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
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
from bzrlib.selftest.testrevision import make_branches
18
from bzrlib.trace import mutter, note
18
from cStringIO import StringIO
21
import bzrlib.errors as errors
22
from bzrlib.errors import InstallFailed, NoSuchRevision, WeaveError
23
from bzrlib.trace import mutter, note, warning
19
24
from bzrlib.branch import Branch
20
25
from bzrlib.progress import ProgressBar
26
from bzrlib.xml5 import serializer_v5
27
from bzrlib.osutils import sha_string, split_lines
29
"""Copying of history from one branch to another.
31
The basic plan is that every branch knows the history of everything
32
that has merged into it. As the first step of a merge, pull, or
33
branch operation we copy history from the source into the destination
36
The copying is done in a slightly complicated order. We don't want to
37
add a revision to the store until everything it refers to is also
38
stored, so that if a revision is present we can totally recreate it.
39
However, we can't know what files are included in a revision until we
40
read its inventory. Therefore, we first pull the XML and hold it in
41
memory until we've updated all of the files referenced.
44
# TODO: Avoid repeatedly opening weaves so many times.
46
# XXX: This doesn't handle ghost (not present in branch) revisions at
47
# all yet. I'm not sure they really should be supported.
49
# NOTE: This doesn't copy revisions which may be present but not
50
# merged into the last revision. I'm not sure we want to do that.
52
# - get a list of revisions that need to be pulled in
53
# - for each one, pull in that revision file
54
# and get the inventory, and store the inventory with right
56
# - and get the ancestry, and store that with right parents too
57
# - and keep a note of all file ids and version seen
58
# - then go through all files; for each one get the weave,
59
# and add in all file versions
24
63
def greedy_fetch(to_branch, from_branch, revision=None, pb=None):
25
"""Copy a revision and all available ancestors from one branch to another
26
If no revision is specified, uses the last revision in the source branch's
64
f = Fetcher(to_branch, from_branch, revision, pb)
65
return f.count_copied, f.failed_revisions
69
class Fetcher(object):
70
"""Pull revisions and texts from one branch to another.
72
This doesn't update the destination's history; that can be done
73
separately if desired.
76
If set, pull only up to this revision_id.
80
last_revision -- if last_revision
81
is given it will be that, otherwise the last revision of
84
count_copied -- number of revisions copied
86
count_weaves -- number of file weaves copied
29
from_history = from_branch.revision_history()
30
required_revisions = set(from_history)
32
if revision is not None:
33
required_revisions.add(revision)
35
rev_index = from_history.index(revision)
38
if rev_index is not None:
39
from_history = from_history[:rev_index + 1]
41
from_history = [revision]
42
to_history = to_branch.revision_history()
44
for rev_id in from_history:
45
if not has_revision(to_branch, rev_id):
46
missing.append(rev_id)
49
while len(missing) > 0:
50
installed, failed = to_branch.install_revisions(from_branch,
54
required_failed = failed.intersection(required_revisions)
55
if len(required_failed) > 0:
56
raise bzrlib.errors.InstallFailed(required_failed)
58
note("Failed to install %s" % rev_id)
59
all_failed.update(failed)
61
for rev_id in missing:
63
revision = from_branch.get_revision(rev_id)
64
except bzrlib.errors.NoSuchRevision:
65
if revision in from_history:
69
for parent in [p.revision_id for p in revision.parents]:
70
if not has_revision(to_branch, parent):
71
new_missing.append(parent)
73
return count, all_failed
76
from testsweet import InTempDir
77
from bzrlib.commit import commit
78
def has_revision(branch, revision_id):
80
branch.get_revision_xml(revision_id)
82
except bzrlib.errors.NoSuchRevision:
85
class TestFetch(InTempDir):
89
return Branch(name, init=True)
88
def __init__(self, to_branch, from_branch, last_revision=None, pb=None):
89
if to_branch == from_branch:
90
raise Exception("can't fetch from a branch to itself")
91
self.to_branch = to_branch
92
self.to_weaves = to_branch.weave_store
93
self.to_control = to_branch.control_weaves
94
self.from_branch = from_branch
95
self.from_weaves = from_branch.weave_store
96
self.from_control = from_branch.control_weaves
97
self.failed_revisions = []
100
self.count_weaves = 0
101
self.copied_file_ids = set()
103
self.pb = bzrlib.ui.ui_factory.progress_bar()
106
self.from_branch.lock_read()
108
self._fetch_revisions(last_revision)
110
self.from_branch.unlock()
113
def _fetch_revisions(self, last_revision):
114
self.last_revision = self._find_last_revision(last_revision)
115
mutter('fetch up to rev {%s}', self.last_revision)
116
if (self.last_revision is not None and
117
self.to_branch.has_revision(self.last_revision)):
120
revs_to_fetch = self._compare_ancestries()
122
raise InstallFailed([self.last_revision])
123
self._copy_revisions(revs_to_fetch)
124
self.new_ancestry = revs_to_fetch
126
def _find_last_revision(self, last_revision):
127
"""Find the limiting source revision.
129
Every ancestor of that revision will be merged across.
131
Returns the revision_id, or returns None if there's no history
132
in the source branch."""
135
self.pb.update('get source history')
136
from_history = self.from_branch.revision_history()
137
self.pb.update('get destination history')
139
return from_history[-1]
141
return None # no history in the source branch
91
#highest indices a: 5, b: 7
92
br_a, br_b = make_branches()
93
assert not has_revision(br_b, br_a.revision_history()[3])
94
assert has_revision(br_b, br_a.revision_history()[2])
95
assert len(br_b.revision_history()) == 7
96
assert greedy_fetch(br_b, br_a, br_a.revision_history()[2])[0] == 0
98
# greedy_fetch is not supposed to alter the revision history
99
assert len(br_b.revision_history()) == 7
100
assert not has_revision(br_b, br_a.revision_history()[3])
102
assert len(br_b.revision_history()) == 7
103
assert greedy_fetch(br_b, br_a, br_a.revision_history()[3])[0] == 1
104
assert has_revision(br_b, br_a.revision_history()[3])
105
assert not has_revision(br_a, br_b.revision_history()[3])
106
assert not has_revision(br_a, br_b.revision_history()[4])
108
# When a non-branch ancestor is missing, it should be a failure, not
110
br_a4 = new_branch('br_a4')
111
count, failures = greedy_fetch(br_a4, br_a)
113
assert failures == set((br_b.revision_history()[4],
114
br_b.revision_history()[5]))
116
assert greedy_fetch(br_a, br_b)[0] == 4
117
assert has_revision(br_a, br_b.revision_history()[3])
118
assert has_revision(br_a, br_b.revision_history()[4])
120
br_b2 = new_branch('br_b2')
121
assert greedy_fetch(br_b2, br_b)[0] == 7
122
assert has_revision(br_b2, br_b.revision_history()[4])
123
assert has_revision(br_b2, br_a.revision_history()[2])
124
assert not has_revision(br_b2, br_a.revision_history()[3])
126
br_a2 = new_branch('br_a2')
127
assert greedy_fetch(br_a2, br_a)[0] == 9
128
assert has_revision(br_a2, br_b.revision_history()[4])
129
assert has_revision(br_a2, br_a.revision_history()[3])
131
br_a3 = new_branch('br_a3')
132
assert greedy_fetch(br_a3, br_a2)[0] == 0
133
for revno in range(4):
134
assert not has_revision(br_a3, br_a.revision_history()[revno])
135
assert greedy_fetch(br_a3, br_a2, br_a.revision_history()[2])[0] == 3
136
fetched = greedy_fetch(br_a3, br_a2, br_a.revision_history()[3])[0]
137
assert fetched == 3, "fetched %d instead of 3" % fetched
138
# InstallFailed should be raised if the branch is missing the revision
139
# that was requested.
140
self.assertRaises(bzrlib.errors.InstallFailed, greedy_fetch, br_a3,
142
# InstallFailed should be raised if the branch is missing a revision
143
# from its own revision history
144
br_a2.append_revision('a-b-c')
145
self.assertRaises(bzrlib.errors.InstallFailed, greedy_fetch, br_a3,
150
if __name__ == '__main__':
152
sys.exit(run_suite(unittest.makeSuite()))
144
def _compare_ancestries(self):
145
"""Get a list of revisions that must be copied.
147
That is, every revision that's in the ancestry of the source
148
branch and not in the destination branch."""
149
self.pb.update('get source ancestry')
150
self.from_ancestry = self.from_branch.get_ancestry(self.last_revision)
152
dest_last_rev = self.to_branch.last_revision()
153
self.pb.update('get destination ancestry')
155
dest_ancestry = self.to_branch.get_ancestry(dest_last_rev)
158
ss = set(dest_ancestry)
160
for rev_id in self.from_ancestry:
162
to_fetch.append(rev_id)
163
mutter('need to get revision {%s}', rev_id)
164
mutter('need to get %d revisions in total', len(to_fetch))
165
self.count_total = len(to_fetch)
168
def _copy_revisions(self, revs_to_fetch):
170
for rev_id in revs_to_fetch:
174
if self.to_branch.has_revision(rev_id):
176
self.pb.update('fetch revision', i, self.count_total)
177
self._copy_one_revision(rev_id)
178
self.count_copied += 1
181
def _copy_one_revision(self, rev_id):
182
"""Copy revision and everything referenced by it."""
183
mutter('copying revision {%s}', rev_id)
184
rev_xml = self.from_branch.get_revision_xml(rev_id)
185
inv_xml = self.from_branch.get_inventory_xml(rev_id)
186
rev = serializer_v5.read_revision_from_string(rev_xml)
187
inv = serializer_v5.read_inventory_from_string(inv_xml)
188
assert rev.revision_id == rev_id
189
assert rev.inventory_sha1 == sha_string(inv_xml)
190
mutter(' commiter %s, %d parents',
193
self._copy_new_texts(rev_id, inv)
194
parents = rev.parent_ids
195
for parent in parents:
196
if not self.to_branch.has_revision(parent):
197
parents.pop(parents.index(parent))
198
self._copy_inventory(rev_id, inv_xml, parents)
199
self.to_branch.revision_store.add(StringIO(rev_xml), rev_id)
200
mutter('copied revision %s', rev_id)
203
def _copy_inventory(self, rev_id, inv_xml, parent_ids):
204
self.to_control.add_text('inventory', rev_id,
205
split_lines(inv_xml), parent_ids,
206
self.to_branch.get_transaction())
208
def _copy_new_texts(self, rev_id, inv):
209
"""Copy any new texts occuring in this revision."""
210
# TODO: Rather than writing out weaves every time, hold them
211
# in memory until everything's done? But this way is nicer
212
# if it's interrupted.
213
for path, ie in inv.iter_entries():
214
if ie.revision != rev_id:
216
mutter('%s {%s} is changed in this revision',
218
self._copy_one_weave(rev_id, ie.file_id)
221
def _copy_one_weave(self, rev_id, file_id):
222
"""Copy one file weave."""
223
mutter('copy file {%s} modified in {%s}', file_id, rev_id)
224
if file_id in self.copied_file_ids:
225
mutter('file {%s} already copied', file_id)
227
from_weave = self.from_weaves.get_weave(file_id,
228
self.from_branch.get_transaction())
229
to_weave = self.to_weaves.get_weave_or_empty(file_id,
230
self.to_branch.get_transaction())
232
to_weave.join(from_weave)
233
except errors.WeaveParentMismatch:
234
to_weave.reweave(from_weave)
235
self.to_weaves.put_weave(file_id, to_weave,
236
self.to_branch.get_transaction())
237
self.count_weaves += 1
238
self.copied_file_ids.add(file_id)
239
mutter('copied file {%s}', file_id)
added
removed
bzrlib/fetch.py
