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.
71
from cStringIO import StringIO
74
import bzrlib.errors as errors
75
from bzrlib.errors import (BzrError, PointlessCommit,
79
from bzrlib.osutils import (kind_marker, isdir,isfile, is_inside_any,
80
is_inside_or_parent_of_any,
81
quotefn, sha_file, split_lines)
82
from bzrlib.testament import Testament
83
from bzrlib.trace import mutter, note, warning
84
from bzrlib.xml5 import serializer_v5
85
from bzrlib.inventory import Inventory, ROOT_ID, InventoryEntry
86
from bzrlib.symbol_versioning import (deprecated_passed,
90
from bzrlib.workingtree import WorkingTree
93
@deprecated_function(zero_seven)
94
def commit(*args, **kwargs):
95
"""Commit a new revision to a branch.
97
Function-style interface for convenience of old callers.
99
New code should use the Commit class instead.
101
## XXX: Remove this in favor of WorkingTree.commit?
102
Commit().commit(*args, **kwargs)
105
class NullCommitReporter(object):
106
"""I report on progress of a commit."""
108
def snapshot_change(self, change, path):
111
def completed(self, revno, rev_id):
114
def deleted(self, file_id):
117
def escaped(self, escape_count, message):
120
def missing(self, path):
123
def renamed(self, change, old_path, new_path):
127
class ReportCommitToLog(NullCommitReporter):
129
# this may be more useful if 'note' was replaced by an overridable
130
# method on self, which would allow more trivial subclassing.
131
# alternative, a callable could be passed in, allowing really trivial
132
# reuse for some uis. RBC 20060511
134
def snapshot_change(self, change, path):
135
if change == 'unchanged':
137
note("%s %s", change, path)
139
def completed(self, revno, rev_id):
140
note('Committed revision %d.', revno)
142
def deleted(self, file_id):
143
note('deleted %s', file_id)
145
def escaped(self, escape_count, message):
146
note("replaced %d control characters in message", escape_count)
148
def missing(self, path):
149
note('missing %s', path)
151
def renamed(self, change, old_path, new_path):
152
note('%s %s => %s', change, old_path, new_path)
155
class Commit(object):
156
"""Task of committing a new revision.
158
This is a MethodObject: it accumulates state as the commit is
159
prepared, and then it is discarded. It doesn't represent
160
historical revisions, just the act of recording a new one.
163
Modified to hold a list of files that have been deleted from
164
the working directory; these should be removed from the
170
if reporter is not None:
171
self.reporter = reporter
173
self.reporter = NullCommitReporter()
174
if config is not None:
180
branch=DEPRECATED_PARAMETER, message=None,
186
allow_pointless=True,
194
"""Commit working copy as a new revision.
196
branch -- the deprecated branch to commit to. New callers should pass in
199
message -- the commit message, a mandatory parameter
201
timestamp -- if not None, seconds-since-epoch for a
202
postdated/predated commit.
204
specific_files -- If true, commit only those files.
206
rev_id -- If set, use this as the new revision id.
207
Useful for test or import commands that need to tightly
208
control what revisions are assigned. If you duplicate
209
a revision id that exists elsewhere it is your own fault.
210
If null (default), a time/random revision id is generated.
212
allow_pointless -- If true (default), commit even if nothing
213
has changed and no merges are recorded.
215
strict -- If true, don't allow a commit if the working tree
216
contains unknown files.
218
revprops -- Properties for new revision
219
:param local: Perform a local only commit.
221
mutter('preparing to commit')
223
if deprecated_passed(branch):
224
warnings.warn("Commit.commit (branch, ...): The branch parameter is "
225
"deprecated as of bzr 0.8. Please use working_tree= instead.",
226
DeprecationWarning, stacklevel=2)
228
self.work_tree = self.branch.bzrdir.open_workingtree()
229
elif working_tree is None:
230
raise BzrError("One of branch and working_tree must be passed into commit().")
232
self.work_tree = working_tree
233
self.branch = self.work_tree.branch
235
raise BzrError("The message keyword parameter is required for commit().")
237
self.bound_branch = None
239
self.master_branch = None
240
self.master_locked = False
242
self.specific_files = specific_files
243
self.allow_pointless = allow_pointless
245
if reporter is None and self.reporter is None:
246
self.reporter = NullCommitReporter()
247
elif reporter is not None:
248
self.reporter = reporter
250
self.work_tree.lock_write()
251
self.pb = bzrlib.ui.ui_factory.nested_progress_bar()
253
# Cannot commit with conflicts present.
254
if len(self.work_tree.conflicts())>0:
255
raise ConflictsInTree
257
# setup the bound branch variables as needed.
258
self._check_bound_branch()
260
# check for out of date working trees
261
# if we are bound, then self.branch is the master branch and this
262
# test is thus all we need.
263
master_last = self.master_branch.last_revision()
264
if (master_last is not None and
265
master_last != self.work_tree.last_revision()):
266
raise errors.OutOfDateTree(self.work_tree)
269
# raise an exception as soon as we find a single unknown.
270
for unknown in self.work_tree.unknowns():
271
raise StrictCommitFailed()
273
if self.config is None:
274
self.config = self.branch.get_config()
276
if isinstance(message, str):
277
message = message.decode(bzrlib.user_encoding)
278
assert isinstance(message, unicode), type(message)
279
self.message = message
280
self._escape_commit_message()
282
self.work_inv = self.work_tree.inventory
283
self.basis_tree = self.work_tree.basis_tree()
284
self.basis_inv = self.basis_tree.inventory
285
# one to finish, one for rev and inventory, and one for each
286
# inventory entry, and the same for the new inventory.
287
# note that this estimate is too long when we do a partial tree
288
# commit which excludes some new files from being considered.
289
# The estimate is corrected when we populate the new inv.
290
self.pb_total = len(self.work_inv) + 5
293
self._gather_parents()
294
if len(self.parents) > 1 and self.specific_files:
295
raise NotImplementedError('selected-file commit of merges is not supported yet: files %r',
298
self.builder = self.branch.get_commit_builder(self.parents,
299
self.config, timestamp, timezone, committer, revprops, rev_id)
301
self._remove_deleted()
302
self._populate_new_inv()
303
self._report_deletes()
305
if not (self.allow_pointless
306
or len(self.parents) > 1
307
or self.builder.new_inventory != self.basis_inv):
308
raise PointlessCommit()
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
self.work_tree.set_pending_merges([])
336
self.work_tree.set_last_revision(self.rev_id)
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 _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
396
"""Cleanup any open locks, progress bars etc."""
397
cleanups = [self._cleanup_bound_branch,
398
self.work_tree.unlock,
400
found_exception = None
401
for cleanup in cleanups:
404
# we want every cleanup to run no matter what.
405
# so we have a catchall here, but we will raise the
406
# last encountered exception up the stack: and
407
# typically this will be useful enough.
410
if found_exception is not None:
411
# don't do a plan raise, because the last exception may have been
412
# trashed, e is our sure-to-work exception even though it loses the
413
# full traceback. XXX: RBC 20060421 perhaps we could check the
414
# exc_info and if its the same one do a plain raise otherwise
415
# 'raise e' as we do now.
418
def _cleanup_bound_branch(self):
419
"""Executed at the end of a try/finally to cleanup a bound branch.
421
If the branch wasn't bound, this is a no-op.
422
If it was, it resents self.branch to the local branch, instead
425
if not self.bound_branch:
427
if self.master_locked:
428
self.master_branch.unlock()
430
def _escape_commit_message(self):
431
"""Replace xml-incompatible control characters."""
432
# FIXME: RBC 20060419 this should be done by the revision
433
# serialiser not by commit. Then we can also add an unescaper
434
# in the deserializer and start roundtripping revision messages
435
# precisely. See repository_implementations/test_repository.py
437
# Python strings can include characters that can't be
438
# represented in well-formed XML; escape characters that
439
# aren't listed in the XML specification
440
# (http://www.w3.org/TR/REC-xml/#NT-Char).
441
self.message, escape_count = re.subn(
442
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
443
lambda match: match.group(0).encode('unicode_escape'),
446
self.reporter.escaped(escape_count, self.message)
448
def _gather_parents(self):
449
"""Record the parents of a merge for merge detection."""
450
# TODO: Make sure that this list doesn't contain duplicate
451
# entries and the order is preserved when doing this.
452
self.parents = self.work_tree.get_parent_ids()
453
self.parent_invs = []
454
for revision in self.parents:
455
if self.branch.repository.has_revision(revision):
456
mutter('commit parent revision {%s}', revision)
457
inventory = self.branch.repository.get_inventory(revision)
458
self.parent_invs.append(inventory)
460
mutter('commit parent ghost revision {%s}', revision)
462
def _remove_deleted(self):
463
"""Remove deleted files from the working inventories.
465
This is done prior to taking the working inventory as the
466
basis for the new committed inventory.
468
This returns true if any files
469
*that existed in the basis inventory* were deleted.
470
Files that were added and deleted
471
in the working copy don't matter.
473
specific = self.specific_files
475
for path, ie in self.work_inv.iter_entries():
476
if specific and not is_inside_any(specific, path):
478
if not self.work_tree.has_filename(path):
479
self.reporter.missing(path)
480
deleted_ids.append((path, ie.file_id))
482
deleted_ids.sort(reverse=True)
483
for path, file_id in deleted_ids:
484
del self.work_inv[file_id]
485
self.work_tree._write_inventory(self.work_inv)
487
def _populate_new_inv(self):
488
"""Build revision inventory.
490
This creates a new empty inventory. Depending on
491
which files are selected for commit, and what is present in the
492
current tree, the new inventory is populated. inventory entries
493
which are candidates for modification have their revision set to
494
None; inventory entries that are carried over untouched have their
495
revision set to their prior value.
497
# ESEPARATIONOFCONCERNS: this function is diffing and using the diff
498
# results to create a new inventory at the same time, which results
499
# in bugs like #46635. Any reason not to use/enhance Tree.changes_from?
501
mutter("Selecting files for commit with filter %s", self.specific_files)
502
entries = self.work_inv.iter_entries()
503
if not self.builder.record_root_entry:
504
warnings.warn('CommitBuilders should support recording the root'
505
' entry as of bzr 0.10.', DeprecationWarning, stacklevel=1)
506
self.builder.new_inventory.add(self.basis_inv.root.copy())
508
self._emit_progress_update()
509
for path, new_ie in entries:
510
self._emit_progress_update()
511
file_id = new_ie.file_id
512
# mutter('check %s {%s}', path, file_id)
513
if (not self.specific_files or
514
is_inside_or_parent_of_any(self.specific_files, path)):
515
# mutter('%s selected for commit', path)
519
# mutter('%s not selected for commit', path)
520
if self.basis_inv.has_id(file_id):
521
ie = self.basis_inv[file_id].copy()
523
# this entry is new and not being committed
526
self.builder.record_entry_contents(ie, self.parent_invs,
527
path, self.work_tree)
528
# describe the nature of the change that has occurred relative to
529
# the basis inventory.
530
if (self.basis_inv.has_id(ie.file_id)):
531
basis_ie = self.basis_inv[ie.file_id]
534
change = ie.describe_change(basis_ie, ie)
535
if change in (InventoryEntry.RENAMED,
536
InventoryEntry.MODIFIED_AND_RENAMED):
537
old_path = self.basis_inv.id2path(ie.file_id)
538
self.reporter.renamed(change, old_path, path)
540
self.reporter.snapshot_change(change, path)
542
if not self.specific_files:
545
# ignore removals that don't match filespec
546
for path, new_ie in self.basis_inv.iter_entries():
547
if new_ie.file_id in self.work_inv:
549
if is_inside_any(self.specific_files, path):
553
self.builder.record_entry_contents(ie, self.parent_invs, path,
556
def _emit_progress_update(self):
557
"""Emit an update to the progress bar."""
558
self.pb.update("Committing", self.pb_count, self.pb_total)
561
def _report_deletes(self):
562
for path, ie in self.basis_inv.iter_entries():
563
if ie.file_id not in self.builder.new_inventory:
564
self.reporter.deleted(path)
added
removed
bzrlib/commit.py
