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
# 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...')
89
# print 'looking for changes...'
90
# disabled; should be done at a higher level
93
pending_merges = branch.pending_merges()
95
missing_ids, new_inv, any_changes = \
96
_gather_commit(branch,
103
if not (any_changes or allow_pointless or pending_merges):
104
raise PointlessCommit()
106
for file_id in missing_ids:
107
# Any files that have been deleted are now removed from the
108
# working inventory. Files that were not selected for commit
109
# are left as they were in the working inventory and ommitted
110
# from the revision inventory.
112
# have to do this later so we don't mess up the iterator.
113
# since parents may be removed before their children we
116
# FIXME: There's probably a better way to do this; perhaps
117
# the workingtree should know how to filter itbranch.
118
if work_inv.has_id(file_id):
119
del work_inv[file_id]
122
rev_id = _gen_revision_id(branch, time.time())
125
inv_tmp = tempfile.TemporaryFile()
126
pack_xml(new_inv, inv_tmp)
128
branch.inventory_store.add(inv_tmp, inv_id)
129
mutter('new inventory_id is {%s}' % inv_id)
131
# We could also just sha hash the inv_tmp file
132
# however, in the case that branch.inventory_store.add()
133
# ever actually does anything special
134
inv_sha1 = branch.get_inventory_sha1(inv_id)
136
branch._write_inventory(work_inv)
138
if timestamp == None:
139
timestamp = time.time()
141
if committer == None:
142
committer = username(branch)
145
timezone = local_time_offset()
147
mutter("building commit log message")
148
rev = Revision(timestamp=timestamp,
153
inventory_sha1=inv_sha1,
157
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
note("%s %s", change, path)
133
def completed(self, revno, rev_id):
134
note('committed r%d {%s}', revno, rev_id)
136
def deleted(self, file_id):
137
note('deleted %s', file_id)
139
def escaped(self, escape_count, message):
140
note("replaced %d control characters in message", escape_count)
142
def missing(self, path):
143
note('missing %s', path)
146
class Commit(object):
147
"""Task of committing a new revision.
149
This is a MethodObject: it accumulates state as the commit is
150
prepared, and then it is discarded. It doesn't represent
151
historical revisions, just the act of recording a new one.
154
Modified to hold a list of files that have been deleted from
155
the working directory; these should be removed from the
161
if reporter is not None:
162
self.reporter = reporter
164
self.reporter = NullCommitReporter()
165
if config is not None:
171
branch=DEPRECATED_PARAMETER, message=None,
177
allow_pointless=True,
183
"""Commit working copy as a new revision.
185
branch -- the deprecated branch to commit to. New callers should pass in
188
message -- the commit message, a mandatory parameter
190
timestamp -- if not None, seconds-since-epoch for a
191
postdated/predated commit.
193
specific_files -- If true, commit only those files.
195
rev_id -- If set, use this as the new revision id.
196
Useful for test or import commands that need to tightly
197
control what revisions are assigned. If you duplicate
198
a revision id that exists elsewhere it is your own fault.
199
If null (default), a time/random revision id is generated.
201
allow_pointless -- If true (default), commit even if nothing
202
has changed and no merges are recorded.
204
strict -- If true, don't allow a commit if the working tree
205
contains unknown files.
207
revprops -- Properties for new revision
208
:param local: Perform a local only commit.
210
mutter('preparing to commit')
212
if deprecated_passed(branch):
213
warn("Commit.commit (branch, ...): The branch parameter is "
214
"deprecated as of bzr 0.8. Please use working_tree= instead.",
215
DeprecationWarning, stacklevel=2)
217
self.work_tree = self.branch.bzrdir.open_workingtree()
218
elif working_tree is None:
219
raise BzrError("One of branch and working_tree must be passed into commit().")
221
self.work_tree = working_tree
222
self.branch = self.work_tree.branch
224
raise BzrError("The message keyword parameter is required for commit().")
226
self.weave_store = self.branch.repository.weave_store
227
self.bound_branch = None
229
self.master_branch = None
231
self.specific_files = specific_files
232
self.allow_pointless = allow_pointless
233
self.revprops = {'branch-nick': self.branch.nick}
235
self.revprops.update(revprops)
237
self.work_tree.lock_write()
239
# setup the bound branch variables as needed.
240
self._check_bound_branch()
242
# check for out of date working trees
243
# if we are bound, then self.branch is the master branch and this
244
# test is thus all we need.
245
if self.work_tree.last_revision() != self.master_branch.last_revision():
246
raise errors.OutOfDateTree(self.work_tree)
249
# raise an exception as soon as we find a single unknown.
250
for unknown in self.work_tree.unknowns():
251
raise StrictCommitFailed()
253
if timestamp is None:
254
self.timestamp = time.time()
256
self.timestamp = long(timestamp)
258
if self.config is None:
259
self.config = bzrlib.config.BranchConfig(self.branch)
262
self.rev_id = _gen_revision_id(self.config, self.timestamp)
266
if committer is None:
267
self.committer = self.config.username()
269
assert isinstance(committer, basestring), type(committer)
270
self.committer = committer
273
self.timezone = local_time_offset()
275
self.timezone = int(timezone)
277
if isinstance(message, str):
278
message = message.decode(bzrlib.user_encoding)
279
assert isinstance(message, unicode), type(message)
280
self.message = message
281
self._escape_commit_message()
283
self.work_inv = self.work_tree.inventory
284
self.basis_tree = self.work_tree.basis_tree()
285
self.basis_inv = self.basis_tree.inventory
287
self._gather_parents()
288
if len(self.parents) > 1 and self.specific_files:
289
raise NotImplementedError('selected-file commit of merges is not supported yet')
290
self._check_parents_present()
292
self._remove_deleted()
293
self._populate_new_inv()
294
self._store_snapshot()
295
self._report_deletes()
297
if not (self.allow_pointless
298
or len(self.parents) > 1
299
or self.new_inv != self.basis_inv):
300
raise PointlessCommit()
302
if len(list(self.work_tree.iter_conflicts()))>0:
303
raise ConflictsInTree
305
self.inv_sha1 = self.branch.repository.add_inventory(
310
self._make_revision()
311
# revision data is in the local branch now.
313
# upload revision data to the master.
314
# this will propogate merged revisions too if needed.
315
if self.bound_branch:
316
self.master_branch.repository.fetch(self.branch.repository,
317
revision_id=self.rev_id)
318
# now the master has the revision data
319
# 'commit' to the master first so a timeout here causes the local
320
# branch to be out of date
321
self.master_branch.append_revision(self.rev_id)
323
# and now do the commit locally.
324
self.branch.append_revision(self.rev_id)
326
self.work_tree.set_pending_merges([])
327
if len(self.parents):
328
precursor = self.parents[0]
331
self.work_tree.set_last_revision(self.rev_id, precursor)
332
# now the work tree is up to date with the branch
334
self.reporter.completed(self.branch.revno()+1, 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})
344
self._cleanup_bound_branch()
345
self.work_tree.unlock()
347
def _check_bound_branch(self):
348
"""Check to see if the local branch is bound.
350
If it is bound, then most of the commit will actually be
351
done using the remote branch as the target branch.
352
Only at the end will the local branch be updated.
354
if self.local and not self.branch.get_bound_location():
355
raise errors.LocalRequiresBoundBranch()
358
self.master_branch = self.branch.get_master_branch()
360
if not self.master_branch:
361
# make this branch the reference branch for out of date checks.
362
self.master_branch = self.branch
365
# If the master branch is bound, we must fail
366
master_bound_location = self.master_branch.get_bound_location()
367
if master_bound_location:
368
raise errors.CommitToDoubleBoundBranch(self.branch,
369
self.master_branch, master_bound_location)
371
# TODO: jam 20051230 We could automatically push local
372
# commits to the remote branch if they would fit.
373
# But for now, just require remote to be identical
376
# Make sure the local branch is identical to the master
377
master_rh = self.master_branch.revision_history()
378
local_rh = self.branch.revision_history()
379
if local_rh != master_rh:
380
raise errors.BoundBranchOutOfDate(self.branch,
383
# Now things are ready to change the master branch
385
self.bound_branch = self.branch
386
self.master_branch.lock_write()
388
#### # Check to see if we have any pending merges. If we do
389
#### # those need to be pushed into the master branch
390
#### pending_merges = self.work_tree.pending_merges()
391
#### if pending_merges:
392
#### for revision_id in pending_merges:
393
#### self.master_branch.repository.fetch(self.bound_branch.repository,
394
#### revision_id=revision_id)
396
def _cleanup_bound_branch(self):
397
"""Executed at the end of a try/finally to cleanup a bound branch.
399
If the branch wasn't bound, this is a no-op.
400
If it was, it resents self.branch to the local branch, instead
403
if not self.bound_branch:
405
self.master_branch.unlock()
407
def _escape_commit_message(self):
408
"""Replace xml-incompatible control characters."""
409
# Python strings can include characters that can't be
410
# represented in well-formed XML; escape characters that
411
# aren't listed in the XML specification
412
# (http://www.w3.org/TR/REC-xml/#NT-Char).
413
self.message, escape_count = re.subn(
414
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
415
lambda match: match.group(0).encode('unicode_escape'),
418
self.reporter.escaped(escape_count, self.message)
420
def _gather_parents(self):
421
"""Record the parents of a merge for merge detection."""
422
pending_merges = self.work_tree.pending_merges()
424
self.parent_invs = []
425
self.present_parents = []
426
precursor_id = self.branch.last_revision()
159
precursor_sha1 = branch.get_revision_sha1(precursor_id)
160
rev.parents.append(RevisionReference(precursor_id, precursor_sha1))
161
for merge_rev in pending_merges:
162
rev.parents.append(RevisionReference(merge_rev))
164
rev_tmp = tempfile.TemporaryFile()
165
pack_xml(rev, rev_tmp)
167
branch.revision_store.add(rev_tmp, rev_id)
168
mutter("new revision_id is {%s}" % rev_id)
170
## XXX: Everything up to here can simply be orphaned if we abort
171
## the commit; it will leave junk files behind but that doesn't
174
## TODO: Read back the just-generated changeset, and make sure it
175
## applies and recreates the right state.
177
## TODO: Also calculate and store the inventory SHA1
178
mutter("committing patch r%d" % (branch.revno() + 1))
180
branch.append_revision(rev_id)
182
branch.set_pending_merges([])
185
# disabled; should go through logging
186
# note("commited r%d" % branch.revno())
187
# print ("commited r%d" % branch.revno())
194
def _gen_revision_id(branch, when):
428
self.parents.append(precursor_id)
429
self.parents += pending_merges
430
for revision in self.parents:
431
if self.branch.repository.has_revision(revision):
432
inventory = self.branch.repository.get_inventory(revision)
433
self.parent_invs.append(inventory)
434
self.present_parents.append(revision)
436
def _check_parents_present(self):
437
for parent_id in self.parents:
438
mutter('commit parent revision {%s}', parent_id)
439
if not self.branch.repository.has_revision(parent_id):
440
if parent_id == self.branch.last_revision():
441
warning("parent is missing %r", parent_id)
442
raise HistoryMissing(self.branch, 'revision', parent_id)
444
mutter("commit will ghost revision %r", parent_id)
446
def _make_revision(self):
447
"""Record a new revision object for this commit."""
448
rev = Revision(timestamp=self.timestamp,
449
timezone=self.timezone,
450
committer=self.committer,
451
message=self.message,
452
inventory_sha1=self.inv_sha1,
453
revision_id=self.rev_id,
454
properties=self.revprops)
455
rev.parent_ids = self.parents
456
self.branch.repository.add_revision(self.rev_id, rev, self.new_inv, self.config)
458
def _remove_deleted(self):
459
"""Remove deleted files from the working inventories.
461
This is done prior to taking the working inventory as the
462
basis for the new committed inventory.
464
This returns true if any files
465
*that existed in the basis inventory* were deleted.
466
Files that were added and deleted
467
in the working copy don't matter.
469
specific = self.specific_files
471
for path, ie in self.work_inv.iter_entries():
472
if specific and not is_inside_any(specific, path):
474
if not self.work_tree.has_filename(path):
475
self.reporter.missing(path)
476
deleted_ids.append((path, ie.file_id))
478
deleted_ids.sort(reverse=True)
479
for path, file_id in deleted_ids:
480
del self.work_inv[file_id]
481
self.work_tree._write_inventory(self.work_inv)
483
def _store_snapshot(self):
484
"""Pass over inventory and record a snapshot.
486
Entries get a new revision when they are modified in
487
any way, which includes a merge with a new set of
488
parents that have the same entry.
490
# XXX: Need to think more here about when the user has
491
# made a specific decision on a particular value -- c.f.
493
for path, ie in self.new_inv.iter_entries():
494
previous_entries = ie.find_previous_heads(
496
self.weave_store.get_weave_or_empty(ie.file_id,
497
self.branch.get_transaction()))
498
if ie.revision is None:
499
change = ie.snapshot(self.rev_id, path, previous_entries,
500
self.work_tree, self.weave_store,
501
self.branch.get_transaction())
504
self.reporter.snapshot_change(change, path)
506
def _populate_new_inv(self):
507
"""Build revision inventory.
509
This creates a new empty inventory. Depending on
510
which files are selected for commit, and what is present in the
511
current tree, the new inventory is populated. inventory entries
512
which are candidates for modification have their revision set to
513
None; inventory entries that are carried over untouched have their
514
revision set to their prior value.
516
mutter("Selecting files for commit with filter %s", self.specific_files)
517
self.new_inv = Inventory()
518
for path, new_ie in self.work_inv.iter_entries():
519
file_id = new_ie.file_id
520
mutter('check %s {%s}', path, new_ie.file_id)
521
if self.specific_files:
522
if not is_inside_any(self.specific_files, path):
523
mutter('%s not selected for commit', path)
524
self._carry_entry(file_id)
527
# this is selected, ensure its parents are too.
528
parent_id = new_ie.parent_id
529
while parent_id != ROOT_ID:
530
if not self.new_inv.has_id(parent_id):
531
ie = self._select_entry(self.work_inv[parent_id])
532
mutter('%s selected for commit because of %s',
533
self.new_inv.id2path(parent_id), path)
535
ie = self.new_inv[parent_id]
536
if ie.revision is not None:
538
mutter('%s selected for commit because of %s',
539
self.new_inv.id2path(parent_id), path)
540
parent_id = ie.parent_id
541
mutter('%s selected for commit', path)
542
self._select_entry(new_ie)
544
def _select_entry(self, new_ie):
545
"""Make new_ie be considered for committing."""
551
def _carry_entry(self, file_id):
552
"""Carry the file unchanged from the basis revision."""
553
if self.basis_inv.has_id(file_id):
554
self.new_inv.add(self.basis_inv[file_id].copy())
556
def _report_deletes(self):
557
for file_id in self.basis_inv:
558
if file_id not in self.new_inv:
559
self.reporter.deleted(self.basis_inv.id2path(file_id))
561
def _gen_revision_id(config, when):
195
562
"""Return new revision-id."""
196
from binascii import hexlify
197
from bzrlib.osutils import rand_bytes, compact_date, user_email
199
s = '%s-%s-' % (user_email(branch), compact_date(when))
563
s = '%s-%s-' % (config.user_email(), compact_date(when))
200
564
s += hexlify(rand_bytes(8))
204
def _gather_commit(branch, work_tree, work_inv, basis_inv, specific_files,
206
"""Build inventory preparatory to commit.
208
Returns missing_ids, new_inv, any_changes.
210
This adds any changed files into the text store, and sets their
211
test-id, sha and size in the returned inventory appropriately.
214
Modified to hold a list of files that have been deleted from
215
the working directory; these should be removed from the
218
from bzrlib.inventory import Inventory
219
from bzrlib.osutils import isdir, isfile, sha_string, quotefn, \
220
local_time_offset, username, kind_marker, is_inside_any
222
from bzrlib.branch import gen_file_id
223
from bzrlib.errors import BzrError
224
from bzrlib.revision import Revision
225
from bzrlib.trace import mutter, note
228
inv = Inventory(work_inv.root.file_id)
231
for path, entry in work_inv.iter_entries():
232
## TODO: Check that the file kind has not changed from the previous
233
## revision of this file (if any).
235
p = branch.abspath(path)
236
file_id = entry.file_id
237
mutter('commit prep file %s, id %r ' % (p, file_id))
239
if specific_files and not is_inside_any(specific_files, path):
240
mutter(' skipping file excluded from commit')
241
if basis_inv.has_id(file_id):
242
# carry over with previous state
243
inv.add(basis_inv[file_id].copy())
245
# omit this from committed inventory
249
if not work_tree.has_id(file_id):
251
print('deleted %s%s' % (path, kind_marker(entry.kind)))
253
mutter(" file is missing, removing from inventory")
254
missing_ids.append(file_id)
257
# this is present in the new inventory; may be new, modified or
259
old_ie = basis_inv.has_id(file_id) and basis_inv[file_id]
265
old_kind = old_ie.kind
266
if old_kind != entry.kind:
267
raise BzrError("entry %r changed kind from %r to %r"
268
% (file_id, old_kind, entry.kind))
270
if entry.kind == 'directory':
272
raise BzrError("%s is entered as directory but not a directory"
274
elif entry.kind == 'file':
276
raise BzrError("%s is entered as file but is not a file" % quotefn(p))
278
new_sha1 = work_tree.get_file_sha1(file_id)
281
and old_ie.text_sha1 == new_sha1):
282
## assert content == basis.get_file(file_id).read()
283
entry.text_id = old_ie.text_id
284
entry.text_sha1 = new_sha1
285
entry.text_size = old_ie.text_size
286
mutter(' unchanged from previous text_id {%s}' %
289
content = file(p, 'rb').read()
291
# calculate the sha again, just in case the file contents
292
# changed since we updated the cache
293
entry.text_sha1 = sha_string(content)
294
entry.text_size = len(content)
296
entry.text_id = gen_file_id(entry.name)
297
branch.text_store.add(content, entry.text_id)
298
mutter(' stored with text_id {%s}' % entry.text_id)
301
marked = path + kind_marker(entry.kind)
303
print 'added', marked
305
elif old_ie == entry:
307
elif (old_ie.name == entry.name
308
and old_ie.parent_id == entry.parent_id):
309
print 'modified', marked
312
print 'renamed', marked
315
return missing_ids, inv, any_changes
added
removed
bzrlib/commit.py
