1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
|
# Copyright (C) 2005 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
"""Testament - a summary of a revision for signing.
A testament can be defined as "something that serves as tangible
proof or evidence." In bzr we use them to allow people to certify
particular revisions as authentic.
The goal is that if two revisions are semantically equal, then they will
have a byte-for-byte equal testament. We can define different versions of
"semantically equal" by using different testament classes; e.g. one that
includes or ignores file-ids.
We sign a testament rather than the revision XML itself for several reasons.
The most important is that the form in which the revision is stored
internally is designed for that purpose, and contains information which need
not be attested to by the signer. For example the inventory contains the
last-changed revision for a file, but this is not necessarily something the
user cares to sign.
Having unnecessary fields signed makes the signatures brittle when the same
revision is stored in different branches or when the format is upgraded.
Handling upgrades is another motivation for using testaments separate from
the stored revision. We would like to be able to compare a signature
generated from an old-format tree to newer tree, or vice versa. This could
be done by comparing the revisions but that makes it unclear about exactly
what is being compared or not.
Different signing keys might indicate different levels of trust; we can in
the future extend this to allow signatures indicating not just that a
particular version is authentic but that it has other properties.
The signature can be applied to either the full testament or to just a
hash of it.
Testament format 1
~~~~~~~~~~~~~~~~~~
* timestamps are given as integers to avoid rounding errors
* parents given in lexicographical order
* indented-text form similar to log; intended to be human readable
* paths are given with forward slashes
* files are named using paths for ease of comparison/debugging
* the testament uses unix line-endings (\n)
"""
# XXX: At the moment, clients trust that the graph described in a weave
# is accurate, but that's not covered by the testament. Perhaps the best
# fix is when verifying a revision to make sure that every file mentioned
# in the revision has compatible ancestry links.
# TODO: perhaps write timestamp in a more readable form
# TODO: Perhaps these should just be different formats in which inventories/
# revisions can be serialized.
from copy import copy
from bzrlib.osutils import (
contains_whitespace,
contains_linebreaks,
sha,
)
class Testament(object):
"""Reduced summary of a revision.
Testaments can be
- produced from a revision
- written to a stream
- loaded from a stream
- compared to a revision
"""
long_header = 'bazaar-ng testament version 1\n'
short_header = 'bazaar-ng testament short form 1\n'
@classmethod
def from_revision(cls, repository, revision_id):
"""Produce a new testament from a historical revision"""
rev = repository.get_revision(revision_id)
inventory = repository.get_inventory(revision_id)
return cls(rev, inventory)
def __init__(self, rev, inventory):
"""Create a new testament for rev using inventory."""
self.revision_id = rev.revision_id
self.committer = rev.committer
self.timezone = rev.timezone or 0
self.timestamp = rev.timestamp
self.message = rev.message
self.parent_ids = rev.parent_ids[:]
self.inventory = inventory
self.revprops = copy(rev.properties)
if contains_whitespace(self.revision_id):
raise ValueError(self.revision_id)
if contains_linebreaks(self.committer):
raise ValueError(self.committer)
def as_text_lines(self):
"""Yield text form as a sequence of lines.
The result is returned in utf-8, because it should be signed or
hashed in that encoding.
"""
r = []
a = r.append
a(self.long_header)
a('revision-id: %s\n' % self.revision_id)
a('committer: %s\n' % self.committer)
a('timestamp: %d\n' % self.timestamp)
a('timezone: %d\n' % self.timezone)
# inventory length contains the root, which is not shown here
a('parents:\n')
for parent_id in sorted(self.parent_ids):
if contains_whitespace(parent_id):
raise ValueError(parent_id)
a(' %s\n' % parent_id)
a('message:\n')
for l in self.message.splitlines():
a(' %s\n' % l)
a('inventory:\n')
for path, ie in self._get_entries():
a(self._entry_to_line(path, ie))
r.extend(self._revprops_to_lines())
return [line.encode('utf-8') for line in r]
def _get_entries(self):
entries = self.inventory.iter_entries()
entries.next()
return entries
def _escape_path(self, path):
if contains_linebreaks(path):
raise ValueError(path)
return unicode(path.replace('\\', '/').replace(' ', '\ '))
def _entry_to_line(self, path, ie):
"""Turn an inventory entry into a testament line"""
if contains_whitespace(ie.file_id):
raise ValueError(ie.file_id)
content = ''
content_spacer=''
if ie.kind == 'file':
# TODO: avoid switching on kind
if not ie.text_sha1:
raise AssertionError()
content = ie.text_sha1
content_spacer = ' '
elif ie.kind == 'symlink':
if not ie.symlink_target:
raise AssertionError()
content = self._escape_path(ie.symlink_target)
content_spacer = ' '
l = u' %s %s %s%s%s\n' % (ie.kind, self._escape_path(path),
ie.file_id.decode('utf8'),
content_spacer, content)
return l
def as_text(self):
return ''.join(self.as_text_lines())
def as_short_text(self):
"""Return short digest-based testament."""
return (self.short_header +
'revision-id: %s\n'
'sha1: %s\n'
% (self.revision_id, self.as_sha1()))
def _revprops_to_lines(self):
"""Pack up revision properties."""
if not self.revprops:
return []
r = ['properties:\n']
for name, value in sorted(self.revprops.items()):
if contains_whitespace(name):
raise ValueError(name)
r.append(' %s:\n' % name)
for line in value.splitlines():
r.append(u' %s\n' % line)
return r
def as_sha1(self):
s = sha()
map(s.update, self.as_text_lines())
return s.hexdigest()
class StrictTestament(Testament):
"""This testament format is for use as a checksum in bundle format 0.8"""
long_header = 'bazaar-ng testament version 2.1\n'
short_header = 'bazaar-ng testament short form 2.1\n'
def _entry_to_line(self, path, ie):
l = Testament._entry_to_line(self, path, ie)[:-1]
l += ' ' + ie.revision
l += {True: ' yes\n', False: ' no\n'}[ie.executable]
return l
class StrictTestament3(StrictTestament):
"""This testament format is for use as a checksum in bundle format 0.9+
It differs from StrictTestament by including data about the tree root.
"""
long_header = 'bazaar testament version 3 strict\n'
short_header = 'bazaar testament short form 3 strict\n'
def _get_entries(self):
return self.inventory.iter_entries()
def _escape_path(self, path):
if contains_linebreaks(path):
raise ValueError(path)
if path == '':
path = '.'
return unicode(path.replace('\\', '/').replace(' ', '\ '))
|