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
19
def commit(branch, message,
26
"""Commit working copy as a new revision.
28
The basic approach is to add all the file texts into the
29
store, then the inventory, then make a new revision pointing
30
to that inventory and store that.
32
This is not quite safe if the working copy changes during the
33
commit; for the moment that is simply not allowed. A better
34
approach is to make a temporary copy of the files before
35
computing their hashes, and then add those hashes in turn to
36
the inventory. This should mean at least that there are no
37
broken hash pointers. There is no way we can get a snapshot
38
of the whole directory at an instant. This would also have to
39
be robust against files disappearing, moving, etc. So the
40
whole thing is a bit hard.
42
timestamp -- if not None, seconds-since-epoch for a
43
postdated/predated commit.
46
If true, commit only those files.
49
If set, use this as the new revision id.
50
Useful for test or import commands that need to tightly
51
control what revisions are assigned. If you duplicate
52
a revision id that exists elsewhere it is your own fault.
53
If null (default), a time/random revision id is generated.
58
from bzrlib.osutils import local_time_offset, username
59
from bzrlib.branch import gen_file_id
60
from bzrlib.errors import BzrError
61
from bzrlib.revision import Revision, RevisionReference
62
from bzrlib.trace import mutter, note
67
# First walk over the working inventory; and both update that
68
# and also build a new revision inventory. The revision
69
# inventory needs to hold the text-id, sha1 and size of the
70
# actual file versions committed in the revision. (These are
71
# not present in the working inventory.) We also need to
72
# detect missing/deleted files, and remove them from the
75
work_tree = branch.working_tree()
76
work_inv = work_tree.inventory
77
basis = branch.basis_tree()
78
basis_inv = basis.inventory
81
note('looking for changes...')
83
missing_ids, new_inv = _gather_commit(branch,
90
for file_id in missing_ids:
91
# Any files that have been deleted are now removed from the
92
# working inventory. Files that were not selected for commit
93
# are left as they were in the working inventory and ommitted
94
# from the revision inventory.
96
# have to do this later so we don't mess up the iterator.
97
# since parents may be removed before their children we
100
# FIXME: There's probably a better way to do this; perhaps
101
# the workingtree should know how to filter itbranch.
102
if work_inv.has_id(file_id):
103
del work_inv[file_id]
107
rev_id = _gen_revision_id(time.time())
110
inv_tmp = tempfile.TemporaryFile()
111
new_inv.write_xml(inv_tmp)
113
branch.inventory_store.add(inv_tmp, inv_id)
114
mutter('new inventory_id is {%s}' % inv_id)
116
# We could also just sha hash the inv_tmp file
117
# however, in the case that branch.inventory_store.add()
118
# ever actually does anything special
119
inv_sha1 = branch.get_inventory_sha1(inv_id)
121
branch._write_inventory(work_inv)
123
if timestamp == None:
124
timestamp = time.time()
126
if committer == None:
127
committer = username()
130
timezone = local_time_offset()
132
mutter("building commit log message")
133
rev = Revision(timestamp=timestamp,
138
inventory_sha1=inv_sha1,
141
precursor_id = branch.last_patch()
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
if len(self.parents):
338
precursor = self.parents[0]
341
self.work_tree.set_last_revision(self.rev_id, precursor)
342
# now the work tree is up to date with the branch
344
self.reporter.completed(self.branch.revno(), self.rev_id)
345
if self.config.post_commit() is not None:
346
hooks = self.config.post_commit().split(' ')
347
# this would be nicer with twisted.python.reflect.namedAny
349
result = eval(hook + '(branch, rev_id)',
350
{'branch':self.branch,
352
'rev_id':self.rev_id})
354
self._cleanup_bound_branch()
355
self.work_tree.unlock()
357
def _check_bound_branch(self):
358
"""Check to see if the local branch is bound.
360
If it is bound, then most of the commit will actually be
361
done using the remote branch as the target branch.
362
Only at the end will the local branch be updated.
364
if self.local and not self.branch.get_bound_location():
365
raise errors.LocalRequiresBoundBranch()
368
self.master_branch = self.branch.get_master_branch()
370
if not self.master_branch:
371
# make this branch the reference branch for out of date checks.
372
self.master_branch = self.branch
375
# If the master branch is bound, we must fail
376
master_bound_location = self.master_branch.get_bound_location()
377
if master_bound_location:
378
raise errors.CommitToDoubleBoundBranch(self.branch,
379
self.master_branch, master_bound_location)
381
# TODO: jam 20051230 We could automatically push local
382
# commits to the remote branch if they would fit.
383
# But for now, just require remote to be identical
386
# Make sure the local branch is identical to the master
387
master_rh = self.master_branch.revision_history()
388
local_rh = self.branch.revision_history()
389
if local_rh != master_rh:
390
raise errors.BoundBranchOutOfDate(self.branch,
393
# Now things are ready to change the master branch
395
self.bound_branch = self.branch
396
self.master_branch.lock_write()
397
self.master_locked = True
399
#### # Check to see if we have any pending merges. If we do
400
#### # those need to be pushed into the master branch
401
#### pending_merges = self.work_tree.pending_merges()
402
#### if pending_merges:
403
#### for revision_id in pending_merges:
404
#### self.master_branch.repository.fetch(self.bound_branch.repository,
405
#### revision_id=revision_id)
407
def _cleanup_bound_branch(self):
408
"""Executed at the end of a try/finally to cleanup a bound branch.
410
If the branch wasn't bound, this is a no-op.
411
If it was, it resents self.branch to the local branch, instead
414
if not self.bound_branch:
416
if self.master_locked:
417
self.master_branch.unlock()
419
def _escape_commit_message(self):
420
"""Replace xml-incompatible control characters."""
421
# Python strings can include characters that can't be
422
# represented in well-formed XML; escape characters that
423
# aren't listed in the XML specification
424
# (http://www.w3.org/TR/REC-xml/#NT-Char).
425
self.message, escape_count = re.subn(
426
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
427
lambda match: match.group(0).encode('unicode_escape'),
430
self.reporter.escaped(escape_count, self.message)
432
def _gather_parents(self):
433
"""Record the parents of a merge for merge detection."""
434
pending_merges = self.work_tree.pending_merges()
436
self.parent_invs = []
437
self.present_parents = []
438
precursor_id = self.branch.last_revision()
143
precursor_sha1 = branch.get_revision_sha1(precursor_id)
144
rev.parents = [RevisionReference(precursor_id, precursor_sha1)]
146
rev_tmp = tempfile.TemporaryFile()
147
rev.write_xml(rev_tmp)
149
branch.revision_store.add(rev_tmp, rev_id)
150
mutter("new revision_id is {%s}" % rev_id)
152
## XXX: Everything up to here can simply be orphaned if we abort
153
## the commit; it will leave junk files behind but that doesn't
156
## TODO: Read back the just-generated changeset, and make sure it
157
## applies and recreates the right state.
159
## TODO: Also calculate and store the inventory SHA1
160
mutter("committing patch r%d" % (branch.revno() + 1))
162
branch.append_revision(rev_id)
165
note("commited r%d" % branch.revno())
171
def _gen_revision_id(when):
440
self.parents.append(precursor_id)
441
self.parents += pending_merges
442
for revision in self.parents:
443
if self.branch.repository.has_revision(revision):
444
inventory = self.branch.repository.get_inventory(revision)
445
self.parent_invs.append(inventory)
446
self.present_parents.append(revision)
448
def _check_parents_present(self):
449
for parent_id in self.parents:
450
mutter('commit parent revision {%s}', parent_id)
451
if not self.branch.repository.has_revision(parent_id):
452
if parent_id == self.branch.last_revision():
453
warning("parent is missing %r", parent_id)
454
raise HistoryMissing(self.branch, 'revision', parent_id)
456
mutter("commit will ghost revision %r", parent_id)
458
def _make_revision(self):
459
"""Record a new revision object for this commit."""
460
rev = Revision(timestamp=self.timestamp,
461
timezone=self.timezone,
462
committer=self.committer,
463
message=self.message,
464
inventory_sha1=self.inv_sha1,
465
revision_id=self.rev_id,
466
properties=self.revprops)
467
rev.parent_ids = self.parents
468
self.branch.repository.add_revision(self.rev_id, rev, self.new_inv, self.config)
470
def _remove_deleted(self):
471
"""Remove deleted files from the working inventories.
473
This is done prior to taking the working inventory as the
474
basis for the new committed inventory.
476
This returns true if any files
477
*that existed in the basis inventory* were deleted.
478
Files that were added and deleted
479
in the working copy don't matter.
481
specific = self.specific_files
483
for path, ie in self.work_inv.iter_entries():
484
if specific and not is_inside_any(specific, path):
486
if not self.work_tree.has_filename(path):
487
self.reporter.missing(path)
488
deleted_ids.append((path, ie.file_id))
490
deleted_ids.sort(reverse=True)
491
for path, file_id in deleted_ids:
492
del self.work_inv[file_id]
493
self.work_tree._write_inventory(self.work_inv)
495
def _store_snapshot(self):
496
"""Pass over inventory and record a snapshot.
498
Entries get a new revision when they are modified in
499
any way, which includes a merge with a new set of
500
parents that have the same entry.
502
# XXX: Need to think more here about when the user has
503
# made a specific decision on a particular value -- c.f.
505
for path, ie in self.new_inv.iter_entries():
506
previous_entries = ie.find_previous_heads(
509
self.branch.repository.get_transaction())
510
if ie.revision is None:
511
change = ie.snapshot(self.rev_id, path, previous_entries,
512
self.work_tree, self.weave_store,
513
self.branch.get_transaction())
516
self.reporter.snapshot_change(change, path)
518
def _populate_new_inv(self):
519
"""Build revision inventory.
521
This creates a new empty inventory. Depending on
522
which files are selected for commit, and what is present in the
523
current tree, the new inventory is populated. inventory entries
524
which are candidates for modification have their revision set to
525
None; inventory entries that are carried over untouched have their
526
revision set to their prior value.
528
mutter("Selecting files for commit with filter %s", self.specific_files)
529
self.new_inv = Inventory()
530
for path, new_ie in self.work_inv.iter_entries():
531
file_id = new_ie.file_id
532
mutter('check %s {%s}', path, new_ie.file_id)
533
if self.specific_files:
534
if not is_inside_any(self.specific_files, path):
535
mutter('%s not selected for commit', path)
536
self._carry_entry(file_id)
539
# this is selected, ensure its parents are too.
540
parent_id = new_ie.parent_id
541
while parent_id != ROOT_ID:
542
if not self.new_inv.has_id(parent_id):
543
ie = self._select_entry(self.work_inv[parent_id])
544
mutter('%s selected for commit because of %s',
545
self.new_inv.id2path(parent_id), path)
547
ie = self.new_inv[parent_id]
548
if ie.revision is not None:
550
mutter('%s selected for commit because of %s',
551
self.new_inv.id2path(parent_id), path)
552
parent_id = ie.parent_id
553
mutter('%s selected for commit', path)
554
self._select_entry(new_ie)
556
def _select_entry(self, new_ie):
557
"""Make new_ie be considered for committing."""
563
def _carry_entry(self, file_id):
564
"""Carry the file unchanged from the basis revision."""
565
if self.basis_inv.has_id(file_id):
566
self.new_inv.add(self.basis_inv[file_id].copy())
568
def _report_deletes(self):
569
for file_id in self.basis_inv:
570
if file_id not in self.new_inv:
571
self.reporter.deleted(self.basis_inv.id2path(file_id))
573
def _gen_revision_id(config, when):
172
574
"""Return new revision-id."""
173
from binascii import hexlify
174
from osutils import rand_bytes, compact_date, user_email
176
s = '%s-%s-' % (user_email(), compact_date(when))
575
s = '%s-%s-' % (config.user_email(), compact_date(when))
177
576
s += hexlify(rand_bytes(8))
181
def _gather_commit(branch, work_tree, work_inv, basis_inv, specific_files,
183
"""Build inventory preparatory to commit.
185
This adds any changed files into the text store, and sets their
186
test-id, sha and size in the returned inventory appropriately.
189
Modified to hold a list of files that have been deleted from
190
the working directory; these should be removed from the
193
from bzrlib.inventory import Inventory
194
from osutils import isdir, isfile, sha_string, quotefn, \
195
local_time_offset, username, kind_marker, is_inside_any
197
from branch import gen_file_id
198
from errors import BzrError
199
from revision import Revision
200
from bzrlib.trace import mutter, note
205
for path, entry in work_inv.iter_entries():
206
## TODO: Check that the file kind has not changed from the previous
207
## revision of this file (if any).
209
p = branch.abspath(path)
210
file_id = entry.file_id
211
mutter('commit prep file %s, id %r ' % (p, file_id))
213
if specific_files and not is_inside_any(specific_files, path):
214
if basis_inv.has_id(file_id):
215
# carry over with previous state
216
inv.add(basis_inv[file_id].copy())
218
# omit this from committed inventory
222
if not work_tree.has_id(file_id):
224
print('deleted %s%s' % (path, kind_marker(entry.kind)))
225
mutter(" file is missing, removing from inventory")
226
missing_ids.append(file_id)
229
# this is present in the new inventory; may be new, modified or
231
old_ie = basis_inv.has_id(file_id) and basis_inv[file_id]
237
old_kind = old_ie.kind
238
if old_kind != entry.kind:
239
raise BzrError("entry %r changed kind from %r to %r"
240
% (file_id, old_kind, entry.kind))
242
if entry.kind == 'directory':
244
raise BzrError("%s is entered as directory but not a directory"
246
elif entry.kind == 'file':
248
raise BzrError("%s is entered as file but is not a file" % quotefn(p))
250
new_sha1 = work_tree.get_file_sha1(file_id)
253
and old_ie.text_sha1 == new_sha1):
254
## assert content == basis.get_file(file_id).read()
255
entry.text_id = old_ie.text_id
256
entry.text_sha1 = new_sha1
257
entry.text_size = old_ie.text_size
258
mutter(' unchanged from previous text_id {%s}' %
261
content = file(p, 'rb').read()
263
# calculate the sha again, just in case the file contents
264
# changed since we updated the cache
265
entry.text_sha1 = sha_string(content)
266
entry.text_size = len(content)
268
entry.text_id = gen_file_id(entry.name)
269
branch.text_store.add(content, entry.text_id)
270
mutter(' stored with text_id {%s}' % entry.text_id)
273
marked = path + kind_marker(entry.kind)
275
print 'added', marked
276
elif old_ie == entry:
278
elif (old_ie.name == entry.name
279
and old_ie.parent_id == entry.parent_id):
280
print 'modified', marked
282
print 'renamed', marked
284
return missing_ids, inv
added
removed
bzrlib/commit.py
