1
# Copyright (C) 2008 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
"""Pyrex extensions to btree node parsing."""
20
cdef extern from "stdlib.h":
21
ctypedef unsigned size_t
23
cdef extern from "Python.h":
24
ctypedef struct PyObject:
26
int PyList_Append(object lst, object item) except -1
28
char *PyString_AsString(object p) except NULL
29
object PyString_FromStringAndSize(char *, Py_ssize_t)
30
int PyString_CheckExact(object s)
31
int PyString_CheckExact_ptr "PyString_CheckExact" (PyObject *)
32
Py_ssize_t PyString_Size(object p)
33
Py_ssize_t PyString_GET_SIZE_ptr "PyString_GET_SIZE" (PyObject *)
34
char * PyString_AS_STRING_ptr "PyString_AS_STRING" (PyObject *)
35
int PyString_AsStringAndSize_ptr(PyObject *, char **buf, Py_ssize_t *len)
36
int PyTuple_CheckExact(object t)
37
Py_ssize_t PyTuple_GET_SIZE(object t)
38
PyObject *PyTuple_GET_ITEM_ptr_object "PyTuple_GET_ITEM" (object tpl, int index)
40
cdef extern from "string.h":
41
void *memcpy(void *dest, void *src, size_t n)
42
void *memchr(void *s, int c, size_t n)
44
# void *memrchr(void *s, int c, size_t n)
45
int strncmp(char *s1, char *s2, size_t n)
48
# TODO: Find some way to import this from _dirstate_helpers
49
cdef void* _my_memrchr(void *s, int c, size_t n):
50
# memrchr seems to be a GNU extension, so we have to implement it ourselves
51
# It is not present in any win32 standard library
63
# TODO: Import this from _dirstate_helpers when it is merged
64
cdef object safe_string_from_size(char *s, Py_ssize_t size):
67
'tried to create a string with an invalid size: %d @0x%x'
69
return PyString_FromStringAndSize(s, size)
72
cdef class BTreeLeafParser:
73
"""Parse the leaf nodes of a BTree index.
75
:ivar bytes: The PyString object containing the uncompressed text for the
77
:ivar key_length: An integer describing how many pieces the keys have for
79
:ivar ref_list_length: An integer describing how many references this index
81
:ivar keys: A PyList of keys found in this node.
83
:ivar _cur_str: A pointer to the start of the next line to parse
84
:ivar _end_str: A pointer to the end of bytes
85
:ivar _start: Pointer to the location within the current line while
87
:ivar _header_found: True when we have parsed the header for this node
92
cdef int ref_list_length
97
# The current start point for parsing
100
cdef int _header_found
102
def __init__(self, bytes, key_length, ref_list_length):
104
self.key_length = key_length
105
self.ref_list_length = ref_list_length
109
self._header_found = 0
111
cdef extract_key(self, char * last):
114
:param last: points at the byte after the last byte permitted for the
118
cdef int loop_counter
122
while loop_counter < self.key_length:
123
loop_counter = loop_counter + 1
125
temp_ptr = <char*>memchr(self._start, c'\0', last - self._start)
127
if loop_counter == self.key_length:
132
failure_string = ("invalid key, wanted segment from " +
133
repr(safe_string_from_size(self._start,
134
last - self._start)))
135
raise AssertionError(failure_string)
136
# capture the key string
137
# TODO: Consider using PyIntern_FromString, the only caveat is that
138
# it assumes a NULL-terminated string, so we have to check if
139
# temp_ptr[0] == c'\0' or some other char.
140
key_element = safe_string_from_size(self._start,
141
temp_ptr - self._start)
142
# advance our pointer
143
self._start = temp_ptr + 1
144
PyList_Append(key_segments, key_element)
145
return tuple(key_segments)
147
cdef int process_line(self) except -1:
148
"""Process a line in the bytes."""
152
cdef char *next_start
153
cdef int loop_counter
155
self._start = self._cur_str
156
# Find the next newline
157
last = <char*>memchr(self._start, c'\n', self._end_str - self._start)
159
# Process until the end of the file
161
self._cur_str = self._end_str
163
# And the next string is right after it
164
self._cur_str = last + 1
165
# The last character is right before the '\n'
168
if last == self._start:
171
if last < self._start:
172
# Unexpected error condition - fail
174
if 0 == self._header_found:
175
# The first line in a leaf node is the header "type=leaf\n"
176
if strncmp("type=leaf", self._start, last - self._start) == 0:
177
self._header_found = 1
180
raise AssertionError('Node did not start with "type=leaf": %r'
181
% (safe_string_from_size(self._start, last - self._start)))
184
key = self.extract_key(last)
185
# find the value area
186
temp_ptr = <char*>_my_memrchr(self._start, c'\0', last - self._start)
191
# capture the value string
192
value = safe_string_from_size(temp_ptr + 1, last - temp_ptr - 1)
193
# shrink the references end point
195
if self.ref_list_length:
198
while loop_counter < self.ref_list_length:
200
# extract a reference list
201
loop_counter = loop_counter + 1
202
if last < self._start:
204
# find the next reference list end point:
205
temp_ptr = <char*>memchr(self._start, c'\t', last - self._start)
207
# Only valid for the last list
208
if loop_counter != self.ref_list_length:
211
raise AssertionError("invalid key")
213
# scan to the end of the ref list area
217
# scan to the end of this ref list
219
next_start = temp_ptr + 1
220
# Now, there may be multiple keys in the ref list.
221
while self._start < ref_ptr:
222
# loop finding keys and extracting them
223
temp_ptr = <char*>memchr(self._start, c'\r',
224
ref_ptr - self._start)
226
# key runs to the end
228
PyList_Append(ref_list, self.extract_key(temp_ptr))
229
PyList_Append(ref_lists, tuple(ref_list))
230
# prepare for the next reference list
231
self._start = next_start
232
ref_lists = tuple(ref_lists)
233
node_value = (value, ref_lists)
235
if last != self._start:
236
# unexpected reference data present
238
node_value = (value, ())
239
PyList_Append(self.keys, (key, node_value))
243
cdef Py_ssize_t byte_count
244
if not PyString_CheckExact(self.bytes):
245
raise AssertionError('self.bytes is not a string.')
246
byte_count = PyString_Size(self.bytes)
247
self._cur_str = PyString_AsString(self.bytes)
248
# This points to the last character in the string
249
self._end_str = self._cur_str + byte_count
250
while self._cur_str < self._end_str:
255
def _parse_leaf_lines(bytes, key_length, ref_list_length):
256
parser = BTreeLeafParser(bytes, key_length, ref_list_length)
257
return parser.parse()
260
def _flatten_node(node, reference_lists):
261
"""Convert a node into the serialized form.
263
:param node: A tuple representing a node:
264
(index, key_tuple, value, references)
265
:param reference_lists: Does this index have reference lists?
266
:return: (string_key, flattened)
267
string_key The serialized key for referencing this node
268
flattened A string with the serialized form for the contents
270
cdef int have_reference_lists
271
cdef Py_ssize_t flat_len
272
cdef Py_ssize_t key_len
273
cdef Py_ssize_t node_len
276
cdef Py_ssize_t value_len
278
cdef Py_ssize_t refs_len
279
cdef Py_ssize_t next_len
280
cdef int first_ref_list
281
cdef int first_reference
283
cdef PyObject *ref_bit
284
cdef Py_ssize_t ref_bit_len
286
if not PyTuple_CheckExact(node):
287
raise TypeError('We expected a tuple() for node not: %s'
289
node_len = PyTuple_GET_SIZE(node)
290
have_reference_lists = reference_lists
291
if have_reference_lists:
293
raise ValueError('With ref_lists, we expected 4 entries not: %s'
296
raise ValueError('Without ref_lists, we need at least 3 entries not: %s'
298
# I don't expect that we can do faster than string.join()
299
string_key = '\0'.join(<object>PyTuple_GET_ITEM_ptr_object(node, 1))
301
# TODO: instead of using string joins, precompute the final string length,
302
# and then malloc a single string and copy everything in.
304
# TODO: We probably want to use PySequenceFast, because we have lists and
305
# tuples, but we aren't sure which we will get.
307
# line := string_key NULL flat_refs NULL value LF
308
# string_key := BYTES (NULL BYTES)*
309
# flat_refs := ref_list (TAB ref_list)*
310
# ref_list := ref (CR ref)*
311
# ref := BYTES (NULL BYTES)*
314
if have_reference_lists:
315
# Figure out how many bytes it will take to store the references
316
ref_lists = <object>PyTuple_GET_ITEM_ptr_object(node, 3)
317
next_len = len(ref_lists) # TODO: use a Py function
319
# If there are no nodes, we don't need to do any work
320
# Otherwise we will need (len - 1) '\t' characters to separate
321
# the reference lists
322
refs_len = refs_len + (next_len - 1)
323
for ref_list in ref_lists:
324
next_len = len(ref_list)
326
# We will need (len - 1) '\r' characters to separate the
328
refs_len = refs_len + (next_len - 1)
329
for reference in ref_list:
330
if not PyTuple_CheckExact(reference):
332
'We expect references to be tuples not: %s'
334
next_len = PyTuple_GET_SIZE(reference)
336
# We will need (len - 1) '\x00' characters to
337
# separate the reference key
338
refs_len = refs_len + (next_len - 1)
339
for i from 0 <= i < next_len:
340
ref_bit = PyTuple_GET_ITEM_ptr_object(reference, i)
341
if not PyString_CheckExact_ptr(ref_bit):
342
raise TypeError('We expect reference bits'
343
' to be strings not: %s'
344
% type(<object>ref_bit))
345
refs_len = refs_len + PyString_GET_SIZE_ptr(ref_bit)
347
# So we have the (key NULL refs NULL value LF)
348
key_len = PyString_Size(string_key)
349
val = PyTuple_GET_ITEM_ptr_object(node, 2)
350
if not PyString_CheckExact_ptr(val):
351
raise TypeError('Expected a plain str for value not: %s'
353
value = PyString_AS_STRING_ptr(val)
354
value_len = PyString_GET_SIZE_ptr(val)
355
flat_len = (key_len + 1 + refs_len + 1 + value_len + 1)
356
line = PyString_FromStringAndSize(NULL, flat_len)
357
# Get a pointer to the new buffer
358
out = PyString_AsString(line)
359
memcpy(out, PyString_AsString(string_key), key_len)
365
for ref_list in ref_lists:
366
if first_ref_list == 0:
371
for reference in ref_list:
372
if first_reference == 0:
376
next_len = PyTuple_GET_SIZE(reference)
377
for i from 0 <= i < next_len:
381
ref_bit = PyTuple_GET_ITEM_ptr_object(reference, i)
382
ref_bit_len = PyString_GET_SIZE_ptr(ref_bit)
383
memcpy(out, PyString_AS_STRING_ptr(ref_bit), ref_bit_len)
384
out = out + ref_bit_len
387
memcpy(out, value, value_len)
388
out = out + value_len
390
return string_key, line
added
removed
bzrlib/_btree_serializer_c.pyx
