1
# Copyright (C) 2005, 2006, 2007 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
17
from cStringIO import StringIO
19
from bzrlib.lazy_import import lazy_import
20
lazy_import(globals(), """
40
revision as _mod_revision,
46
from bzrlib.bundle import serializer
47
from bzrlib.revisiontree import RevisionTree
48
from bzrlib.store.versioned import VersionedFileStore
49
from bzrlib.store.text import TextStore
50
from bzrlib.testament import Testament
51
from bzrlib.util import bencode
54
from bzrlib.decorators import needs_read_lock, needs_write_lock
55
from bzrlib.inter import InterObject
56
from bzrlib.inventory import Inventory, InventoryDirectory, ROOT_ID
57
from bzrlib.symbol_versioning import (
60
from bzrlib.trace import mutter, mutter_callsite, note, warning
63
# Old formats display a warning, but only once
64
_deprecation_warning_done = False
67
class CommitBuilder(object):
68
"""Provides an interface to build up a commit.
70
This allows describing a tree to be committed without needing to
71
know the internals of the format of the repository.
74
# all clients should supply tree roots.
75
record_root_entry = True
76
# the default CommitBuilder does not manage trees whose root is versioned.
77
_versioned_root = False
79
def __init__(self, repository, parents, config, timestamp=None,
80
timezone=None, committer=None, revprops=None,
82
"""Initiate a CommitBuilder.
84
:param repository: Repository to commit to.
85
:param parents: Revision ids of the parents of the new revision.
86
:param config: Configuration to use.
87
:param timestamp: Optional timestamp recorded for commit.
88
:param timezone: Optional timezone for timestamp.
89
:param committer: Optional committer to set for commit.
90
:param revprops: Optional dictionary of revision properties.
91
:param revision_id: Optional revision id.
96
self._committer = self._config.username()
98
assert isinstance(committer, basestring), type(committer)
99
self._committer = committer
101
self.new_inventory = Inventory(None)
102
self._new_revision_id = revision_id
103
self.parents = parents
104
self.repository = repository
107
if revprops is not None:
108
self._revprops.update(revprops)
110
if timestamp is None:
111
timestamp = time.time()
112
# Restrict resolution to 1ms
113
self._timestamp = round(timestamp, 3)
116
self._timezone = osutils.local_time_offset()
118
self._timezone = int(timezone)
120
self._generate_revision_if_needed()
121
self.__heads = graph.HeadsCache(repository.get_graph()).heads
123
def commit(self, message):
124
"""Make the actual commit.
126
:return: The revision id of the recorded revision.
128
rev = _mod_revision.Revision(
129
timestamp=self._timestamp,
130
timezone=self._timezone,
131
committer=self._committer,
133
inventory_sha1=self.inv_sha1,
134
revision_id=self._new_revision_id,
135
properties=self._revprops)
136
rev.parent_ids = self.parents
137
self.repository.add_revision(self._new_revision_id, rev,
138
self.new_inventory, self._config)
139
self.repository.commit_write_group()
140
return self._new_revision_id
143
"""Abort the commit that is being built.
145
self.repository.abort_write_group()
147
def revision_tree(self):
148
"""Return the tree that was just committed.
150
After calling commit() this can be called to get a RevisionTree
151
representing the newly committed tree. This is preferred to
152
calling Repository.revision_tree() because that may require
153
deserializing the inventory, while we already have a copy in
156
return RevisionTree(self.repository, self.new_inventory,
157
self._new_revision_id)
159
def finish_inventory(self):
160
"""Tell the builder that the inventory is finished."""
161
if self.new_inventory.root is None:
162
raise AssertionError('Root entry should be supplied to'
163
' record_entry_contents, as of bzr 0.10.',
164
DeprecationWarning, stacklevel=2)
165
self.new_inventory.add(InventoryDirectory(ROOT_ID, '', None))
166
self.new_inventory.revision_id = self._new_revision_id
167
self.inv_sha1 = self.repository.add_inventory(
168
self._new_revision_id,
173
def _gen_revision_id(self):
174
"""Return new revision-id."""
175
return generate_ids.gen_revision_id(self._config.username(),
178
def _generate_revision_if_needed(self):
179
"""Create a revision id if None was supplied.
181
If the repository can not support user-specified revision ids
182
they should override this function and raise CannotSetRevisionId
183
if _new_revision_id is not None.
185
:raises: CannotSetRevisionId
187
if self._new_revision_id is None:
188
self._new_revision_id = self._gen_revision_id()
189
self.random_revid = True
191
self.random_revid = False
193
def _heads(self, file_id, revision_ids):
194
"""Calculate the graph heads for revision_ids in the graph of file_id.
196
This can use either a per-file graph or a global revision graph as we
197
have an identity relationship between the two graphs.
199
return self.__heads(revision_ids)
201
def _check_root(self, ie, parent_invs, tree):
202
"""Helper for record_entry_contents.
204
:param ie: An entry being added.
205
:param parent_invs: The inventories of the parent revisions of the
207
:param tree: The tree that is being committed.
209
# In this revision format, root entries have no knit or weave When
210
# serializing out to disk and back in root.revision is always
212
ie.revision = self._new_revision_id
214
def _get_delta(self, ie, basis_inv, path):
215
"""Get a delta against the basis inventory for ie."""
216
if ie.file_id not in basis_inv:
218
return (None, path, ie.file_id, ie)
219
elif ie != basis_inv[ie.file_id]:
221
# TODO: avoid tis id2path call.
222
return (basis_inv.id2path(ie.file_id), path, ie.file_id, ie)
227
def record_entry_contents(self, ie, parent_invs, path, tree,
229
"""Record the content of ie from tree into the commit if needed.
231
Side effect: sets ie.revision when unchanged
233
:param ie: An inventory entry present in the commit.
234
:param parent_invs: The inventories of the parent revisions of the
236
:param path: The path the entry is at in the tree.
237
:param tree: The tree which contains this entry and should be used to
239
:param content_summary: Summary data from the tree about the paths
240
content - stat, length, exec, sha/link target. This is only
241
accessed when the entry has a revision of None - that is when it is
242
a candidate to commit.
243
:return: A tuple (change_delta, version_recorded). change_delta is
244
an inventory_delta change for this entry against the basis tree of
245
the commit, or None if no change occured against the basis tree.
246
version_recorded is True if a new version of the entry has been
247
recorded. For instance, committing a merge where a file was only
248
changed on the other side will return (delta, False).
250
if self.new_inventory.root is None:
251
if ie.parent_id is not None:
252
raise errors.RootMissing()
253
self._check_root(ie, parent_invs, tree)
254
if ie.revision is None:
255
kind = content_summary[0]
257
# ie is carried over from a prior commit
259
# XXX: repository specific check for nested tree support goes here - if
260
# the repo doesn't want nested trees we skip it ?
261
if (kind == 'tree-reference' and
262
not self.repository._format.supports_tree_reference):
263
# mismatch between commit builder logic and repository:
264
# this needs the entry creation pushed down into the builder.
265
raise NotImplementedError('Missing repository subtree support.')
266
self.new_inventory.add(ie)
268
# TODO: slow, take it out of the inner loop.
270
basis_inv = parent_invs[0]
272
basis_inv = Inventory(root_id=None)
274
# ie.revision is always None if the InventoryEntry is considered
275
# for committing. We may record the previous parents revision if the
276
# content is actually unchanged against a sole head.
277
if ie.revision is not None:
278
if not self._versioned_root and path == '':
279
# repositories that do not version the root set the root's
280
# revision to the new commit even when no change occurs, and
281
# this masks when a change may have occurred against the basis,
282
# so calculate if one happened.
283
if ie.file_id in basis_inv:
284
delta = (basis_inv.id2path(ie.file_id), path,
288
delta = (None, path, ie.file_id, ie)
291
# we don't need to commit this, because the caller already
292
# determined that an existing revision of this file is
294
return None, (ie.revision == self._new_revision_id)
295
# XXX: Friction: parent_candidates should return a list not a dict
296
# so that we don't have to walk the inventories again.
297
parent_candiate_entries = ie.parent_candidates(parent_invs)
298
head_set = self._heads(ie.file_id, parent_candiate_entries.keys())
300
for inv in parent_invs:
301
if ie.file_id in inv:
302
old_rev = inv[ie.file_id].revision
303
if old_rev in head_set:
304
heads.append(inv[ie.file_id].revision)
305
head_set.remove(inv[ie.file_id].revision)
308
# now we check to see if we need to write a new record to the
310
# We write a new entry unless there is one head to the ancestors, and
311
# the kind-derived content is unchanged.
313
# Cheapest check first: no ancestors, or more the one head in the
314
# ancestors, we write a new node.
318
# There is a single head, look it up for comparison
319
parent_entry = parent_candiate_entries[heads[0]]
320
# if the non-content specific data has changed, we'll be writing a
322
if (parent_entry.parent_id != ie.parent_id or
323
parent_entry.name != ie.name):
325
# now we need to do content specific checks:
327
# if the kind changed the content obviously has
328
if kind != parent_entry.kind:
331
assert content_summary[2] is not None, \
332
"Files must not have executable = None"
334
if (# if the file length changed we have to store:
335
parent_entry.text_size != content_summary[1] or
336
# if the exec bit has changed we have to store:
337
parent_entry.executable != content_summary[2]):
339
elif parent_entry.text_sha1 == content_summary[3]:
340
# all meta and content is unchanged (using a hash cache
341
# hit to check the sha)
342
ie.revision = parent_entry.revision
343
ie.text_size = parent_entry.text_size
344
ie.text_sha1 = parent_entry.text_sha1
345
ie.executable = parent_entry.executable
346
return self._get_delta(ie, basis_inv, path), False
348
# Either there is only a hash change(no hash cache entry,
349
# or same size content change), or there is no change on
351
# Provide the parent's hash to the store layer, so that the
352
# content is unchanged we will not store a new node.
353
nostore_sha = parent_entry.text_sha1
355
# We want to record a new node regardless of the presence or
356
# absence of a content change in the file.
358
ie.executable = content_summary[2]
359
lines = tree.get_file(ie.file_id, path).readlines()
361
ie.text_sha1, ie.text_size = self._add_text_to_weave(
362
ie.file_id, lines, heads, nostore_sha)
363
except errors.ExistingContent:
364
# Turns out that the file content was unchanged, and we were
365
# only going to store a new node if it was changed. Carry over
367
ie.revision = parent_entry.revision
368
ie.text_size = parent_entry.text_size
369
ie.text_sha1 = parent_entry.text_sha1
370
ie.executable = parent_entry.executable
371
return self._get_delta(ie, basis_inv, path), False
372
elif kind == 'directory':
374
# all data is meta here, nothing specific to directory, so
376
ie.revision = parent_entry.revision
377
return self._get_delta(ie, basis_inv, path), False
379
self._add_text_to_weave(ie.file_id, lines, heads, None)
380
elif kind == 'symlink':
381
current_link_target = content_summary[3]
383
# symlink target is not generic metadata, check if it has
385
if current_link_target != parent_entry.symlink_target:
388
# unchanged, carry over.
389
ie.revision = parent_entry.revision
390
ie.symlink_target = parent_entry.symlink_target
391
return self._get_delta(ie, basis_inv, path), False
392
ie.symlink_target = current_link_target
394
self._add_text_to_weave(ie.file_id, lines, heads, None)
395
elif kind == 'tree-reference':
397
if content_summary[3] != parent_entry.reference_revision:
400
# unchanged, carry over.
401
ie.reference_revision = parent_entry.reference_revision
402
ie.revision = parent_entry.revision
403
return self._get_delta(ie, basis_inv, path), False
404
ie.reference_revision = content_summary[3]
406
self._add_text_to_weave(ie.file_id, lines, heads, None)
408
raise NotImplementedError('unknown kind')
409
ie.revision = self._new_revision_id
410
return self._get_delta(ie, basis_inv, path), True
412
def _add_text_to_weave(self, file_id, new_lines, parents, nostore_sha):
413
versionedfile = self.repository.weave_store.get_weave_or_empty(
414
file_id, self.repository.get_transaction())
415
# Don't change this to add_lines - add_lines_with_ghosts is cheaper
416
# than add_lines, and allows committing when a parent is ghosted for
418
# Note: as we read the content directly from the tree, we know its not
419
# been turned into unicode or badly split - but a broken tree
420
# implementation could give us bad output from readlines() so this is
421
# not a guarantee of safety. What would be better is always checking
422
# the content during test suite execution. RBC 20070912
424
return versionedfile.add_lines_with_ghosts(
425
self._new_revision_id, parents, new_lines,
426
nostore_sha=nostore_sha, random_id=self.random_revid,
427
check_content=False)[0:2]
429
versionedfile.clear_cache()
432
class RootCommitBuilder(CommitBuilder):
433
"""This commitbuilder actually records the root id"""
435
# the root entry gets versioned properly by this builder.
436
_versioned_root = True
438
def _check_root(self, ie, parent_invs, tree):
439
"""Helper for record_entry_contents.
441
:param ie: An entry being added.
442
:param parent_invs: The inventories of the parent revisions of the
444
:param tree: The tree that is being committed.
448
######################################################################
451
class Repository(object):
452
"""Repository holding history for one or more branches.
454
The repository holds and retrieves historical information including
455
revisions and file history. It's normally accessed only by the Branch,
456
which views a particular line of development through that history.
458
The Repository builds on top of Stores and a Transport, which respectively
459
describe the disk data format and the way of accessing the (possibly
463
# What class to use for a CommitBuilder. Often its simpler to change this
464
# in a Repository class subclass rather than to override
465
# get_commit_builder.
466
_commit_builder_class = CommitBuilder
467
# The search regex used by xml based repositories to determine what things
468
# where changed in a single commit.
469
_file_ids_altered_regex = lazy_regex.lazy_compile(
470
r'file_id="(?P<file_id>[^"]+)"'
471
r'.* revision="(?P<revision_id>[^"]+)"'
474
def abort_write_group(self):
475
"""Commit the contents accrued within the current write group.
477
:seealso: start_write_group.
479
if self._write_group is not self.get_transaction():
480
# has an unlock or relock occured ?
481
raise errors.BzrError('mismatched lock context and write group.')
482
self._abort_write_group()
483
self._write_group = None
485
def _abort_write_group(self):
486
"""Template method for per-repository write group cleanup.
488
This is called during abort before the write group is considered to be
489
finished and should cleanup any internal state accrued during the write
490
group. There is no requirement that data handed to the repository be
491
*not* made available - this is not a rollback - but neither should any
492
attempt be made to ensure that data added is fully commited. Abort is
493
invoked when an error has occured so futher disk or network operations
494
may not be possible or may error and if possible should not be
498
def add_inventory(self, revision_id, inv, parents):
499
"""Add the inventory inv to the repository as revision_id.
501
:param parents: The revision ids of the parents that revision_id
502
is known to have and are in the repository already.
504
:returns: The validator(which is a sha1 digest, though what is sha'd is
505
repository format specific) of the serialized inventory.
507
assert self.is_in_write_group()
508
_mod_revision.check_not_reserved_id(revision_id)
509
assert inv.revision_id is None or inv.revision_id == revision_id, \
510
"Mismatch between inventory revision" \
511
" id and insertion revid (%r, %r)" % (inv.revision_id, revision_id)
512
assert inv.root is not None
513
inv_lines = self._serialise_inventory_to_lines(inv)
514
inv_vf = self.get_inventory_weave()
515
return self._inventory_add_lines(inv_vf, revision_id, parents,
516
inv_lines, check_content=False)
518
def _inventory_add_lines(self, inv_vf, revision_id, parents, lines,
520
"""Store lines in inv_vf and return the sha1 of the inventory."""
522
for parent in parents:
524
final_parents.append(parent)
525
return inv_vf.add_lines(revision_id, final_parents, lines,
526
check_content=check_content)[0]
528
def add_revision(self, revision_id, rev, inv=None, config=None):
529
"""Add rev to the revision store as revision_id.
531
:param revision_id: the revision id to use.
532
:param rev: The revision object.
533
:param inv: The inventory for the revision. if None, it will be looked
534
up in the inventory storer
535
:param config: If None no digital signature will be created.
536
If supplied its signature_needed method will be used
537
to determine if a signature should be made.
539
# TODO: jam 20070210 Shouldn't we check rev.revision_id and
541
_mod_revision.check_not_reserved_id(revision_id)
542
if config is not None and config.signature_needed():
544
inv = self.get_inventory(revision_id)
545
plaintext = Testament(rev, inv).as_short_text()
546
self.store_revision_signature(
547
gpg.GPGStrategy(config), plaintext, revision_id)
548
if not revision_id in self.get_inventory_weave():
550
raise errors.WeaveRevisionNotPresent(revision_id,
551
self.get_inventory_weave())
553
# yes, this is not suitable for adding with ghosts.
554
self.add_inventory(revision_id, inv, rev.parent_ids)
555
self._revision_store.add_revision(rev, self.get_transaction())
557
def _add_revision_text(self, revision_id, text):
558
revision = self._revision_store._serializer.read_revision_from_string(
560
self._revision_store._add_revision(revision, StringIO(text),
561
self.get_transaction())
563
def all_revision_ids(self):
564
"""Returns a list of all the revision ids in the repository.
566
This is deprecated because code should generally work on the graph
567
reachable from a particular revision, and ignore any other revisions
568
that might be present. There is no direct replacement method.
570
if 'evil' in debug.debug_flags:
571
mutter_callsite(2, "all_revision_ids is linear with history.")
572
return self._all_revision_ids()
574
def _all_revision_ids(self):
575
"""Returns a list of all the revision ids in the repository.
577
These are in as much topological order as the underlying store can
580
raise NotImplementedError(self._all_revision_ids)
582
def break_lock(self):
583
"""Break a lock if one is present from another instance.
585
Uses the ui factory to ask for confirmation if the lock may be from
588
self.control_files.break_lock()
591
def _eliminate_revisions_not_present(self, revision_ids):
592
"""Check every revision id in revision_ids to see if we have it.
594
Returns a set of the present revisions.
597
for id in revision_ids:
598
if self.has_revision(id):
603
def create(a_bzrdir):
604
"""Construct the current default format repository in a_bzrdir."""
605
return RepositoryFormat.get_default_format().initialize(a_bzrdir)
607
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
608
"""instantiate a Repository.
610
:param _format: The format of the repository on disk.
611
:param a_bzrdir: The BzrDir of the repository.
613
In the future we will have a single api for all stores for
614
getting file texts, inventories and revisions, then
615
this construct will accept instances of those things.
617
super(Repository, self).__init__()
618
self._format = _format
619
# the following are part of the public API for Repository:
620
self.bzrdir = a_bzrdir
621
self.control_files = control_files
622
self._revision_store = _revision_store
623
# backwards compatibility
624
self.weave_store = text_store
626
self._reconcile_does_inventory_gc = True
627
self._reconcile_fixes_text_parents = False
628
self._reconcile_backsup_inventory = True
629
# not right yet - should be more semantically clear ?
631
self.control_store = control_store
632
self.control_weaves = control_store
633
# TODO: make sure to construct the right store classes, etc, depending
634
# on whether escaping is required.
635
self._warn_if_deprecated()
636
self._write_group = None
637
self.base = control_files._transport.base
640
return '%s(%r)' % (self.__class__.__name__,
643
def has_same_location(self, other):
644
"""Returns a boolean indicating if this repository is at the same
645
location as another repository.
647
This might return False even when two repository objects are accessing
648
the same physical repository via different URLs.
650
if self.__class__ is not other.__class__:
652
return (self.control_files._transport.base ==
653
other.control_files._transport.base)
655
def is_in_write_group(self):
656
"""Return True if there is an open write group.
658
:seealso: start_write_group.
660
return self._write_group is not None
663
return self.control_files.is_locked()
665
def is_write_locked(self):
666
"""Return True if this object is write locked."""
667
return self.is_locked() and self.control_files._lock_mode == 'w'
669
def lock_write(self, token=None):
670
"""Lock this repository for writing.
672
This causes caching within the repository obejct to start accumlating
673
data during reads, and allows a 'write_group' to be obtained. Write
674
groups must be used for actual data insertion.
676
:param token: if this is already locked, then lock_write will fail
677
unless the token matches the existing lock.
678
:returns: a token if this instance supports tokens, otherwise None.
679
:raises TokenLockingNotSupported: when a token is given but this
680
instance doesn't support using token locks.
681
:raises MismatchedToken: if the specified token doesn't match the token
682
of the existing lock.
683
:seealso: start_write_group.
685
A token should be passed in if you know that you have locked the object
686
some other way, and need to synchronise this object's state with that
689
XXX: this docstring is duplicated in many places, e.g. lockable_files.py
691
result = self.control_files.lock_write(token=token)
696
self.control_files.lock_read()
699
def get_physical_lock_status(self):
700
return self.control_files.get_physical_lock_status()
702
def leave_lock_in_place(self):
703
"""Tell this repository not to release the physical lock when this
706
If lock_write doesn't return a token, then this method is not supported.
708
self.control_files.leave_in_place()
710
def dont_leave_lock_in_place(self):
711
"""Tell this repository to release the physical lock when this
712
object is unlocked, even if it didn't originally acquire it.
714
If lock_write doesn't return a token, then this method is not supported.
716
self.control_files.dont_leave_in_place()
719
def gather_stats(self, revid=None, committers=None):
720
"""Gather statistics from a revision id.
722
:param revid: The revision id to gather statistics from, if None, then
723
no revision specific statistics are gathered.
724
:param committers: Optional parameter controlling whether to grab
725
a count of committers from the revision specific statistics.
726
:return: A dictionary of statistics. Currently this contains:
727
committers: The number of committers if requested.
728
firstrev: A tuple with timestamp, timezone for the penultimate left
729
most ancestor of revid, if revid is not the NULL_REVISION.
730
latestrev: A tuple with timestamp, timezone for revid, if revid is
731
not the NULL_REVISION.
732
revisions: The total revision count in the repository.
733
size: An estimate disk size of the repository in bytes.
736
if revid and committers:
737
result['committers'] = 0
738
if revid and revid != _mod_revision.NULL_REVISION:
740
all_committers = set()
741
revisions = self.get_ancestry(revid)
742
# pop the leading None
744
first_revision = None
746
# ignore the revisions in the middle - just grab first and last
747
revisions = revisions[0], revisions[-1]
748
for revision in self.get_revisions(revisions):
749
if not first_revision:
750
first_revision = revision
752
all_committers.add(revision.committer)
753
last_revision = revision
755
result['committers'] = len(all_committers)
756
result['firstrev'] = (first_revision.timestamp,
757
first_revision.timezone)
758
result['latestrev'] = (last_revision.timestamp,
759
last_revision.timezone)
761
# now gather global repository information
762
if self.bzrdir.root_transport.listable():
763
c, t = self._revision_store.total_size(self.get_transaction())
764
result['revisions'] = c
768
def find_branches(self, using=False):
769
"""Find branches underneath this repository.
771
This will include branches inside other branches.
773
:param using: If True, list only branches using this repository.
775
if using and not self.is_shared():
777
return [self.bzrdir.open_branch()]
778
except errors.NotBranchError:
780
class Evaluator(object):
783
self.first_call = True
785
def __call__(self, bzrdir):
786
# On the first call, the parameter is always the bzrdir
787
# containing the current repo.
788
if not self.first_call:
790
repository = bzrdir.open_repository()
791
except errors.NoRepositoryPresent:
794
return False, (None, repository)
795
self.first_call = False
797
value = (bzrdir.open_branch(), None)
798
except errors.NotBranchError:
803
for branch, repository in bzrdir.BzrDir.find_bzrdirs(
804
self.bzrdir.root_transport, evaluate=Evaluator()):
805
if branch is not None:
806
branches.append(branch)
807
if not using and repository is not None:
808
branches.extend(repository.find_branches())
811
def get_data_stream(self, revision_ids):
812
raise NotImplementedError(self.get_data_stream)
814
def get_data_stream_for_search(self, search_result):
815
"""Get a data stream that can be inserted to a repository.
817
:param search_result: A bzrlib.graph.SearchResult selecting the
819
:return: A data stream that can be inserted into a repository using
822
raise NotImplementedError(self.get_data_stream_for_search)
824
def insert_data_stream(self, stream):
825
"""XXX What does this really do?
827
Is it a substitute for fetch?
828
Should it manage its own write group ?
830
for item_key, bytes in stream:
831
if item_key[0] == 'file':
832
(file_id,) = item_key[1:]
833
knit = self.weave_store.get_weave_or_empty(
834
file_id, self.get_transaction())
835
elif item_key == ('inventory',):
836
knit = self.get_inventory_weave()
837
elif item_key == ('revisions',):
838
knit = self._revision_store.get_revision_file(
839
self.get_transaction())
840
elif item_key == ('signatures',):
841
knit = self._revision_store.get_signature_file(
842
self.get_transaction())
844
raise errors.RepositoryDataStreamError(
845
"Unrecognised data stream key '%s'" % (item_key,))
846
decoded_list = bencode.bdecode(bytes)
847
format = decoded_list.pop(0)
850
for version, options, parents, some_bytes in decoded_list:
851
data_list.append((version, options, len(some_bytes), parents))
852
knit_bytes += some_bytes
853
buffer = StringIO(knit_bytes)
854
def reader_func(count):
858
return buffer.read(count)
859
knit.insert_data_stream(
860
(format, data_list, reader_func))
863
def search_missing_revision_ids(self, other, revision_id=None, find_ghosts=True):
864
"""Return the revision ids that other has that this does not.
866
These are returned in topological order.
868
revision_id: only return revision ids included by revision_id.
870
return InterRepository.get(other, self).search_missing_revision_ids(
871
revision_id, find_ghosts)
873
@deprecated_method(symbol_versioning.one_two)
875
def missing_revision_ids(self, other, revision_id=None, find_ghosts=True):
876
"""Return the revision ids that other has that this does not.
878
These are returned in topological order.
880
revision_id: only return revision ids included by revision_id.
882
keys = self.search_missing_revision_ids(
883
other, revision_id, find_ghosts).get_keys()
886
parents = other.get_graph().get_parent_map(keys)
889
return tsort.topo_sort(parents)
893
"""Open the repository rooted at base.
895
For instance, if the repository is at URL/.bzr/repository,
896
Repository.open(URL) -> a Repository instance.
898
control = bzrdir.BzrDir.open(base)
899
return control.open_repository()
901
def copy_content_into(self, destination, revision_id=None):
902
"""Make a complete copy of the content in self into destination.
904
This is a destructive operation! Do not use it on existing
907
return InterRepository.get(self, destination).copy_content(revision_id)
909
def commit_write_group(self):
910
"""Commit the contents accrued within the current write group.
912
:seealso: start_write_group.
914
if self._write_group is not self.get_transaction():
915
# has an unlock or relock occured ?
916
raise errors.BzrError('mismatched lock context %r and '
918
(self.get_transaction(), self._write_group))
919
self._commit_write_group()
920
self._write_group = None
922
def _commit_write_group(self):
923
"""Template method for per-repository write group cleanup.
925
This is called before the write group is considered to be
926
finished and should ensure that all data handed to the repository
927
for writing during the write group is safely committed (to the
928
extent possible considering file system caching etc).
931
def fetch(self, source, revision_id=None, pb=None, find_ghosts=False):
932
"""Fetch the content required to construct revision_id from source.
934
If revision_id is None all content is copied.
935
:param find_ghosts: Find and copy revisions in the source that are
936
ghosts in the target (and not reachable directly by walking out to
937
the first-present revision in target from revision_id).
939
# fast path same-url fetch operations
940
if self.has_same_location(source):
941
# check that last_revision is in 'from' and then return a
943
if (revision_id is not None and
944
not _mod_revision.is_null(revision_id)):
945
self.get_revision(revision_id)
947
inter = InterRepository.get(source, self)
949
return inter.fetch(revision_id=revision_id, pb=pb, find_ghosts=find_ghosts)
950
except NotImplementedError:
951
raise errors.IncompatibleRepositories(source, self)
953
def create_bundle(self, target, base, fileobj, format=None):
954
return serializer.write_bundle(self, target, base, fileobj, format)
956
def get_commit_builder(self, branch, parents, config, timestamp=None,
957
timezone=None, committer=None, revprops=None,
959
"""Obtain a CommitBuilder for this repository.
961
:param branch: Branch to commit to.
962
:param parents: Revision ids of the parents of the new revision.
963
:param config: Configuration to use.
964
:param timestamp: Optional timestamp recorded for commit.
965
:param timezone: Optional timezone for timestamp.
966
:param committer: Optional committer to set for commit.
967
:param revprops: Optional dictionary of revision properties.
968
:param revision_id: Optional revision id.
970
result = self._commit_builder_class(self, parents, config,
971
timestamp, timezone, committer, revprops, revision_id)
972
self.start_write_group()
976
if (self.control_files._lock_count == 1 and
977
self.control_files._lock_mode == 'w'):
978
if self._write_group is not None:
979
self.abort_write_group()
980
self.control_files.unlock()
981
raise errors.BzrError(
982
'Must end write groups before releasing write locks.')
983
self.control_files.unlock()
986
def clone(self, a_bzrdir, revision_id=None):
987
"""Clone this repository into a_bzrdir using the current format.
989
Currently no check is made that the format of this repository and
990
the bzrdir format are compatible. FIXME RBC 20060201.
992
:return: The newly created destination repository.
994
# TODO: deprecate after 0.16; cloning this with all its settings is
995
# probably not very useful -- mbp 20070423
996
dest_repo = self._create_sprouting_repo(a_bzrdir, shared=self.is_shared())
997
self.copy_content_into(dest_repo, revision_id)
1000
def start_write_group(self):
1001
"""Start a write group in the repository.
1003
Write groups are used by repositories which do not have a 1:1 mapping
1004
between file ids and backend store to manage the insertion of data from
1005
both fetch and commit operations.
1007
A write lock is required around the start_write_group/commit_write_group
1008
for the support of lock-requiring repository formats.
1010
One can only insert data into a repository inside a write group.
1014
if not self.is_write_locked():
1015
raise errors.NotWriteLocked(self)
1016
if self._write_group:
1017
raise errors.BzrError('already in a write group')
1018
self._start_write_group()
1019
# so we can detect unlock/relock - the write group is now entered.
1020
self._write_group = self.get_transaction()
1022
def _start_write_group(self):
1023
"""Template method for per-repository write group startup.
1025
This is called before the write group is considered to be
1030
def sprout(self, to_bzrdir, revision_id=None):
1031
"""Create a descendent repository for new development.
1033
Unlike clone, this does not copy the settings of the repository.
1035
dest_repo = self._create_sprouting_repo(to_bzrdir, shared=False)
1036
dest_repo.fetch(self, revision_id=revision_id)
1039
def _create_sprouting_repo(self, a_bzrdir, shared):
1040
if not isinstance(a_bzrdir._format, self.bzrdir._format.__class__):
1041
# use target default format.
1042
dest_repo = a_bzrdir.create_repository()
1044
# Most control formats need the repository to be specifically
1045
# created, but on some old all-in-one formats it's not needed
1047
dest_repo = self._format.initialize(a_bzrdir, shared=shared)
1048
except errors.UninitializableFormat:
1049
dest_repo = a_bzrdir.open_repository()
1053
def has_revision(self, revision_id):
1054
"""True if this repository has a copy of the revision."""
1055
return revision_id in self.has_revisions((revision_id,))
1057
def has_revisions(self, revision_ids):
1058
"""Probe to find out the presence of multiple revisions.
1060
:param revision_ids: An iterable of revision_ids.
1061
:return: A set of the revision_ids that were present.
1063
raise NotImplementedError(self.has_revisions)
1065
return self._revision_store.has_revision_id(revision_id,
1066
self.get_transaction())
1069
def get_revision(self, revision_id):
1070
"""Return the Revision object for a named revision."""
1071
return self.get_revisions([revision_id])[0]
1074
def get_revision_reconcile(self, revision_id):
1075
"""'reconcile' helper routine that allows access to a revision always.
1077
This variant of get_revision does not cross check the weave graph
1078
against the revision one as get_revision does: but it should only
1079
be used by reconcile, or reconcile-alike commands that are correcting
1080
or testing the revision graph.
1082
return self._get_revisions([revision_id])[0]
1085
def get_revisions(self, revision_ids):
1086
"""Get many revisions at once."""
1087
return self._get_revisions(revision_ids)
1090
def _get_revisions(self, revision_ids):
1091
"""Core work logic to get many revisions without sanity checks."""
1092
for rev_id in revision_ids:
1093
if not rev_id or not isinstance(rev_id, basestring):
1094
raise errors.InvalidRevisionId(revision_id=rev_id, branch=self)
1095
revs = self._revision_store.get_revisions(revision_ids,
1096
self.get_transaction())
1098
assert not isinstance(rev.revision_id, unicode)
1099
for parent_id in rev.parent_ids:
1100
assert not isinstance(parent_id, unicode)
1104
def get_revision_xml(self, revision_id):
1105
# TODO: jam 20070210 This shouldn't be necessary since get_revision
1106
# would have already do it.
1107
# TODO: jam 20070210 Just use _serializer.write_revision_to_string()
1108
rev = self.get_revision(revision_id)
1109
rev_tmp = StringIO()
1110
# the current serializer..
1111
self._revision_store._serializer.write_revision(rev, rev_tmp)
1113
return rev_tmp.getvalue()
1116
def get_deltas_for_revisions(self, revisions):
1117
"""Produce a generator of revision deltas.
1119
Note that the input is a sequence of REVISIONS, not revision_ids.
1120
Trees will be held in memory until the generator exits.
1121
Each delta is relative to the revision's lefthand predecessor.
1123
required_trees = set()
1124
for revision in revisions:
1125
required_trees.add(revision.revision_id)
1126
required_trees.update(revision.parent_ids[:1])
1127
trees = dict((t.get_revision_id(), t) for
1128
t in self.revision_trees(required_trees))
1129
for revision in revisions:
1130
if not revision.parent_ids:
1131
old_tree = self.revision_tree(None)
1133
old_tree = trees[revision.parent_ids[0]]
1134
yield trees[revision.revision_id].changes_from(old_tree)
1137
def get_revision_delta(self, revision_id):
1138
"""Return the delta for one revision.
1140
The delta is relative to the left-hand predecessor of the
1143
r = self.get_revision(revision_id)
1144
return list(self.get_deltas_for_revisions([r]))[0]
1147
def store_revision_signature(self, gpg_strategy, plaintext, revision_id):
1148
signature = gpg_strategy.sign(plaintext)
1149
self.add_signature_text(revision_id, signature)
1152
def add_signature_text(self, revision_id, signature):
1153
self._revision_store.add_revision_signature_text(revision_id,
1155
self.get_transaction())
1157
def find_text_key_references(self):
1158
"""Find the text key references within the repository.
1160
:return: a dictionary mapping (file_id, revision_id) tuples to altered file-ids to an iterable of
1161
revision_ids. Each altered file-ids has the exact revision_ids that
1162
altered it listed explicitly.
1163
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
1164
to whether they were referred to by the inventory of the
1165
revision_id that they contain. The inventory texts from all present
1166
revision ids are assessed to generate this report.
1168
revision_ids = self.all_revision_ids()
1169
w = self.get_inventory_weave()
1170
pb = ui.ui_factory.nested_progress_bar()
1172
return self._find_text_key_references_from_xml_inventory_lines(
1173
w.iter_lines_added_or_present_in_versions(revision_ids, pb=pb))
1177
def _find_text_key_references_from_xml_inventory_lines(self,
1179
"""Core routine for extracting references to texts from inventories.
1181
This performs the translation of xml lines to revision ids.
1183
:param line_iterator: An iterator of lines, origin_version_id
1184
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
1185
to whether they were referred to by the inventory of the
1186
revision_id that they contain. Note that if that revision_id was
1187
not part of the line_iterator's output then False will be given -
1188
even though it may actually refer to that key.
1190
if not self._serializer.support_altered_by_hack:
1191
raise AssertionError(
1192
"_find_text_key_references_from_xml_inventory_lines only "
1193
"supported for branches which store inventory as unnested xml"
1194
", not on %r" % self)
1197
# this code needs to read every new line in every inventory for the
1198
# inventories [revision_ids]. Seeing a line twice is ok. Seeing a line
1199
# not present in one of those inventories is unnecessary but not
1200
# harmful because we are filtering by the revision id marker in the
1201
# inventory lines : we only select file ids altered in one of those
1202
# revisions. We don't need to see all lines in the inventory because
1203
# only those added in an inventory in rev X can contain a revision=X
1205
unescape_revid_cache = {}
1206
unescape_fileid_cache = {}
1208
# jam 20061218 In a big fetch, this handles hundreds of thousands
1209
# of lines, so it has had a lot of inlining and optimizing done.
1210
# Sorry that it is a little bit messy.
1211
# Move several functions to be local variables, since this is a long
1213
search = self._file_ids_altered_regex.search
1214
unescape = _unescape_xml
1215
setdefault = result.setdefault
1216
for line, version_id in line_iterator:
1217
match = search(line)
1220
# One call to match.group() returning multiple items is quite a
1221
# bit faster than 2 calls to match.group() each returning 1
1222
file_id, revision_id = match.group('file_id', 'revision_id')
1224
# Inlining the cache lookups helps a lot when you make 170,000
1225
# lines and 350k ids, versus 8.4 unique ids.
1226
# Using a cache helps in 2 ways:
1227
# 1) Avoids unnecessary decoding calls
1228
# 2) Re-uses cached strings, which helps in future set and
1230
# (2) is enough that removing encoding entirely along with
1231
# the cache (so we are using plain strings) results in no
1232
# performance improvement.
1234
revision_id = unescape_revid_cache[revision_id]
1236
unescaped = unescape(revision_id)
1237
unescape_revid_cache[revision_id] = unescaped
1238
revision_id = unescaped
1240
# Note that unconditionally unescaping means that we deserialise
1241
# every fileid, which for general 'pull' is not great, but we don't
1242
# really want to have some many fulltexts that this matters anyway.
1245
file_id = unescape_fileid_cache[file_id]
1247
unescaped = unescape(file_id)
1248
unescape_fileid_cache[file_id] = unescaped
1251
key = (file_id, revision_id)
1252
setdefault(key, False)
1253
if revision_id == version_id:
1257
def _find_file_ids_from_xml_inventory_lines(self, line_iterator,
1259
"""Helper routine for fileids_altered_by_revision_ids.
1261
This performs the translation of xml lines to revision ids.
1263
:param line_iterator: An iterator of lines, origin_version_id
1264
:param revision_ids: The revision ids to filter for. This should be a
1265
set or other type which supports efficient __contains__ lookups, as
1266
the revision id from each parsed line will be looked up in the
1267
revision_ids filter.
1268
:return: a dictionary mapping altered file-ids to an iterable of
1269
revision_ids. Each altered file-ids has the exact revision_ids that
1270
altered it listed explicitly.
1273
setdefault = result.setdefault
1274
for file_id, revision_id in \
1275
self._find_text_key_references_from_xml_inventory_lines(
1276
line_iterator).iterkeys():
1277
# once data is all ensured-consistent; then this is
1278
# if revision_id == version_id
1279
if revision_id in revision_ids:
1280
setdefault(file_id, set()).add(revision_id)
1283
def fileids_altered_by_revision_ids(self, revision_ids):
1284
"""Find the file ids and versions affected by revisions.
1286
:param revisions: an iterable containing revision ids.
1287
:return: a dictionary mapping altered file-ids to an iterable of
1288
revision_ids. Each altered file-ids has the exact revision_ids that
1289
altered it listed explicitly.
1291
selected_revision_ids = set(revision_ids)
1292
w = self.get_inventory_weave()
1293
pb = ui.ui_factory.nested_progress_bar()
1295
return self._find_file_ids_from_xml_inventory_lines(
1296
w.iter_lines_added_or_present_in_versions(
1297
selected_revision_ids, pb=pb),
1298
selected_revision_ids)
1302
def iter_files_bytes(self, desired_files):
1303
"""Iterate through file versions.
1305
Files will not necessarily be returned in the order they occur in
1306
desired_files. No specific order is guaranteed.
1308
Yields pairs of identifier, bytes_iterator. identifier is an opaque
1309
value supplied by the caller as part of desired_files. It should
1310
uniquely identify the file version in the caller's context. (Examples:
1311
an index number or a TreeTransform trans_id.)
1313
bytes_iterator is an iterable of bytestrings for the file. The
1314
kind of iterable and length of the bytestrings are unspecified, but for
1315
this implementation, it is a list of lines produced by
1316
VersionedFile.get_lines().
1318
:param desired_files: a list of (file_id, revision_id, identifier)
1321
transaction = self.get_transaction()
1322
for file_id, revision_id, callable_data in desired_files:
1324
weave = self.weave_store.get_weave(file_id, transaction)
1325
except errors.NoSuchFile:
1326
raise errors.NoSuchIdInRepository(self, file_id)
1327
yield callable_data, weave.get_lines(revision_id)
1329
def _generate_text_key_index(self, text_key_references=None,
1331
"""Generate a new text key index for the repository.
1333
This is an expensive function that will take considerable time to run.
1335
:return: A dict mapping text keys ((file_id, revision_id) tuples) to a
1336
list of parents, also text keys. When a given key has no parents,
1337
the parents list will be [NULL_REVISION].
1339
# All revisions, to find inventory parents.
1340
if ancestors is None:
1341
graph = self.get_graph()
1342
ancestors = graph.get_parent_map(self.all_revision_ids())
1343
if text_key_references is None:
1344
text_key_references = self.find_text_key_references()
1345
pb = ui.ui_factory.nested_progress_bar()
1347
return self._do_generate_text_key_index(ancestors,
1348
text_key_references, pb)
1352
def _do_generate_text_key_index(self, ancestors, text_key_references, pb):
1353
"""Helper for _generate_text_key_index to avoid deep nesting."""
1354
revision_order = tsort.topo_sort(ancestors)
1355
invalid_keys = set()
1357
for revision_id in revision_order:
1358
revision_keys[revision_id] = set()
1359
text_count = len(text_key_references)
1360
# a cache of the text keys to allow reuse; costs a dict of all the
1361
# keys, but saves a 2-tuple for every child of a given key.
1363
for text_key, valid in text_key_references.iteritems():
1365
invalid_keys.add(text_key)
1367
revision_keys[text_key[1]].add(text_key)
1368
text_key_cache[text_key] = text_key
1369
del text_key_references
1371
text_graph = graph.Graph(graph.DictParentsProvider(text_index))
1372
NULL_REVISION = _mod_revision.NULL_REVISION
1373
# Set a cache with a size of 10 - this suffices for bzr.dev but may be
1374
# too small for large or very branchy trees. However, for 55K path
1375
# trees, it would be easy to use too much memory trivially. Ideally we
1376
# could gauge this by looking at available real memory etc, but this is
1377
# always a tricky proposition.
1378
inventory_cache = lru_cache.LRUCache(10)
1379
batch_size = 10 # should be ~150MB on a 55K path tree
1380
batch_count = len(revision_order) / batch_size + 1
1382
pb.update("Calculating text parents.", processed_texts, text_count)
1383
for offset in xrange(batch_count):
1384
to_query = revision_order[offset * batch_size:(offset + 1) *
1388
for rev_tree in self.revision_trees(to_query):
1389
revision_id = rev_tree.get_revision_id()
1390
parent_ids = ancestors[revision_id]
1391
for text_key in revision_keys[revision_id]:
1392
pb.update("Calculating text parents.", processed_texts)
1393
processed_texts += 1
1394
candidate_parents = []
1395
for parent_id in parent_ids:
1396
parent_text_key = (text_key[0], parent_id)
1398
check_parent = parent_text_key not in \
1399
revision_keys[parent_id]
1401
# the parent parent_id is a ghost:
1402
check_parent = False
1403
# truncate the derived graph against this ghost.
1404
parent_text_key = None
1406
# look at the parent commit details inventories to
1407
# determine possible candidates in the per file graph.
1410
inv = inventory_cache[parent_id]
1412
inv = self.revision_tree(parent_id).inventory
1413
inventory_cache[parent_id] = inv
1414
parent_entry = inv._byid.get(text_key[0], None)
1415
if parent_entry is not None:
1417
text_key[0], parent_entry.revision)
1419
parent_text_key = None
1420
if parent_text_key is not None:
1421
candidate_parents.append(
1422
text_key_cache[parent_text_key])
1423
parent_heads = text_graph.heads(candidate_parents)
1424
new_parents = list(parent_heads)
1425
new_parents.sort(key=lambda x:candidate_parents.index(x))
1426
if new_parents == []:
1427
new_parents = [NULL_REVISION]
1428
text_index[text_key] = new_parents
1430
for text_key in invalid_keys:
1431
text_index[text_key] = [NULL_REVISION]
1434
def item_keys_introduced_by(self, revision_ids, _files_pb=None):
1435
"""Get an iterable listing the keys of all the data introduced by a set
1438
The keys will be ordered so that the corresponding items can be safely
1439
fetched and inserted in that order.
1441
:returns: An iterable producing tuples of (knit-kind, file-id,
1442
versions). knit-kind is one of 'file', 'inventory', 'signatures',
1443
'revisions'. file-id is None unless knit-kind is 'file'.
1445
# XXX: it's a bit weird to control the inventory weave caching in this
1446
# generator. Ideally the caching would be done in fetch.py I think. Or
1447
# maybe this generator should explicitly have the contract that it
1448
# should not be iterated until the previously yielded item has been
1451
inv_w = self.get_inventory_weave()
1452
inv_w.enable_cache()
1454
# file ids that changed
1455
file_ids = self.fileids_altered_by_revision_ids(revision_ids)
1457
num_file_ids = len(file_ids)
1458
for file_id, altered_versions in file_ids.iteritems():
1459
if _files_pb is not None:
1460
_files_pb.update("fetch texts", count, num_file_ids)
1462
yield ("file", file_id, altered_versions)
1463
# We're done with the files_pb. Note that it finished by the caller,
1464
# just as it was created by the caller.
1468
yield ("inventory", None, revision_ids)
1472
revisions_with_signatures = set()
1473
for rev_id in revision_ids:
1475
self.get_signature_text(rev_id)
1476
except errors.NoSuchRevision:
1480
revisions_with_signatures.add(rev_id)
1482
yield ("signatures", None, revisions_with_signatures)
1485
yield ("revisions", None, revision_ids)
1488
def get_inventory_weave(self):
1489
return self.control_weaves.get_weave('inventory',
1490
self.get_transaction())
1493
def get_inventory(self, revision_id):
1494
"""Get Inventory object by revision id."""
1495
return self.iter_inventories([revision_id]).next()
1497
def iter_inventories(self, revision_ids):
1498
"""Get many inventories by revision_ids.
1500
This will buffer some or all of the texts used in constructing the
1501
inventories in memory, but will only parse a single inventory at a
1504
:return: An iterator of inventories.
1506
assert None not in revision_ids
1507
assert _mod_revision.NULL_REVISION not in revision_ids
1508
return self._iter_inventories(revision_ids)
1510
def _iter_inventories(self, revision_ids):
1511
"""single-document based inventory iteration."""
1512
texts = self.get_inventory_weave().get_texts(revision_ids)
1513
for text, revision_id in zip(texts, revision_ids):
1514
yield self.deserialise_inventory(revision_id, text)
1516
def deserialise_inventory(self, revision_id, xml):
1517
"""Transform the xml into an inventory object.
1519
:param revision_id: The expected revision id of the inventory.
1520
:param xml: A serialised inventory.
1522
result = self._serializer.read_inventory_from_string(xml, revision_id)
1523
if result.revision_id != revision_id:
1524
raise AssertionError('revision id mismatch %s != %s' % (
1525
result.revision_id, revision_id))
1528
def serialise_inventory(self, inv):
1529
return self._serializer.write_inventory_to_string(inv)
1531
def _serialise_inventory_to_lines(self, inv):
1532
return self._serializer.write_inventory_to_lines(inv)
1534
def get_serializer_format(self):
1535
return self._serializer.format_num
1538
def get_inventory_xml(self, revision_id):
1539
"""Get inventory XML as a file object."""
1541
assert isinstance(revision_id, str), type(revision_id)
1542
iw = self.get_inventory_weave()
1543
return iw.get_text(revision_id)
1545
raise errors.HistoryMissing(self, 'inventory', revision_id)
1548
def get_inventory_sha1(self, revision_id):
1549
"""Return the sha1 hash of the inventory entry
1551
return self.get_revision(revision_id).inventory_sha1
1554
@deprecated_method(symbol_versioning.one_four)
1555
def get_revision_graph(self, revision_id=None):
1556
"""Return a dictionary containing the revision graph.
1558
NB: This method should not be used as it accesses the entire graph all
1559
at once, which is much more data than most operations should require.
1561
:param revision_id: The revision_id to get a graph from. If None, then
1562
the entire revision graph is returned. This is a deprecated mode of
1563
operation and will be removed in the future.
1564
:return: a dictionary of revision_id->revision_parents_list.
1566
raise NotImplementedError(self.get_revision_graph)
1569
@deprecated_method(symbol_versioning.one_three)
1570
def get_revision_graph_with_ghosts(self, revision_ids=None):
1571
"""Return a graph of the revisions with ghosts marked as applicable.
1573
:param revision_ids: an iterable of revisions to graph or None for all.
1574
:return: a Graph object with the graph reachable from revision_ids.
1576
if 'evil' in debug.debug_flags:
1578
"get_revision_graph_with_ghosts scales with size of history.")
1579
result = deprecated_graph.Graph()
1580
if not revision_ids:
1581
pending = set(self.all_revision_ids())
1584
pending = set(revision_ids)
1585
# special case NULL_REVISION
1586
if _mod_revision.NULL_REVISION in pending:
1587
pending.remove(_mod_revision.NULL_REVISION)
1588
required = set(pending)
1591
revision_id = pending.pop()
1593
rev = self.get_revision(revision_id)
1594
except errors.NoSuchRevision:
1595
if revision_id in required:
1598
result.add_ghost(revision_id)
1600
for parent_id in rev.parent_ids:
1601
# is this queued or done ?
1602
if (parent_id not in pending and
1603
parent_id not in done):
1605
pending.add(parent_id)
1606
result.add_node(revision_id, rev.parent_ids)
1607
done.add(revision_id)
1610
def iter_reverse_revision_history(self, revision_id):
1611
"""Iterate backwards through revision ids in the lefthand history
1613
:param revision_id: The revision id to start with. All its lefthand
1614
ancestors will be traversed.
1616
graph = self.get_graph()
1617
next_id = revision_id
1619
if next_id in (None, _mod_revision.NULL_REVISION):
1622
parents = graph.get_parent_map([next_id])[next_id]
1623
if len(parents) == 0:
1626
next_id = parents[0]
1629
def get_revision_inventory(self, revision_id):
1630
"""Return inventory of a past revision."""
1631
# TODO: Unify this with get_inventory()
1632
# bzr 0.0.6 and later imposes the constraint that the inventory_id
1633
# must be the same as its revision, so this is trivial.
1634
if revision_id is None:
1635
# This does not make sense: if there is no revision,
1636
# then it is the current tree inventory surely ?!
1637
# and thus get_root_id() is something that looks at the last
1638
# commit on the branch, and the get_root_id is an inventory check.
1639
raise NotImplementedError
1640
# return Inventory(self.get_root_id())
1642
return self.get_inventory(revision_id)
1645
def is_shared(self):
1646
"""Return True if this repository is flagged as a shared repository."""
1647
raise NotImplementedError(self.is_shared)
1650
def reconcile(self, other=None, thorough=False):
1651
"""Reconcile this repository."""
1652
from bzrlib.reconcile import RepoReconciler
1653
reconciler = RepoReconciler(self, thorough=thorough)
1654
reconciler.reconcile()
1657
def _refresh_data(self):
1658
"""Helper called from lock_* to ensure coherency with disk.
1660
The default implementation does nothing; it is however possible
1661
for repositories to maintain loaded indices across multiple locks
1662
by checking inside their implementation of this method to see
1663
whether their indices are still valid. This depends of course on
1664
the disk format being validatable in this manner.
1668
def revision_tree(self, revision_id):
1669
"""Return Tree for a revision on this branch.
1671
`revision_id` may be None for the empty tree revision.
1673
# TODO: refactor this to use an existing revision object
1674
# so we don't need to read it in twice.
1675
if revision_id is None or revision_id == _mod_revision.NULL_REVISION:
1676
return RevisionTree(self, Inventory(root_id=None),
1677
_mod_revision.NULL_REVISION)
1679
inv = self.get_revision_inventory(revision_id)
1680
return RevisionTree(self, inv, revision_id)
1683
def revision_trees(self, revision_ids):
1684
"""Return Tree for a revision on this branch.
1686
`revision_id` may not be None or 'null:'"""
1687
inventories = self.iter_inventories(revision_ids)
1688
for inv in inventories:
1689
yield RevisionTree(self, inv, inv.revision_id)
1692
def get_ancestry(self, revision_id, topo_sorted=True):
1693
"""Return a list of revision-ids integrated by a revision.
1695
The first element of the list is always None, indicating the origin
1696
revision. This might change when we have history horizons, or
1697
perhaps we should have a new API.
1699
This is topologically sorted.
1701
if _mod_revision.is_null(revision_id):
1703
if not self.has_revision(revision_id):
1704
raise errors.NoSuchRevision(self, revision_id)
1705
w = self.get_inventory_weave()
1706
candidates = w.get_ancestry(revision_id, topo_sorted)
1707
return [None] + candidates # self._eliminate_revisions_not_present(candidates)
1710
"""Compress the data within the repository.
1712
This operation only makes sense for some repository types. For other
1713
types it should be a no-op that just returns.
1715
This stub method does not require a lock, but subclasses should use
1716
@needs_write_lock as this is a long running call its reasonable to
1717
implicitly lock for the user.
1721
def print_file(self, file, revision_id):
1722
"""Print `file` to stdout.
1724
FIXME RBC 20060125 as John Meinel points out this is a bad api
1725
- it writes to stdout, it assumes that that is valid etc. Fix
1726
by creating a new more flexible convenience function.
1728
tree = self.revision_tree(revision_id)
1729
# use inventory as it was in that revision
1730
file_id = tree.inventory.path2id(file)
1732
# TODO: jam 20060427 Write a test for this code path
1733
# it had a bug in it, and was raising the wrong
1735
raise errors.BzrError("%r is not present in revision %s" % (file, revision_id))
1736
tree.print_file(file_id)
1738
def get_transaction(self):
1739
return self.control_files.get_transaction()
1741
def revision_parents(self, revision_id):
1742
return self.get_inventory_weave().parent_names(revision_id)
1744
@deprecated_method(symbol_versioning.one_one)
1745
def get_parents(self, revision_ids):
1746
"""See StackedParentsProvider.get_parents"""
1747
parent_map = self.get_parent_map(revision_ids)
1748
return [parent_map.get(r, None) for r in revision_ids]
1750
def get_parent_map(self, keys):
1751
"""See graph._StackedParentsProvider.get_parent_map"""
1753
for revision_id in keys:
1754
if revision_id == _mod_revision.NULL_REVISION:
1755
parent_map[revision_id] = ()
1758
parent_id_list = self.get_revision(revision_id).parent_ids
1759
except errors.NoSuchRevision:
1762
if len(parent_id_list) == 0:
1763
parent_ids = (_mod_revision.NULL_REVISION,)
1765
parent_ids = tuple(parent_id_list)
1766
parent_map[revision_id] = parent_ids
1769
def _make_parents_provider(self):
1772
def get_graph(self, other_repository=None):
1773
"""Return the graph walker for this repository format"""
1774
parents_provider = self._make_parents_provider()
1775
if (other_repository is not None and
1776
not self.has_same_location(other_repository)):
1777
parents_provider = graph._StackedParentsProvider(
1778
[parents_provider, other_repository._make_parents_provider()])
1779
return graph.Graph(parents_provider)
1781
def _get_versioned_file_checker(self):
1782
"""Return an object suitable for checking versioned files."""
1783
return _VersionedFileChecker(self)
1785
def revision_ids_to_search_result(self, result_set):
1786
"""Convert a set of revision ids to a graph SearchResult."""
1787
result_parents = set()
1788
for parents in self.get_graph().get_parent_map(
1789
result_set).itervalues():
1790
result_parents.update(parents)
1791
included_keys = result_set.intersection(result_parents)
1792
start_keys = result_set.difference(included_keys)
1793
exclude_keys = result_parents.difference(result_set)
1794
result = graph.SearchResult(start_keys, exclude_keys,
1795
len(result_set), result_set)
1799
def set_make_working_trees(self, new_value):
1800
"""Set the policy flag for making working trees when creating branches.
1802
This only applies to branches that use this repository.
1804
The default is 'True'.
1805
:param new_value: True to restore the default, False to disable making
1808
raise NotImplementedError(self.set_make_working_trees)
1810
def make_working_trees(self):
1811
"""Returns the policy for making working trees on new branches."""
1812
raise NotImplementedError(self.make_working_trees)
1815
def sign_revision(self, revision_id, gpg_strategy):
1816
plaintext = Testament.from_revision(self, revision_id).as_short_text()
1817
self.store_revision_signature(gpg_strategy, plaintext, revision_id)
1820
def has_signature_for_revision_id(self, revision_id):
1821
"""Query for a revision signature for revision_id in the repository."""
1822
return self._revision_store.has_signature(revision_id,
1823
self.get_transaction())
1826
def get_signature_text(self, revision_id):
1827
"""Return the text for a signature."""
1828
return self._revision_store.get_signature_text(revision_id,
1829
self.get_transaction())
1832
def check(self, revision_ids=None):
1833
"""Check consistency of all history of given revision_ids.
1835
Different repository implementations should override _check().
1837
:param revision_ids: A non-empty list of revision_ids whose ancestry
1838
will be checked. Typically the last revision_id of a branch.
1840
return self._check(revision_ids)
1842
def _check(self, revision_ids):
1843
result = check.Check(self)
1847
def _warn_if_deprecated(self):
1848
global _deprecation_warning_done
1849
if _deprecation_warning_done:
1851
_deprecation_warning_done = True
1852
warning("Format %s for %s is deprecated - please use 'bzr upgrade' to get better performance"
1853
% (self._format, self.bzrdir.transport.base))
1855
def supports_rich_root(self):
1856
return self._format.rich_root_data
1858
def _check_ascii_revisionid(self, revision_id, method):
1859
"""Private helper for ascii-only repositories."""
1860
# weave repositories refuse to store revisionids that are non-ascii.
1861
if revision_id is not None:
1862
# weaves require ascii revision ids.
1863
if isinstance(revision_id, unicode):
1865
revision_id.encode('ascii')
1866
except UnicodeEncodeError:
1867
raise errors.NonAsciiRevisionId(method, self)
1870
revision_id.decode('ascii')
1871
except UnicodeDecodeError:
1872
raise errors.NonAsciiRevisionId(method, self)
1874
def revision_graph_can_have_wrong_parents(self):
1875
"""Is it possible for this repository to have a revision graph with
1878
If True, then this repository must also implement
1879
_find_inconsistent_revision_parents so that check and reconcile can
1880
check for inconsistencies before proceeding with other checks that may
1881
depend on the revision index being consistent.
1883
raise NotImplementedError(self.revision_graph_can_have_wrong_parents)
1886
# remove these delegates a while after bzr 0.15
1887
def __make_delegated(name, from_module):
1888
def _deprecated_repository_forwarder():
1889
symbol_versioning.warn('%s moved to %s in bzr 0.15'
1890
% (name, from_module),
1893
m = __import__(from_module, globals(), locals(), [name])
1895
return getattr(m, name)
1896
except AttributeError:
1897
raise AttributeError('module %s has no name %s'
1899
globals()[name] = _deprecated_repository_forwarder
1902
'AllInOneRepository',
1903
'WeaveMetaDirRepository',
1904
'PreSplitOutRepositoryFormat',
1905
'RepositoryFormat4',
1906
'RepositoryFormat5',
1907
'RepositoryFormat6',
1908
'RepositoryFormat7',
1910
__make_delegated(_name, 'bzrlib.repofmt.weaverepo')
1914
'RepositoryFormatKnit',
1915
'RepositoryFormatKnit1',
1917
__make_delegated(_name, 'bzrlib.repofmt.knitrepo')
1920
def install_revision(repository, rev, revision_tree):
1921
"""Install all revision data into a repository."""
1922
install_revisions(repository, [(rev, revision_tree, None)])
1925
def install_revisions(repository, iterable, num_revisions=None, pb=None):
1926
"""Install all revision data into a repository.
1928
Accepts an iterable of revision, tree, signature tuples. The signature
1931
repository.start_write_group()
1933
for n, (revision, revision_tree, signature) in enumerate(iterable):
1934
_install_revision(repository, revision, revision_tree, signature)
1936
pb.update('Transferring revisions', n + 1, num_revisions)
1938
repository.abort_write_group()
1941
repository.commit_write_group()
1944
def _install_revision(repository, rev, revision_tree, signature):
1945
"""Install all revision data into a repository."""
1946
present_parents = []
1948
for p_id in rev.parent_ids:
1949
if repository.has_revision(p_id):
1950
present_parents.append(p_id)
1951
parent_trees[p_id] = repository.revision_tree(p_id)
1953
parent_trees[p_id] = repository.revision_tree(None)
1955
inv = revision_tree.inventory
1956
entries = inv.iter_entries()
1957
# backwards compatibility hack: skip the root id.
1958
if not repository.supports_rich_root():
1959
path, root = entries.next()
1960
if root.revision != rev.revision_id:
1961
raise errors.IncompatibleRevision(repr(repository))
1962
# Add the texts that are not already present
1963
for path, ie in entries:
1964
w = repository.weave_store.get_weave_or_empty(ie.file_id,
1965
repository.get_transaction())
1966
if ie.revision not in w:
1968
# FIXME: TODO: The following loop *may* be overlapping/duplicate
1969
# with InventoryEntry.find_previous_heads(). if it is, then there
1970
# is a latent bug here where the parents may have ancestors of each
1972
for revision, tree in parent_trees.iteritems():
1973
if ie.file_id not in tree:
1975
parent_id = tree.inventory[ie.file_id].revision
1976
if parent_id in text_parents:
1978
text_parents.append(parent_id)
1980
vfile = repository.weave_store.get_weave_or_empty(ie.file_id,
1981
repository.get_transaction())
1982
lines = revision_tree.get_file(ie.file_id).readlines()
1983
vfile.add_lines(rev.revision_id, text_parents, lines)
1985
# install the inventory
1986
repository.add_inventory(rev.revision_id, inv, present_parents)
1987
except errors.RevisionAlreadyPresent:
1989
if signature is not None:
1990
repository.add_signature_text(rev.revision_id, signature)
1991
repository.add_revision(rev.revision_id, rev, inv)
1994
class MetaDirRepository(Repository):
1995
"""Repositories in the new meta-dir layout."""
1997
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
1998
super(MetaDirRepository, self).__init__(_format,
2004
dir_mode = self.control_files._dir_mode
2005
file_mode = self.control_files._file_mode
2008
def is_shared(self):
2009
"""Return True if this repository is flagged as a shared repository."""
2010
return self.control_files._transport.has('shared-storage')
2013
def set_make_working_trees(self, new_value):
2014
"""Set the policy flag for making working trees when creating branches.
2016
This only applies to branches that use this repository.
2018
The default is 'True'.
2019
:param new_value: True to restore the default, False to disable making
2024
self.control_files._transport.delete('no-working-trees')
2025
except errors.NoSuchFile:
2028
self.control_files.put_utf8('no-working-trees', '')
2030
def make_working_trees(self):
2031
"""Returns the policy for making working trees on new branches."""
2032
return not self.control_files._transport.has('no-working-trees')
2035
class RepositoryFormatRegistry(registry.Registry):
2036
"""Registry of RepositoryFormats."""
2038
def get(self, format_string):
2039
r = registry.Registry.get(self, format_string)
2045
format_registry = RepositoryFormatRegistry()
2046
"""Registry of formats, indexed by their identifying format string.
2048
This can contain either format instances themselves, or classes/factories that
2049
can be called to obtain one.
2053
#####################################################################
2054
# Repository Formats
2056
class RepositoryFormat(object):
2057
"""A repository format.
2059
Formats provide three things:
2060
* An initialization routine to construct repository data on disk.
2061
* a format string which is used when the BzrDir supports versioned
2063
* an open routine which returns a Repository instance.
2065
There is one and only one Format subclass for each on-disk format. But
2066
there can be one Repository subclass that is used for several different
2067
formats. The _format attribute on a Repository instance can be used to
2068
determine the disk format.
2070
Formats are placed in an dict by their format string for reference
2071
during opening. These should be subclasses of RepositoryFormat
2074
Once a format is deprecated, just deprecate the initialize and open
2075
methods on the format class. Do not deprecate the object, as the
2076
object will be created every system load.
2078
Common instance attributes:
2079
_matchingbzrdir - the bzrdir format that the repository format was
2080
originally written to work with. This can be used if manually
2081
constructing a bzrdir and repository, or more commonly for test suite
2085
# Set to True or False in derived classes. True indicates that the format
2086
# supports ghosts gracefully.
2087
supports_ghosts = None
2088
# Can this repository be given external locations to lookup additional
2089
# data. Set to True or False in derived classes.
2090
supports_external_lookups = None
2093
return "<%s>" % self.__class__.__name__
2095
def __eq__(self, other):
2096
# format objects are generally stateless
2097
return isinstance(other, self.__class__)
2099
def __ne__(self, other):
2100
return not self == other
2103
def find_format(klass, a_bzrdir):
2104
"""Return the format for the repository object in a_bzrdir.
2106
This is used by bzr native formats that have a "format" file in
2107
the repository. Other methods may be used by different types of
2111
transport = a_bzrdir.get_repository_transport(None)
2112
format_string = transport.get("format").read()
2113
return format_registry.get(format_string)
2114
except errors.NoSuchFile:
2115
raise errors.NoRepositoryPresent(a_bzrdir)
2117
raise errors.UnknownFormatError(format=format_string,
2121
def register_format(klass, format):
2122
format_registry.register(format.get_format_string(), format)
2125
def unregister_format(klass, format):
2126
format_registry.remove(format.get_format_string())
2129
def get_default_format(klass):
2130
"""Return the current default format."""
2131
from bzrlib import bzrdir
2132
return bzrdir.format_registry.make_bzrdir('default').repository_format
2134
def _get_control_store(self, repo_transport, control_files):
2135
"""Return the control store for this repository."""
2136
raise NotImplementedError(self._get_control_store)
2138
def get_format_string(self):
2139
"""Return the ASCII format string that identifies this format.
2141
Note that in pre format ?? repositories the format string is
2142
not permitted nor written to disk.
2144
raise NotImplementedError(self.get_format_string)
2146
def get_format_description(self):
2147
"""Return the short description for this format."""
2148
raise NotImplementedError(self.get_format_description)
2150
def _get_revision_store(self, repo_transport, control_files):
2151
"""Return the revision store object for this a_bzrdir."""
2152
raise NotImplementedError(self._get_revision_store)
2154
def _get_text_rev_store(self,
2161
"""Common logic for getting a revision store for a repository.
2163
see self._get_revision_store for the subclass-overridable method to
2164
get the store for a repository.
2166
from bzrlib.store.revision.text import TextRevisionStore
2167
dir_mode = control_files._dir_mode
2168
file_mode = control_files._file_mode
2169
text_store = TextStore(transport.clone(name),
2171
compressed=compressed,
2173
file_mode=file_mode)
2174
_revision_store = TextRevisionStore(text_store, serializer)
2175
return _revision_store
2177
# TODO: this shouldn't be in the base class, it's specific to things that
2178
# use weaves or knits -- mbp 20070207
2179
def _get_versioned_file_store(self,
2184
versionedfile_class=None,
2185
versionedfile_kwargs={},
2187
if versionedfile_class is None:
2188
versionedfile_class = self._versionedfile_class
2189
weave_transport = control_files._transport.clone(name)
2190
dir_mode = control_files._dir_mode
2191
file_mode = control_files._file_mode
2192
return VersionedFileStore(weave_transport, prefixed=prefixed,
2194
file_mode=file_mode,
2195
versionedfile_class=versionedfile_class,
2196
versionedfile_kwargs=versionedfile_kwargs,
2199
def initialize(self, a_bzrdir, shared=False):
2200
"""Initialize a repository of this format in a_bzrdir.
2202
:param a_bzrdir: The bzrdir to put the new repository in it.
2203
:param shared: The repository should be initialized as a sharable one.
2204
:returns: The new repository object.
2206
This may raise UninitializableFormat if shared repository are not
2207
compatible the a_bzrdir.
2209
raise NotImplementedError(self.initialize)
2211
def is_supported(self):
2212
"""Is this format supported?
2214
Supported formats must be initializable and openable.
2215
Unsupported formats may not support initialization or committing or
2216
some other features depending on the reason for not being supported.
2220
def check_conversion_target(self, target_format):
2221
raise NotImplementedError(self.check_conversion_target)
2223
def open(self, a_bzrdir, _found=False):
2224
"""Return an instance of this format for the bzrdir a_bzrdir.
2226
_found is a private parameter, do not use it.
2228
raise NotImplementedError(self.open)
2231
class MetaDirRepositoryFormat(RepositoryFormat):
2232
"""Common base class for the new repositories using the metadir layout."""
2234
rich_root_data = False
2235
supports_tree_reference = False
2236
supports_external_lookups = False
2237
_matchingbzrdir = bzrdir.BzrDirMetaFormat1()
2240
super(MetaDirRepositoryFormat, self).__init__()
2242
def _create_control_files(self, a_bzrdir):
2243
"""Create the required files and the initial control_files object."""
2244
# FIXME: RBC 20060125 don't peek under the covers
2245
# NB: no need to escape relative paths that are url safe.
2246
repository_transport = a_bzrdir.get_repository_transport(self)
2247
control_files = lockable_files.LockableFiles(repository_transport,
2248
'lock', lockdir.LockDir)
2249
control_files.create_lock()
2250
return control_files
2252
def _upload_blank_content(self, a_bzrdir, dirs, files, utf8_files, shared):
2253
"""Upload the initial blank content."""
2254
control_files = self._create_control_files(a_bzrdir)
2255
control_files.lock_write()
2257
control_files._transport.mkdir_multi(dirs,
2258
mode=control_files._dir_mode)
2259
for file, content in files:
2260
control_files.put(file, content)
2261
for file, content in utf8_files:
2262
control_files.put_utf8(file, content)
2264
control_files.put_utf8('shared-storage', '')
2266
control_files.unlock()
2269
# formats which have no format string are not discoverable
2270
# and not independently creatable, so are not registered. They're
2271
# all in bzrlib.repofmt.weaverepo now. When an instance of one of these is
2272
# needed, it's constructed directly by the BzrDir. Non-native formats where
2273
# the repository is not separately opened are similar.
2275
format_registry.register_lazy(
2276
'Bazaar-NG Repository format 7',
2277
'bzrlib.repofmt.weaverepo',
2281
format_registry.register_lazy(
2282
'Bazaar-NG Knit Repository Format 1',
2283
'bzrlib.repofmt.knitrepo',
2284
'RepositoryFormatKnit1',
2287
format_registry.register_lazy(
2288
'Bazaar Knit Repository Format 3 (bzr 0.15)\n',
2289
'bzrlib.repofmt.knitrepo',
2290
'RepositoryFormatKnit3',
2293
format_registry.register_lazy(
2294
'Bazaar Knit Repository Format 4 (bzr 1.0)\n',
2295
'bzrlib.repofmt.knitrepo',
2296
'RepositoryFormatKnit4',
2299
# Pack-based formats. There is one format for pre-subtrees, and one for
2300
# post-subtrees to allow ease of testing.
2301
# NOTE: These are experimental in 0.92. Stable in 1.0 and above
2302
format_registry.register_lazy(
2303
'Bazaar pack repository format 1 (needs bzr 0.92)\n',
2304
'bzrlib.repofmt.pack_repo',
2305
'RepositoryFormatKnitPack1',
2307
format_registry.register_lazy(
2308
'Bazaar pack repository format 1 with subtree support (needs bzr 0.92)\n',
2309
'bzrlib.repofmt.pack_repo',
2310
'RepositoryFormatKnitPack3',
2312
format_registry.register_lazy(
2313
'Bazaar pack repository format 1 with rich root (needs bzr 1.0)\n',
2314
'bzrlib.repofmt.pack_repo',
2315
'RepositoryFormatKnitPack4',
2317
# Development formats.
2319
# development 0 - stub to introduce development versioning scheme.
2320
format_registry.register_lazy(
2321
"Bazaar development format 0 (needs bzr.dev from before 1.3)\n",
2322
'bzrlib.repofmt.pack_repo',
2323
'RepositoryFormatPackDevelopment0',
2325
format_registry.register_lazy(
2326
("Bazaar development format 0 with subtree support "
2327
"(needs bzr.dev from before 1.3)\n"),
2328
'bzrlib.repofmt.pack_repo',
2329
'RepositoryFormatPackDevelopment0Subtree',
2331
# 1.3->1.4 go below here
2334
class InterRepository(InterObject):
2335
"""This class represents operations taking place between two repositories.
2337
Its instances have methods like copy_content and fetch, and contain
2338
references to the source and target repositories these operations can be
2341
Often we will provide convenience methods on 'repository' which carry out
2342
operations with another repository - they will always forward to
2343
InterRepository.get(other).method_name(parameters).
2347
"""The available optimised InterRepository types."""
2349
def copy_content(self, revision_id=None):
2350
raise NotImplementedError(self.copy_content)
2352
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2353
"""Fetch the content required to construct revision_id.
2355
The content is copied from self.source to self.target.
2357
:param revision_id: if None all content is copied, if NULL_REVISION no
2359
:param pb: optional progress bar to use for progress reports. If not
2360
provided a default one will be created.
2362
Returns the copied revision count and the failed revisions in a tuple:
2365
raise NotImplementedError(self.fetch)
2367
def _walk_to_common_revisions(self, revision_ids):
2368
"""Walk out from revision_ids in source to revisions target has.
2370
:param revision_ids: The start point for the search.
2371
:return: A set of revision ids.
2373
graph = self.source.get_graph()
2374
missing_revs = set()
2375
# ensure we don't pay silly lookup costs.
2376
revision_ids = frozenset(revision_ids)
2377
searcher = graph._make_breadth_first_searcher(revision_ids)
2378
null_set = frozenset([_mod_revision.NULL_REVISION])
2381
next_revs, ghosts = searcher.next_with_ghosts()
2382
except StopIteration:
2384
if revision_ids.intersection(ghosts):
2385
absent_ids = set(revision_ids.intersection(ghosts))
2386
# If all absent_ids are present in target, no error is needed.
2387
absent_ids.difference_update(
2388
self.target.has_revisions(absent_ids))
2390
raise errors.NoSuchRevision(self.source, absent_ids.pop())
2391
# we don't care about other ghosts as we can't fetch them and
2392
# haven't been asked to.
2393
next_revs = set(next_revs)
2394
# we always have NULL_REVISION present.
2395
have_revs = self.target.has_revisions(next_revs).union(null_set)
2396
missing_revs.update(next_revs - have_revs)
2397
searcher.stop_searching_any(have_revs)
2398
return searcher.get_result()
2400
@deprecated_method(symbol_versioning.one_two)
2402
def missing_revision_ids(self, revision_id=None, find_ghosts=True):
2403
"""Return the revision ids that source has that target does not.
2405
These are returned in topological order.
2407
:param revision_id: only return revision ids included by this
2409
:param find_ghosts: If True find missing revisions in deep history
2410
rather than just finding the surface difference.
2412
return list(self.search_missing_revision_ids(
2413
revision_id, find_ghosts).get_keys())
2416
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2417
"""Return the revision ids that source has that target does not.
2419
:param revision_id: only return revision ids included by this
2421
:param find_ghosts: If True find missing revisions in deep history
2422
rather than just finding the surface difference.
2423
:return: A bzrlib.graph.SearchResult.
2425
# stop searching at found target revisions.
2426
if not find_ghosts and revision_id is not None:
2427
return self._walk_to_common_revisions([revision_id])
2428
# generic, possibly worst case, slow code path.
2429
target_ids = set(self.target.all_revision_ids())
2430
if revision_id is not None:
2431
source_ids = self.source.get_ancestry(revision_id)
2432
assert source_ids[0] is None
2435
source_ids = self.source.all_revision_ids()
2436
result_set = set(source_ids).difference(target_ids)
2437
return self.source.revision_ids_to_search_result(result_set)
2440
def _same_model(source, target):
2441
"""True if source and target have the same data representation."""
2442
if source.supports_rich_root() != target.supports_rich_root():
2444
if source._serializer != target._serializer:
2449
class InterSameDataRepository(InterRepository):
2450
"""Code for converting between repositories that represent the same data.
2452
Data format and model must match for this to work.
2456
def _get_repo_format_to_test(self):
2457
"""Repository format for testing with.
2459
InterSameData can pull from subtree to subtree and from non-subtree to
2460
non-subtree, so we test this with the richest repository format.
2462
from bzrlib.repofmt import knitrepo
2463
return knitrepo.RepositoryFormatKnit3()
2466
def is_compatible(source, target):
2467
return InterRepository._same_model(source, target)
2470
def copy_content(self, revision_id=None):
2471
"""Make a complete copy of the content in self into destination.
2473
This copies both the repository's revision data, and configuration information
2474
such as the make_working_trees setting.
2476
This is a destructive operation! Do not use it on existing
2479
:param revision_id: Only copy the content needed to construct
2480
revision_id and its parents.
2483
self.target.set_make_working_trees(self.source.make_working_trees())
2484
except NotImplementedError:
2486
# but don't bother fetching if we have the needed data now.
2487
if (revision_id not in (None, _mod_revision.NULL_REVISION) and
2488
self.target.has_revision(revision_id)):
2490
self.target.fetch(self.source, revision_id=revision_id)
2493
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2494
"""See InterRepository.fetch()."""
2495
from bzrlib.fetch import GenericRepoFetcher
2496
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2497
self.source, self.source._format, self.target,
2498
self.target._format)
2499
f = GenericRepoFetcher(to_repository=self.target,
2500
from_repository=self.source,
2501
last_revision=revision_id,
2502
pb=pb, find_ghosts=find_ghosts)
2503
return f.count_copied, f.failed_revisions
2506
class InterWeaveRepo(InterSameDataRepository):
2507
"""Optimised code paths between Weave based repositories.
2509
This should be in bzrlib/repofmt/weaverepo.py but we have not yet
2510
implemented lazy inter-object optimisation.
2514
def _get_repo_format_to_test(self):
2515
from bzrlib.repofmt import weaverepo
2516
return weaverepo.RepositoryFormat7()
2519
def is_compatible(source, target):
2520
"""Be compatible with known Weave formats.
2522
We don't test for the stores being of specific types because that
2523
could lead to confusing results, and there is no need to be
2526
from bzrlib.repofmt.weaverepo import (
2532
return (isinstance(source._format, (RepositoryFormat5,
2534
RepositoryFormat7)) and
2535
isinstance(target._format, (RepositoryFormat5,
2537
RepositoryFormat7)))
2538
except AttributeError:
2542
def copy_content(self, revision_id=None):
2543
"""See InterRepository.copy_content()."""
2544
# weave specific optimised path:
2546
self.target.set_make_working_trees(self.source.make_working_trees())
2547
except NotImplementedError:
2549
# FIXME do not peek!
2550
if self.source.control_files._transport.listable():
2551
pb = ui.ui_factory.nested_progress_bar()
2553
self.target.weave_store.copy_all_ids(
2554
self.source.weave_store,
2556
from_transaction=self.source.get_transaction(),
2557
to_transaction=self.target.get_transaction())
2558
pb.update('copying inventory', 0, 1)
2559
self.target.control_weaves.copy_multi(
2560
self.source.control_weaves, ['inventory'],
2561
from_transaction=self.source.get_transaction(),
2562
to_transaction=self.target.get_transaction())
2563
self.target._revision_store.text_store.copy_all_ids(
2564
self.source._revision_store.text_store,
2569
self.target.fetch(self.source, revision_id=revision_id)
2572
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2573
"""See InterRepository.fetch()."""
2574
from bzrlib.fetch import GenericRepoFetcher
2575
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2576
self.source, self.source._format, self.target, self.target._format)
2577
f = GenericRepoFetcher(to_repository=self.target,
2578
from_repository=self.source,
2579
last_revision=revision_id,
2580
pb=pb, find_ghosts=find_ghosts)
2581
return f.count_copied, f.failed_revisions
2584
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2585
"""See InterRepository.missing_revision_ids()."""
2586
# we want all revisions to satisfy revision_id in source.
2587
# but we don't want to stat every file here and there.
2588
# we want then, all revisions other needs to satisfy revision_id
2589
# checked, but not those that we have locally.
2590
# so the first thing is to get a subset of the revisions to
2591
# satisfy revision_id in source, and then eliminate those that
2592
# we do already have.
2593
# this is slow on high latency connection to self, but as as this
2594
# disk format scales terribly for push anyway due to rewriting
2595
# inventory.weave, this is considered acceptable.
2597
if revision_id is not None:
2598
source_ids = self.source.get_ancestry(revision_id)
2599
assert source_ids[0] is None
2602
source_ids = self.source._all_possible_ids()
2603
source_ids_set = set(source_ids)
2604
# source_ids is the worst possible case we may need to pull.
2605
# now we want to filter source_ids against what we actually
2606
# have in target, but don't try to check for existence where we know
2607
# we do not have a revision as that would be pointless.
2608
target_ids = set(self.target._all_possible_ids())
2609
possibly_present_revisions = target_ids.intersection(source_ids_set)
2610
actually_present_revisions = set(
2611
self.target._eliminate_revisions_not_present(possibly_present_revisions))
2612
required_revisions = source_ids_set.difference(actually_present_revisions)
2613
if revision_id is not None:
2614
# we used get_ancestry to determine source_ids then we are assured all
2615
# revisions referenced are present as they are installed in topological order.
2616
# and the tip revision was validated by get_ancestry.
2617
result_set = required_revisions
2619
# if we just grabbed the possibly available ids, then
2620
# we only have an estimate of whats available and need to validate
2621
# that against the revision records.
2623
self.source._eliminate_revisions_not_present(required_revisions))
2624
return self.source.revision_ids_to_search_result(result_set)
2627
class InterKnitRepo(InterSameDataRepository):
2628
"""Optimised code paths between Knit based repositories."""
2631
def _get_repo_format_to_test(self):
2632
from bzrlib.repofmt import knitrepo
2633
return knitrepo.RepositoryFormatKnit1()
2636
def is_compatible(source, target):
2637
"""Be compatible with known Knit formats.
2639
We don't test for the stores being of specific types because that
2640
could lead to confusing results, and there is no need to be
2643
from bzrlib.repofmt.knitrepo import RepositoryFormatKnit
2645
are_knits = (isinstance(source._format, RepositoryFormatKnit) and
2646
isinstance(target._format, RepositoryFormatKnit))
2647
except AttributeError:
2649
return are_knits and InterRepository._same_model(source, target)
2652
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2653
"""See InterRepository.fetch()."""
2654
from bzrlib.fetch import KnitRepoFetcher
2655
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2656
self.source, self.source._format, self.target, self.target._format)
2657
f = KnitRepoFetcher(to_repository=self.target,
2658
from_repository=self.source,
2659
last_revision=revision_id,
2660
pb=pb, find_ghosts=find_ghosts)
2661
return f.count_copied, f.failed_revisions
2664
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2665
"""See InterRepository.missing_revision_ids()."""
2666
if revision_id is not None:
2667
source_ids = self.source.get_ancestry(revision_id)
2668
assert source_ids[0] is None
2671
source_ids = self.source.all_revision_ids()
2672
source_ids_set = set(source_ids)
2673
# source_ids is the worst possible case we may need to pull.
2674
# now we want to filter source_ids against what we actually
2675
# have in target, but don't try to check for existence where we know
2676
# we do not have a revision as that would be pointless.
2677
target_ids = set(self.target.all_revision_ids())
2678
possibly_present_revisions = target_ids.intersection(source_ids_set)
2679
actually_present_revisions = set(
2680
self.target._eliminate_revisions_not_present(possibly_present_revisions))
2681
required_revisions = source_ids_set.difference(actually_present_revisions)
2682
if revision_id is not None:
2683
# we used get_ancestry to determine source_ids then we are assured all
2684
# revisions referenced are present as they are installed in topological order.
2685
# and the tip revision was validated by get_ancestry.
2686
result_set = required_revisions
2688
# if we just grabbed the possibly available ids, then
2689
# we only have an estimate of whats available and need to validate
2690
# that against the revision records.
2692
self.source._eliminate_revisions_not_present(required_revisions))
2693
return self.source.revision_ids_to_search_result(result_set)
2696
class InterPackRepo(InterSameDataRepository):
2697
"""Optimised code paths between Pack based repositories."""
2700
def _get_repo_format_to_test(self):
2701
from bzrlib.repofmt import pack_repo
2702
return pack_repo.RepositoryFormatKnitPack1()
2705
def is_compatible(source, target):
2706
"""Be compatible with known Pack formats.
2708
We don't test for the stores being of specific types because that
2709
could lead to confusing results, and there is no need to be
2712
from bzrlib.repofmt.pack_repo import RepositoryFormatPack
2714
are_packs = (isinstance(source._format, RepositoryFormatPack) and
2715
isinstance(target._format, RepositoryFormatPack))
2716
except AttributeError:
2718
return are_packs and InterRepository._same_model(source, target)
2721
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2722
"""See InterRepository.fetch()."""
2723
from bzrlib.repofmt.pack_repo import Packer
2724
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2725
self.source, self.source._format, self.target, self.target._format)
2726
self.count_copied = 0
2727
if revision_id is None:
2729
# everything to do - use pack logic
2730
# to fetch from all packs to one without
2731
# inventory parsing etc, IFF nothing to be copied is in the target.
2733
revision_ids = self.source.all_revision_ids()
2734
# implementing the TODO will involve:
2735
# - detecting when all of a pack is selected
2736
# - avoiding as much as possible pre-selection, so the
2737
# more-core routines such as create_pack_from_packs can filter in
2738
# a just-in-time fashion. (though having a HEADS list on a
2739
# repository might make this a lot easier, because we could
2740
# sensibly detect 'new revisions' without doing a full index scan.
2741
elif _mod_revision.is_null(revision_id):
2746
revision_ids = self.search_missing_revision_ids(revision_id,
2747
find_ghosts=find_ghosts).get_keys()
2748
except errors.NoSuchRevision:
2749
raise errors.InstallFailed([revision_id])
2750
packs = self.source._pack_collection.all_packs()
2751
pack = Packer(self.target._pack_collection, packs, '.fetch',
2752
revision_ids).pack()
2753
if pack is not None:
2754
self.target._pack_collection._save_pack_names()
2755
# Trigger an autopack. This may duplicate effort as we've just done
2756
# a pack creation, but for now it is simpler to think about as
2757
# 'upload data, then repack if needed'.
2758
self.target._pack_collection.autopack()
2759
return (pack.get_revision_count(), [])
2764
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2765
"""See InterRepository.missing_revision_ids().
2767
:param find_ghosts: Find ghosts throughout the ancestry of
2770
if not find_ghosts and revision_id is not None:
2771
return self._walk_to_common_revisions([revision_id])
2772
elif revision_id is not None:
2773
source_ids = self.source.get_ancestry(revision_id)
2774
assert source_ids[0] is None
2777
source_ids = self.source.all_revision_ids()
2778
# source_ids is the worst possible case we may need to pull.
2779
# now we want to filter source_ids against what we actually
2780
# have in target, but don't try to check for existence where we know
2781
# we do not have a revision as that would be pointless.
2782
target_ids = set(self.target.all_revision_ids())
2783
result_set = set(source_ids).difference(target_ids)
2784
return self.source.revision_ids_to_search_result(result_set)
2787
class InterModel1and2(InterRepository):
2790
def _get_repo_format_to_test(self):
2794
def is_compatible(source, target):
2795
if not source.supports_rich_root() and target.supports_rich_root():
2801
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2802
"""See InterRepository.fetch()."""
2803
from bzrlib.fetch import Model1toKnit2Fetcher
2804
f = Model1toKnit2Fetcher(to_repository=self.target,
2805
from_repository=self.source,
2806
last_revision=revision_id,
2807
pb=pb, find_ghosts=find_ghosts)
2808
return f.count_copied, f.failed_revisions
2811
def copy_content(self, revision_id=None):
2812
"""Make a complete copy of the content in self into destination.
2814
This is a destructive operation! Do not use it on existing
2817
:param revision_id: Only copy the content needed to construct
2818
revision_id and its parents.
2821
self.target.set_make_working_trees(self.source.make_working_trees())
2822
except NotImplementedError:
2824
# but don't bother fetching if we have the needed data now.
2825
if (revision_id not in (None, _mod_revision.NULL_REVISION) and
2826
self.target.has_revision(revision_id)):
2828
self.target.fetch(self.source, revision_id=revision_id)
2831
class InterKnit1and2(InterKnitRepo):
2834
def _get_repo_format_to_test(self):
2838
def is_compatible(source, target):
2839
"""Be compatible with Knit1 source and Knit3 target"""
2840
from bzrlib.repofmt.knitrepo import RepositoryFormatKnit3
2842
from bzrlib.repofmt.knitrepo import (RepositoryFormatKnit1,
2843
RepositoryFormatKnit3)
2844
from bzrlib.repofmt.pack_repo import (
2845
RepositoryFormatKnitPack1,
2846
RepositoryFormatKnitPack3,
2847
RepositoryFormatPackDevelopment0,
2848
RepositoryFormatPackDevelopment0Subtree,
2851
RepositoryFormatKnit1,
2852
RepositoryFormatKnitPack1,
2853
RepositoryFormatPackDevelopment0,
2856
RepositoryFormatKnit3,
2857
RepositoryFormatKnitPack3,
2858
RepositoryFormatPackDevelopment0Subtree,
2860
return (isinstance(source._format, nosubtrees) and
2861
isinstance(target._format, subtrees))
2862
except AttributeError:
2866
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2867
"""See InterRepository.fetch()."""
2868
from bzrlib.fetch import Knit1to2Fetcher
2869
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2870
self.source, self.source._format, self.target,
2871
self.target._format)
2872
f = Knit1to2Fetcher(to_repository=self.target,
2873
from_repository=self.source,
2874
last_revision=revision_id,
2875
pb=pb, find_ghosts=find_ghosts)
2876
return f.count_copied, f.failed_revisions
2879
class InterDifferingSerializer(InterKnitRepo):
2882
def _get_repo_format_to_test(self):
2886
def is_compatible(source, target):
2887
"""Be compatible with Knit2 source and Knit3 target"""
2888
if source.supports_rich_root() != target.supports_rich_root():
2890
# Ideally, we'd support fetching if the source had no tree references
2891
# even if it supported them...
2892
if (getattr(source, '_format.supports_tree_reference', False) and
2893
not getattr(target, '_format.supports_tree_reference', False)):
2898
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2899
"""See InterRepository.fetch()."""
2900
revision_ids = self.target.search_missing_revision_ids(self.source,
2901
revision_id, find_ghosts=find_ghosts).get_keys()
2902
revision_ids = tsort.topo_sort(
2903
self.source.get_graph().get_parent_map(revision_ids))
2904
def revisions_iterator():
2905
for current_revision_id in revision_ids:
2906
revision = self.source.get_revision(current_revision_id)
2907
tree = self.source.revision_tree(current_revision_id)
2909
signature = self.source.get_signature_text(
2910
current_revision_id)
2911
except errors.NoSuchRevision:
2913
yield revision, tree, signature
2915
my_pb = ui.ui_factory.nested_progress_bar()
2920
install_revisions(self.target, revisions_iterator(),
2921
len(revision_ids), pb)
2923
if my_pb is not None:
2925
return len(revision_ids), 0
2928
class InterRemoteToOther(InterRepository):
2930
def __init__(self, source, target):
2931
InterRepository.__init__(self, source, target)
2932
self._real_inter = None
2935
def is_compatible(source, target):
2936
if not isinstance(source, remote.RemoteRepository):
2938
# Is source's model compatible with target's model?
2939
source._ensure_real()
2940
real_source = source._real_repository
2941
assert not isinstance(real_source, remote.RemoteRepository), (
2942
"We don't support remote repos backed by remote repos yet.")
2943
return InterRepository._same_model(real_source, target)
2946
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2947
"""See InterRepository.fetch()."""
2948
from bzrlib.fetch import RemoteToOtherFetcher
2949
mutter("Using fetch logic to copy between %s(remote) and %s(%s)",
2950
self.source, self.target, self.target._format)
2951
# TODO: jam 20070210 This should be an assert, not a translate
2952
revision_id = osutils.safe_revision_id(revision_id)
2953
f = RemoteToOtherFetcher(to_repository=self.target,
2954
from_repository=self.source,
2955
last_revision=revision_id,
2956
pb=pb, find_ghosts=find_ghosts)
2957
return f.count_copied, f.failed_revisions
2960
def _get_repo_format_to_test(self):
2964
class InterOtherToRemote(InterRepository):
2966
def __init__(self, source, target):
2967
InterRepository.__init__(self, source, target)
2968
self._real_inter = None
2971
def is_compatible(source, target):
2972
if isinstance(target, remote.RemoteRepository):
2976
def _ensure_real_inter(self):
2977
if self._real_inter is None:
2978
self.target._ensure_real()
2979
real_target = self.target._real_repository
2980
self._real_inter = InterRepository.get(self.source, real_target)
2982
def copy_content(self, revision_id=None):
2983
self._ensure_real_inter()
2984
self._real_inter.copy_content(revision_id=revision_id)
2986
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2987
self._ensure_real_inter()
2988
self._real_inter.fetch(revision_id=revision_id, pb=pb,
2989
find_ghosts=find_ghosts)
2992
def _get_repo_format_to_test(self):
2996
InterRepository.register_optimiser(InterDifferingSerializer)
2997
InterRepository.register_optimiser(InterSameDataRepository)
2998
InterRepository.register_optimiser(InterWeaveRepo)
2999
InterRepository.register_optimiser(InterKnitRepo)
3000
InterRepository.register_optimiser(InterModel1and2)
3001
InterRepository.register_optimiser(InterKnit1and2)
3002
InterRepository.register_optimiser(InterPackRepo)
3003
InterRepository.register_optimiser(InterRemoteToOther)
3004
InterRepository.register_optimiser(InterOtherToRemote)
3007
class CopyConverter(object):
3008
"""A repository conversion tool which just performs a copy of the content.
3010
This is slow but quite reliable.
3013
def __init__(self, target_format):
3014
"""Create a CopyConverter.
3016
:param target_format: The format the resulting repository should be.
3018
self.target_format = target_format
3020
def convert(self, repo, pb):
3021
"""Perform the conversion of to_convert, giving feedback via pb.
3023
:param to_convert: The disk object to convert.
3024
:param pb: a progress bar to use for progress information.
3029
# this is only useful with metadir layouts - separated repo content.
3030
# trigger an assertion if not such
3031
repo._format.get_format_string()
3032
self.repo_dir = repo.bzrdir
3033
self.step('Moving repository to repository.backup')
3034
self.repo_dir.transport.move('repository', 'repository.backup')
3035
backup_transport = self.repo_dir.transport.clone('repository.backup')
3036
repo._format.check_conversion_target(self.target_format)
3037
self.source_repo = repo._format.open(self.repo_dir,
3039
_override_transport=backup_transport)
3040
self.step('Creating new repository')
3041
converted = self.target_format.initialize(self.repo_dir,
3042
self.source_repo.is_shared())
3043
converted.lock_write()
3045
self.step('Copying content into repository.')
3046
self.source_repo.copy_content_into(converted)
3049
self.step('Deleting old repository content.')
3050
self.repo_dir.transport.delete_tree('repository.backup')
3051
self.pb.note('repository converted')
3053
def step(self, message):
3054
"""Update the pb by a step."""
3056
self.pb.update(message, self.count, self.total)
3068
def _unescaper(match, _map=_unescape_map):
3069
code = match.group(1)
3073
if not code.startswith('#'):
3075
return unichr(int(code[1:])).encode('utf8')
3081
def _unescape_xml(data):
3082
"""Unescape predefined XML entities in a string of data."""
3084
if _unescape_re is None:
3085
_unescape_re = re.compile('\&([^;]*);')
3086
return _unescape_re.sub(_unescaper, data)
3089
class _VersionedFileChecker(object):
3091
def __init__(self, repository):
3092
self.repository = repository
3093
self.text_index = self.repository._generate_text_key_index()
3095
def calculate_file_version_parents(self, revision_id, file_id):
3096
"""Calculate the correct parents for a file version according to
3099
parent_keys = self.text_index[(file_id, revision_id)]
3100
if parent_keys == [_mod_revision.NULL_REVISION]:
3102
# strip the file_id, for the weave api
3103
return tuple([revision_id for file_id, revision_id in parent_keys])
3105
def check_file_version_parents(self, weave, file_id):
3106
"""Check the parents stored in a versioned file are correct.
3108
It also detects file versions that are not referenced by their
3109
corresponding revision's inventory.
3111
:returns: A tuple of (wrong_parents, dangling_file_versions).
3112
wrong_parents is a dict mapping {revision_id: (stored_parents,
3113
correct_parents)} for each revision_id where the stored parents
3114
are not correct. dangling_file_versions is a set of (file_id,
3115
revision_id) tuples for versions that are present in this versioned
3116
file, but not used by the corresponding inventory.
3119
unused_versions = set()
3120
versions = weave.versions()
3121
parent_map = weave.get_parent_map(versions)
3122
for num, revision_id in enumerate(versions):
3124
correct_parents = self.calculate_file_version_parents(
3125
revision_id, file_id)
3127
# The version is not part of the used keys.
3128
unused_versions.add(revision_id)
3131
knit_parents = tuple(parent_map[revision_id])
3132
except errors.RevisionNotPresent:
3134
if correct_parents != knit_parents:
3135
wrong_parents[revision_id] = (knit_parents, correct_parents)
3136
return wrong_parents, unused_versions
added
removed
bzrlib/repository.py
