1
# Copyright (C) 2008, 2009, 2010 Canonical Ltd
1
# Copyright (C) 2008 Canonical Ltd
3
3
# This program is free software; you can redistribute it and/or modify
4
4
# it under the terms of the GNU General Public License as published by
13
13
# You should have received a copy of the GNU General Public License
14
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
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
18
"""Pyrex extensions to btree node parsing."""
33
33
char *PyString_AsString(object p) except NULL
34
34
object PyString_FromStringAndSize(char *, Py_ssize_t)
35
PyObject *PyString_FromStringAndSize_ptr "PyString_FromStringAndSize" (char *, Py_ssize_t)
36
35
int PyString_CheckExact(object s)
37
36
int PyString_CheckExact_ptr "PyString_CheckExact" (PyObject *)
38
37
Py_ssize_t PyString_Size(object p)
39
38
Py_ssize_t PyString_GET_SIZE_ptr "PyString_GET_SIZE" (PyObject *)
40
39
char * PyString_AS_STRING_ptr "PyString_AS_STRING" (PyObject *)
41
char * PyString_AS_STRING(object)
42
Py_ssize_t PyString_GET_SIZE(object)
43
40
int PyString_AsStringAndSize_ptr(PyObject *, char **buf, Py_ssize_t *len)
44
void PyString_InternInPlace(PyObject **)
45
41
int PyTuple_CheckExact(object t)
46
object PyTuple_New(Py_ssize_t n_entries)
47
void PyTuple_SET_ITEM(object, Py_ssize_t offset, object) # steals the ref
48
42
Py_ssize_t PyTuple_GET_SIZE(object t)
49
43
PyObject *PyTuple_GET_ITEM_ptr_object "PyTuple_GET_ITEM" (object tpl, int index)
50
void Py_INCREF(object)
51
void Py_DECREF_ptr "Py_DECREF" (PyObject *)
53
45
cdef extern from "string.h":
54
46
void *memcpy(void *dest, void *src, size_t n)
57
49
# void *memrchr(void *s, int c, size_t n)
58
50
int strncmp(char *s1, char *s2, size_t n)
60
# It seems we need to import the definitions so that the pyrex compiler has
61
# local names to access them.
62
from _static_tuple_c cimport StaticTuple, \
63
import_static_tuple_c, StaticTuple_New, \
64
StaticTuple_Intern, StaticTuple_SET_ITEM, StaticTuple_CheckExact
67
53
# TODO: Find some way to import this from _dirstate_helpers
68
cdef void* _my_memrchr(void *s, int c, size_t n): # cannot_raise
54
cdef void* _my_memrchr(void *s, int c, size_t n):
69
55
# memrchr seems to be a GNU extension, so we have to implement it ourselves
70
56
# It is not present in any win32 standard library
89
74
return PyString_FromStringAndSize(s, size)
92
cdef object safe_interned_string_from_size(char *s, Py_ssize_t size):
96
'tried to create a string with an invalid size: %d @0x%x'
98
py_str = PyString_FromStringAndSize_ptr(s, size)
99
PyString_InternInPlace(&py_str)
100
result = <object>py_str
101
# Casting a PyObject* to an <object> triggers an INCREF from Pyrex, so we
102
# DECREF it to avoid geting immortal strings
103
Py_DECREF_ptr(py_str)
106
from bzrlib import _static_tuple_c
107
# This sets up the StaticTuple C_API functionality
108
import_static_tuple_c()
111
77
cdef class BTreeLeafParser:
112
78
"""Parse the leaf nodes of a BTree index.
157
122
cdef char *temp_ptr
158
123
cdef int loop_counter
161
key = StaticTuple_New(self.key_length)
162
for loop_counter from 0 <= loop_counter < self.key_length:
127
while loop_counter < self.key_length:
128
loop_counter = loop_counter + 1
163
129
# grab a key segment
164
130
temp_ptr = <char*>memchr(self._start, c'\0', last - self._start)
165
131
if temp_ptr == NULL:
166
if loop_counter + 1 == self.key_length:
132
if loop_counter == self.key_length:
167
133
# capture to last
173
139
last - self._start)))
174
140
raise AssertionError(failure_string)
175
141
# capture the key string
176
if (self.key_length == 1
177
and (temp_ptr - self._start) == 45
178
and strncmp(self._start, 'sha1:', 5) == 0):
179
key_element = safe_string_from_size(self._start,
180
temp_ptr - self._start)
182
key_element = safe_interned_string_from_size(self._start,
183
temp_ptr - self._start)
142
# TODO: Consider using PyIntern_FromString, the only caveat is that
143
# it assumes a NULL-terminated string, so we have to check if
144
# temp_ptr[0] == c'\0' or some other char.
145
key_element = safe_string_from_size(self._start,
146
temp_ptr - self._start)
184
147
# advance our pointer
185
148
self._start = temp_ptr + 1
186
Py_INCREF(key_element)
187
StaticTuple_SET_ITEM(key, loop_counter, key_element)
188
key = StaticTuple_Intern(key)
149
PyList_Append(key_segments, key_element)
150
return tuple(key_segments)
191
152
cdef int process_line(self) except -1:
192
153
"""Process a line in the bytes."""
208
168
# And the next string is right after it
209
169
self._cur_str = last + 1
210
170
# The last character is right before the '\n'
212
173
if last == self._start:
215
176
if last < self._start:
216
177
# Unexpected error condition - fail
217
raise AssertionError("last < self._start")
218
179
if 0 == self._header_found:
219
180
# The first line in a leaf node is the header "type=leaf\n"
220
181
if strncmp("type=leaf", self._start, last - self._start) == 0:
224
185
raise AssertionError('Node did not start with "type=leaf": %r'
225
186
% (safe_string_from_size(self._start, last - self._start)))
227
189
key = self.extract_key(last)
228
190
# find the value area
229
191
temp_ptr = <char*>_my_memrchr(self._start, c'\0', last - self._start)
230
192
if temp_ptr == NULL:
232
raise AssertionError("Failed to find the value area")
234
# Because of how conversions were done, we ended up with *lots* of
235
# values that are identical. These are all of the 0-length nodes
236
# that are referred to by the TREE_ROOT (and likely some other
237
# directory nodes.) For example, bzr has 25k references to
238
# something like '12607215 328306 0 0', which ends up consuming 1MB
239
# of memory, just for those strings.
240
str_len = last - temp_ptr - 1
242
and strncmp(" 0 0", last - 4, 4) == 0):
243
# This drops peak mem for bzr.dev from 87.4MB => 86.2MB
244
# For Launchpad 236MB => 232MB
245
value = safe_interned_string_from_size(temp_ptr + 1, str_len)
247
value = safe_string_from_size(temp_ptr + 1, str_len)
196
# capture the value string
197
value = safe_string_from_size(temp_ptr + 1, last - temp_ptr - 1)
248
198
# shrink the references end point
251
200
if self.ref_list_length:
252
ref_lists = StaticTuple_New(self.ref_list_length)
254
203
while loop_counter < self.ref_list_length:
256
205
# extract a reference list
257
206
loop_counter = loop_counter + 1
258
207
if last < self._start:
259
raise AssertionError("last < self._start")
260
209
# find the next reference list end point:
261
210
temp_ptr = <char*>memchr(self._start, c'\t', last - self._start)
262
211
if temp_ptr == NULL:
263
212
# Only valid for the last list
264
213
if loop_counter != self.ref_list_length:
266
raise AssertionError(
267
"invalid key, loop_counter != self.ref_list_length")
216
raise AssertionError("invalid key")
269
218
# scan to the end of the ref list area
281
230
if temp_ptr == NULL:
282
231
# key runs to the end
283
232
temp_ptr = ref_ptr
285
233
PyList_Append(ref_list, self.extract_key(temp_ptr))
286
ref_list = StaticTuple_Intern(StaticTuple(*ref_list))
288
StaticTuple_SET_ITEM(ref_lists, loop_counter - 1, ref_list)
234
PyList_Append(ref_lists, tuple(ref_list))
289
235
# prepare for the next reference list
290
236
self._start = next_start
291
node_value = StaticTuple(value, ref_lists)
237
ref_lists = tuple(ref_lists)
238
node_value = (value, ref_lists)
293
240
if last != self._start:
294
241
# unexpected reference data present
295
raise AssertionError("unexpected reference data present")
296
node_value = StaticTuple(value, StaticTuple())
297
PyList_Append(self.keys, StaticTuple(key, node_value))
243
node_value = (value, ())
244
PyList_Append(self.keys, (key, node_value))
337
285
cdef int first_ref_list
338
286
cdef int first_reference
288
cdef PyObject *ref_bit
340
289
cdef Py_ssize_t ref_bit_len
342
if not PyTuple_CheckExact(node) and not StaticTuple_CheckExact(node):
343
raise TypeError('We expected a tuple() or StaticTuple() for node not: %s'
291
if not PyTuple_CheckExact(node):
292
raise TypeError('We expected a tuple() for node not: %s'
294
node_len = PyTuple_GET_SIZE(node)
346
295
have_reference_lists = reference_lists
347
296
if have_reference_lists:
348
297
if node_len != 4:
351
300
elif node_len < 3:
352
301
raise ValueError('Without ref_lists, we need at least 3 entries not: %s'
354
# TODO: We can probably do better than string.join(), namely
355
# when key has only 1 item, we can just grab that string
356
# And when there are 2 items, we could do a single malloc + len() + 1
357
# also, doing .join() requires a PyObject_GetAttrString call, which
358
# we could also avoid.
359
# TODO: Note that pyrex 0.9.6 generates fairly crummy code here, using the
360
# python object interface, versus 0.9.8+ which uses a helper that
361
# checks if this supports the sequence interface.
362
# We *could* do more work on our own, and grab the actual items
363
# lists. For now, just ask people to use a better compiler. :)
364
string_key = '\0'.join(node[1])
303
# I don't expect that we can do faster than string.join()
304
string_key = '\0'.join(<object>PyTuple_GET_ITEM_ptr_object(node, 1))
366
306
# TODO: instead of using string joins, precompute the final string length,
367
307
# and then malloc a single string and copy everything in.
379
319
if have_reference_lists:
380
320
# Figure out how many bytes it will take to store the references
321
ref_lists = <object>PyTuple_GET_ITEM_ptr_object(node, 3)
382
322
next_len = len(ref_lists) # TODO: use a Py function
384
324
# If there are no nodes, we don't need to do any work
393
333
refs_len = refs_len + (next_len - 1)
394
334
for reference in ref_list:
395
if (not PyTuple_CheckExact(reference)
396
and not StaticTuple_CheckExact(reference)):
335
if not PyTuple_CheckExact(reference):
398
337
'We expect references to be tuples not: %s'
399
338
% type(reference))
400
next_len = len(reference)
339
next_len = PyTuple_GET_SIZE(reference)
402
341
# We will need (len - 1) '\x00' characters to
403
342
# separate the reference key
404
343
refs_len = refs_len + (next_len - 1)
405
for ref_bit in reference:
406
if not PyString_CheckExact(ref_bit):
344
for i from 0 <= i < next_len:
345
ref_bit = PyTuple_GET_ITEM_ptr_object(reference, i)
346
if not PyString_CheckExact_ptr(ref_bit):
407
347
raise TypeError('We expect reference bits'
408
348
' to be strings not: %s'
409
349
% type(<object>ref_bit))
410
refs_len = refs_len + PyString_GET_SIZE(ref_bit)
350
refs_len = refs_len + PyString_GET_SIZE_ptr(ref_bit)
412
352
# So we have the (key NULL refs NULL value LF)
413
353
key_len = PyString_Size(string_key)
415
if not PyString_CheckExact(val):
354
val = PyTuple_GET_ITEM_ptr_object(node, 2)
355
if not PyString_CheckExact_ptr(val):
416
356
raise TypeError('Expected a plain str for value not: %s'
418
value = PyString_AS_STRING(val)
419
value_len = PyString_GET_SIZE(val)
358
value = PyString_AS_STRING_ptr(val)
359
value_len = PyString_GET_SIZE_ptr(val)
420
360
flat_len = (key_len + 1 + refs_len + 1 + value_len + 1)
421
361
line = PyString_FromStringAndSize(NULL, flat_len)
422
362
# Get a pointer to the new buffer
440
380
first_reference = 0
441
next_len = len(reference)
381
next_len = PyTuple_GET_SIZE(reference)
442
382
for i from 0 <= i < next_len:
446
ref_bit = reference[i]
447
ref_bit_len = PyString_GET_SIZE(ref_bit)
448
memcpy(out, PyString_AS_STRING(ref_bit), ref_bit_len)
386
ref_bit = PyTuple_GET_ITEM_ptr_object(reference, i)
387
ref_bit_len = PyString_GET_SIZE_ptr(ref_bit)
388
memcpy(out, PyString_AS_STRING_ptr(ref_bit), ref_bit_len)
449
389
out = out + ref_bit_len
added
removed
bzrlib/_btree_serializer_c.pyx
