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
19
from cStringIO import StringIO
22
import bzrlib.errors as errors
23
from bzrlib.errors import (InstallFailed, NoSuchRevision, WeaveError,
25
from bzrlib.trace import mutter, note, warning
21
from bzrlib.trace import mutter, note
26
22
from bzrlib.branch import Branch
27
23
from bzrlib.progress import ProgressBar
28
from bzrlib.xml5 import serializer_v5
29
from bzrlib.osutils import sha_string, split_lines
31
"""Copying of history from one branch to another.
33
The basic plan is that every branch knows the history of everything
34
that has merged into it. As the first step of a merge, pull, or
35
branch operation we copy history from the source into the destination
38
The copying is done in a slightly complicated order. We don't want to
39
add a revision to the store until everything it refers to is also
40
stored, so that if a revision is present we can totally recreate it.
41
However, we can't know what files are included in a revision until we
42
read its inventory. Therefore, we first pull the XML and hold it in
43
memory until we've updated all of the files referenced.
46
# TODO: Avoid repeatedly opening weaves so many times.
48
# XXX: This doesn't handle ghost (not present in branch) revisions at
49
# all yet. I'm not sure they really should be supported.
51
# NOTE: This doesn't copy revisions which may be present but not
52
# merged into the last revision. I'm not sure we want to do that.
54
# - get a list of revisions that need to be pulled in
55
# - for each one, pull in that revision file
56
# and get the inventory, and store the inventory with right
58
# - and get the ancestry, and store that with right parents too
59
# - and keep a note of all file ids and version seen
60
# - then go through all files; for each one get the weave,
61
# and add in all file versions
65
def greedy_fetch(to_branch, from_branch, revision=None, pb=None):
26
def greedy_fetch(to_branch, from_branch, revision, pb):
66
27
f = Fetcher(to_branch, from_branch, revision, pb)
67
28
return f.count_copied, f.failed_revisions
71
31
class Fetcher(object):
72
"""Pull revisions and texts from one branch to another.
74
This doesn't update the destination's history; that can be done
75
separately if desired.
78
If set, pull only up to this revision_id.
82
last_revision -- if last_revision
83
is given it will be that, otherwise the last revision of
86
count_copied -- number of revisions copied
88
count_weaves -- number of file weaves copied
90
def __init__(self, to_branch, from_branch, last_revision=None, pb=None):
91
if to_branch == from_branch:
92
raise Exception("can't fetch from a branch to itself")
32
"""Pull history from one branch to another."""
33
def __init__(self, to_branch, from_branch, revision_limit=None, pb=None):
93
34
self.to_branch = to_branch
94
self.to_weaves = to_branch.weave_store
95
self.to_control = to_branch.control_weaves
96
35
self.from_branch = from_branch
97
self.from_weaves = from_branch.weave_store
98
self.from_control = from_branch.control_weaves
99
self.failed_revisions = []
100
self.count_copied = 0
102
self.count_weaves = 0
103
self.copied_file_ids = set()
104
self.file_ids_names = {}
36
self.revision_limit = revision_limit
106
38
self.pb = bzrlib.ui.ui_factory.progress_bar()
109
self.from_branch.lock_read()
111
self._fetch_revisions(last_revision)
113
self.from_branch.unlock()
116
def _fetch_revisions(self, last_revision):
117
self.last_revision = self._find_last_revision(last_revision)
118
mutter('fetch up to rev {%s}', self.last_revision)
119
if (self.last_revision is not None and
120
self.to_branch.has_revision(self.last_revision)):
123
revs_to_fetch = self._compare_ancestries()
125
raise InstallFailed([self.last_revision])
126
self._copy_revisions(revs_to_fetch)
127
self.new_ancestry = revs_to_fetch
129
def _find_last_revision(self, last_revision):
130
"""Find the limiting source revision.
132
Every ancestor of that revision will be merged across.
134
Returns the revision_id, or returns None if there's no history
135
in the source branch."""
138
self.pb.update('get source history')
139
from_history = self.from_branch.revision_history()
140
self.pb.update('get destination history')
142
return from_history[-1]
144
return None # no history in the source branch
147
def _compare_ancestries(self):
148
"""Get a list of revisions that must be copied.
150
That is, every revision that's in the ancestry of the source
151
branch and not in the destination branch."""
152
self.pb.update('get source ancestry')
153
self.from_ancestry = self.from_branch.get_ancestry(self.last_revision)
155
dest_last_rev = self.to_branch.last_revision()
156
self.pb.update('get destination ancestry')
158
dest_ancestry = self.to_branch.get_ancestry(dest_last_rev)
161
ss = set(dest_ancestry)
163
for rev_id in self.from_ancestry:
165
to_fetch.append(rev_id)
41
self._scan_histories()
42
self.failed_revisions = []
47
def _scan_histories(self):
48
self.from_history = from_branch.revision_history()
49
self.required_revisions = set(from_history)
50
self.to_history = to_branch.revision_history()
51
if self.revision_limit:
52
raise NotImplementedError('sorry, revision_limit not handled yet')
53
self.need_revisions = []
54
for rev_id in self.from_history:
55
if not has_revision(self.to_branch):
56
self.need_revisions.append(rev_id)
166
57
mutter('need to get revision {%s}', rev_id)
167
mutter('need to get %d revisions in total', len(to_fetch))
168
self.count_total = len(to_fetch)
171
def _copy_revisions(self, revs_to_fetch):
173
for rev_id in revs_to_fetch:
177
if self.to_branch.has_revision(rev_id):
179
self.pb.update('copy revision', i, self.count_total)
180
self._copy_one_revision(rev_id)
181
self.count_copied += 1
184
def _copy_one_revision(self, rev_id):
185
"""Copy revision and everything referenced by it."""
186
mutter('copying revision {%s}', rev_id)
187
rev_xml = self.from_branch.get_revision_xml(rev_id)
188
inv_xml = self.from_branch.get_inventory_xml(rev_id)
189
rev = serializer_v5.read_revision_from_string(rev_xml)
190
inv = serializer_v5.read_inventory_from_string(inv_xml)
191
assert rev.revision_id == rev_id
192
assert rev.inventory_sha1 == sha_string(inv_xml)
193
mutter(' commiter %s, %d parents',
196
self._copy_new_texts(rev_id, inv)
197
parents = rev.parent_ids
198
new_parents = copy(parents)
199
for parent in parents:
200
if not self.to_branch.has_revision(parent):
201
new_parents.pop(new_parents.index(parent))
202
self._copy_inventory(rev_id, inv_xml, new_parents)
203
self.to_branch.revision_store.add(StringIO(rev_xml), rev_id)
204
mutter('copied revision %s', rev_id)
206
def _copy_inventory(self, rev_id, inv_xml, parent_ids):
207
self.to_control.add_text('inventory', rev_id,
208
split_lines(inv_xml), parent_ids,
209
self.to_branch.get_transaction())
211
def _copy_new_texts(self, rev_id, inv):
212
"""Copy any new texts occuring in this revision."""
213
# TODO: Rather than writing out weaves every time, hold them
214
# in memory until everything's done? But this way is nicer
215
# if it's interrupted.
216
for path, ie in inv.iter_entries():
217
self._copy_one_weave(rev_id, ie.file_id, ie.revision)
219
def _copy_one_weave(self, rev_id, file_id, text_revision):
220
"""Copy one file weave, esuring the result contains text_revision."""
221
# check if the revision is already there
222
if file_id in self.file_ids_names.keys( ) and \
223
text_revision in self.file_ids_names[file_id]:
225
to_weave = self.to_weaves.get_weave_or_empty(file_id,
226
self.to_branch.get_transaction())
227
if not file_id in self.file_ids_names.keys( ):
228
self.file_ids_names[file_id] = to_weave.names( )
229
if text_revision in to_weave:
231
from_weave = self.from_weaves.get_weave(file_id,
232
self.from_branch.get_transaction())
233
if text_revision not in from_weave:
234
raise MissingText(self.from_branch, text_revision, file_id)
235
mutter('copy file {%s} modified in {%s}', file_id, rev_id)
237
if to_weave.numversions() > 0:
238
# destination has contents, must merge
61
while self.need_revisions:
62
rev_id = self.need_revisions.pop()
63
mutter('try to get revision {%s}', rev_id)
69
def has_revision(branch, revision_id):
71
branch.get_revision_xml(revision_id)
73
except bzrlib.errors.NoSuchRevision:
77
def old_greedy_fetch(to_branch, from_branch, revision=None, pb=None):
78
"""Copy all history from one branch to another.
81
If set, copy only up to this point in the source branch.
83
@returns: number copied, missing ids
85
from_history = from_branch.revision_history()
86
required_revisions = set(from_history)
88
if revision is not None:
89
required_revisions.add(revision)
91
rev_index = from_history.index(revision)
94
if rev_index is not None:
95
from_history = from_history[:rev_index + 1]
97
from_history = [revision]
98
to_history = to_branch.revision_history()
100
for rev_id in from_history:
101
if not has_revision(to_branch, rev_id):
102
missing.append(rev_id)
104
# recurse down through the revision graph, looking for things that
107
while len(missing) > 0:
108
installed, failed = to_branch.install_revisions(from_branch,
109
revision_ids=missing,
112
required_failed = failed.intersection(required_revisions)
113
if len(required_failed) > 0:
114
raise bzrlib.errors.InstallFailed(required_failed)
115
for rev_id in failed:
116
note("Failed to install %s" % rev_id)
117
all_failed.update(failed)
119
for rev_id in missing:
240
to_weave.join(from_weave)
241
except errors.WeaveParentMismatch:
242
to_weave.reweave(from_weave)
244
# destination is empty, just replace it
245
to_weave = from_weave.copy( )
246
self.to_weaves.put_weave(file_id, to_weave,
247
self.to_branch.get_transaction())
248
self.count_weaves += 1
249
self.copied_file_ids.add(file_id)
250
self.file_ids_names[file_id] = to_weave.names()
251
mutter('copied file {%s}', file_id)
121
revision = from_branch.get_revision(rev_id)
122
except bzrlib.errors.NoSuchRevision:
123
if revision in from_history:
127
for parent in [p.revision_id for p in revision.parents]:
128
if not has_revision(to_branch, parent):
129
new_missing.add(parent)
130
missing = new_missing
131
return count, all_failed
134
def old_install_revisions(branch, other, revision_ids, pb):
135
"""Copy revisions from other branch into branch.
137
This is a lower-level function used by a pull or a merge. It
138
incorporates some history from one branch into another, but
139
does not update the revision history or operate on the working
143
Sequence of revisions to copy.
146
Progress bar for copying.
149
if hasattr(other.revision_store, "prefetch"):
150
other.revision_store.prefetch(revision_ids)
151
if hasattr(other.inventory_store, "prefetch"):
152
other.inventory_store.prefetch(revision_ids)
155
pb = bzrlib.ui.ui_factory.progress_bar()
162
for i, rev_id in enumerate(revision_ids):
163
pb.update('fetching revision', i+1, len(revision_ids))
165
rev = other.get_revision(rev_id)
166
except bzrlib.errors.NoSuchRevision:
170
revisions.append(rev)
171
inv = other.get_inventory(rev_id)
172
for key, entry in inv.iter_entries():
173
if entry.text_id is None:
175
if entry.text_id not in branch.text_store:
176
needed_texts.add(entry.text_id)
180
count, cp_fail = branch.text_store.copy_multi(other.text_store,
182
count, cp_fail = branch.inventory_store.copy_multi(other.inventory_store,
184
count, cp_fail = branch.revision_store.copy_multi(other.revision_store,
187
assert len(cp_fail) == 0
188
return count, failures
added
removed
bzrlib/fetch.py
