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...')
90
pending_merges = branch.pending_merges()
92
missing_ids, new_inv, any_changes = \
93
_gather_commit(branch,
100
if not (any_changes or allow_pointless or pending_merges):
101
raise PointlessCommit()
103
for file_id in missing_ids:
104
# Any files that have been deleted are now removed from the
105
# working inventory. Files that were not selected for commit
106
# are left as they were in the working inventory and ommitted
107
# from the revision inventory.
109
# have to do this later so we don't mess up the iterator.
110
# since parents may be removed before their children we
113
# FIXME: There's probably a better way to do this; perhaps
114
# the workingtree should know how to filter itbranch.
115
if work_inv.has_id(file_id):
116
del work_inv[file_id]
120
rev_id = _gen_revision_id(time.time())
123
inv_tmp = tempfile.TemporaryFile()
124
pack_xml(new_inv, inv_tmp)
126
branch.inventory_store.add(inv_tmp, inv_id)
127
mutter('new inventory_id is {%s}' % inv_id)
129
# We could also just sha hash the inv_tmp file
130
# however, in the case that branch.inventory_store.add()
131
# ever actually does anything special
132
inv_sha1 = branch.get_inventory_sha1(inv_id)
134
branch._write_inventory(work_inv)
136
if timestamp == None:
137
timestamp = time.time()
139
if committer == None:
140
committer = username()
143
timezone = local_time_offset()
145
mutter("building commit log message")
146
rev = Revision(timestamp=timestamp,
151
inventory_sha1=inv_sha1,
155
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.weave import Weave
95
from bzrlib.weavefile import read_weave, write_weave_v5
96
from bzrlib.workingtree import WorkingTree
99
@deprecated_function(zero_seven)
100
def commit(*args, **kwargs):
101
"""Commit a new revision to a branch.
103
Function-style interface for convenience of old callers.
105
New code should use the Commit class instead.
107
## XXX: Remove this in favor of Branch.commit?
108
Commit().commit(*args, **kwargs)
111
class NullCommitReporter(object):
112
"""I report on progress of a commit."""
114
def snapshot_change(self, change, path):
117
def completed(self, revno, rev_id):
120
def deleted(self, file_id):
123
def escaped(self, escape_count, message):
126
def missing(self, path):
130
class ReportCommitToLog(NullCommitReporter):
132
def snapshot_change(self, change, path):
133
note("%s %s", change, path)
135
def completed(self, revno, rev_id):
136
note('committed r%d {%s}', revno, rev_id)
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,
184
"""Commit working copy as a new revision.
186
branch -- the deprecated branch to commit to. New callers should pass in
189
message -- the commit message, a mandatory parameter
191
timestamp -- if not None, seconds-since-epoch for a
192
postdated/predated commit.
194
specific_files -- If true, commit only those files.
196
rev_id -- If set, use this as the new revision id.
197
Useful for test or import commands that need to tightly
198
control what revisions are assigned. If you duplicate
199
a revision id that exists elsewhere it is your own fault.
200
If null (default), a time/random revision id is generated.
202
allow_pointless -- If true (default), commit even if nothing
203
has changed and no merges are recorded.
205
strict -- If true, don't allow a commit if the working tree
206
contains unknown files.
208
revprops -- Properties for new revision
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
228
self.master_branch = None
230
self.specific_files = specific_files
231
self.allow_pointless = allow_pointless
232
self.revprops = {'branch-nick': self.branch.nick}
234
self.revprops.update(revprops)
236
self.work_tree.lock_write()
238
# setup the bound branch variables as needed.
239
self._check_bound_branch()
241
# check for out of date working trees
242
# if we are bound, then self.branch is the master branch and this
243
# test is thus all we need.
244
if self.work_tree.last_revision() != self.branch.last_revision():
245
raise errors.OutOfDateTree(self.work_tree)
248
# raise an exception as soon as we find a single unknown.
249
for unknown in self.work_tree.unknowns():
250
raise StrictCommitFailed()
252
if timestamp is None:
253
self.timestamp = time.time()
255
self.timestamp = long(timestamp)
257
if self.config is None:
258
self.config = bzrlib.config.BranchConfig(self.branch)
261
self.rev_id = _gen_revision_id(self.config, self.timestamp)
265
if committer is None:
266
self.committer = self.config.username()
268
assert isinstance(committer, basestring), type(committer)
269
self.committer = committer
272
self.timezone = local_time_offset()
274
self.timezone = int(timezone)
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
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')
289
self._check_parents_present()
291
self._remove_deleted()
292
self._populate_new_inv()
293
self._store_snapshot()
294
self._report_deletes()
296
if not (self.allow_pointless
297
or len(self.parents) > 1
298
or self.new_inv != self.basis_inv):
299
raise PointlessCommit()
301
if len(list(self.work_tree.iter_conflicts()))>0:
302
raise ConflictsInTree
304
self.inv_sha1 = self.branch.repository.add_inventory(
309
self._make_revision()
310
# revision is in the master branch now.
312
self.branch.append_revision(self.rev_id)
313
# now its in the master branch history.
315
self._update_bound_branch()
316
# now the local branch is up to date
318
self.work_tree.set_pending_merges([])
319
if len(self.parents):
320
precursor = self.parents[0]
323
self.work_tree.set_last_revision(self.rev_id, precursor)
324
# now the work tree is up to date with the branch
326
self.reporter.completed(self.branch.revno()+1, self.rev_id)
327
if self.config.post_commit() is not None:
328
hooks = self.config.post_commit().split(' ')
329
# this would be nicer with twisted.python.reflect.namedAny
331
result = eval(hook + '(branch, rev_id)',
332
{'branch':self.branch,
334
'rev_id':self.rev_id})
336
self._cleanup_bound_branch()
337
self.work_tree.unlock()
339
def _check_bound_branch(self):
340
"""Check to see if the local branch is bound.
342
If it is bound, then most of the commit will actually be
343
done using the remote branch as the target branch.
344
Only at the end will the local branch be updated.
346
# TODO: jam 20051230 Consider a special error for the case
347
# where the local branch is bound, and can't access the
349
self.master_branch = self.branch.get_master_branch()
350
if not self.master_branch:
353
# If the master branch is bound, we must fail
354
master_bound_location = self.master_branch.get_bound_location()
355
if master_bound_location:
356
raise errors.CommitToDoubleBoundBranch(self.branch,
357
self.master_branch, master_bound_location)
359
# TODO: jam 20051230 We could automatically push local
360
# commits to the remote branch if they would fit.
361
# But for now, just require remote to be identical
364
# Make sure the local branch is identical to the master
365
master_rh = self.master_branch.revision_history()
366
local_rh = self.branch.revision_history()
367
if local_rh != master_rh:
368
raise errors.BoundBranchOutOfDate(self.branch,
371
# Now things are ready to change the master branch
373
self.bound_branch = self.branch
374
self.master_branch.lock_write()
375
self.branch = self.master_branch
377
# Check to see if we have any pending merges. If we do
378
# those need to be pushed into the master branch
379
pending_merges = self.work_tree.pending_merges()
381
for revision_id in pending_merges:
382
self.master_branch.repository.fetch(self.bound_branch.repository,
383
revision_id=revision_id)
385
def _cleanup_bound_branch(self):
386
"""Executed at the end of a try/finally to cleanup a bound branch.
388
If the branch wasn't bound, this is a no-op.
389
If it was, it resents self.branch to the local branch, instead
392
if not self.bound_branch:
394
self.branch = self.bound_branch
395
self.master_branch.unlock()
397
def _update_bound_branch(self):
398
"""Update the local bound branch, after commit.
400
This only runs if the commit to the master branch succeeds.
402
if not self.bound_branch:
404
# We always want the local branch to look like the remote one
405
# TODO: jam 20051231 We might want overwrite=True here, but
406
# the local branch should be a prefix of master anyway
407
self.bound_branch.pull(self.master_branch)
409
# TODO: jam 20051231 At this point we probably
410
# want to merge any changes into master branch's
413
def _escape_commit_message(self):
414
"""Replace xml-incompatible control characters."""
415
# Python strings can include characters that can't be
416
# represented in well-formed XML; escape characters that
417
# aren't listed in the XML specification
418
# (http://www.w3.org/TR/REC-xml/#NT-Char).
419
self.message, escape_count = re.subn(
420
u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
421
lambda match: match.group(0).encode('unicode_escape'),
424
self.reporter.escaped(escape_count, self.message)
426
def _gather_parents(self):
427
"""Record the parents of a merge for merge detection."""
428
pending_merges = self.work_tree.pending_merges()
430
self.parent_invs = []
431
self.present_parents = []
432
precursor_id = self.branch.last_revision()
157
precursor_sha1 = branch.get_revision_sha1(precursor_id)
158
rev.parents.append(RevisionReference(precursor_id, precursor_sha1))
159
for merge_rev in pending_merges:
160
rev.parents.append(RevisionReference(merge_rev))
162
rev_tmp = tempfile.TemporaryFile()
163
pack_xml(rev, rev_tmp)
165
branch.revision_store.add(rev_tmp, rev_id)
166
mutter("new revision_id is {%s}" % rev_id)
168
## XXX: Everything up to here can simply be orphaned if we abort
169
## the commit; it will leave junk files behind but that doesn't
172
## TODO: Read back the just-generated changeset, and make sure it
173
## applies and recreates the right state.
175
## TODO: Also calculate and store the inventory SHA1
176
mutter("committing patch r%d" % (branch.revno() + 1))
178
branch.append_revision(rev_id)
180
branch.set_pending_merges([])
183
note("commited r%d" % branch.revno())
189
def _gen_revision_id(when):
434
self.parents.append(precursor_id)
435
self.parents += pending_merges
436
for revision in self.parents:
437
if self.branch.repository.has_revision(revision):
438
inventory = self.branch.repository.get_inventory(revision)
439
self.parent_invs.append(inventory)
440
self.present_parents.append(revision)
442
def _check_parents_present(self):
443
for parent_id in self.parents:
444
mutter('commit parent revision {%s}', parent_id)
445
if not self.branch.repository.has_revision(parent_id):
446
if parent_id == self.branch.last_revision():
447
warning("parent is missing %r", parent_id)
448
raise HistoryMissing(self.branch, 'revision', parent_id)
450
mutter("commit will ghost revision %r", parent_id)
452
def _make_revision(self):
453
"""Record a new revision object for this commit."""
454
rev = Revision(timestamp=self.timestamp,
455
timezone=self.timezone,
456
committer=self.committer,
457
message=self.message,
458
inventory_sha1=self.inv_sha1,
459
revision_id=self.rev_id,
460
properties=self.revprops)
461
rev.parent_ids = self.parents
462
self.branch.repository.add_revision(self.rev_id, rev, self.new_inv, self.config)
464
def _remove_deleted(self):
465
"""Remove deleted files from the working inventories.
467
This is done prior to taking the working inventory as the
468
basis for the new committed inventory.
470
This returns true if any files
471
*that existed in the basis inventory* were deleted.
472
Files that were added and deleted
473
in the working copy don't matter.
475
specific = self.specific_files
477
for path, ie in self.work_inv.iter_entries():
478
if specific and not is_inside_any(specific, path):
480
if not self.work_tree.has_filename(path):
481
self.reporter.missing(path)
482
deleted_ids.append((path, ie.file_id))
484
deleted_ids.sort(reverse=True)
485
for path, file_id in deleted_ids:
486
del self.work_inv[file_id]
487
self.work_tree._write_inventory(self.work_inv)
489
def _store_snapshot(self):
490
"""Pass over inventory and record a snapshot.
492
Entries get a new revision when they are modified in
493
any way, which includes a merge with a new set of
494
parents that have the same entry.
496
# XXX: Need to think more here about when the user has
497
# made a specific decision on a particular value -- c.f.
499
for path, ie in self.new_inv.iter_entries():
500
previous_entries = ie.find_previous_heads(
502
self.weave_store.get_weave_prelude_or_empty(ie.file_id,
503
self.branch.get_transaction()))
504
if ie.revision is None:
505
change = ie.snapshot(self.rev_id, path, previous_entries,
506
self.work_tree, self.weave_store,
507
self.branch.get_transaction())
510
self.reporter.snapshot_change(change, path)
512
def _populate_new_inv(self):
513
"""Build revision inventory.
515
This creates a new empty inventory. Depending on
516
which files are selected for commit, and what is present in the
517
current tree, the new inventory is populated. inventory entries
518
which are candidates for modification have their revision set to
519
None; inventory entries that are carried over untouched have their
520
revision set to their prior value.
522
mutter("Selecting files for commit with filter %s", self.specific_files)
523
self.new_inv = Inventory()
524
for path, new_ie in self.work_inv.iter_entries():
525
file_id = new_ie.file_id
526
mutter('check %s {%s}', path, new_ie.file_id)
527
if self.specific_files:
528
if not is_inside_any(self.specific_files, path):
529
mutter('%s not selected for commit', path)
530
self._carry_entry(file_id)
533
# this is selected, ensure its parents are too.
534
parent_id = new_ie.parent_id
535
while parent_id != ROOT_ID:
536
if not self.new_inv.has_id(parent_id):
537
ie = self._select_entry(self.work_inv[parent_id])
538
mutter('%s selected for commit because of %s',
539
self.new_inv.id2path(parent_id), path)
541
ie = self.new_inv[parent_id]
542
if ie.revision is not None:
544
mutter('%s selected for commit because of %s',
545
self.new_inv.id2path(parent_id), path)
546
parent_id = ie.parent_id
547
mutter('%s selected for commit', path)
548
self._select_entry(new_ie)
550
def _select_entry(self, new_ie):
551
"""Make new_ie be considered for committing."""
557
def _carry_entry(self, file_id):
558
"""Carry the file unchanged from the basis revision."""
559
if self.basis_inv.has_id(file_id):
560
self.new_inv.add(self.basis_inv[file_id].copy())
562
def _report_deletes(self):
563
for file_id in self.basis_inv:
564
if file_id not in self.new_inv:
565
self.reporter.deleted(self.basis_inv.id2path(file_id))
567
def _gen_revision_id(config, when):
190
568
"""Return new revision-id."""
191
from binascii import hexlify
192
from osutils import rand_bytes, compact_date, user_email
194
s = '%s-%s-' % (user_email(), compact_date(when))
569
s = '%s-%s-' % (config.user_email(), compact_date(when))
195
570
s += hexlify(rand_bytes(8))
199
def _gather_commit(branch, work_tree, work_inv, basis_inv, specific_files,
201
"""Build inventory preparatory to commit.
203
Returns missing_ids, new_inv, any_changes.
205
This adds any changed files into the text store, and sets their
206
test-id, sha and size in the returned inventory appropriately.
209
Modified to hold a list of files that have been deleted from
210
the working directory; these should be removed from the
213
from bzrlib.inventory import Inventory
214
from bzrlib.osutils import isdir, isfile, sha_string, quotefn, \
215
local_time_offset, username, kind_marker, is_inside_any
217
from bzrlib.branch import gen_file_id
218
from bzrlib.errors import BzrError
219
from bzrlib.revision import Revision
220
from bzrlib.trace import mutter, note
223
inv = Inventory(work_inv.root.file_id)
226
for path, entry in work_inv.iter_entries():
227
## TODO: Check that the file kind has not changed from the previous
228
## revision of this file (if any).
230
p = branch.abspath(path)
231
file_id = entry.file_id
232
mutter('commit prep file %s, id %r ' % (p, file_id))
234
if specific_files and not is_inside_any(specific_files, path):
235
mutter(' skipping file excluded from commit')
236
if basis_inv.has_id(file_id):
237
# carry over with previous state
238
inv.add(basis_inv[file_id].copy())
240
# omit this from committed inventory
244
if not work_tree.has_id(file_id):
246
print('deleted %s%s' % (path, kind_marker(entry.kind)))
248
mutter(" file is missing, removing from inventory")
249
missing_ids.append(file_id)
252
# this is present in the new inventory; may be new, modified or
254
old_ie = basis_inv.has_id(file_id) and basis_inv[file_id]
260
old_kind = old_ie.kind
261
if old_kind != entry.kind:
262
raise BzrError("entry %r changed kind from %r to %r"
263
% (file_id, old_kind, entry.kind))
265
if entry.kind == 'directory':
267
raise BzrError("%s is entered as directory but not a directory"
269
elif entry.kind == 'file':
271
raise BzrError("%s is entered as file but is not a file" % quotefn(p))
273
new_sha1 = work_tree.get_file_sha1(file_id)
276
and old_ie.text_sha1 == new_sha1):
277
## assert content == basis.get_file(file_id).read()
278
entry.text_id = old_ie.text_id
279
entry.text_sha1 = new_sha1
280
entry.text_size = old_ie.text_size
281
mutter(' unchanged from previous text_id {%s}' %
284
content = file(p, 'rb').read()
286
# calculate the sha again, just in case the file contents
287
# changed since we updated the cache
288
entry.text_sha1 = sha_string(content)
289
entry.text_size = len(content)
291
entry.text_id = gen_file_id(entry.name)
292
branch.text_store.add(content, entry.text_id)
293
mutter(' stored with text_id {%s}' % entry.text_id)
296
marked = path + kind_marker(entry.kind)
298
print 'added', marked
300
elif old_ie == entry:
302
elif (old_ie.name == entry.name
303
and old_ie.parent_id == entry.parent_id):
304
print 'modified', marked
307
print 'renamed', marked
310
return missing_ids, inv, any_changes
added
removed
bzrlib/commit.py
