1
1
# Copyright (C) 2005 Canonical Ltd
3
3
# This program is free software; you can redistribute it and/or modify
4
4
# it under the terms of the GNU General Public License as published by
5
5
# the Free Software Foundation; either version 2 of the License, or
6
6
# (at your option) any later version.
8
8
# This program is distributed in the hope that it will be useful,
9
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
11
# GNU General Public License for more details.
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
18
# XXX: Can we do any better about making interrupted commits change
19
# nothing? Perhaps the best approach is to integrate commit of
20
# AtomicFiles with releasing the lock on the Branch.
22
# TODO: Separate 'prepare' phase where we find a list of potentially
23
# committed files. We then can then pause the commit to prompt for a
24
# commit message, knowing the summary will be the same as what's
25
# actually used for the commit. (But perhaps simpler to simply get
26
# the tree status, then use that for a selective commit?)
28
# The newly committed revision is going to have a shape corresponding
29
# to that of the working inventory. Files that are not in the
30
# working tree and that were in the predecessor are reported as
31
# removed --- this can include files that were either removed from the
32
# inventory or deleted in the working tree. If they were only
33
# deleted from disk, they are removed from the working inventory.
35
# We then consider the remaining entries, which will be in the new
36
# version. Directory entries are simply copied across. File entries
37
# must be checked to see if a new version of the file should be
38
# recorded. For each parent revision inventory, we check to see what
39
# version of the file was present. If the file was present in at
40
# least one tree, and if it was the same version in all the trees,
41
# then we can just refer to that version. Otherwise, a new version
42
# representing the merger of the file versions must be added.
44
# TODO: Update hashcache before and after - or does the WorkingTree
47
# TODO: Rather than mashing together the ancestry and storing it back,
48
# perhaps the weave should have single method which does it all in one
49
# go, avoiding a lot of redundant work.
51
# TODO: Perhaps give a warning if one of the revisions marked as
52
# merged is already in the ancestry, and then don't record it as a
55
# TODO: If the file is newly merged but unchanged from the version it
56
# merges from, then it should still be reported as newly added
57
# relative to the basis revision.
59
# TODO: Do checks that the tree can be committed *before* running the
60
# editor; this should include checks for a pointless commit and for
61
# unknown or missing files.
63
# TODO: If commit fails, leave the message in a file somewhere.
72
from binascii import hexlify
73
from cStringIO import StringIO
75
from bzrlib.atomicfile import AtomicFile
76
from bzrlib.osutils import (local_time_offset,
77
rand_bytes, compact_date,
78
kind_marker, is_inside_any, quotefn,
79
sha_file, isdir, isfile,
82
import bzrlib.errors as errors
83
from bzrlib.errors import (BzrError, PointlessCommit,
88
from bzrlib.revision import Revision
89
from bzrlib.testament import Testament
90
from bzrlib.trace import mutter, note, warning
91
from bzrlib.xml5 import serializer_v5
92
from bzrlib.inventory import Inventory, ROOT_ID
93
from bzrlib.symbol_versioning import *
94
from bzrlib.workingtree import WorkingTree
97
@deprecated_function(zero_seven)
98
def commit(*args, **kwargs):
99
"""Commit a new revision to a branch.
101
Function-style interface for convenience of old callers.
103
New code should use the Commit class instead.
105
## XXX: Remove this in favor of Branch.commit?
106
Commit().commit(*args, **kwargs)
109
class NullCommitReporter(object):
110
"""I report on progress of a commit."""
112
def snapshot_change(self, change, path):
115
def completed(self, revno, rev_id):
118
def deleted(self, file_id):
121
def escaped(self, escape_count, message):
124
def missing(self, path):
128
class ReportCommitToLog(NullCommitReporter):
130
def snapshot_change(self, change, path):
131
note("%s %s", change, path)
133
def completed(self, revno, rev_id):
134
note('Committed revision %d.', revno)
136
def deleted(self, file_id):
137
note('deleted %s', file_id)
139
def escaped(self, escape_count, message):
140
note("replaced %d control characters in message", escape_count)
142
def missing(self, path):
143
note('missing %s', path)
146
class Commit(object):
147
"""Task of committing a new revision.
149
This is a MethodObject: it accumulates state as the commit is
150
prepared, and then it is discarded. It doesn't represent
151
historical revisions, just the act of recording a new one.
154
Modified to hold a list of files that have been deleted from
155
the working directory; these should be removed from the
161
if reporter is not None:
162
self.reporter = reporter
164
self.reporter = NullCommitReporter()
165
if config is not None:
171
branch=DEPRECATED_PARAMETER, message=None,
177
allow_pointless=True,
185
"""Commit working copy as a new revision.
187
branch -- the deprecated branch to commit to. New callers should pass in
190
message -- the commit message, a mandatory parameter
192
timestamp -- if not None, seconds-since-epoch for a
193
postdated/predated commit.
195
specific_files -- If true, commit only those files.
197
rev_id -- If set, use this as the new revision id.
198
Useful for test or import commands that need to tightly
199
control what revisions are assigned. If you duplicate
200
a revision id that exists elsewhere it is your own fault.
201
If null (default), a time/random revision id is generated.
203
allow_pointless -- If true (default), commit even if nothing
204
has changed and no merges are recorded.
206
strict -- If true, don't allow a commit if the working tree
207
contains unknown files.
209
revprops -- Properties for new revision
210
:param local: Perform a local only commit.
212
mutter('preparing to commit')
214
if deprecated_passed(branch):
215
warn("Commit.commit (branch, ...): The branch parameter is "
216
"deprecated as of bzr 0.8. Please use working_tree= instead.",
217
DeprecationWarning, stacklevel=2)
219
self.work_tree = self.branch.bzrdir.open_workingtree()
220
elif working_tree is None:
221
raise BzrError("One of branch and working_tree must be passed into commit().")
223
self.work_tree = working_tree
224
self.branch = self.work_tree.branch
226
raise BzrError("The message keyword parameter is required for commit().")
228
self.weave_store = self.branch.repository.weave_store
229
self.bound_branch = None
231
self.master_branch = None
233
self.specific_files = specific_files
234
self.allow_pointless = allow_pointless
236
if revprops is not None:
237
self.revprops.update(revprops)
239
if reporter is None and self.reporter is None:
240
self.reporter = NullCommitReporter()
241
elif reporter is not None:
242
self.reporter = reporter
244
self.work_tree.lock_write()
246
# setup the bound branch variables as needed.
247
self._check_bound_branch()
249
# check for out of date working trees
250
# if we are bound, then self.branch is the master branch and this
251
# test is thus all we need.
252
if self.work_tree.last_revision() != self.master_branch.last_revision():
253
raise errors.OutOfDateTree(self.work_tree)
256
# raise an exception as soon as we find a single unknown.
257
for unknown in self.work_tree.unknowns():
258
raise StrictCommitFailed()
260
if timestamp is None:
261
self.timestamp = time.time()
263
self.timestamp = long(timestamp)
265
if self.config is None:
266
self.config = bzrlib.config.BranchConfig(self.branch)
269
self.rev_id = _gen_revision_id(self.config, self.timestamp)
273
if committer is None:
274
self.committer = self.config.username()
276
assert isinstance(committer, basestring), type(committer)
277
self.committer = committer
280
self.timezone = local_time_offset()
282
self.timezone = int(timezone)
284
if isinstance(message, str):
285
message = message.decode(bzrlib.user_encoding)
286
assert isinstance(message, unicode), type(message)
287
self.message = message
288
self._escape_commit_message()
290
self.work_inv = self.work_tree.inventory
291
self.basis_tree = self.work_tree.basis_tree()
292
self.basis_inv = self.basis_tree.inventory
294
self._gather_parents()
295
if len(self.parents) > 1 and self.specific_files:
296
raise NotImplementedError('selected-file commit of merges is not supported yet')
297
self._check_parents_present()
299
self._remove_deleted()
300
self._populate_new_inv()
301
self._store_snapshot()
302
self._report_deletes()
304
if not (self.allow_pointless
305
or len(self.parents) > 1
306
or self.new_inv != self.basis_inv):
307
raise PointlessCommit()
309
if len(list(self.work_tree.iter_conflicts()))>0:
310
raise ConflictsInTree
312
self.inv_sha1 = self.branch.repository.add_inventory(
317
self._make_revision()
318
# revision data is in the local branch now.
320
# upload revision data to the master.
321
# this will propogate merged revisions too if needed.
322
if self.bound_branch:
323
self.master_branch.repository.fetch(self.branch.repository,
324
revision_id=self.rev_id)
325
# now the master has the revision data
326
# 'commit' to the master first so a timeout here causes the local
327
# branch to be out of date
328
self.master_branch.append_revision(self.rev_id)
330
# and now do the commit locally.
331
self.branch.append_revision(self.rev_id)
333
self.work_tree.set_pending_merges([])
334
if len(self.parents):
335
precursor = self.parents[0]
338
self.work_tree.set_last_revision(self.rev_id, precursor)
339
# now the work tree is up to date with the branch
341
self.reporter.completed(self.branch.revno(), self.rev_id)
342
if self.config.post_commit() is not None:
343
hooks = self.config.post_commit().split(' ')
344
# this would be nicer with twisted.python.reflect.namedAny
346
result = eval(hook + '(branch, rev_id)',
347
{'branch':self.branch,
349
'rev_id':self.rev_id})
351
self._cleanup_bound_branch()
352
self.work_tree.unlock()
354
def _check_bound_branch(self):
355
"""Check to see if the local branch is bound.
357
If it is bound, then most of the commit will actually be
358
done using the remote branch as the target branch.
359
Only at the end will the local branch be updated.
361
if self.local and not self.branch.get_bound_location():
362
raise errors.LocalRequiresBoundBranch()
365
self.master_branch = self.branch.get_master_branch()
367
if not self.master_branch:
368
# make this branch the reference branch for out of date checks.
369
self.master_branch = self.branch
372
# If the master branch is bound, we must fail
373
master_bound_location = self.master_branch.get_bound_location()
374
if master_bound_location:
375
raise errors.CommitToDoubleBoundBranch(self.branch,
376
self.master_branch, master_bound_location)
378
# TODO: jam 20051230 We could automatically push local
379
# commits to the remote branch if they would fit.
380
# But for now, just require remote to be identical
383
# Make sure the local branch is identical to the master
384
master_rh = self.master_branch.revision_history()
385
local_rh = self.branch.revision_history()
386
if local_rh != master_rh:
387
raise errors.BoundBranchOutOfDate(self.branch,
390
# Now things are ready to change the master branch
392
self.bound_branch = self.branch
393
self.master_branch.lock_write()
395
#### # Check to see if we have any pending merges. If we do
396
#### # those need to be pushed into the master branch
397
#### pending_merges = self.work_tree.pending_merges()
398
#### if pending_merges:
399
#### for revision_id in pending_merges:
400
#### self.master_branch.repository.fetch(self.bound_branch.repository,
401
#### revision_id=revision_id)
403
def _cleanup_bound_branch(self):
404
"""Executed at the end of a try/finally to cleanup a bound branch.
406
If the branch wasn't bound, this is a no-op.
407
If it was, it resents self.branch to the local branch, instead
410
if not self.bound_branch:
412
self.master_branch.unlock()
414
def _escape_commit_message(self):
415
"""Replace xml-incompatible control characters."""
416
# Python strings can include characters that can't be
417
# represented in well-formed XML; escape characters that
418
# aren't listed in the XML specification
419
# (http://www.w3.org/TR/REC-xml/#NT-Char).
420
self.message, escape_count = re.subn(
421
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
422
lambda match: match.group(0).encode('unicode_escape'),
425
self.reporter.escaped(escape_count, self.message)
427
def _gather_parents(self):
428
"""Record the parents of a merge for merge detection."""
429
pending_merges = self.work_tree.pending_merges()
431
self.parent_invs = []
432
self.present_parents = []
433
precursor_id = self.branch.last_revision()
19
# FIXME: "bzr commit doc/format" commits doc/format.txt!
21
def commit(branch, message,
28
allow_pointless=True):
29
"""Commit working copy as a new revision.
31
The basic approach is to add all the file texts into the
32
store, then the inventory, then make a new revision pointing
33
to that inventory and store that.
35
This is not quite safe if the working copy changes during the
36
commit; for the moment that is simply not allowed. A better
37
approach is to make a temporary copy of the files before
38
computing their hashes, and then add those hashes in turn to
39
the inventory. This should mean at least that there are no
40
broken hash pointers. There is no way we can get a snapshot
41
of the whole directory at an instant. This would also have to
42
be robust against files disappearing, moving, etc. So the
43
whole thing is a bit hard.
45
This raises PointlessCommit if there are no changes, no new merges,
46
and allow_pointless is false.
48
timestamp -- if not None, seconds-since-epoch for a
49
postdated/predated commit.
52
If true, commit only those files.
55
If set, use this as the new revision id.
56
Useful for test or import commands that need to tightly
57
control what revisions are assigned. If you duplicate
58
a revision id that exists elsewhere it is your own fault.
59
If null (default), a time/random revision id is generated.
64
from bzrlib.osutils import local_time_offset, username
65
from bzrlib.branch import gen_file_id
66
from bzrlib.errors import BzrError, PointlessCommit
67
from bzrlib.revision import Revision, RevisionReference
68
from bzrlib.trace import mutter, note
69
from bzrlib.xml import pack_xml
74
# First walk over the working inventory; and both update that
75
# and also build a new revision inventory. The revision
76
# inventory needs to hold the text-id, sha1 and size of the
77
# actual file versions committed in the revision. (These are
78
# not present in the working inventory.) We also need to
79
# detect missing/deleted files, and remove them from the
82
work_tree = branch.working_tree()
83
work_inv = work_tree.inventory
84
basis = branch.basis_tree()
85
basis_inv = basis.inventory
88
note('looking for changes...')
90
pending_merges = branch.pending_merges()
92
missing_ids, new_inv, any_changes = \
93
_gather_commit(branch,
100
if not (any_changes or allow_pointless or pending_merges):
101
raise PointlessCommit()
103
for file_id in missing_ids:
104
# Any files that have been deleted are now removed from the
105
# working inventory. Files that were not selected for commit
106
# are left as they were in the working inventory and ommitted
107
# from the revision inventory.
109
# have to do this later so we don't mess up the iterator.
110
# since parents may be removed before their children we
113
# FIXME: There's probably a better way to do this; perhaps
114
# the workingtree should know how to filter itbranch.
115
if work_inv.has_id(file_id):
116
del work_inv[file_id]
120
rev_id = _gen_revision_id(time.time())
123
inv_tmp = tempfile.TemporaryFile()
124
pack_xml(new_inv, inv_tmp)
126
branch.inventory_store.add(inv_tmp, inv_id)
127
mutter('new inventory_id is {%s}' % inv_id)
129
# We could also just sha hash the inv_tmp file
130
# however, in the case that branch.inventory_store.add()
131
# ever actually does anything special
132
inv_sha1 = branch.get_inventory_sha1(inv_id)
134
branch._write_inventory(work_inv)
136
if timestamp == None:
137
timestamp = time.time()
139
if committer == None:
140
committer = username()
143
timezone = local_time_offset()
145
mutter("building commit log message")
146
rev = Revision(timestamp=timestamp,
151
inventory_sha1=inv_sha1,
155
precursor_id = branch.last_patch()
435
self.parents.append(precursor_id)
436
self.parents += pending_merges
437
for revision in self.parents:
438
if self.branch.repository.has_revision(revision):
439
inventory = self.branch.repository.get_inventory(revision)
440
self.parent_invs.append(inventory)
441
self.present_parents.append(revision)
443
def _check_parents_present(self):
444
for parent_id in self.parents:
445
mutter('commit parent revision {%s}', parent_id)
446
if not self.branch.repository.has_revision(parent_id):
447
if parent_id == self.branch.last_revision():
448
warning("parent is missing %r", parent_id)
449
raise HistoryMissing(self.branch, 'revision', parent_id)
451
mutter("commit will ghost revision %r", parent_id)
453
def _make_revision(self):
454
"""Record a new revision object for this commit."""
455
rev = Revision(timestamp=self.timestamp,
456
timezone=self.timezone,
457
committer=self.committer,
458
message=self.message,
459
inventory_sha1=self.inv_sha1,
460
revision_id=self.rev_id,
461
properties=self.revprops)
462
rev.parent_ids = self.parents
463
self.branch.repository.add_revision(self.rev_id, rev, self.new_inv, self.config)
465
def _remove_deleted(self):
466
"""Remove deleted files from the working inventories.
468
This is done prior to taking the working inventory as the
469
basis for the new committed inventory.
471
This returns true if any files
472
*that existed in the basis inventory* were deleted.
473
Files that were added and deleted
474
in the working copy don't matter.
476
specific = self.specific_files
478
for path, ie in self.work_inv.iter_entries():
479
if specific and not is_inside_any(specific, path):
481
if not self.work_tree.has_filename(path):
482
self.reporter.missing(path)
483
deleted_ids.append((path, ie.file_id))
485
deleted_ids.sort(reverse=True)
486
for path, file_id in deleted_ids:
487
del self.work_inv[file_id]
488
self.work_tree._write_inventory(self.work_inv)
490
def _store_snapshot(self):
491
"""Pass over inventory and record a snapshot.
493
Entries get a new revision when they are modified in
494
any way, which includes a merge with a new set of
495
parents that have the same entry.
497
# XXX: Need to think more here about when the user has
498
# made a specific decision on a particular value -- c.f.
500
for path, ie in self.new_inv.iter_entries():
501
previous_entries = ie.find_previous_heads(
504
self.branch.repository.get_transaction())
505
if ie.revision is None:
506
change = ie.snapshot(self.rev_id, path, previous_entries,
507
self.work_tree, self.weave_store,
508
self.branch.get_transaction())
511
self.reporter.snapshot_change(change, path)
513
def _populate_new_inv(self):
514
"""Build revision inventory.
516
This creates a new empty inventory. Depending on
517
which files are selected for commit, and what is present in the
518
current tree, the new inventory is populated. inventory entries
519
which are candidates for modification have their revision set to
520
None; inventory entries that are carried over untouched have their
521
revision set to their prior value.
523
mutter("Selecting files for commit with filter %s", self.specific_files)
524
self.new_inv = Inventory()
525
for path, new_ie in self.work_inv.iter_entries():
526
file_id = new_ie.file_id
527
mutter('check %s {%s}', path, new_ie.file_id)
528
if self.specific_files:
529
if not is_inside_any(self.specific_files, path):
530
mutter('%s not selected for commit', path)
531
self._carry_entry(file_id)
534
# this is selected, ensure its parents are too.
535
parent_id = new_ie.parent_id
536
while parent_id != ROOT_ID:
537
if not self.new_inv.has_id(parent_id):
538
ie = self._select_entry(self.work_inv[parent_id])
539
mutter('%s selected for commit because of %s',
540
self.new_inv.id2path(parent_id), path)
542
ie = self.new_inv[parent_id]
543
if ie.revision is not None:
545
mutter('%s selected for commit because of %s',
546
self.new_inv.id2path(parent_id), path)
547
parent_id = ie.parent_id
548
mutter('%s selected for commit', path)
549
self._select_entry(new_ie)
551
def _select_entry(self, new_ie):
552
"""Make new_ie be considered for committing."""
558
def _carry_entry(self, file_id):
559
"""Carry the file unchanged from the basis revision."""
560
if self.basis_inv.has_id(file_id):
561
self.new_inv.add(self.basis_inv[file_id].copy())
563
def _report_deletes(self):
564
for file_id in self.basis_inv:
565
if file_id not in self.new_inv:
566
self.reporter.deleted(self.basis_inv.id2path(file_id))
568
def _gen_revision_id(config, when):
157
precursor_sha1 = branch.get_revision_sha1(precursor_id)
158
rev.parents.append(RevisionReference(precursor_id, precursor_sha1))
159
for merge_rev in pending_merges:
160
rev.parents.append(RevisionReference(merge_rev))
162
rev_tmp = tempfile.TemporaryFile()
163
pack_xml(rev, rev_tmp)
165
branch.revision_store.add(rev_tmp, rev_id)
166
mutter("new revision_id is {%s}" % rev_id)
168
## XXX: Everything up to here can simply be orphaned if we abort
169
## the commit; it will leave junk files behind but that doesn't
172
## TODO: Read back the just-generated changeset, and make sure it
173
## applies and recreates the right state.
175
## TODO: Also calculate and store the inventory SHA1
176
mutter("committing patch r%d" % (branch.revno() + 1))
178
branch.append_revision(rev_id)
180
branch.set_pending_merges([])
183
note("commited r%d" % branch.revno())
189
def _gen_revision_id(when):
569
190
"""Return new revision-id."""
570
s = '%s-%s-' % (config.user_email(), compact_date(when))
191
from binascii import hexlify
192
from osutils import rand_bytes, compact_date, user_email
194
s = '%s-%s-' % (user_email(), compact_date(when))
571
195
s += hexlify(rand_bytes(8))
199
def _gather_commit(branch, work_tree, work_inv, basis_inv, specific_files,
201
"""Build inventory preparatory to commit.
203
Returns missing_ids, new_inv, any_changes.
205
This adds any changed files into the text store, and sets their
206
test-id, sha and size in the returned inventory appropriately.
209
Modified to hold a list of files that have been deleted from
210
the working directory; these should be removed from the
213
from bzrlib.inventory import Inventory
214
from bzrlib.osutils import isdir, isfile, sha_string, quotefn, \
215
local_time_offset, username, kind_marker, is_inside_any
217
from bzrlib.branch import gen_file_id
218
from bzrlib.errors import BzrError
219
from bzrlib.revision import Revision
220
from bzrlib.trace import mutter, note
223
inv = Inventory(work_inv.root.file_id)
226
for path, entry in work_inv.iter_entries():
227
## TODO: Check that the file kind has not changed from the previous
228
## revision of this file (if any).
230
p = branch.abspath(path)
231
file_id = entry.file_id
232
mutter('commit prep file %s, id %r ' % (p, file_id))
234
if specific_files and not is_inside_any(specific_files, path):
235
mutter(' skipping file excluded from commit')
236
if basis_inv.has_id(file_id):
237
# carry over with previous state
238
inv.add(basis_inv[file_id].copy())
240
# omit this from committed inventory
244
if not work_tree.has_id(file_id):
246
print('deleted %s%s' % (path, kind_marker(entry.kind)))
248
mutter(" file is missing, removing from inventory")
249
missing_ids.append(file_id)
252
# this is present in the new inventory; may be new, modified or
254
old_ie = basis_inv.has_id(file_id) and basis_inv[file_id]
260
old_kind = old_ie.kind
261
if old_kind != entry.kind:
262
raise BzrError("entry %r changed kind from %r to %r"
263
% (file_id, old_kind, entry.kind))
265
if entry.kind == 'directory':
267
raise BzrError("%s is entered as directory but not a directory"
269
elif entry.kind == 'file':
271
raise BzrError("%s is entered as file but is not a file" % quotefn(p))
273
new_sha1 = work_tree.get_file_sha1(file_id)
276
and old_ie.text_sha1 == new_sha1):
277
## assert content == basis.get_file(file_id).read()
278
entry.text_id = old_ie.text_id
279
entry.text_sha1 = new_sha1
280
entry.text_size = old_ie.text_size
281
mutter(' unchanged from previous text_id {%s}' %
284
content = file(p, 'rb').read()
286
# calculate the sha again, just in case the file contents
287
# changed since we updated the cache
288
entry.text_sha1 = sha_string(content)
289
entry.text_size = len(content)
291
entry.text_id = gen_file_id(entry.name)
292
branch.text_store.add(content, entry.text_id)
293
mutter(' stored with text_id {%s}' % entry.text_id)
296
marked = path + kind_marker(entry.kind)
298
print 'added', marked
300
elif old_ie == entry:
302
elif (old_ie.name == entry.name
303
and old_ie.parent_id == entry.parent_id):
304
print 'modified', marked
307
print 'renamed', marked
310
return missing_ids, inv, any_changes
added
removed
bzrlib/commit.py
