1
# Copyright (C) 2005, 2006 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
# The newly committed revision is going to have a shape corresponding
19
# to that of the working inventory. Files that are not in the
20
# working tree and that were in the predecessor are reported as
21
# removed --- this can include files that were either removed from the
22
# inventory or deleted in the working tree. If they were only
23
# deleted from disk, they are removed from the working inventory.
25
# We then consider the remaining entries, which will be in the new
26
# version. Directory entries are simply copied across. File entries
27
# must be checked to see if a new version of the file should be
28
# recorded. For each parent revision inventory, we check to see what
29
# version of the file was present. If the file was present in at
30
# least one tree, and if it was the same version in all the trees,
31
# then we can just refer to that version. Otherwise, a new version
32
# representing the merger of the file versions must be added.
34
# TODO: Update hashcache before and after - or does the WorkingTree
37
# TODO: Rather than mashing together the ancestry and storing it back,
38
# perhaps the weave should have single method which does it all in one
39
# go, avoiding a lot of redundant work.
41
# TODO: Perhaps give a warning if one of the revisions marked as
42
# merged is already in the ancestry, and then don't record it as a
45
# TODO: If the file is newly merged but unchanged from the version it
46
# merges from, then it should still be reported as newly added
47
# relative to the basis revision.
49
# TODO: Change the parameter 'rev_id' to 'revision_id' to be consistent with
50
# the rest of the code; add a deprecation of the old name.
57
from cStringIO import StringIO
64
from bzrlib.errors import (BzrError, PointlessCommit,
68
from bzrlib.osutils import (kind_marker, isdir,isfile, is_inside_any,
69
is_inside_or_parent_of_any,
70
quotefn, sha_file, split_lines)
71
from bzrlib.testament import Testament
72
from bzrlib.trace import mutter, note, warning
73
from bzrlib.xml5 import serializer_v5
74
from bzrlib.inventory import Inventory, InventoryEntry
75
from bzrlib import symbol_versioning
76
from bzrlib.symbol_versioning import (deprecated_passed,
79
from bzrlib.workingtree import WorkingTree
83
class NullCommitReporter(object):
84
"""I report on progress of a commit."""
86
def snapshot_change(self, change, path):
89
def completed(self, revno, rev_id):
92
def deleted(self, file_id):
95
def escaped(self, escape_count, message):
98
def missing(self, path):
101
def renamed(self, change, old_path, new_path):
105
class ReportCommitToLog(NullCommitReporter):
107
# this may be more useful if 'note' was replaced by an overridable
108
# method on self, which would allow more trivial subclassing.
109
# alternative, a callable could be passed in, allowing really trivial
110
# reuse for some uis. RBC 20060511
112
def snapshot_change(self, change, path):
113
if change == 'unchanged':
115
if change == 'added' and path == '':
117
note("%s %s", change, path)
119
def completed(self, revno, rev_id):
120
note('Committed revision %d.', revno)
122
def deleted(self, file_id):
123
note('deleted %s', file_id)
125
def escaped(self, escape_count, message):
126
note("replaced %d control characters in message", escape_count)
128
def missing(self, path):
129
note('missing %s', path)
131
def renamed(self, change, old_path, new_path):
132
note('%s %s => %s', change, old_path, new_path)
135
class Commit(object):
136
"""Task of committing a new revision.
138
This is a MethodObject: it accumulates state as the commit is
139
prepared, and then it is discarded. It doesn't represent
140
historical revisions, just the act of recording a new one.
143
Modified to hold a list of files that have been deleted from
144
the working directory; these should be removed from the
150
if reporter is not None:
151
self.reporter = reporter
153
self.reporter = NullCommitReporter()
154
if config is not None:
160
branch=DEPRECATED_PARAMETER, message=None,
166
allow_pointless=True,
174
message_callback=None):
175
"""Commit working copy as a new revision.
177
branch -- the deprecated branch to commit to. New callers should pass in
180
message -- the commit message (it or message_callback is required)
182
timestamp -- if not None, seconds-since-epoch for a
183
postdated/predated commit.
185
specific_files -- If true, commit only those files.
187
rev_id -- If set, use this as the new revision id.
188
Useful for test or import commands that need to tightly
189
control what revisions are assigned. If you duplicate
190
a revision id that exists elsewhere it is your own fault.
191
If null (default), a time/random revision id is generated.
193
allow_pointless -- If true (default), commit even if nothing
194
has changed and no merges are recorded.
196
strict -- If true, don't allow a commit if the working tree
197
contains unknown files.
199
revprops -- Properties for new revision
200
:param local: Perform a local only commit.
202
mutter('preparing to commit')
204
if deprecated_passed(branch):
205
symbol_versioning.warn("Commit.commit (branch, ...): The branch parameter is "
206
"deprecated as of bzr 0.8. Please use working_tree= instead.",
207
DeprecationWarning, stacklevel=2)
209
self.work_tree = self.branch.bzrdir.open_workingtree()
210
elif working_tree is None:
211
raise BzrError("One of branch and working_tree must be passed into commit().")
213
self.work_tree = working_tree
214
self.branch = self.work_tree.branch
215
if message_callback is None:
216
if message is not None:
217
if isinstance(message, str):
218
message = message.decode(bzrlib.user_encoding)
219
message_callback = lambda x: message
221
raise BzrError("The message or message_callback keyword"
222
" parameter is required for commit().")
224
self.bound_branch = None
226
self.master_branch = None
227
self.master_locked = False
229
self.specific_files = specific_files
230
self.allow_pointless = allow_pointless
232
if reporter is None and self.reporter is None:
233
self.reporter = NullCommitReporter()
234
elif reporter is not None:
235
self.reporter = reporter
237
self.work_tree.lock_write()
238
self.pb = bzrlib.ui.ui_factory.nested_progress_bar()
239
self.basis_tree = self.work_tree.basis_tree()
240
self.basis_tree.lock_read()
242
# Cannot commit with conflicts present.
243
if len(self.work_tree.conflicts())>0:
244
raise ConflictsInTree
246
# setup the bound branch variables as needed.
247
self._check_bound_branch()
249
# check for out of date working trees
251
first_tree_parent = self.work_tree.get_parent_ids()[0]
253
# if there are no parents, treat our parent as 'None'
254
# this is so that we still consier the master branch
255
# - in a checkout scenario the tree may have no
256
# parents but the branch may do.
257
first_tree_parent = None
258
master_last = self.master_branch.last_revision()
259
if (master_last is not None and
260
master_last != first_tree_parent):
261
raise errors.OutOfDateTree(self.work_tree)
264
# raise an exception as soon as we find a single unknown.
265
for unknown in self.work_tree.unknowns():
266
raise StrictCommitFailed()
268
if self.config is None:
269
self.config = self.branch.get_config()
271
self.work_inv = self.work_tree.inventory
272
self.basis_inv = self.basis_tree.inventory
273
if specific_files is not None:
274
# Ensure specified files are versioned
275
# (We don't actually need the ids here)
276
tree.find_ids_across_trees(specific_files,
277
[self.basis_tree, self.work_tree])
278
# one to finish, one for rev and inventory, and one for each
279
# inventory entry, and the same for the new inventory.
280
# note that this estimate is too long when we do a partial tree
281
# commit which excludes some new files from being considered.
282
# The estimate is corrected when we populate the new inv.
283
self.pb_total = len(self.work_inv) + 5
286
self._gather_parents()
287
if len(self.parents) > 1 and self.specific_files:
288
raise NotImplementedError('selected-file commit of merges is not supported yet: files %r',
291
self.builder = self.branch.get_commit_builder(self.parents,
292
self.config, timestamp, timezone, committer, revprops, rev_id)
294
self._remove_deleted()
295
self._populate_new_inv()
296
self._report_deletes()
298
self._check_pointless()
300
self._emit_progress_update()
301
# TODO: Now the new inventory is known, check for conflicts and
302
# prompt the user for a commit message.
303
# ADHB 2006-08-08: If this is done, populate_new_inv should not add
304
# weave lines, because nothing should be recorded until it is known
305
# that commit will succeed.
306
self.builder.finish_inventory()
307
self._emit_progress_update()
308
message = message_callback(self)
309
assert isinstance(message, unicode), type(message)
310
self.message = message
311
self._escape_commit_message()
313
self.rev_id = self.builder.commit(self.message)
314
self._emit_progress_update()
315
# revision data is in the local branch now.
317
# upload revision data to the master.
318
# this will propagate merged revisions too if needed.
319
if self.bound_branch:
320
self.master_branch.repository.fetch(self.branch.repository,
321
revision_id=self.rev_id)
322
# now the master has the revision data
323
# 'commit' to the master first so a timeout here causes the local
324
# branch to be out of date
325
self.master_branch.append_revision(self.rev_id)
327
# and now do the commit locally.
328
self.branch.append_revision(self.rev_id)
330
rev_tree = self.builder.revision_tree()
331
self.work_tree.set_parent_trees([(self.rev_id, rev_tree)])
332
# now the work tree is up to date with the branch
334
self.reporter.completed(self.branch.revno(), self.rev_id)
335
if self.config.post_commit() is not None:
336
hooks = self.config.post_commit().split(' ')
337
# this would be nicer with twisted.python.reflect.namedAny
339
result = eval(hook + '(branch, rev_id)',
340
{'branch':self.branch,
342
'rev_id':self.rev_id})
343
self._emit_progress_update()
348
def _any_real_changes(self):
349
"""Are there real changes between new_inventory and basis?
351
For trees without rich roots, inv.root.revision changes every commit.
352
But if that is the only change, we want to treat it as though there
355
new_entries = self.builder.new_inventory.iter_entries()
356
basis_entries = self.basis_inv.iter_entries()
357
new_path, new_root_ie = new_entries.next()
358
basis_path, basis_root_ie = basis_entries.next()
360
# This is a copy of InventoryEntry.__eq__ only leaving out .revision
361
def ie_equal_no_revision(this, other):
362
return ((this.file_id == other.file_id)
363
and (this.name == other.name)
364
and (this.symlink_target == other.symlink_target)
365
and (this.text_sha1 == other.text_sha1)
366
and (this.text_size == other.text_size)
367
and (this.text_id == other.text_id)
368
and (this.parent_id == other.parent_id)
369
and (this.kind == other.kind)
370
and (this.executable == other.executable)
372
if not ie_equal_no_revision(new_root_ie, basis_root_ie):
375
for new_ie, basis_ie in zip(new_entries, basis_entries):
376
if new_ie != basis_ie:
379
# No actual changes present
382
def _check_pointless(self):
383
if self.allow_pointless:
385
# A merge with no effect on files
386
if len(self.parents) > 1:
388
# work around the fact that a newly-initted tree does differ from its
390
if len(self.basis_inv) == 0 and len(self.builder.new_inventory) == 1:
391
raise PointlessCommit()
392
# Shortcut, if the number of entries changes, then we obviously have
394
if len(self.builder.new_inventory) != len(self.basis_inv):
396
# If length == 1, then we only have the root entry. Which means
397
# that there is no real difference (only the root could be different)
398
if (len(self.builder.new_inventory) != 1 and self._any_real_changes()):
400
raise PointlessCommit()
402
def _check_bound_branch(self):
403
"""Check to see if the local branch is bound.
405
If it is bound, then most of the commit will actually be
406
done using the remote branch as the target branch.
407
Only at the end will the local branch be updated.
409
if self.local and not self.branch.get_bound_location():
410
raise errors.LocalRequiresBoundBranch()
413
self.master_branch = self.branch.get_master_branch()
415
if not self.master_branch:
416
# make this branch the reference branch for out of date checks.
417
self.master_branch = self.branch
420
# If the master branch is bound, we must fail
421
master_bound_location = self.master_branch.get_bound_location()
422
if master_bound_location:
423
raise errors.CommitToDoubleBoundBranch(self.branch,
424
self.master_branch, master_bound_location)
426
# TODO: jam 20051230 We could automatically push local
427
# commits to the remote branch if they would fit.
428
# But for now, just require remote to be identical
431
# Make sure the local branch is identical to the master
432
master_rh = self.master_branch.revision_history()
433
local_rh = self.branch.revision_history()
434
if local_rh != master_rh:
435
raise errors.BoundBranchOutOfDate(self.branch,
438
# Now things are ready to change the master branch
440
self.bound_branch = self.branch
441
self.master_branch.lock_write()
442
self.master_locked = True
445
"""Cleanup any open locks, progress bars etc."""
446
cleanups = [self._cleanup_bound_branch,
447
self.basis_tree.unlock,
448
self.work_tree.unlock,
450
found_exception = None
451
for cleanup in cleanups:
454
# we want every cleanup to run no matter what.
455
# so we have a catchall here, but we will raise the
456
# last encountered exception up the stack: and
457
# typically this will be useful enough.
460
if found_exception is not None:
461
# don't do a plan raise, because the last exception may have been
462
# trashed, e is our sure-to-work exception even though it loses the
463
# full traceback. XXX: RBC 20060421 perhaps we could check the
464
# exc_info and if its the same one do a plain raise otherwise
465
# 'raise e' as we do now.
468
def _cleanup_bound_branch(self):
469
"""Executed at the end of a try/finally to cleanup a bound branch.
471
If the branch wasn't bound, this is a no-op.
472
If it was, it resents self.branch to the local branch, instead
475
if not self.bound_branch:
477
if self.master_locked:
478
self.master_branch.unlock()
480
def _escape_commit_message(self):
481
"""Replace xml-incompatible control characters."""
482
# FIXME: RBC 20060419 this should be done by the revision
483
# serialiser not by commit. Then we can also add an unescaper
484
# in the deserializer and start roundtripping revision messages
485
# precisely. See repository_implementations/test_repository.py
487
# Python strings can include characters that can't be
488
# represented in well-formed XML; escape characters that
489
# aren't listed in the XML specification
490
# (http://www.w3.org/TR/REC-xml/#NT-Char).
491
self.message, escape_count = re.subn(
492
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
493
lambda match: match.group(0).encode('unicode_escape'),
496
self.reporter.escaped(escape_count, self.message)
498
def _gather_parents(self):
499
"""Record the parents of a merge for merge detection."""
500
# TODO: Make sure that this list doesn't contain duplicate
501
# entries and the order is preserved when doing this.
502
self.parents = self.work_tree.get_parent_ids()
503
self.parent_invs = []
504
for revision in self.parents:
505
if self.branch.repository.has_revision(revision):
506
mutter('commit parent revision {%s}', revision)
507
inventory = self.branch.repository.get_inventory(revision)
508
self.parent_invs.append(inventory)
510
mutter('commit parent ghost revision {%s}', revision)
512
def _remove_deleted(self):
513
"""Remove deleted files from the working inventories.
515
This is done prior to taking the working inventory as the
516
basis for the new committed inventory.
518
This returns true if any files
519
*that existed in the basis inventory* were deleted.
520
Files that were added and deleted
521
in the working copy don't matter.
523
specific = self.specific_files
525
deleted_paths = set()
526
for path, ie in self.work_inv.iter_entries():
527
if is_inside_any(deleted_paths, path):
528
# The tree will delete the required ids recursively.
530
if specific and not is_inside_any(specific, path):
532
if not self.work_tree.has_filename(path):
533
deleted_paths.add(path)
534
self.reporter.missing(path)
535
deleted_ids.append(ie.file_id)
536
self.work_tree.unversion(deleted_ids)
538
def _populate_new_inv(self):
539
"""Build revision inventory.
541
This creates a new empty inventory. Depending on
542
which files are selected for commit, and what is present in the
543
current tree, the new inventory is populated. inventory entries
544
which are candidates for modification have their revision set to
545
None; inventory entries that are carried over untouched have their
546
revision set to their prior value.
548
# ESEPARATIONOFCONCERNS: this function is diffing and using the diff
549
# results to create a new inventory at the same time, which results
550
# in bugs like #46635. Any reason not to use/enhance Tree.changes_from?
552
mutter("Selecting files for commit with filter %s", self.specific_files)
553
assert self.work_inv.root is not None
554
entries = self.work_inv.iter_entries()
555
if not self.builder.record_root_entry:
556
symbol_versioning.warn('CommitBuilders should support recording'
557
' the root entry as of bzr 0.10.', DeprecationWarning,
559
self.builder.new_inventory.add(self.basis_inv.root.copy())
561
self._emit_progress_update()
562
for path, new_ie in entries:
563
self._emit_progress_update()
564
file_id = new_ie.file_id
565
# mutter('check %s {%s}', path, file_id)
566
if (not self.specific_files or
567
is_inside_or_parent_of_any(self.specific_files, path)):
568
# mutter('%s selected for commit', path)
572
# mutter('%s not selected for commit', path)
573
if self.basis_inv.has_id(file_id):
574
ie = self.basis_inv[file_id].copy()
576
# this entry is new and not being committed
578
self.builder.record_entry_contents(ie, self.parent_invs,
579
path, self.work_tree)
580
# describe the nature of the change that has occurred relative to
581
# the basis inventory.
582
if (self.basis_inv.has_id(ie.file_id)):
583
basis_ie = self.basis_inv[ie.file_id]
586
change = ie.describe_change(basis_ie, ie)
587
if change in (InventoryEntry.RENAMED,
588
InventoryEntry.MODIFIED_AND_RENAMED):
589
old_path = self.basis_inv.id2path(ie.file_id)
590
self.reporter.renamed(change, old_path, path)
592
self.reporter.snapshot_change(change, path)
594
if not self.specific_files:
597
# ignore removals that don't match filespec
598
for path, new_ie in self.basis_inv.iter_entries():
599
if new_ie.file_id in self.work_inv:
601
if is_inside_any(self.specific_files, path):
605
self.builder.record_entry_contents(ie, self.parent_invs, path,
608
def _emit_progress_update(self):
609
"""Emit an update to the progress bar."""
610
self.pb.update("Committing", self.pb_count, self.pb_total)
613
def _report_deletes(self):
614
for path, ie in self.basis_inv.iter_entries():
615
if ie.file_id not in self.builder.new_inventory:
616
self.reporter.deleted(path)
added
removed
bzrlib/commit.py
