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
# XXX: Can we do any better about making interrupted commits change
21
# TODO: Separate 'prepare' phase where we find a list of potentially
22
# committed files. We then can then pause the commit to prompt for a
23
# commit message, knowing the summary will be the same as what's
24
# actually used for the commit. (But perhaps simpler to simply get
25
# the tree status, then use that for a selective commit?)
27
# The newly committed revision is going to have a shape corresponding
28
# to that of the working inventory. Files that are not in the
29
# working tree and that were in the predecessor are reported as
30
# removed --- this can include files that were either removed from the
31
# inventory or deleted in the working tree. If they were only
32
# deleted from disk, they are removed from the working inventory.
34
# We then consider the remaining entries, which will be in the new
35
# version. Directory entries are simply copied across. File entries
36
# must be checked to see if a new version of the file should be
37
# recorded. For each parent revision inventory, we check to see what
38
# version of the file was present. If the file was present in at
39
# least one tree, and if it was the same version in all the trees,
40
# then we can just refer to that version. Otherwise, a new version
41
# representing the merger of the file versions must be added.
43
# TODO: Update hashcache before and after - or does the WorkingTree
46
# TODO: Rather than mashing together the ancestry and storing it back,
47
# perhaps the weave should have single method which does it all in one
48
# go, avoiding a lot of redundant work.
50
# TODO: Perhaps give a warning if one of the revisions marked as
51
# merged is already in the ancestry, and then don't record it as a
54
# TODO: If the file is newly merged but unchanged from the version it
55
# merges from, then it should still be reported as newly added
56
# relative to the basis revision.
58
# TODO: Do checks that the tree can be committed *before* running the
59
# editor; this should include checks for a pointless commit and for
60
# unknown or missing files.
62
# TODO: If commit fails, leave the message in a file somewhere.
64
# TODO: Change the parameter 'rev_id' to 'revision_id' to be consistent with
65
# the rest of the code; add a deprecation of the old name.
72
from cStringIO import StringIO
79
from bzrlib.errors import (BzrError, PointlessCommit,
83
from bzrlib.osutils import (kind_marker, isdir,isfile, is_inside_any,
84
is_inside_or_parent_of_any,
85
quotefn, sha_file, split_lines)
86
from bzrlib.testament import Testament
87
from bzrlib.trace import mutter, note, warning
88
from bzrlib.xml5 import serializer_v5
89
from bzrlib.inventory import Inventory, ROOT_ID, InventoryEntry
90
from bzrlib import symbol_versioning
91
from bzrlib.symbol_versioning import (deprecated_passed,
94
from bzrlib.workingtree import WorkingTree
97
class NullCommitReporter(object):
98
"""I report on progress of a commit."""
100
def snapshot_change(self, change, path):
103
def completed(self, revno, rev_id):
106
def deleted(self, file_id):
109
def escaped(self, escape_count, message):
112
def missing(self, path):
115
def renamed(self, change, old_path, new_path):
119
class ReportCommitToLog(NullCommitReporter):
121
# this may be more useful if 'note' was replaced by an overridable
122
# method on self, which would allow more trivial subclassing.
123
# alternative, a callable could be passed in, allowing really trivial
124
# reuse for some uis. RBC 20060511
126
def snapshot_change(self, change, path):
127
if change == 'unchanged':
129
note("%s %s", change, path)
131
def completed(self, revno, rev_id):
132
note('Committed revision %d.', revno)
134
def deleted(self, file_id):
135
note('deleted %s', file_id)
137
def escaped(self, escape_count, message):
138
note("replaced %d control characters in message", escape_count)
140
def missing(self, path):
141
note('missing %s', path)
143
def renamed(self, change, old_path, new_path):
144
note('%s %s => %s', change, old_path, new_path)
147
class Commit(object):
148
"""Task of committing a new revision.
150
This is a MethodObject: it accumulates state as the commit is
151
prepared, and then it is discarded. It doesn't represent
152
historical revisions, just the act of recording a new one.
155
Modified to hold a list of files that have been deleted from
156
the working directory; these should be removed from the
162
if reporter is not None:
163
self.reporter = reporter
165
self.reporter = NullCommitReporter()
166
if config is not None:
172
branch=DEPRECATED_PARAMETER, message=None,
178
allow_pointless=True,
186
"""Commit working copy as a new revision.
188
branch -- the deprecated branch to commit to. New callers should pass in
191
message -- the commit message, a mandatory parameter
193
timestamp -- if not None, seconds-since-epoch for a
194
postdated/predated commit.
196
specific_files -- If true, commit only those files.
198
rev_id -- If set, use this as the new revision id.
199
Useful for test or import commands that need to tightly
200
control what revisions are assigned. If you duplicate
201
a revision id that exists elsewhere it is your own fault.
202
If null (default), a time/random revision id is generated.
204
allow_pointless -- If true (default), commit even if nothing
205
has changed and no merges are recorded.
207
strict -- If true, don't allow a commit if the working tree
208
contains unknown files.
210
revprops -- Properties for new revision
211
:param local: Perform a local only commit.
213
mutter('preparing to commit')
215
if deprecated_passed(branch):
216
symbol_versioning.warn("Commit.commit (branch, ...): The branch parameter is "
217
"deprecated as of bzr 0.8. Please use working_tree= instead.",
218
DeprecationWarning, stacklevel=2)
220
self.work_tree = self.branch.bzrdir.open_workingtree()
221
elif working_tree is None:
222
raise BzrError("One of branch and working_tree must be passed into commit().")
224
self.work_tree = working_tree
225
self.branch = self.work_tree.branch
227
raise BzrError("The message keyword parameter is required for commit().")
229
self.bound_branch = None
231
self.master_branch = None
232
self.master_locked = False
234
self.specific_files = specific_files
235
self.allow_pointless = allow_pointless
237
if reporter is None and self.reporter is None:
238
self.reporter = NullCommitReporter()
239
elif reporter is not None:
240
self.reporter = reporter
242
self.work_tree.lock_write()
243
self.pb = bzrlib.ui.ui_factory.nested_progress_bar()
245
# Cannot commit with conflicts present.
246
if len(self.work_tree.conflicts())>0:
247
raise ConflictsInTree
249
# setup the bound branch variables as needed.
250
self._check_bound_branch()
252
# check for out of date working trees
254
first_tree_parent = self.work_tree.get_parent_ids()[0]
256
# if there are no parents, treat our parent as 'None'
257
# this is so that we still consier the master branch
258
# - in a checkout scenario the tree may have no
259
# parents but the branch may do.
260
first_tree_parent = None
261
master_last = self.master_branch.last_revision()
262
if (master_last is not None and
263
master_last != first_tree_parent):
264
raise errors.OutOfDateTree(self.work_tree)
267
# raise an exception as soon as we find a single unknown.
268
for unknown in self.work_tree.unknowns():
269
raise StrictCommitFailed()
271
if self.config is None:
272
self.config = self.branch.get_config()
274
if isinstance(message, str):
275
message = message.decode(bzrlib.user_encoding)
276
assert isinstance(message, unicode), type(message)
277
self.message = message
278
self._escape_commit_message()
280
self.work_inv = self.work_tree.inventory
281
self.basis_tree = self.work_tree.basis_tree()
282
self.basis_inv = self.basis_tree.inventory
283
if specific_files is not None:
284
# Ensure specified files are versioned
285
# (We don't actually need the ids here)
286
tree.find_ids_across_trees(specific_files,
287
[self.basis_tree, self.work_tree])
288
# one to finish, one for rev and inventory, and one for each
289
# inventory entry, and the same for the new inventory.
290
# note that this estimate is too long when we do a partial tree
291
# commit which excludes some new files from being considered.
292
# The estimate is corrected when we populate the new inv.
293
self.pb_total = len(self.work_inv) + 5
296
self._gather_parents()
297
if len(self.parents) > 1 and self.specific_files:
298
raise NotImplementedError('selected-file commit of merges is not supported yet: files %r',
301
self.builder = self.branch.get_commit_builder(self.parents,
302
self.config, timestamp, timezone, committer, revprops, rev_id)
304
self._remove_deleted()
305
self._populate_new_inv()
306
self._report_deletes()
308
self._check_pointless()
310
self._emit_progress_update()
311
# TODO: Now the new inventory is known, check for conflicts and
312
# prompt the user for a commit message.
313
# ADHB 2006-08-08: If this is done, populate_new_inv should not add
314
# weave lines, because nothing should be recorded until it is known
315
# that commit will succeed.
316
self.builder.finish_inventory()
317
self._emit_progress_update()
318
self.rev_id = self.builder.commit(self.message)
319
self._emit_progress_update()
320
# revision data is in the local branch now.
322
# upload revision data to the master.
323
# this will propagate merged revisions too if needed.
324
if self.bound_branch:
325
self.master_branch.repository.fetch(self.branch.repository,
326
revision_id=self.rev_id)
327
# now the master has the revision data
328
# 'commit' to the master first so a timeout here causes the local
329
# branch to be out of date
330
self.master_branch.append_revision(self.rev_id)
332
# and now do the commit locally.
333
self.branch.append_revision(self.rev_id)
335
rev_tree = self.builder.revision_tree()
336
self.work_tree.set_parent_trees([(self.rev_id, rev_tree)])
337
# now the work tree is up to date with the branch
339
self.reporter.completed(self.branch.revno(), self.rev_id)
340
if self.config.post_commit() is not None:
341
hooks = self.config.post_commit().split(' ')
342
# this would be nicer with twisted.python.reflect.namedAny
344
result = eval(hook + '(branch, rev_id)',
345
{'branch':self.branch,
347
'rev_id':self.rev_id})
348
self._emit_progress_update()
353
def _any_real_changes(self):
354
"""Are there real changes between new_inventory and basis?
356
For trees without rich roots, inv.root.revision changes every commit.
357
But if that is the only change, we want to treat it as though there
360
new_entries = self.builder.new_inventory.iter_entries()
361
basis_entries = self.basis_inv.iter_entries()
362
new_path, new_root_ie = new_entries.next()
363
basis_path, basis_root_ie = basis_entries.next()
365
# This is a copy of InventoryEntry.__eq__ only leaving out .revision
366
def ie_equal_no_revision(this, other):
367
return ((this.file_id == other.file_id)
368
and (this.name == other.name)
369
and (this.symlink_target == other.symlink_target)
370
and (this.text_sha1 == other.text_sha1)
371
and (this.text_size == other.text_size)
372
and (this.text_id == other.text_id)
373
and (this.parent_id == other.parent_id)
374
and (this.kind == other.kind)
375
and (this.executable == other.executable)
377
if not ie_equal_no_revision(new_root_ie, basis_root_ie):
380
for new_ie, basis_ie in zip(new_entries, basis_entries):
381
if new_ie != basis_ie:
384
# No actual changes present
387
def _check_pointless(self):
388
if self.allow_pointless:
390
# A merge with no effect on files
391
if len(self.parents) > 1:
393
# Shortcut, if the number of entries changes, then we obviously have
395
if len(self.builder.new_inventory) != len(self.basis_inv):
397
# If length == 1, then we only have the root entry. Which means
398
# that there is no real difference (only the root could be different)
399
if (len(self.builder.new_inventory) != 1 and self._any_real_changes()):
401
raise PointlessCommit()
403
def _check_bound_branch(self):
404
"""Check to see if the local branch is bound.
406
If it is bound, then most of the commit will actually be
407
done using the remote branch as the target branch.
408
Only at the end will the local branch be updated.
410
if self.local and not self.branch.get_bound_location():
411
raise errors.LocalRequiresBoundBranch()
414
self.master_branch = self.branch.get_master_branch()
416
if not self.master_branch:
417
# make this branch the reference branch for out of date checks.
418
self.master_branch = self.branch
421
# If the master branch is bound, we must fail
422
master_bound_location = self.master_branch.get_bound_location()
423
if master_bound_location:
424
raise errors.CommitToDoubleBoundBranch(self.branch,
425
self.master_branch, master_bound_location)
427
# TODO: jam 20051230 We could automatically push local
428
# commits to the remote branch if they would fit.
429
# But for now, just require remote to be identical
432
# Make sure the local branch is identical to the master
433
master_rh = self.master_branch.revision_history()
434
local_rh = self.branch.revision_history()
435
if local_rh != master_rh:
436
raise errors.BoundBranchOutOfDate(self.branch,
439
# Now things are ready to change the master branch
441
self.bound_branch = self.branch
442
self.master_branch.lock_write()
443
self.master_locked = True
446
"""Cleanup any open locks, progress bars etc."""
447
cleanups = [self._cleanup_bound_branch,
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
entries = self.work_inv.iter_entries()
554
if not self.builder.record_root_entry:
555
symbol_versioning.warn('CommitBuilders should support recording'
556
' the root entry as of bzr 0.10.', DeprecationWarning,
558
self.builder.new_inventory.add(self.basis_inv.root.copy())
560
self._emit_progress_update()
561
for path, new_ie in entries:
562
self._emit_progress_update()
563
file_id = new_ie.file_id
564
# mutter('check %s {%s}', path, file_id)
565
if (not self.specific_files or
566
is_inside_or_parent_of_any(self.specific_files, path)):
567
# mutter('%s selected for commit', path)
571
# mutter('%s not selected for commit', path)
572
if self.basis_inv.has_id(file_id):
573
ie = self.basis_inv[file_id].copy()
575
# 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
