1
# Copyright (C) 2007 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
17
"""Container format for Bazaar data.
19
"Containers" and "records" are described in doc/developers/container-format.txt.
24
from bzrlib import errors
27
FORMAT_ONE = "Bazaar pack format 1 (introduced in 0.18)"
30
_whitespace_re = re.compile('[\t\n\x0b\x0c\r ]')
33
def _check_name(name):
34
"""Do some basic checking of 'name'.
36
At the moment, this just checks that there are no whitespace characters in a
39
:raises InvalidRecordError: if name is not valid.
40
:seealso: _check_name_encoding
42
if _whitespace_re.search(name) is not None:
43
raise errors.InvalidRecordError("%r is not a valid name." % (name,))
46
def _check_name_encoding(name):
47
"""Check that 'name' is valid UTF-8.
49
This is separate from _check_name because UTF-8 decoding is relatively
50
expensive, and we usually want to avoid it.
52
:raises InvalidRecordError: if name is not valid UTF-8.
56
except UnicodeDecodeError, e:
57
raise errors.InvalidRecordError(str(e))
60
class ContainerWriter(object):
61
"""A class for writing containers."""
63
def __init__(self, write_func):
66
:param write_func: a callable that will be called when this
67
ContainerWriter needs to write some bytes.
69
self.write_func = write_func
72
"""Begin writing a container."""
73
self.write_func(FORMAT_ONE + "\n")
76
"""Finish writing a container."""
79
def add_bytes_record(self, bytes, names):
80
"""Add a Bytes record with the given names."""
84
self.write_func(str(len(bytes)) + "\n")
87
# Make sure we're writing valid names. Note that we will leave a
88
# half-written record if a name is bad!
90
self.write_func(name + "\n")
93
# Finally, the contents.
94
self.write_func(bytes)
97
class BaseReader(object):
99
def __init__(self, source_file):
102
:param source_file: a file-like object with `read` and `readline`
105
self._source = source_file
107
def reader_func(self, length=None):
108
return self._source.read(length)
110
def _read_line(self):
111
line = self._source.readline()
112
if not line.endswith('\n'):
113
raise errors.UnexpectedEndOfContainerError()
114
return line.rstrip('\n')
117
class ContainerReader(BaseReader):
118
"""A class for reading Bazaar's container format."""
120
def iter_records(self):
121
"""Iterate over the container, yielding each record as it is read.
123
Each yielded record will be a 2-tuple of (names, callable), where names
124
is a ``list`` and bytes is a function that takes one argument,
127
You **must not** call the callable after advancing the interator to the
128
next record. That is, this code is invalid::
130
record_iter = container.iter_records()
131
names1, callable1 = record_iter.next()
132
names2, callable2 = record_iter.next()
133
bytes1 = callable1(None)
135
As it will give incorrect results and invalidate the state of the
138
:raises ContainerError: if any sort of containter corruption is
139
detected, e.g. UnknownContainerFormatError is the format of the
140
container is unrecognised.
141
:seealso: ContainerReader.read
144
return self._iter_records()
146
def iter_record_objects(self):
147
"""Iterate over the container, yielding each record as it is read.
149
Each yielded record will be an object with ``read`` and ``validate``
150
methods. Like with iter_records, it is not safe to use a record object
151
after advancing the iterator to yield next record.
153
:raises ContainerError: if any sort of containter corruption is
154
detected, e.g. UnknownContainerFormatError is the format of the
155
container is unrecognised.
156
:seealso: iter_records
159
return self._iter_record_objects()
161
def _iter_records(self):
162
for record in self._iter_record_objects():
165
def _iter_record_objects(self):
167
record_kind = self.reader_func(1)
168
if record_kind == 'B':
170
reader = BytesRecordReader(self._source)
172
elif record_kind == 'E':
173
# End marker. There are no more records.
175
elif record_kind == '':
176
# End of stream encountered, but no End Marker record seen, so
177
# this container is incomplete.
178
raise errors.UnexpectedEndOfContainerError()
180
# Unknown record type.
181
raise errors.UnknownRecordTypeError(record_kind)
183
def _read_format(self):
184
format = self._read_line()
185
if format != FORMAT_ONE:
186
raise errors.UnknownContainerFormatError(format)
189
"""Validate this container and its records.
191
Validating consumes the data stream just like iter_records and
192
iter_record_objects, so you cannot call it after
193
iter_records/iter_record_objects.
195
:raises ContainerError: if something is invalid.
198
for record_names, read_bytes in self.iter_records():
200
for name in record_names:
201
_check_name_encoding(name)
202
# Check that the name is unique. Note that Python will refuse
203
# to decode non-shortest forms of UTF-8 encoding, so there is no
204
# risk that the same unicode string has been encoded two
206
if name in all_names:
207
raise errors.DuplicateRecordNameError(name)
209
excess_bytes = self.reader_func(1)
210
if excess_bytes != '':
211
raise errors.ContainerHasExcessDataError(excess_bytes)
214
class BytesRecordReader(BaseReader):
219
You can either validate or read a record, you can't do both.
221
:returns: A tuple of (names, callable). The callable can be called
222
repeatedly to obtain the bytes for the record, with a max_length
223
argument. If max_length is None, returns all the bytes. Because
224
records can be arbitrarily large, using None is not recommended
225
unless you have reason to believe the content will fit in memory.
227
# Read the content length.
228
length_line = self._read_line()
230
length = int(length_line)
232
raise errors.InvalidRecordError(
233
"%r is not a valid length." % (length_line,))
235
# Read the list of names.
238
name = self._read_line()
244
self._remaining_length = length
245
return names, self._content_reader
247
def _content_reader(self, max_length):
248
if max_length is None:
249
length_to_read = self._remaining_length
251
length_to_read = min(max_length, self._remaining_length)
252
self._remaining_length -= length_to_read
253
bytes = self.reader_func(length_to_read)
254
if len(bytes) != length_to_read:
255
raise errors.UnexpectedEndOfContainerError()
259
"""Validate this record.
261
You can either validate or read, you can't do both.
263
:raises ContainerError: if this record is invalid.
265
names, read_bytes = self.read()
267
_check_name_encoding(name)
added
removed
bzrlib/pack.py
