS_STRING(bytes)
367
c_bytes_len = PyString_GET_SIZE(bytes)
369
if c_bytes_len < 9 or memcmp(c_bytes, "chknode:\n", 9) != 0:
370
raise ValueError("not a serialised internal node: %r" % bytes)
371
if c_bytes[c_bytes_len - 1] != c'\n':
372
raise ValueError("bytes does not end in a newline")
376
end = c_bytes + c_bytes_len
377
maximum_size = _get_int_from_line(&cur, end, "maximum_size")
378
width = _get_int_from_line(&cur, end, "width")
379
length = _get_int_from_line(&cur, end, "length")
381
next_line = <char *>memchr(cur, c'\n', end - cur)
382
if next_line == NULL:
383
raise ValueError('Missing the prefix line\n')
385
prefix_length = next_line - cur
389
# Find the null separator
390
next_line = <char *>memchr(cur, c'\n', end - cur)
391
if next_line == NULL:
392
raise ValueError('missing trailing newline')
393
next_null = <char *>_my_memrchr(cur, c'\0', next_line - cur)
394
if next_null == NULL:
395
raise ValueError('bad no null')
396
item_prefix = PyString_FromStringAndSize(NULL,
397
prefix_length + next_null - cur)
398
c_item_prefix = PyString_AS_STRING(item_prefix)
400
memcpy(c_item_prefix, prefix, prefix_length)
401
memcpy(c_item_prefix + prefix_length, cur, next_null - cur)
402
flat_key = PyString_FromStringAndSize(next_null + 1,
403
next_line - next_null - 1)
404
flat_key = StaticTuple(flat_key).intern()
405
PyDict_SetItem(items, item_prefix, flat_key)
407
assert len(items) > 0
408
result._items = items
410
result._maximum_size = maximum_size
412
result._key_width = width
413
# XXX: InternalNodes don't really care about their size, and this will
414
# change if we add prefix compression
415
result._raw_size = None # len(bytes)
416
result._node_width = len(item_prefix)
417
result._search_prefix = PyString_FromStringAndSize(prefix, prefix_length)
421
_key_type = StaticTuple
added
removed
bzrlib/_chk_map_pyx.pyx
