1
# Copyright (c) 2012 Canonical Ltd
3
# This program is free software: you can redistribute it and/or modify
4
# it under the terms of the GNU Affero General Public License as published by
5
# the Free Software Foundation, either version 3 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 Affero General Public License for more details.
13
# You should have received a copy of the GNU Affero General Public
14
# License along with this program. If not, see
15
# <http://www.gnu.org/licenses/>.
20
from email.utils import parsedate_tz
27
from pycassa.system_manager import (
32
from pycassa.util import convert_time_to_uuid
34
from grackle.cassandra import workaround_1779
37
def create_schema(host, keyspace, clobber=False, create_keyspace=False):
38
mgr = SystemManager(host)
41
mgr.create_keyspace(keyspace, replication_factor=1)
44
for cf in mgr.get_keyspace_column_families(keyspace):
45
mgr.drop_column_family(keyspace, cf)
49
mgr.create_column_family, keyspace, 'message',
50
key_validation_class=LEXICAL_UUID_TYPE)
52
mgr.create_column_family, keyspace, 'archive_message',
53
comparator_type=TIME_UUID_TYPE,
54
default_validation_class=LEXICAL_UUID_TYPE)
60
def _parse_message(message):
61
"""Get a date and dict of an RFC822 message."""
62
parsed = email.parser.Parser().parsestr(message)
65
for key in ('from', 'to', 'subject', 'message-id'):
66
value = parsed.get(key, None)
68
message_dict[key] = value
70
date = parsed.get('date')
73
pdate = parsedate_tz(date)
74
date = datetime.datetime(
76
tzinfo=dateutil.tz.tzoffset('', pdate[9]))
79
message_dict['date'] = date.isoformat() if date is not None else None
81
return date, message_dict
84
def _utc_datetime(dt):
85
return dt.astimezone(dateutil.tz.tzutc())
88
def _utc_timestamp(dt):
89
return calendar.timegm(_utc_datetime(dt).timetuple())
92
def _utc_timeuuid(dt, lowest_val=True):
93
return convert_time_to_uuid(_utc_timestamp(dt), lowest_val)
96
def _cmp_timeuuid(a, b):
98
return cmp(a.time, b.time)
102
def _bound_timeuuid(a, b, max=False):
103
if a == '' or _cmp_timeuuid(b, a) == (1 if max else -1):
108
def _make_bounds(memo, range_start, range_finish, backward):
111
memo = uuid.UUID(memo)
116
if range_start is not None:
117
start = _bound_timeuuid(
118
start, _utc_timeuuid(range_start), max=True)
119
if range_finish is not None:
120
finish = _bound_timeuuid(
121
finish, _utc_timeuuid(range_finish, lowest_val=False))
122
return memo, start, finish
125
LEGAL_HEADERS = set([
126
'date', 'from', 'subject', 'message-id',
130
def _format_message(message, headers=[], include_raw=False):
134
assert not set(headers).difference(LEGAL_HEADERS)
136
for header in headers:
137
hdict[header] = message.get(header)
138
data['headers'] = hdict
141
data['raw'] = message['raw']
146
class CassandraConnection(object):
148
def __init__(self, keyspace, hosts):
149
self._keyspace = keyspace
151
self._connection = self._connect()
152
self._pool = self._connect()
153
self.messages = self._column_family('message')
154
self.archive_messages = self._column_family('archive_message')
157
return pycassa.pool.ConnectionPool(self._keyspace, self._hosts)
159
def _column_family(self, name):
160
return pycassa.ColumnFamily(self._pool, name)
162
def add_message(self, archive_uuid, message):
163
message_uuid = uuid.uuid4()
164
message_date, message_dict = _parse_message(message)
165
message_dict['raw'] = message
166
message_dict['date_created'] = (
167
datetime.datetime.utcnow().isoformat() + 'Z')
168
self.messages.insert(message_uuid, message_dict)
169
self.archive_messages.insert(
171
{_utc_timestamp(message_date): message_uuid})
173
'Imported %s into %s'
174
% (message_dict.get('message-id', None), archive_uuid))
177
def _trim(self, sequence, end):
178
"""Return the sequence with one of the ends trimmed.
180
:param end: if true, remove the last element. otherwise remove
188
def get_messages(self, archive_uuid, order, count, memo, backward=False,
189
start_date=None, finish_date=None, format='all',
190
headers=['from', 'date', 'subject', 'message-id']):
191
if order in ("date", "-date"):
192
reversed = order[0] == '-'
194
raise AssertionError("Unsupported order.")
196
memo, start, finish = _make_bounds(
197
memo, start_date, finish_date, backward)
199
# Get up to n+1 messages from the memo: the last item of the
200
# previous batch (because that's where the memo starts) + this
202
pairs = self.archive_messages.get(
203
archive_uuid, column_count=count + 1, column_start=start,
204
column_finish=finish, column_reversed=reversed).items()
206
if len(pairs) and memo and pairs[0][0] <= memo:
207
# The memo (from the previous batch) was included in the result.
209
pairs = self._trim(pairs, False ^ backward)
210
elif len(pairs) > count:
211
# There was no memo in the result, so the n+1th element is
212
# unnecessary. Kill it.
213
pairs = self._trim(pairs, True ^ backward)
216
return (None, [], None)
218
assert 0 < len(pairs) <= count
220
# We've narrowed down the message references. Fetch the messages.
221
ids = [v for k, v in pairs]
222
formatter = functools.partial(
223
_format_message, headers=headers, include_raw=True)
224
# XXX: No need to get all columns. Restrict based on format.
225
messages = self.messages.multiget(ids)
229
[formatter(messages[id]) for id in ids],