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
if change == 'unchanged':
133
note("%s %s", change, path)
135
def completed(self, revno, rev_id):
136
note('Committed revision %d.', revno)
138
def deleted(self, file_id):
139
note('deleted %s', file_id)
141
def escaped(self, escape_count, message):
142
note("replaced %d control characters in message", escape_count)
144
def missing(self, path):
145
note('missing %s', path)
148
class Commit(object):
149
"""Task of committing a new revision.
151
This is a MethodObject: it accumulates state as the commit is
152
prepared, and then it is discarded. It doesn't represent
153
historical revisions, just the act of recording a new one.
156
Modified to hold a list of files that have been deleted from
157
the working directory; these should be removed from the
163
if reporter is not None:
164
self.reporter = reporter
166
self.reporter = NullCommitReporter()
167
if config is not None:
173
branch=DEPRECATED_PARAMETER, message=None,
179
allow_pointless=True,
187
"""Commit working copy as a new revision.
189
branch -- the deprecated branch to commit to. New callers should pass in
192
message -- the commit message, a mandatory parameter
194
timestamp -- if not None, seconds-since-epoch for a
195
postdated/predated commit.
197
specific_files -- If true, commit only those files.
199
rev_id -- If set, use this as the new revision id.
200
Useful for test or import commands that need to tightly
201
control what revisions are assigned. If you duplicate
202
a revision id that exists elsewhere it is your own fault.
203
If null (default), a time/random revision id is generated.
205
allow_pointless -- If true (default), commit even if nothing
206
has changed and no merges are recorded.
208
strict -- If true, don't allow a commit if the working tree
209
contains unknown files.
211
revprops -- Properties for new revision
212
:param local: Perform a local only commit.
214
mutter('preparing to commit')
216
if deprecated_passed(branch):
217
warn("Commit.commit (branch, ...): The branch parameter is "
218
"deprecated as of bzr 0.8. Please use working_tree= instead.",
219
DeprecationWarning, stacklevel=2)
221
self.work_tree = self.branch.bzrdir.open_workingtree()
222
elif working_tree is None:
223
raise BzrError("One of branch and working_tree must be passed into commit().")
225
self.work_tree = working_tree
226
self.branch = self.work_tree.branch
228
raise BzrError("The message keyword parameter is required for commit().")
230
self.weave_store = self.branch.repository.weave_store
231
self.bound_branch = None
233
self.master_branch = None
234
self.master_locked = False
236
self.specific_files = specific_files
237
self.allow_pointless = allow_pointless
239
if revprops is not None:
240
self.revprops.update(revprops)
242
if reporter is None and self.reporter is None:
243
self.reporter = NullCommitReporter()
244
elif reporter is not None:
245
self.reporter = reporter
247
self.work_tree.lock_write()
249
# setup the bound branch variables as needed.
250
self._check_bound_branch()
252
# check for out of date working trees
253
# if we are bound, then self.branch is the master branch and this
254
# test is thus all we need.
255
if self.work_tree.last_revision() != self.master_branch.last_revision():
256
raise errors.OutOfDateTree(self.work_tree)
259
# raise an exception as soon as we find a single unknown.
260
for unknown in self.work_tree.unknowns():
261
raise StrictCommitFailed()
263
if timestamp is None:
264
self.timestamp = time.time()
266
self.timestamp = long(timestamp)
268
if self.config is None:
269
self.config = bzrlib.config.BranchConfig(self.branch)
272
self.rev_id = _gen_revision_id(self.config, self.timestamp)
276
if committer is None:
277
self.committer = self.config.username()
279
assert isinstance(committer, basestring), type(committer)
280
self.committer = committer
283
self.timezone = local_time_offset()
285
self.timezone = int(timezone)
287
if isinstance(message, str):
288
message = message.decode(bzrlib.user_encoding)
289
assert isinstance(message, unicode), type(message)
290
self.message = message
291
self._escape_commit_message()
293
self.work_inv = self.work_tree.inventory
294
self.basis_tree = self.work_tree.basis_tree()
295
self.basis_inv = self.basis_tree.inventory
297
self._gather_parents()
298
if len(self.parents) > 1 and self.specific_files:
299
raise NotImplementedError('selected-file commit of merges is not supported yet')
300
self._check_parents_present()
302
self._remove_deleted()
303
self._populate_new_inv()
304
self._store_snapshot()
305
self._report_deletes()
307
if not (self.allow_pointless
308
or len(self.parents) > 1
309
or self.new_inv != self.basis_inv):
310
raise PointlessCommit()
312
if len(list(self.work_tree.iter_conflicts()))>0:
313
raise ConflictsInTree
315
self.inv_sha1 = self.branch.repository.add_inventory(
320
self._make_revision()
321
# revision data is in the local branch now.
323
# upload revision data to the master.
324
# this will propogate merged revisions too if needed.
325
if self.bound_branch:
326
self.master_branch.repository.fetch(self.branch.repository,
327
revision_id=self.rev_id)
328
# now the master has the revision data
329
# 'commit' to the master first so a timeout here causes the local
330
# branch to be out of date
331
self.master_branch.append_revision(self.rev_id)
333
# and now do the commit locally.
334
self.branch.append_revision(self.rev_id)
336
self.work_tree.set_pending_merges([])
337
self.work_tree.set_last_revision(self.rev_id)
338
# now the work tree is up to date with the branch
340
self.reporter.completed(self.branch.revno(), self.rev_id)
341
if self.config.post_commit() is not None:
342
hooks = self.config.post_commit().split(' ')
343
# this would be nicer with twisted.python.reflect.namedAny
345
result = eval(hook + '(branch, rev_id)',
346
{'branch':self.branch,
348
'rev_id':self.rev_id})
350
self._cleanup_bound_branch()
351
self.work_tree.unlock()
353
def _check_bound_branch(self):
354
"""Check to see if the local branch is bound.
356
If it is bound, then most of the commit will actually be
357
done using the remote branch as the target branch.
358
Only at the end will the local branch be updated.
360
if self.local and not self.branch.get_bound_location():
361
raise errors.LocalRequiresBoundBranch()
364
self.master_branch = self.branch.get_master_branch()
366
if not self.master_branch:
367
# make this branch the reference branch for out of date checks.
368
self.master_branch = self.branch
371
# If the master branch is bound, we must fail
372
master_bound_location = self.master_branch.get_bound_location()
373
if master_bound_location:
374
raise errors.CommitToDoubleBoundBranch(self.branch,
375
self.master_branch, master_bound_location)
377
# TODO: jam 20051230 We could automatically push local
378
# commits to the remote branch if they would fit.
379
# But for now, just require remote to be identical
382
# Make sure the local branch is identical to the master
383
master_rh = self.master_branch.revision_history()
384
local_rh = self.branch.revision_history()
385
if local_rh != master_rh:
386
raise errors.BoundBranchOutOfDate(self.branch,
389
# Now things are ready to change the master branch
391
self.bound_branch = self.branch
392
self.master_branch.lock_write()
393
self.master_locked = True
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
if self.master_locked:
413
self.master_branch.unlock()
415
def _escape_commit_message(self):
416
"""Replace xml-incompatible control characters."""
417
# Python strings can include characters that can't be
418
# represented in well-formed XML; escape characters that
419
# aren't listed in the XML specification
420
# (http://www.w3.org/TR/REC-xml/#NT-Char).
421
self.message, escape_count = re.subn(
422
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
423
lambda match: match.group(0).encode('unicode_escape'),
426
self.reporter.escaped(escape_count, self.message)
428
def _gather_parents(self):
429
"""Record the parents of a merge for merge detection."""
430
pending_merges = self.work_tree.pending_merges()
432
self.parent_invs = []
433
self.present_parents = []
434
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]
119
rev_id = _gen_revision_id(branch, time.time())
122
inv_tmp = tempfile.TemporaryFile()
123
pack_xml(new_inv, inv_tmp)
125
branch.inventory_store.add(inv_tmp, inv_id)
126
mutter('new inventory_id is {%s}' % inv_id)
128
# We could also just sha hash the inv_tmp file
129
# however, in the case that branch.inventory_store.add()
130
# ever actually does anything special
131
inv_sha1 = branch.get_inventory_sha1(inv_id)
133
branch._write_inventory(work_inv)
135
if timestamp == None:
136
timestamp = time.time()
138
if committer == None:
139
committer = username(branch)
142
timezone = local_time_offset()
144
mutter("building commit log message")
145
rev = Revision(timestamp=timestamp,
150
inventory_sha1=inv_sha1,
154
precursor_id = branch.last_patch()
436
self.parents.append(precursor_id)
437
self.parents += pending_merges
438
for revision in self.parents:
439
if self.branch.repository.has_revision(revision):
440
inventory = self.branch.repository.get_inventory(revision)
441
self.parent_invs.append(inventory)
442
self.present_parents.append(revision)
444
def _check_parents_present(self):
445
for parent_id in self.parents:
446
mutter('commit parent revision {%s}', parent_id)
447
if not self.branch.repository.has_revision(parent_id):
448
if parent_id == self.branch.last_revision():
449
warning("parent is missing %r", parent_id)
450
raise HistoryMissing(self.branch, 'revision', parent_id)
452
mutter("commit will ghost revision %r", parent_id)
454
def _make_revision(self):
455
"""Record a new revision object for this commit."""
456
rev = Revision(timestamp=self.timestamp,
457
timezone=self.timezone,
458
committer=self.committer,
459
message=self.message,
460
inventory_sha1=self.inv_sha1,
461
revision_id=self.rev_id,
462
properties=self.revprops)
463
rev.parent_ids = self.parents
464
self.branch.repository.add_revision(self.rev_id, rev, self.new_inv, self.config)
466
def _remove_deleted(self):
467
"""Remove deleted files from the working inventories.
469
This is done prior to taking the working inventory as the
470
basis for the new committed inventory.
472
This returns true if any files
473
*that existed in the basis inventory* were deleted.
474
Files that were added and deleted
475
in the working copy don't matter.
477
specific = self.specific_files
479
for path, ie in self.work_inv.iter_entries():
480
if specific and not is_inside_any(specific, path):
482
if not self.work_tree.has_filename(path):
483
self.reporter.missing(path)
484
deleted_ids.append((path, ie.file_id))
486
deleted_ids.sort(reverse=True)
487
for path, file_id in deleted_ids:
488
del self.work_inv[file_id]
489
self.work_tree._write_inventory(self.work_inv)
491
def _store_snapshot(self):
492
"""Pass over inventory and record a snapshot.
494
Entries get a new revision when they are modified in
495
any way, which includes a merge with a new set of
496
parents that have the same entry.
498
# XXX: Need to think more here about when the user has
499
# made a specific decision on a particular value -- c.f.
501
for path, ie in self.new_inv.iter_entries():
502
previous_entries = ie.find_previous_heads(
505
self.branch.repository.get_transaction())
506
if ie.revision is None:
507
change = ie.snapshot(self.rev_id, path, previous_entries,
508
self.work_tree, self.weave_store,
509
self.branch.get_transaction())
512
self.reporter.snapshot_change(change, path)
514
def _populate_new_inv(self):
515
"""Build revision inventory.
517
This creates a new empty inventory. Depending on
518
which files are selected for commit, and what is present in the
519
current tree, the new inventory is populated. inventory entries
520
which are candidates for modification have their revision set to
521
None; inventory entries that are carried over untouched have their
522
revision set to their prior value.
524
mutter("Selecting files for commit with filter %s", self.specific_files)
525
self.new_inv = Inventory(revision_id=self.rev_id)
526
for path, new_ie in self.work_inv.iter_entries():
527
file_id = new_ie.file_id
528
mutter('check %s {%s}', path, new_ie.file_id)
529
if self.specific_files:
530
if not is_inside_any(self.specific_files, path):
531
mutter('%s not selected for commit', path)
532
self._carry_entry(file_id)
535
# this is selected, ensure its parents are too.
536
parent_id = new_ie.parent_id
537
while parent_id != ROOT_ID:
538
if not self.new_inv.has_id(parent_id):
539
ie = self._select_entry(self.work_inv[parent_id])
540
mutter('%s selected for commit because of %s',
541
self.new_inv.id2path(parent_id), path)
543
ie = self.new_inv[parent_id]
544
if ie.revision is not None:
546
mutter('%s selected for commit because of %s',
547
self.new_inv.id2path(parent_id), path)
548
parent_id = ie.parent_id
549
mutter('%s selected for commit', path)
550
self._select_entry(new_ie)
552
def _select_entry(self, new_ie):
553
"""Make new_ie be considered for committing."""
559
def _carry_entry(self, file_id):
560
"""Carry the file unchanged from the basis revision."""
561
if self.basis_inv.has_id(file_id):
562
self.new_inv.add(self.basis_inv[file_id].copy())
564
def _report_deletes(self):
565
for file_id in self.basis_inv:
566
if file_id not in self.new_inv:
567
self.reporter.deleted(self.basis_inv.id2path(file_id))
569
def _gen_revision_id(config, when):
156
precursor_sha1 = branch.get_revision_sha1(precursor_id)
157
rev.parents.append(RevisionReference(precursor_id, precursor_sha1))
158
for merge_rev in pending_merges:
159
rev.parents.append(RevisionReference(merge_rev))
161
rev_tmp = tempfile.TemporaryFile()
162
pack_xml(rev, rev_tmp)
164
branch.revision_store.add(rev_tmp, rev_id)
165
mutter("new revision_id is {%s}" % rev_id)
167
## XXX: Everything up to here can simply be orphaned if we abort
168
## the commit; it will leave junk files behind but that doesn't
171
## TODO: Read back the just-generated changeset, and make sure it
172
## applies and recreates the right state.
174
## TODO: Also calculate and store the inventory SHA1
175
mutter("committing patch r%d" % (branch.revno() + 1))
177
branch.append_revision(rev_id)
179
branch.set_pending_merges([])
182
note("commited r%d" % branch.revno())
188
def _gen_revision_id(branch, when):
570
189
"""Return new revision-id."""
571
s = '%s-%s-' % (config.user_email(), compact_date(when))
190
from binascii import hexlify
191
from bzrlib.osutils import rand_bytes, compact_date, user_email
193
s = '%s-%s-' % (user_email(branch), compact_date(when))
572
194
s += hexlify(rand_bytes(8))
198
def _gather_commit(branch, work_tree, work_inv, basis_inv, specific_files,
200
"""Build inventory preparatory to commit.
202
Returns missing_ids, new_inv, any_changes.
204
This adds any changed files into the text store, and sets their
205
test-id, sha and size in the returned inventory appropriately.
208
Modified to hold a list of files that have been deleted from
209
the working directory; these should be removed from the
212
from bzrlib.inventory import Inventory
213
from bzrlib.osutils import isdir, isfile, sha_string, quotefn, \
214
local_time_offset, username, kind_marker, is_inside_any
216
from bzrlib.branch import gen_file_id
217
from bzrlib.errors import BzrError
218
from bzrlib.revision import Revision
219
from bzrlib.trace import mutter, note
222
inv = Inventory(work_inv.root.file_id)
225
for path, entry in work_inv.iter_entries():
226
## TODO: Check that the file kind has not changed from the previous
227
## revision of this file (if any).
229
p = branch.abspath(path)
230
file_id = entry.file_id
231
mutter('commit prep file %s, id %r ' % (p, file_id))
233
if specific_files and not is_inside_any(specific_files, path):
234
mutter(' skipping file excluded from commit')
235
if basis_inv.has_id(file_id):
236
# carry over with previous state
237
inv.add(basis_inv[file_id].copy())
239
# omit this from committed inventory
243
if not work_tree.has_id(file_id):
245
print('deleted %s%s' % (path, kind_marker(entry.kind)))
247
mutter(" file is missing, removing from inventory")
248
missing_ids.append(file_id)
251
# this is present in the new inventory; may be new, modified or
253
old_ie = basis_inv.has_id(file_id) and basis_inv[file_id]
259
old_kind = old_ie.kind
260
if old_kind != entry.kind:
261
raise BzrError("entry %r changed kind from %r to %r"
262
% (file_id, old_kind, entry.kind))
264
if entry.kind == 'directory':
266
raise BzrError("%s is entered as directory but not a directory"
268
elif entry.kind == 'file':
270
raise BzrError("%s is entered as file but is not a file" % quotefn(p))
272
new_sha1 = work_tree.get_file_sha1(file_id)
275
and old_ie.text_sha1 == new_sha1):
276
## assert content == basis.get_file(file_id).read()
277
entry.text_id = old_ie.text_id
278
entry.text_sha1 = new_sha1
279
entry.text_size = old_ie.text_size
280
mutter(' unchanged from previous text_id {%s}' %
283
content = file(p, 'rb').read()
285
# calculate the sha again, just in case the file contents
286
# changed since we updated the cache
287
entry.text_sha1 = sha_string(content)
288
entry.text_size = len(content)
290
entry.text_id = gen_file_id(entry.name)
291
branch.text_store.add(content, entry.text_id)
292
mutter(' stored with text_id {%s}' % entry.text_id)
295
marked = path + kind_marker(entry.kind)
297
print 'added', marked
299
elif old_ie == entry:
301
elif (old_ie.name == entry.name
302
and old_ie.parent_id == entry.parent_id):
303
print 'modified', marked
306
print 'renamed', marked
309
return missing_ids, inv, any_changes
added
removed
bzrlib/commit.py
