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
rev.inventory_sha1 = self.add_inventory(revision_id, inv,
556
self._revision_store.add_revision(rev, self.get_transaction())
558
def _add_revision_text(self, revision_id, text):
559
revision = self._revision_store._serializer.read_revision_from_string(
561
self._revision_store._add_revision(revision, StringIO(text),
562
self.get_transaction())
564
def all_revision_ids(self):
565
"""Returns a list of all the revision ids in the repository.
567
This is deprecated because code should generally work on the graph
568
reachable from a particular revision, and ignore any other revisions
569
that might be present. There is no direct replacement method.
571
if 'evil' in debug.debug_flags:
572
mutter_callsite(2, "all_revision_ids is linear with history.")
573
return self._all_revision_ids()
575
def _all_revision_ids(self):
576
"""Returns a list of all the revision ids in the repository.
578
These are in as much topological order as the underlying store can
581
raise NotImplementedError(self._all_revision_ids)
583
def break_lock(self):
584
"""Break a lock if one is present from another instance.
586
Uses the ui factory to ask for confirmation if the lock may be from
589
self.control_files.break_lock()
592
def _eliminate_revisions_not_present(self, revision_ids):
593
"""Check every revision id in revision_ids to see if we have it.
595
Returns a set of the present revisions.
598
for id in revision_ids:
599
if self.has_revision(id):
604
def create(a_bzrdir):
605
"""Construct the current default format repository in a_bzrdir."""
606
return RepositoryFormat.get_default_format().initialize(a_bzrdir)
608
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
609
"""instantiate a Repository.
611
:param _format: The format of the repository on disk.
612
:param a_bzrdir: The BzrDir of the repository.
614
In the future we will have a single api for all stores for
615
getting file texts, inventories and revisions, then
616
this construct will accept instances of those things.
618
super(Repository, self).__init__()
619
self._format = _format
620
# the following are part of the public API for Repository:
621
self.bzrdir = a_bzrdir
622
self.control_files = control_files
623
self._revision_store = _revision_store
624
# backwards compatibility
625
self.weave_store = text_store
627
self._reconcile_does_inventory_gc = True
628
self._reconcile_fixes_text_parents = False
629
self._reconcile_backsup_inventory = True
630
# not right yet - should be more semantically clear ?
632
self.control_store = control_store
633
self.control_weaves = control_store
634
# TODO: make sure to construct the right store classes, etc, depending
635
# on whether escaping is required.
636
self._warn_if_deprecated()
637
self._write_group = None
638
self.base = control_files._transport.base
641
return '%s(%r)' % (self.__class__.__name__,
644
def has_same_location(self, other):
645
"""Returns a boolean indicating if this repository is at the same
646
location as another repository.
648
This might return False even when two repository objects are accessing
649
the same physical repository via different URLs.
651
if self.__class__ is not other.__class__:
653
return (self.control_files._transport.base ==
654
other.control_files._transport.base)
656
def is_in_write_group(self):
657
"""Return True if there is an open write group.
659
:seealso: start_write_group.
661
return self._write_group is not None
664
return self.control_files.is_locked()
666
def is_write_locked(self):
667
"""Return True if this object is write locked."""
668
return self.is_locked() and self.control_files._lock_mode == 'w'
670
def lock_write(self, token=None):
671
"""Lock this repository for writing.
673
This causes caching within the repository obejct to start accumlating
674
data during reads, and allows a 'write_group' to be obtained. Write
675
groups must be used for actual data insertion.
677
:param token: if this is already locked, then lock_write will fail
678
unless the token matches the existing lock.
679
:returns: a token if this instance supports tokens, otherwise None.
680
:raises TokenLockingNotSupported: when a token is given but this
681
instance doesn't support using token locks.
682
:raises MismatchedToken: if the specified token doesn't match the token
683
of the existing lock.
684
:seealso: start_write_group.
686
A token should be passed in if you know that you have locked the object
687
some other way, and need to synchronise this object's state with that
690
XXX: this docstring is duplicated in many places, e.g. lockable_files.py
692
result = self.control_files.lock_write(token=token)
697
self.control_files.lock_read()
700
def get_physical_lock_status(self):
701
return self.control_files.get_physical_lock_status()
703
def leave_lock_in_place(self):
704
"""Tell this repository not to release the physical lock when this
707
If lock_write doesn't return a token, then this method is not supported.
709
self.control_files.leave_in_place()
711
def dont_leave_lock_in_place(self):
712
"""Tell this repository to release the physical lock when this
713
object is unlocked, even if it didn't originally acquire it.
715
If lock_write doesn't return a token, then this method is not supported.
717
self.control_files.dont_leave_in_place()
720
def gather_stats(self, revid=None, committers=None):
721
"""Gather statistics from a revision id.
723
:param revid: The revision id to gather statistics from, if None, then
724
no revision specific statistics are gathered.
725
:param committers: Optional parameter controlling whether to grab
726
a count of committers from the revision specific statistics.
727
:return: A dictionary of statistics. Currently this contains:
728
committers: The number of committers if requested.
729
firstrev: A tuple with timestamp, timezone for the penultimate left
730
most ancestor of revid, if revid is not the NULL_REVISION.
731
latestrev: A tuple with timestamp, timezone for revid, if revid is
732
not the NULL_REVISION.
733
revisions: The total revision count in the repository.
734
size: An estimate disk size of the repository in bytes.
737
if revid and committers:
738
result['committers'] = 0
739
if revid and revid != _mod_revision.NULL_REVISION:
741
all_committers = set()
742
revisions = self.get_ancestry(revid)
743
# pop the leading None
745
first_revision = None
747
# ignore the revisions in the middle - just grab first and last
748
revisions = revisions[0], revisions[-1]
749
for revision in self.get_revisions(revisions):
750
if not first_revision:
751
first_revision = revision
753
all_committers.add(revision.committer)
754
last_revision = revision
756
result['committers'] = len(all_committers)
757
result['firstrev'] = (first_revision.timestamp,
758
first_revision.timezone)
759
result['latestrev'] = (last_revision.timestamp,
760
last_revision.timezone)
762
# now gather global repository information
763
if self.bzrdir.root_transport.listable():
764
c, t = self._revision_store.total_size(self.get_transaction())
765
result['revisions'] = c
769
def find_branches(self, using=False):
770
"""Find branches underneath this repository.
772
This will include branches inside other branches.
774
:param using: If True, list only branches using this repository.
776
if using and not self.is_shared():
778
return [self.bzrdir.open_branch()]
779
except errors.NotBranchError:
781
class Evaluator(object):
784
self.first_call = True
786
def __call__(self, bzrdir):
787
# On the first call, the parameter is always the bzrdir
788
# containing the current repo.
789
if not self.first_call:
791
repository = bzrdir.open_repository()
792
except errors.NoRepositoryPresent:
795
return False, (None, repository)
796
self.first_call = False
798
value = (bzrdir.open_branch(), None)
799
except errors.NotBranchError:
804
for branch, repository in bzrdir.BzrDir.find_bzrdirs(
805
self.bzrdir.root_transport, evaluate=Evaluator()):
806
if branch is not None:
807
branches.append(branch)
808
if not using and repository is not None:
809
branches.extend(repository.find_branches())
812
def get_data_stream(self, revision_ids):
813
raise NotImplementedError(self.get_data_stream)
815
def get_data_stream_for_search(self, search_result):
816
"""Get a data stream that can be inserted to a repository.
818
:param search_result: A bzrlib.graph.SearchResult selecting the
820
:return: A data stream that can be inserted into a repository using
823
raise NotImplementedError(self.get_data_stream_for_search)
825
def insert_data_stream(self, stream):
826
"""XXX What does this really do?
828
Is it a substitute for fetch?
829
Should it manage its own write group ?
831
for item_key, bytes in stream:
832
if item_key[0] == 'file':
833
(file_id,) = item_key[1:]
834
knit = self.weave_store.get_weave_or_empty(
835
file_id, self.get_transaction())
836
elif item_key == ('inventory',):
837
knit = self.get_inventory_weave()
838
elif item_key == ('revisions',):
839
knit = self._revision_store.get_revision_file(
840
self.get_transaction())
841
elif item_key == ('signatures',):
842
knit = self._revision_store.get_signature_file(
843
self.get_transaction())
845
raise errors.RepositoryDataStreamError(
846
"Unrecognised data stream key '%s'" % (item_key,))
847
decoded_list = bencode.bdecode(bytes)
848
format = decoded_list.pop(0)
851
for version, options, parents, some_bytes in decoded_list:
852
data_list.append((version, options, len(some_bytes), parents))
853
knit_bytes += some_bytes
854
buffer = StringIO(knit_bytes)
855
def reader_func(count):
859
return buffer.read(count)
860
knit.insert_data_stream(
861
(format, data_list, reader_func))
864
def search_missing_revision_ids(self, other, revision_id=None, find_ghosts=True):
865
"""Return the revision ids that other has that this does not.
867
These are returned in topological order.
869
revision_id: only return revision ids included by revision_id.
871
return InterRepository.get(other, self).search_missing_revision_ids(
872
revision_id, find_ghosts)
874
@deprecated_method(symbol_versioning.one_two)
876
def missing_revision_ids(self, other, revision_id=None, find_ghosts=True):
877
"""Return the revision ids that other has that this does not.
879
These are returned in topological order.
881
revision_id: only return revision ids included by revision_id.
883
keys = self.search_missing_revision_ids(
884
other, revision_id, find_ghosts).get_keys()
887
parents = other.get_graph().get_parent_map(keys)
890
return tsort.topo_sort(parents)
894
"""Open the repository rooted at base.
896
For instance, if the repository is at URL/.bzr/repository,
897
Repository.open(URL) -> a Repository instance.
899
control = bzrdir.BzrDir.open(base)
900
return control.open_repository()
902
def copy_content_into(self, destination, revision_id=None):
903
"""Make a complete copy of the content in self into destination.
905
This is a destructive operation! Do not use it on existing
908
return InterRepository.get(self, destination).copy_content(revision_id)
910
def commit_write_group(self):
911
"""Commit the contents accrued within the current write group.
913
:seealso: start_write_group.
915
if self._write_group is not self.get_transaction():
916
# has an unlock or relock occured ?
917
raise errors.BzrError('mismatched lock context %r and '
919
(self.get_transaction(), self._write_group))
920
self._commit_write_group()
921
self._write_group = None
923
def _commit_write_group(self):
924
"""Template method for per-repository write group cleanup.
926
This is called before the write group is considered to be
927
finished and should ensure that all data handed to the repository
928
for writing during the write group is safely committed (to the
929
extent possible considering file system caching etc).
932
def fetch(self, source, revision_id=None, pb=None, find_ghosts=False):
933
"""Fetch the content required to construct revision_id from source.
935
If revision_id is None all content is copied.
936
:param find_ghosts: Find and copy revisions in the source that are
937
ghosts in the target (and not reachable directly by walking out to
938
the first-present revision in target from revision_id).
940
# fast path same-url fetch operations
941
if self.has_same_location(source):
942
# check that last_revision is in 'from' and then return a
944
if (revision_id is not None and
945
not _mod_revision.is_null(revision_id)):
946
self.get_revision(revision_id)
948
inter = InterRepository.get(source, self)
950
return inter.fetch(revision_id=revision_id, pb=pb, find_ghosts=find_ghosts)
951
except NotImplementedError:
952
raise errors.IncompatibleRepositories(source, self)
954
def create_bundle(self, target, base, fileobj, format=None):
955
return serializer.write_bundle(self, target, base, fileobj, format)
957
def get_commit_builder(self, branch, parents, config, timestamp=None,
958
timezone=None, committer=None, revprops=None,
960
"""Obtain a CommitBuilder for this repository.
962
:param branch: Branch to commit to.
963
:param parents: Revision ids of the parents of the new revision.
964
:param config: Configuration to use.
965
:param timestamp: Optional timestamp recorded for commit.
966
:param timezone: Optional timezone for timestamp.
967
:param committer: Optional committer to set for commit.
968
:param revprops: Optional dictionary of revision properties.
969
:param revision_id: Optional revision id.
971
result = self._commit_builder_class(self, parents, config,
972
timestamp, timezone, committer, revprops, revision_id)
973
self.start_write_group()
977
if (self.control_files._lock_count == 1 and
978
self.control_files._lock_mode == 'w'):
979
if self._write_group is not None:
980
self.abort_write_group()
981
self.control_files.unlock()
982
raise errors.BzrError(
983
'Must end write groups before releasing write locks.')
984
self.control_files.unlock()
987
def clone(self, a_bzrdir, revision_id=None):
988
"""Clone this repository into a_bzrdir using the current format.
990
Currently no check is made that the format of this repository and
991
the bzrdir format are compatible. FIXME RBC 20060201.
993
:return: The newly created destination repository.
995
# TODO: deprecate after 0.16; cloning this with all its settings is
996
# probably not very useful -- mbp 20070423
997
dest_repo = self._create_sprouting_repo(a_bzrdir, shared=self.is_shared())
998
self.copy_content_into(dest_repo, revision_id)
1001
def start_write_group(self):
1002
"""Start a write group in the repository.
1004
Write groups are used by repositories which do not have a 1:1 mapping
1005
between file ids and backend store to manage the insertion of data from
1006
both fetch and commit operations.
1008
A write lock is required around the start_write_group/commit_write_group
1009
for the support of lock-requiring repository formats.
1011
One can only insert data into a repository inside a write group.
1015
if not self.is_write_locked():
1016
raise errors.NotWriteLocked(self)
1017
if self._write_group:
1018
raise errors.BzrError('already in a write group')
1019
self._start_write_group()
1020
# so we can detect unlock/relock - the write group is now entered.
1021
self._write_group = self.get_transaction()
1023
def _start_write_group(self):
1024
"""Template method for per-repository write group startup.
1026
This is called before the write group is considered to be
1031
def sprout(self, to_bzrdir, revision_id=None):
1032
"""Create a descendent repository for new development.
1034
Unlike clone, this does not copy the settings of the repository.
1036
dest_repo = self._create_sprouting_repo(to_bzrdir, shared=False)
1037
dest_repo.fetch(self, revision_id=revision_id)
1040
def _create_sprouting_repo(self, a_bzrdir, shared):
1041
if not isinstance(a_bzrdir._format, self.bzrdir._format.__class__):
1042
# use target default format.
1043
dest_repo = a_bzrdir.create_repository()
1045
# Most control formats need the repository to be specifically
1046
# created, but on some old all-in-one formats it's not needed
1048
dest_repo = self._format.initialize(a_bzrdir, shared=shared)
1049
except errors.UninitializableFormat:
1050
dest_repo = a_bzrdir.open_repository()
1054
def has_revision(self, revision_id):
1055
"""True if this repository has a copy of the revision."""
1056
return revision_id in self.has_revisions((revision_id,))
1058
def has_revisions(self, revision_ids):
1059
"""Probe to find out the presence of multiple revisions.
1061
:param revision_ids: An iterable of revision_ids.
1062
:return: A set of the revision_ids that were present.
1064
raise NotImplementedError(self.has_revisions)
1066
return self._revision_store.has_revision_id(revision_id,
1067
self.get_transaction())
1070
def get_revision(self, revision_id):
1071
"""Return the Revision object for a named revision."""
1072
return self.get_revisions([revision_id])[0]
1075
def get_revision_reconcile(self, revision_id):
1076
"""'reconcile' helper routine that allows access to a revision always.
1078
This variant of get_revision does not cross check the weave graph
1079
against the revision one as get_revision does: but it should only
1080
be used by reconcile, or reconcile-alike commands that are correcting
1081
or testing the revision graph.
1083
return self._get_revisions([revision_id])[0]
1086
def get_revisions(self, revision_ids):
1087
"""Get many revisions at once."""
1088
return self._get_revisions(revision_ids)
1091
def _get_revisions(self, revision_ids):
1092
"""Core work logic to get many revisions without sanity checks."""
1093
for rev_id in revision_ids:
1094
if not rev_id or not isinstance(rev_id, basestring):
1095
raise errors.InvalidRevisionId(revision_id=rev_id, branch=self)
1096
revs = self._revision_store.get_revisions(revision_ids,
1097
self.get_transaction())
1099
assert not isinstance(rev.revision_id, unicode)
1100
for parent_id in rev.parent_ids:
1101
assert not isinstance(parent_id, unicode)
1105
def get_revision_xml(self, revision_id):
1106
# TODO: jam 20070210 This shouldn't be necessary since get_revision
1107
# would have already do it.
1108
# TODO: jam 20070210 Just use _serializer.write_revision_to_string()
1109
rev = self.get_revision(revision_id)
1110
rev_tmp = StringIO()
1111
# the current serializer..
1112
self._revision_store._serializer.write_revision(rev, rev_tmp)
1114
return rev_tmp.getvalue()
1117
def get_deltas_for_revisions(self, revisions):
1118
"""Produce a generator of revision deltas.
1120
Note that the input is a sequence of REVISIONS, not revision_ids.
1121
Trees will be held in memory until the generator exits.
1122
Each delta is relative to the revision's lefthand predecessor.
1124
required_trees = set()
1125
for revision in revisions:
1126
required_trees.add(revision.revision_id)
1127
required_trees.update(revision.parent_ids[:1])
1128
trees = dict((t.get_revision_id(), t) for
1129
t in self.revision_trees(required_trees))
1130
for revision in revisions:
1131
if not revision.parent_ids:
1132
old_tree = self.revision_tree(None)
1134
old_tree = trees[revision.parent_ids[0]]
1135
yield trees[revision.revision_id].changes_from(old_tree)
1138
def get_revision_delta(self, revision_id):
1139
"""Return the delta for one revision.
1141
The delta is relative to the left-hand predecessor of the
1144
r = self.get_revision(revision_id)
1145
return list(self.get_deltas_for_revisions([r]))[0]
1148
def store_revision_signature(self, gpg_strategy, plaintext, revision_id):
1149
signature = gpg_strategy.sign(plaintext)
1150
self.add_signature_text(revision_id, signature)
1153
def add_signature_text(self, revision_id, signature):
1154
self._revision_store.add_revision_signature_text(revision_id,
1156
self.get_transaction())
1158
def find_text_key_references(self):
1159
"""Find the text key references within the repository.
1161
:return: a dictionary mapping (file_id, revision_id) tuples to altered file-ids to an iterable of
1162
revision_ids. Each altered file-ids has the exact revision_ids that
1163
altered it listed explicitly.
1164
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
1165
to whether they were referred to by the inventory of the
1166
revision_id that they contain. The inventory texts from all present
1167
revision ids are assessed to generate this report.
1169
revision_ids = self.all_revision_ids()
1170
w = self.get_inventory_weave()
1171
pb = ui.ui_factory.nested_progress_bar()
1173
return self._find_text_key_references_from_xml_inventory_lines(
1174
w.iter_lines_added_or_present_in_versions(revision_ids, pb=pb))
1178
def _find_text_key_references_from_xml_inventory_lines(self,
1180
"""Core routine for extracting references to texts from inventories.
1182
This performs the translation of xml lines to revision ids.
1184
:param line_iterator: An iterator of lines, origin_version_id
1185
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
1186
to whether they were referred to by the inventory of the
1187
revision_id that they contain. Note that if that revision_id was
1188
not part of the line_iterator's output then False will be given -
1189
even though it may actually refer to that key.
1191
if not self._serializer.support_altered_by_hack:
1192
raise AssertionError(
1193
"_find_text_key_references_from_xml_inventory_lines only "
1194
"supported for branches which store inventory as unnested xml"
1195
", not on %r" % self)
1198
# this code needs to read every new line in every inventory for the
1199
# inventories [revision_ids]. Seeing a line twice is ok. Seeing a line
1200
# not present in one of those inventories is unnecessary but not
1201
# harmful because we are filtering by the revision id marker in the
1202
# inventory lines : we only select file ids altered in one of those
1203
# revisions. We don't need to see all lines in the inventory because
1204
# only those added in an inventory in rev X can contain a revision=X
1206
unescape_revid_cache = {}
1207
unescape_fileid_cache = {}
1209
# jam 20061218 In a big fetch, this handles hundreds of thousands
1210
# of lines, so it has had a lot of inlining and optimizing done.
1211
# Sorry that it is a little bit messy.
1212
# Move several functions to be local variables, since this is a long
1214
search = self._file_ids_altered_regex.search
1215
unescape = _unescape_xml
1216
setdefault = result.setdefault
1217
for line, version_id in line_iterator:
1218
match = search(line)
1221
# One call to match.group() returning multiple items is quite a
1222
# bit faster than 2 calls to match.group() each returning 1
1223
file_id, revision_id = match.group('file_id', 'revision_id')
1225
# Inlining the cache lookups helps a lot when you make 170,000
1226
# lines and 350k ids, versus 8.4 unique ids.
1227
# Using a cache helps in 2 ways:
1228
# 1) Avoids unnecessary decoding calls
1229
# 2) Re-uses cached strings, which helps in future set and
1231
# (2) is enough that removing encoding entirely along with
1232
# the cache (so we are using plain strings) results in no
1233
# performance improvement.
1235
revision_id = unescape_revid_cache[revision_id]
1237
unescaped = unescape(revision_id)
1238
unescape_revid_cache[revision_id] = unescaped
1239
revision_id = unescaped
1241
# Note that unconditionally unescaping means that we deserialise
1242
# every fileid, which for general 'pull' is not great, but we don't
1243
# really want to have some many fulltexts that this matters anyway.
1246
file_id = unescape_fileid_cache[file_id]
1248
unescaped = unescape(file_id)
1249
unescape_fileid_cache[file_id] = unescaped
1252
key = (file_id, revision_id)
1253
setdefault(key, False)
1254
if revision_id == version_id:
1258
def _find_file_ids_from_xml_inventory_lines(self, line_iterator,
1260
"""Helper routine for fileids_altered_by_revision_ids.
1262
This performs the translation of xml lines to revision ids.
1264
:param line_iterator: An iterator of lines, origin_version_id
1265
:param revision_ids: The revision ids to filter for. This should be a
1266
set or other type which supports efficient __contains__ lookups, as
1267
the revision id from each parsed line will be looked up in the
1268
revision_ids filter.
1269
:return: a dictionary mapping altered file-ids to an iterable of
1270
revision_ids. Each altered file-ids has the exact revision_ids that
1271
altered it listed explicitly.
1274
setdefault = result.setdefault
1275
for file_id, revision_id in \
1276
self._find_text_key_references_from_xml_inventory_lines(
1277
line_iterator).iterkeys():
1278
# once data is all ensured-consistent; then this is
1279
# if revision_id == version_id
1280
if revision_id in revision_ids:
1281
setdefault(file_id, set()).add(revision_id)
1284
def fileids_altered_by_revision_ids(self, revision_ids):
1285
"""Find the file ids and versions affected by revisions.
1287
:param revisions: an iterable containing revision ids.
1288
:return: a dictionary mapping altered file-ids to an iterable of
1289
revision_ids. Each altered file-ids has the exact revision_ids that
1290
altered it listed explicitly.
1292
selected_revision_ids = set(revision_ids)
1293
w = self.get_inventory_weave()
1294
pb = ui.ui_factory.nested_progress_bar()
1296
return self._find_file_ids_from_xml_inventory_lines(
1297
w.iter_lines_added_or_present_in_versions(
1298
selected_revision_ids, pb=pb),
1299
selected_revision_ids)
1303
def iter_files_bytes(self, desired_files):
1304
"""Iterate through file versions.
1306
Files will not necessarily be returned in the order they occur in
1307
desired_files. No specific order is guaranteed.
1309
Yields pairs of identifier, bytes_iterator. identifier is an opaque
1310
value supplied by the caller as part of desired_files. It should
1311
uniquely identify the file version in the caller's context. (Examples:
1312
an index number or a TreeTransform trans_id.)
1314
bytes_iterator is an iterable of bytestrings for the file. The
1315
kind of iterable and length of the bytestrings are unspecified, but for
1316
this implementation, it is a list of lines produced by
1317
VersionedFile.get_lines().
1319
:param desired_files: a list of (file_id, revision_id, identifier)
1322
transaction = self.get_transaction()
1323
for file_id, revision_id, callable_data in desired_files:
1325
weave = self.weave_store.get_weave(file_id, transaction)
1326
except errors.NoSuchFile:
1327
raise errors.NoSuchIdInRepository(self, file_id)
1328
yield callable_data, weave.get_lines(revision_id)
1330
def _generate_text_key_index(self, text_key_references=None,
1332
"""Generate a new text key index for the repository.
1334
This is an expensive function that will take considerable time to run.
1336
:return: A dict mapping text keys ((file_id, revision_id) tuples) to a
1337
list of parents, also text keys. When a given key has no parents,
1338
the parents list will be [NULL_REVISION].
1340
# All revisions, to find inventory parents.
1341
if ancestors is None:
1342
graph = self.get_graph()
1343
ancestors = graph.get_parent_map(self.all_revision_ids())
1344
if text_key_references is None:
1345
text_key_references = self.find_text_key_references()
1346
pb = ui.ui_factory.nested_progress_bar()
1348
return self._do_generate_text_key_index(ancestors,
1349
text_key_references, pb)
1353
def _do_generate_text_key_index(self, ancestors, text_key_references, pb):
1354
"""Helper for _generate_text_key_index to avoid deep nesting."""
1355
revision_order = tsort.topo_sort(ancestors)
1356
invalid_keys = set()
1358
for revision_id in revision_order:
1359
revision_keys[revision_id] = set()
1360
text_count = len(text_key_references)
1361
# a cache of the text keys to allow reuse; costs a dict of all the
1362
# keys, but saves a 2-tuple for every child of a given key.
1364
for text_key, valid in text_key_references.iteritems():
1366
invalid_keys.add(text_key)
1368
revision_keys[text_key[1]].add(text_key)
1369
text_key_cache[text_key] = text_key
1370
del text_key_references
1372
text_graph = graph.Graph(graph.DictParentsProvider(text_index))
1373
NULL_REVISION = _mod_revision.NULL_REVISION
1374
# Set a cache with a size of 10 - this suffices for bzr.dev but may be
1375
# too small for large or very branchy trees. However, for 55K path
1376
# trees, it would be easy to use too much memory trivially. Ideally we
1377
# could gauge this by looking at available real memory etc, but this is
1378
# always a tricky proposition.
1379
inventory_cache = lru_cache.LRUCache(10)
1380
batch_size = 10 # should be ~150MB on a 55K path tree
1381
batch_count = len(revision_order) / batch_size + 1
1383
pb.update("Calculating text parents.", processed_texts, text_count)
1384
for offset in xrange(batch_count):
1385
to_query = revision_order[offset * batch_size:(offset + 1) *
1389
for rev_tree in self.revision_trees(to_query):
1390
revision_id = rev_tree.get_revision_id()
1391
parent_ids = ancestors[revision_id]
1392
for text_key in revision_keys[revision_id]:
1393
pb.update("Calculating text parents.", processed_texts)
1394
processed_texts += 1
1395
candidate_parents = []
1396
for parent_id in parent_ids:
1397
parent_text_key = (text_key[0], parent_id)
1399
check_parent = parent_text_key not in \
1400
revision_keys[parent_id]
1402
# the parent parent_id is a ghost:
1403
check_parent = False
1404
# truncate the derived graph against this ghost.
1405
parent_text_key = None
1407
# look at the parent commit details inventories to
1408
# determine possible candidates in the per file graph.
1411
inv = inventory_cache[parent_id]
1413
inv = self.revision_tree(parent_id).inventory
1414
inventory_cache[parent_id] = inv
1415
parent_entry = inv._byid.get(text_key[0], None)
1416
if parent_entry is not None:
1418
text_key[0], parent_entry.revision)
1420
parent_text_key = None
1421
if parent_text_key is not None:
1422
candidate_parents.append(
1423
text_key_cache[parent_text_key])
1424
parent_heads = text_graph.heads(candidate_parents)
1425
new_parents = list(parent_heads)
1426
new_parents.sort(key=lambda x:candidate_parents.index(x))
1427
if new_parents == []:
1428
new_parents = [NULL_REVISION]
1429
text_index[text_key] = new_parents
1431
for text_key in invalid_keys:
1432
text_index[text_key] = [NULL_REVISION]
1435
def item_keys_introduced_by(self, revision_ids, _files_pb=None):
1436
"""Get an iterable listing the keys of all the data introduced by a set
1439
The keys will be ordered so that the corresponding items can be safely
1440
fetched and inserted in that order.
1442
:returns: An iterable producing tuples of (knit-kind, file-id,
1443
versions). knit-kind is one of 'file', 'inventory', 'signatures',
1444
'revisions'. file-id is None unless knit-kind is 'file'.
1446
# XXX: it's a bit weird to control the inventory weave caching in this
1447
# generator. Ideally the caching would be done in fetch.py I think. Or
1448
# maybe this generator should explicitly have the contract that it
1449
# should not be iterated until the previously yielded item has been
1452
inv_w = self.get_inventory_weave()
1453
inv_w.enable_cache()
1455
# file ids that changed
1456
file_ids = self.fileids_altered_by_revision_ids(revision_ids)
1458
num_file_ids = len(file_ids)
1459
for file_id, altered_versions in file_ids.iteritems():
1460
if _files_pb is not None:
1461
_files_pb.update("fetch texts", count, num_file_ids)
1463
yield ("file", file_id, altered_versions)
1464
# We're done with the files_pb. Note that it finished by the caller,
1465
# just as it was created by the caller.
1469
yield ("inventory", None, revision_ids)
1473
revisions_with_signatures = set()
1474
for rev_id in revision_ids:
1476
self.get_signature_text(rev_id)
1477
except errors.NoSuchRevision:
1481
revisions_with_signatures.add(rev_id)
1483
yield ("signatures", None, revisions_with_signatures)
1486
yield ("revisions", None, revision_ids)
1489
def get_inventory_weave(self):
1490
return self.control_weaves.get_weave('inventory',
1491
self.get_transaction())
1494
def get_inventory(self, revision_id):
1495
"""Get Inventory object by revision id."""
1496
return self.iter_inventories([revision_id]).next()
1498
def iter_inventories(self, revision_ids):
1499
"""Get many inventories by revision_ids.
1501
This will buffer some or all of the texts used in constructing the
1502
inventories in memory, but will only parse a single inventory at a
1505
:return: An iterator of inventories.
1507
assert None not in revision_ids
1508
assert _mod_revision.NULL_REVISION not in revision_ids
1509
return self._iter_inventories(revision_ids)
1511
def _iter_inventories(self, revision_ids):
1512
"""single-document based inventory iteration."""
1513
texts = self.get_inventory_weave().get_texts(revision_ids)
1514
for text, revision_id in zip(texts, revision_ids):
1515
yield self.deserialise_inventory(revision_id, text)
1517
def deserialise_inventory(self, revision_id, xml):
1518
"""Transform the xml into an inventory object.
1520
:param revision_id: The expected revision id of the inventory.
1521
:param xml: A serialised inventory.
1523
result = self._serializer.read_inventory_from_string(xml, revision_id)
1524
if result.revision_id != revision_id:
1525
raise AssertionError('revision id mismatch %s != %s' % (
1526
result.revision_id, revision_id))
1529
def serialise_inventory(self, inv):
1530
return self._serializer.write_inventory_to_string(inv)
1532
def _serialise_inventory_to_lines(self, inv):
1533
return self._serializer.write_inventory_to_lines(inv)
1535
def get_serializer_format(self):
1536
return self._serializer.format_num
1539
def get_inventory_xml(self, revision_id):
1540
"""Get inventory XML as a file object."""
1542
assert isinstance(revision_id, str), type(revision_id)
1543
iw = self.get_inventory_weave()
1544
return iw.get_text(revision_id)
1546
raise errors.HistoryMissing(self, 'inventory', revision_id)
1549
def get_inventory_sha1(self, revision_id):
1550
"""Return the sha1 hash of the inventory entry
1552
return self.get_revision(revision_id).inventory_sha1
1555
@deprecated_method(symbol_versioning.one_four)
1556
def get_revision_graph(self, revision_id=None):
1557
"""Return a dictionary containing the revision graph.
1559
NB: This method should not be used as it accesses the entire graph all
1560
at once, which is much more data than most operations should require.
1562
:param revision_id: The revision_id to get a graph from. If None, then
1563
the entire revision graph is returned. This is a deprecated mode of
1564
operation and will be removed in the future.
1565
:return: a dictionary of revision_id->revision_parents_list.
1567
raise NotImplementedError(self.get_revision_graph)
1570
@deprecated_method(symbol_versioning.one_three)
1571
def get_revision_graph_with_ghosts(self, revision_ids=None):
1572
"""Return a graph of the revisions with ghosts marked as applicable.
1574
:param revision_ids: an iterable of revisions to graph or None for all.
1575
:return: a Graph object with the graph reachable from revision_ids.
1577
if 'evil' in debug.debug_flags:
1579
"get_revision_graph_with_ghosts scales with size of history.")
1580
result = deprecated_graph.Graph()
1581
if not revision_ids:
1582
pending = set(self.all_revision_ids())
1585
pending = set(revision_ids)
1586
# special case NULL_REVISION
1587
if _mod_revision.NULL_REVISION in pending:
1588
pending.remove(_mod_revision.NULL_REVISION)
1589
required = set(pending)
1592
revision_id = pending.pop()
1594
rev = self.get_revision(revision_id)
1595
except errors.NoSuchRevision:
1596
if revision_id in required:
1599
result.add_ghost(revision_id)
1601
for parent_id in rev.parent_ids:
1602
# is this queued or done ?
1603
if (parent_id not in pending and
1604
parent_id not in done):
1606
pending.add(parent_id)
1607
result.add_node(revision_id, rev.parent_ids)
1608
done.add(revision_id)
1611
def iter_reverse_revision_history(self, revision_id):
1612
"""Iterate backwards through revision ids in the lefthand history
1614
:param revision_id: The revision id to start with. All its lefthand
1615
ancestors will be traversed.
1617
graph = self.get_graph()
1618
next_id = revision_id
1620
if next_id in (None, _mod_revision.NULL_REVISION):
1623
# Note: The following line may raise KeyError in the event of
1624
# truncated history. We decided not to have a try:except:raise
1625
# RevisionNotPresent here until we see a use for it, because of the
1626
# cost in an inner loop that is by its very nature O(history).
1627
# Robert Collins 20080326
1628
parents = graph.get_parent_map([next_id])[next_id]
1629
if len(parents) == 0:
1632
next_id = parents[0]
1635
def get_revision_inventory(self, revision_id):
1636
"""Return inventory of a past revision."""
1637
# TODO: Unify this with get_inventory()
1638
# bzr 0.0.6 and later imposes the constraint that the inventory_id
1639
# must be the same as its revision, so this is trivial.
1640
if revision_id is None:
1641
# This does not make sense: if there is no revision,
1642
# then it is the current tree inventory surely ?!
1643
# and thus get_root_id() is something that looks at the last
1644
# commit on the branch, and the get_root_id is an inventory check.
1645
raise NotImplementedError
1646
# return Inventory(self.get_root_id())
1648
return self.get_inventory(revision_id)
1651
def is_shared(self):
1652
"""Return True if this repository is flagged as a shared repository."""
1653
raise NotImplementedError(self.is_shared)
1656
def reconcile(self, other=None, thorough=False):
1657
"""Reconcile this repository."""
1658
from bzrlib.reconcile import RepoReconciler
1659
reconciler = RepoReconciler(self, thorough=thorough)
1660
reconciler.reconcile()
1663
def _refresh_data(self):
1664
"""Helper called from lock_* to ensure coherency with disk.
1666
The default implementation does nothing; it is however possible
1667
for repositories to maintain loaded indices across multiple locks
1668
by checking inside their implementation of this method to see
1669
whether their indices are still valid. This depends of course on
1670
the disk format being validatable in this manner.
1674
def revision_tree(self, revision_id):
1675
"""Return Tree for a revision on this branch.
1677
`revision_id` may be None for the empty tree revision.
1679
# TODO: refactor this to use an existing revision object
1680
# so we don't need to read it in twice.
1681
if revision_id is None or revision_id == _mod_revision.NULL_REVISION:
1682
return RevisionTree(self, Inventory(root_id=None),
1683
_mod_revision.NULL_REVISION)
1685
inv = self.get_revision_inventory(revision_id)
1686
return RevisionTree(self, inv, revision_id)
1689
def revision_trees(self, revision_ids):
1690
"""Return Tree for a revision on this branch.
1692
`revision_id` may not be None or 'null:'"""
1693
inventories = self.iter_inventories(revision_ids)
1694
for inv in inventories:
1695
yield RevisionTree(self, inv, inv.revision_id)
1698
def get_ancestry(self, revision_id, topo_sorted=True):
1699
"""Return a list of revision-ids integrated by a revision.
1701
The first element of the list is always None, indicating the origin
1702
revision. This might change when we have history horizons, or
1703
perhaps we should have a new API.
1705
This is topologically sorted.
1707
if _mod_revision.is_null(revision_id):
1709
if not self.has_revision(revision_id):
1710
raise errors.NoSuchRevision(self, revision_id)
1711
w = self.get_inventory_weave()
1712
candidates = w.get_ancestry(revision_id, topo_sorted)
1713
return [None] + candidates # self._eliminate_revisions_not_present(candidates)
1716
"""Compress the data within the repository.
1718
This operation only makes sense for some repository types. For other
1719
types it should be a no-op that just returns.
1721
This stub method does not require a lock, but subclasses should use
1722
@needs_write_lock as this is a long running call its reasonable to
1723
implicitly lock for the user.
1727
def print_file(self, file, revision_id):
1728
"""Print `file` to stdout.
1730
FIXME RBC 20060125 as John Meinel points out this is a bad api
1731
- it writes to stdout, it assumes that that is valid etc. Fix
1732
by creating a new more flexible convenience function.
1734
tree = self.revision_tree(revision_id)
1735
# use inventory as it was in that revision
1736
file_id = tree.inventory.path2id(file)
1738
# TODO: jam 20060427 Write a test for this code path
1739
# it had a bug in it, and was raising the wrong
1741
raise errors.BzrError("%r is not present in revision %s" % (file, revision_id))
1742
tree.print_file(file_id)
1744
def get_transaction(self):
1745
return self.control_files.get_transaction()
1747
def revision_parents(self, revision_id):
1748
return self.get_inventory_weave().parent_names(revision_id)
1750
@deprecated_method(symbol_versioning.one_one)
1751
def get_parents(self, revision_ids):
1752
"""See StackedParentsProvider.get_parents"""
1753
parent_map = self.get_parent_map(revision_ids)
1754
return [parent_map.get(r, None) for r in revision_ids]
1756
def get_parent_map(self, keys):
1757
"""See graph._StackedParentsProvider.get_parent_map"""
1759
for revision_id in keys:
1760
if revision_id == _mod_revision.NULL_REVISION:
1761
parent_map[revision_id] = ()
1764
parent_id_list = self.get_revision(revision_id).parent_ids
1765
except errors.NoSuchRevision:
1768
if len(parent_id_list) == 0:
1769
parent_ids = (_mod_revision.NULL_REVISION,)
1771
parent_ids = tuple(parent_id_list)
1772
parent_map[revision_id] = parent_ids
1775
def _make_parents_provider(self):
1778
def get_graph(self, other_repository=None):
1779
"""Return the graph walker for this repository format"""
1780
parents_provider = self._make_parents_provider()
1781
if (other_repository is not None and
1782
not self.has_same_location(other_repository)):
1783
parents_provider = graph._StackedParentsProvider(
1784
[parents_provider, other_repository._make_parents_provider()])
1785
return graph.Graph(parents_provider)
1787
def _get_versioned_file_checker(self):
1788
"""Return an object suitable for checking versioned files."""
1789
return _VersionedFileChecker(self)
1791
def revision_ids_to_search_result(self, result_set):
1792
"""Convert a set of revision ids to a graph SearchResult."""
1793
result_parents = set()
1794
for parents in self.get_graph().get_parent_map(
1795
result_set).itervalues():
1796
result_parents.update(parents)
1797
included_keys = result_set.intersection(result_parents)
1798
start_keys = result_set.difference(included_keys)
1799
exclude_keys = result_parents.difference(result_set)
1800
result = graph.SearchResult(start_keys, exclude_keys,
1801
len(result_set), result_set)
1805
def set_make_working_trees(self, new_value):
1806
"""Set the policy flag for making working trees when creating branches.
1808
This only applies to branches that use this repository.
1810
The default is 'True'.
1811
:param new_value: True to restore the default, False to disable making
1814
raise NotImplementedError(self.set_make_working_trees)
1816
def make_working_trees(self):
1817
"""Returns the policy for making working trees on new branches."""
1818
raise NotImplementedError(self.make_working_trees)
1821
def sign_revision(self, revision_id, gpg_strategy):
1822
plaintext = Testament.from_revision(self, revision_id).as_short_text()
1823
self.store_revision_signature(gpg_strategy, plaintext, revision_id)
1826
def has_signature_for_revision_id(self, revision_id):
1827
"""Query for a revision signature for revision_id in the repository."""
1828
return self._revision_store.has_signature(revision_id,
1829
self.get_transaction())
1832
def get_signature_text(self, revision_id):
1833
"""Return the text for a signature."""
1834
return self._revision_store.get_signature_text(revision_id,
1835
self.get_transaction())
1838
def check(self, revision_ids=None):
1839
"""Check consistency of all history of given revision_ids.
1841
Different repository implementations should override _check().
1843
:param revision_ids: A non-empty list of revision_ids whose ancestry
1844
will be checked. Typically the last revision_id of a branch.
1846
return self._check(revision_ids)
1848
def _check(self, revision_ids):
1849
result = check.Check(self)
1853
def _warn_if_deprecated(self):
1854
global _deprecation_warning_done
1855
if _deprecation_warning_done:
1857
_deprecation_warning_done = True
1858
warning("Format %s for %s is deprecated - please use 'bzr upgrade' to get better performance"
1859
% (self._format, self.bzrdir.transport.base))
1861
def supports_rich_root(self):
1862
return self._format.rich_root_data
1864
def _check_ascii_revisionid(self, revision_id, method):
1865
"""Private helper for ascii-only repositories."""
1866
# weave repositories refuse to store revisionids that are non-ascii.
1867
if revision_id is not None:
1868
# weaves require ascii revision ids.
1869
if isinstance(revision_id, unicode):
1871
revision_id.encode('ascii')
1872
except UnicodeEncodeError:
1873
raise errors.NonAsciiRevisionId(method, self)
1876
revision_id.decode('ascii')
1877
except UnicodeDecodeError:
1878
raise errors.NonAsciiRevisionId(method, self)
1880
def revision_graph_can_have_wrong_parents(self):
1881
"""Is it possible for this repository to have a revision graph with
1884
If True, then this repository must also implement
1885
_find_inconsistent_revision_parents so that check and reconcile can
1886
check for inconsistencies before proceeding with other checks that may
1887
depend on the revision index being consistent.
1889
raise NotImplementedError(self.revision_graph_can_have_wrong_parents)
1892
# remove these delegates a while after bzr 0.15
1893
def __make_delegated(name, from_module):
1894
def _deprecated_repository_forwarder():
1895
symbol_versioning.warn('%s moved to %s in bzr 0.15'
1896
% (name, from_module),
1899
m = __import__(from_module, globals(), locals(), [name])
1901
return getattr(m, name)
1902
except AttributeError:
1903
raise AttributeError('module %s has no name %s'
1905
globals()[name] = _deprecated_repository_forwarder
1908
'AllInOneRepository',
1909
'WeaveMetaDirRepository',
1910
'PreSplitOutRepositoryFormat',
1911
'RepositoryFormat4',
1912
'RepositoryFormat5',
1913
'RepositoryFormat6',
1914
'RepositoryFormat7',
1916
__make_delegated(_name, 'bzrlib.repofmt.weaverepo')
1920
'RepositoryFormatKnit',
1921
'RepositoryFormatKnit1',
1923
__make_delegated(_name, 'bzrlib.repofmt.knitrepo')
1926
def install_revision(repository, rev, revision_tree):
1927
"""Install all revision data into a repository."""
1928
install_revisions(repository, [(rev, revision_tree, None)])
1931
def install_revisions(repository, iterable, num_revisions=None, pb=None):
1932
"""Install all revision data into a repository.
1934
Accepts an iterable of revision, tree, signature tuples. The signature
1937
repository.start_write_group()
1939
for n, (revision, revision_tree, signature) in enumerate(iterable):
1940
_install_revision(repository, revision, revision_tree, signature)
1942
pb.update('Transferring revisions', n + 1, num_revisions)
1944
repository.abort_write_group()
1947
repository.commit_write_group()
1950
def _install_revision(repository, rev, revision_tree, signature):
1951
"""Install all revision data into a repository."""
1952
present_parents = []
1954
for p_id in rev.parent_ids:
1955
if repository.has_revision(p_id):
1956
present_parents.append(p_id)
1957
parent_trees[p_id] = repository.revision_tree(p_id)
1959
parent_trees[p_id] = repository.revision_tree(None)
1961
inv = revision_tree.inventory
1962
entries = inv.iter_entries()
1963
# backwards compatibility hack: skip the root id.
1964
if not repository.supports_rich_root():
1965
path, root = entries.next()
1966
if root.revision != rev.revision_id:
1967
raise errors.IncompatibleRevision(repr(repository))
1968
# Add the texts that are not already present
1969
for path, ie in entries:
1970
w = repository.weave_store.get_weave_or_empty(ie.file_id,
1971
repository.get_transaction())
1972
if ie.revision not in w:
1974
# FIXME: TODO: The following loop *may* be overlapping/duplicate
1975
# with InventoryEntry.find_previous_heads(). if it is, then there
1976
# is a latent bug here where the parents may have ancestors of each
1978
for revision, tree in parent_trees.iteritems():
1979
if ie.file_id not in tree:
1981
parent_id = tree.inventory[ie.file_id].revision
1982
if parent_id in text_parents:
1984
text_parents.append(parent_id)
1986
vfile = repository.weave_store.get_weave_or_empty(ie.file_id,
1987
repository.get_transaction())
1988
lines = revision_tree.get_file(ie.file_id).readlines()
1989
vfile.add_lines(rev.revision_id, text_parents, lines)
1991
# install the inventory
1992
repository.add_inventory(rev.revision_id, inv, present_parents)
1993
except errors.RevisionAlreadyPresent:
1995
if signature is not None:
1996
repository.add_signature_text(rev.revision_id, signature)
1997
repository.add_revision(rev.revision_id, rev, inv)
2000
class MetaDirRepository(Repository):
2001
"""Repositories in the new meta-dir layout."""
2003
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
2004
super(MetaDirRepository, self).__init__(_format,
2010
dir_mode = self.control_files._dir_mode
2011
file_mode = self.control_files._file_mode
2014
def is_shared(self):
2015
"""Return True if this repository is flagged as a shared repository."""
2016
return self.control_files._transport.has('shared-storage')
2019
def set_make_working_trees(self, new_value):
2020
"""Set the policy flag for making working trees when creating branches.
2022
This only applies to branches that use this repository.
2024
The default is 'True'.
2025
:param new_value: True to restore the default, False to disable making
2030
self.control_files._transport.delete('no-working-trees')
2031
except errors.NoSuchFile:
2034
self.control_files.put_utf8('no-working-trees', '')
2036
def make_working_trees(self):
2037
"""Returns the policy for making working trees on new branches."""
2038
return not self.control_files._transport.has('no-working-trees')
2041
class MetaDirVersionedFileRepository(MetaDirRepository):
2042
"""Repositories in a meta-dir, that work via versioned file objects."""
2044
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
2045
super(MetaDirVersionedFileRepository, self).__init__(_format,
2051
_revision_store.get_scope = self.get_transaction
2052
control_store.get_scope = self.get_transaction
2053
text_store.get_scope = self.get_transaction
2056
class RepositoryFormatRegistry(registry.Registry):
2057
"""Registry of RepositoryFormats."""
2059
def get(self, format_string):
2060
r = registry.Registry.get(self, format_string)
2066
format_registry = RepositoryFormatRegistry()
2067
"""Registry of formats, indexed by their identifying format string.
2069
This can contain either format instances themselves, or classes/factories that
2070
can be called to obtain one.
2074
#####################################################################
2075
# Repository Formats
2077
class RepositoryFormat(object):
2078
"""A repository format.
2080
Formats provide three things:
2081
* An initialization routine to construct repository data on disk.
2082
* a format string which is used when the BzrDir supports versioned
2084
* an open routine which returns a Repository instance.
2086
There is one and only one Format subclass for each on-disk format. But
2087
there can be one Repository subclass that is used for several different
2088
formats. The _format attribute on a Repository instance can be used to
2089
determine the disk format.
2091
Formats are placed in an dict by their format string for reference
2092
during opening. These should be subclasses of RepositoryFormat
2095
Once a format is deprecated, just deprecate the initialize and open
2096
methods on the format class. Do not deprecate the object, as the
2097
object will be created every system load.
2099
Common instance attributes:
2100
_matchingbzrdir - the bzrdir format that the repository format was
2101
originally written to work with. This can be used if manually
2102
constructing a bzrdir and repository, or more commonly for test suite
2106
# Set to True or False in derived classes. True indicates that the format
2107
# supports ghosts gracefully.
2108
supports_ghosts = None
2109
# Can this repository be given external locations to lookup additional
2110
# data. Set to True or False in derived classes.
2111
supports_external_lookups = None
2114
return "<%s>" % self.__class__.__name__
2116
def __eq__(self, other):
2117
# format objects are generally stateless
2118
return isinstance(other, self.__class__)
2120
def __ne__(self, other):
2121
return not self == other
2124
def find_format(klass, a_bzrdir):
2125
"""Return the format for the repository object in a_bzrdir.
2127
This is used by bzr native formats that have a "format" file in
2128
the repository. Other methods may be used by different types of
2132
transport = a_bzrdir.get_repository_transport(None)
2133
format_string = transport.get("format").read()
2134
return format_registry.get(format_string)
2135
except errors.NoSuchFile:
2136
raise errors.NoRepositoryPresent(a_bzrdir)
2138
raise errors.UnknownFormatError(format=format_string,
2142
def register_format(klass, format):
2143
format_registry.register(format.get_format_string(), format)
2146
def unregister_format(klass, format):
2147
format_registry.remove(format.get_format_string())
2150
def get_default_format(klass):
2151
"""Return the current default format."""
2152
from bzrlib import bzrdir
2153
return bzrdir.format_registry.make_bzrdir('default').repository_format
2155
def _get_control_store(self, repo_transport, control_files):
2156
"""Return the control store for this repository."""
2157
raise NotImplementedError(self._get_control_store)
2159
def get_format_string(self):
2160
"""Return the ASCII format string that identifies this format.
2162
Note that in pre format ?? repositories the format string is
2163
not permitted nor written to disk.
2165
raise NotImplementedError(self.get_format_string)
2167
def get_format_description(self):
2168
"""Return the short description for this format."""
2169
raise NotImplementedError(self.get_format_description)
2171
def _get_revision_store(self, repo_transport, control_files):
2172
"""Return the revision store object for this a_bzrdir."""
2173
raise NotImplementedError(self._get_revision_store)
2175
def _get_text_rev_store(self,
2182
"""Common logic for getting a revision store for a repository.
2184
see self._get_revision_store for the subclass-overridable method to
2185
get the store for a repository.
2187
from bzrlib.store.revision.text import TextRevisionStore
2188
dir_mode = control_files._dir_mode
2189
file_mode = control_files._file_mode
2190
text_store = TextStore(transport.clone(name),
2192
compressed=compressed,
2194
file_mode=file_mode)
2195
_revision_store = TextRevisionStore(text_store, serializer)
2196
return _revision_store
2198
# TODO: this shouldn't be in the base class, it's specific to things that
2199
# use weaves or knits -- mbp 20070207
2200
def _get_versioned_file_store(self,
2205
versionedfile_class=None,
2206
versionedfile_kwargs={},
2208
if versionedfile_class is None:
2209
versionedfile_class = self._versionedfile_class
2210
weave_transport = control_files._transport.clone(name)
2211
dir_mode = control_files._dir_mode
2212
file_mode = control_files._file_mode
2213
return VersionedFileStore(weave_transport, prefixed=prefixed,
2215
file_mode=file_mode,
2216
versionedfile_class=versionedfile_class,
2217
versionedfile_kwargs=versionedfile_kwargs,
2220
def initialize(self, a_bzrdir, shared=False):
2221
"""Initialize a repository of this format in a_bzrdir.
2223
:param a_bzrdir: The bzrdir to put the new repository in it.
2224
:param shared: The repository should be initialized as a sharable one.
2225
:returns: The new repository object.
2227
This may raise UninitializableFormat if shared repository are not
2228
compatible the a_bzrdir.
2230
raise NotImplementedError(self.initialize)
2232
def is_supported(self):
2233
"""Is this format supported?
2235
Supported formats must be initializable and openable.
2236
Unsupported formats may not support initialization or committing or
2237
some other features depending on the reason for not being supported.
2241
def check_conversion_target(self, target_format):
2242
raise NotImplementedError(self.check_conversion_target)
2244
def open(self, a_bzrdir, _found=False):
2245
"""Return an instance of this format for the bzrdir a_bzrdir.
2247
_found is a private parameter, do not use it.
2249
raise NotImplementedError(self.open)
2252
class MetaDirRepositoryFormat(RepositoryFormat):
2253
"""Common base class for the new repositories using the metadir layout."""
2255
rich_root_data = False
2256
supports_tree_reference = False
2257
supports_external_lookups = False
2258
_matchingbzrdir = bzrdir.BzrDirMetaFormat1()
2261
super(MetaDirRepositoryFormat, self).__init__()
2263
def _create_control_files(self, a_bzrdir):
2264
"""Create the required files and the initial control_files object."""
2265
# FIXME: RBC 20060125 don't peek under the covers
2266
# NB: no need to escape relative paths that are url safe.
2267
repository_transport = a_bzrdir.get_repository_transport(self)
2268
control_files = lockable_files.LockableFiles(repository_transport,
2269
'lock', lockdir.LockDir)
2270
control_files.create_lock()
2271
return control_files
2273
def _upload_blank_content(self, a_bzrdir, dirs, files, utf8_files, shared):
2274
"""Upload the initial blank content."""
2275
control_files = self._create_control_files(a_bzrdir)
2276
control_files.lock_write()
2278
control_files._transport.mkdir_multi(dirs,
2279
mode=control_files._dir_mode)
2280
for file, content in files:
2281
control_files.put(file, content)
2282
for file, content in utf8_files:
2283
control_files.put_utf8(file, content)
2285
control_files.put_utf8('shared-storage', '')
2287
control_files.unlock()
2290
# formats which have no format string are not discoverable
2291
# and not independently creatable, so are not registered. They're
2292
# all in bzrlib.repofmt.weaverepo now. When an instance of one of these is
2293
# needed, it's constructed directly by the BzrDir. Non-native formats where
2294
# the repository is not separately opened are similar.
2296
format_registry.register_lazy(
2297
'Bazaar-NG Repository format 7',
2298
'bzrlib.repofmt.weaverepo',
2302
format_registry.register_lazy(
2303
'Bazaar-NG Knit Repository Format 1',
2304
'bzrlib.repofmt.knitrepo',
2305
'RepositoryFormatKnit1',
2308
format_registry.register_lazy(
2309
'Bazaar Knit Repository Format 3 (bzr 0.15)\n',
2310
'bzrlib.repofmt.knitrepo',
2311
'RepositoryFormatKnit3',
2314
format_registry.register_lazy(
2315
'Bazaar Knit Repository Format 4 (bzr 1.0)\n',
2316
'bzrlib.repofmt.knitrepo',
2317
'RepositoryFormatKnit4',
2320
# Pack-based formats. There is one format for pre-subtrees, and one for
2321
# post-subtrees to allow ease of testing.
2322
# NOTE: These are experimental in 0.92. Stable in 1.0 and above
2323
format_registry.register_lazy(
2324
'Bazaar pack repository format 1 (needs bzr 0.92)\n',
2325
'bzrlib.repofmt.pack_repo',
2326
'RepositoryFormatKnitPack1',
2328
format_registry.register_lazy(
2329
'Bazaar pack repository format 1 with subtree support (needs bzr 0.92)\n',
2330
'bzrlib.repofmt.pack_repo',
2331
'RepositoryFormatKnitPack3',
2333
format_registry.register_lazy(
2334
'Bazaar pack repository format 1 with rich root (needs bzr 1.0)\n',
2335
'bzrlib.repofmt.pack_repo',
2336
'RepositoryFormatKnitPack4',
2338
# Development formats.
2340
# development 0 - stub to introduce development versioning scheme.
2341
format_registry.register_lazy(
2342
"Bazaar development format 0 (needs bzr.dev from before 1.3)\n",
2343
'bzrlib.repofmt.pack_repo',
2344
'RepositoryFormatPackDevelopment0',
2346
format_registry.register_lazy(
2347
("Bazaar development format 0 with subtree support "
2348
"(needs bzr.dev from before 1.3)\n"),
2349
'bzrlib.repofmt.pack_repo',
2350
'RepositoryFormatPackDevelopment0Subtree',
2352
# 1.3->1.4 go below here
2355
class InterRepository(InterObject):
2356
"""This class represents operations taking place between two repositories.
2358
Its instances have methods like copy_content and fetch, and contain
2359
references to the source and target repositories these operations can be
2362
Often we will provide convenience methods on 'repository' which carry out
2363
operations with another repository - they will always forward to
2364
InterRepository.get(other).method_name(parameters).
2368
"""The available optimised InterRepository types."""
2370
def copy_content(self, revision_id=None):
2371
raise NotImplementedError(self.copy_content)
2373
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2374
"""Fetch the content required to construct revision_id.
2376
The content is copied from self.source to self.target.
2378
:param revision_id: if None all content is copied, if NULL_REVISION no
2380
:param pb: optional progress bar to use for progress reports. If not
2381
provided a default one will be created.
2383
Returns the copied revision count and the failed revisions in a tuple:
2386
raise NotImplementedError(self.fetch)
2388
def _walk_to_common_revisions(self, revision_ids):
2389
"""Walk out from revision_ids in source to revisions target has.
2391
:param revision_ids: The start point for the search.
2392
:return: A set of revision ids.
2394
graph = self.source.get_graph()
2395
missing_revs = set()
2396
# ensure we don't pay silly lookup costs.
2397
revision_ids = frozenset(revision_ids)
2398
searcher = graph._make_breadth_first_searcher(revision_ids)
2399
null_set = frozenset([_mod_revision.NULL_REVISION])
2402
next_revs, ghosts = searcher.next_with_ghosts()
2403
except StopIteration:
2405
if revision_ids.intersection(ghosts):
2406
absent_ids = set(revision_ids.intersection(ghosts))
2407
# If all absent_ids are present in target, no error is needed.
2408
absent_ids.difference_update(
2409
self.target.has_revisions(absent_ids))
2411
raise errors.NoSuchRevision(self.source, absent_ids.pop())
2412
# we don't care about other ghosts as we can't fetch them and
2413
# haven't been asked to.
2414
next_revs = set(next_revs)
2415
# we always have NULL_REVISION present.
2416
have_revs = self.target.has_revisions(next_revs).union(null_set)
2417
missing_revs.update(next_revs - have_revs)
2418
searcher.stop_searching_any(have_revs)
2419
return searcher.get_result()
2421
@deprecated_method(symbol_versioning.one_two)
2423
def missing_revision_ids(self, revision_id=None, find_ghosts=True):
2424
"""Return the revision ids that source has that target does not.
2426
These are returned in topological order.
2428
:param revision_id: only return revision ids included by this
2430
:param find_ghosts: If True find missing revisions in deep history
2431
rather than just finding the surface difference.
2433
return list(self.search_missing_revision_ids(
2434
revision_id, find_ghosts).get_keys())
2437
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2438
"""Return the revision ids that source has that target does not.
2440
:param revision_id: only return revision ids included by this
2442
:param find_ghosts: If True find missing revisions in deep history
2443
rather than just finding the surface difference.
2444
:return: A bzrlib.graph.SearchResult.
2446
# stop searching at found target revisions.
2447
if not find_ghosts and revision_id is not None:
2448
return self._walk_to_common_revisions([revision_id])
2449
# generic, possibly worst case, slow code path.
2450
target_ids = set(self.target.all_revision_ids())
2451
if revision_id is not None:
2452
source_ids = self.source.get_ancestry(revision_id)
2453
assert source_ids[0] is None
2456
source_ids = self.source.all_revision_ids()
2457
result_set = set(source_ids).difference(target_ids)
2458
return self.source.revision_ids_to_search_result(result_set)
2461
def _same_model(source, target):
2462
"""True if source and target have the same data representation."""
2463
if source.supports_rich_root() != target.supports_rich_root():
2465
if source._serializer != target._serializer:
2470
class InterSameDataRepository(InterRepository):
2471
"""Code for converting between repositories that represent the same data.
2473
Data format and model must match for this to work.
2477
def _get_repo_format_to_test(self):
2478
"""Repository format for testing with.
2480
InterSameData can pull from subtree to subtree and from non-subtree to
2481
non-subtree, so we test this with the richest repository format.
2483
from bzrlib.repofmt import knitrepo
2484
return knitrepo.RepositoryFormatKnit3()
2487
def is_compatible(source, target):
2488
return InterRepository._same_model(source, target)
2491
def copy_content(self, revision_id=None):
2492
"""Make a complete copy of the content in self into destination.
2494
This copies both the repository's revision data, and configuration information
2495
such as the make_working_trees setting.
2497
This is a destructive operation! Do not use it on existing
2500
:param revision_id: Only copy the content needed to construct
2501
revision_id and its parents.
2504
self.target.set_make_working_trees(self.source.make_working_trees())
2505
except NotImplementedError:
2507
# but don't bother fetching if we have the needed data now.
2508
if (revision_id not in (None, _mod_revision.NULL_REVISION) and
2509
self.target.has_revision(revision_id)):
2511
self.target.fetch(self.source, revision_id=revision_id)
2514
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2515
"""See InterRepository.fetch()."""
2516
from bzrlib.fetch import GenericRepoFetcher
2517
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2518
self.source, self.source._format, self.target,
2519
self.target._format)
2520
f = GenericRepoFetcher(to_repository=self.target,
2521
from_repository=self.source,
2522
last_revision=revision_id,
2523
pb=pb, find_ghosts=find_ghosts)
2524
return f.count_copied, f.failed_revisions
2527
class InterWeaveRepo(InterSameDataRepository):
2528
"""Optimised code paths between Weave based repositories.
2530
This should be in bzrlib/repofmt/weaverepo.py but we have not yet
2531
implemented lazy inter-object optimisation.
2535
def _get_repo_format_to_test(self):
2536
from bzrlib.repofmt import weaverepo
2537
return weaverepo.RepositoryFormat7()
2540
def is_compatible(source, target):
2541
"""Be compatible with known Weave formats.
2543
We don't test for the stores being of specific types because that
2544
could lead to confusing results, and there is no need to be
2547
from bzrlib.repofmt.weaverepo import (
2553
return (isinstance(source._format, (RepositoryFormat5,
2555
RepositoryFormat7)) and
2556
isinstance(target._format, (RepositoryFormat5,
2558
RepositoryFormat7)))
2559
except AttributeError:
2563
def copy_content(self, revision_id=None):
2564
"""See InterRepository.copy_content()."""
2565
# weave specific optimised path:
2567
self.target.set_make_working_trees(self.source.make_working_trees())
2568
except NotImplementedError:
2570
# FIXME do not peek!
2571
if self.source.control_files._transport.listable():
2572
pb = ui.ui_factory.nested_progress_bar()
2574
self.target.weave_store.copy_all_ids(
2575
self.source.weave_store,
2577
from_transaction=self.source.get_transaction(),
2578
to_transaction=self.target.get_transaction())
2579
pb.update('copying inventory', 0, 1)
2580
self.target.control_weaves.copy_multi(
2581
self.source.control_weaves, ['inventory'],
2582
from_transaction=self.source.get_transaction(),
2583
to_transaction=self.target.get_transaction())
2584
self.target._revision_store.text_store.copy_all_ids(
2585
self.source._revision_store.text_store,
2590
self.target.fetch(self.source, revision_id=revision_id)
2593
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2594
"""See InterRepository.fetch()."""
2595
from bzrlib.fetch import GenericRepoFetcher
2596
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2597
self.source, self.source._format, self.target, self.target._format)
2598
f = GenericRepoFetcher(to_repository=self.target,
2599
from_repository=self.source,
2600
last_revision=revision_id,
2601
pb=pb, find_ghosts=find_ghosts)
2602
return f.count_copied, f.failed_revisions
2605
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2606
"""See InterRepository.missing_revision_ids()."""
2607
# we want all revisions to satisfy revision_id in source.
2608
# but we don't want to stat every file here and there.
2609
# we want then, all revisions other needs to satisfy revision_id
2610
# checked, but not those that we have locally.
2611
# so the first thing is to get a subset of the revisions to
2612
# satisfy revision_id in source, and then eliminate those that
2613
# we do already have.
2614
# this is slow on high latency connection to self, but as as this
2615
# disk format scales terribly for push anyway due to rewriting
2616
# inventory.weave, this is considered acceptable.
2618
if revision_id is not None:
2619
source_ids = self.source.get_ancestry(revision_id)
2620
assert source_ids[0] is None
2623
source_ids = self.source._all_possible_ids()
2624
source_ids_set = set(source_ids)
2625
# source_ids is the worst possible case we may need to pull.
2626
# now we want to filter source_ids against what we actually
2627
# have in target, but don't try to check for existence where we know
2628
# we do not have a revision as that would be pointless.
2629
target_ids = set(self.target._all_possible_ids())
2630
possibly_present_revisions = target_ids.intersection(source_ids_set)
2631
actually_present_revisions = set(
2632
self.target._eliminate_revisions_not_present(possibly_present_revisions))
2633
required_revisions = source_ids_set.difference(actually_present_revisions)
2634
if revision_id is not None:
2635
# we used get_ancestry to determine source_ids then we are assured all
2636
# revisions referenced are present as they are installed in topological order.
2637
# and the tip revision was validated by get_ancestry.
2638
result_set = required_revisions
2640
# if we just grabbed the possibly available ids, then
2641
# we only have an estimate of whats available and need to validate
2642
# that against the revision records.
2644
self.source._eliminate_revisions_not_present(required_revisions))
2645
return self.source.revision_ids_to_search_result(result_set)
2648
class InterKnitRepo(InterSameDataRepository):
2649
"""Optimised code paths between Knit based repositories."""
2652
def _get_repo_format_to_test(self):
2653
from bzrlib.repofmt import knitrepo
2654
return knitrepo.RepositoryFormatKnit1()
2657
def is_compatible(source, target):
2658
"""Be compatible with known Knit formats.
2660
We don't test for the stores being of specific types because that
2661
could lead to confusing results, and there is no need to be
2664
from bzrlib.repofmt.knitrepo import RepositoryFormatKnit
2666
are_knits = (isinstance(source._format, RepositoryFormatKnit) and
2667
isinstance(target._format, RepositoryFormatKnit))
2668
except AttributeError:
2670
return are_knits and InterRepository._same_model(source, target)
2673
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2674
"""See InterRepository.fetch()."""
2675
from bzrlib.fetch import KnitRepoFetcher
2676
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2677
self.source, self.source._format, self.target, self.target._format)
2678
f = KnitRepoFetcher(to_repository=self.target,
2679
from_repository=self.source,
2680
last_revision=revision_id,
2681
pb=pb, find_ghosts=find_ghosts)
2682
return f.count_copied, f.failed_revisions
2685
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2686
"""See InterRepository.missing_revision_ids()."""
2687
if revision_id is not None:
2688
source_ids = self.source.get_ancestry(revision_id)
2689
assert source_ids[0] is None
2692
source_ids = self.source.all_revision_ids()
2693
source_ids_set = set(source_ids)
2694
# source_ids is the worst possible case we may need to pull.
2695
# now we want to filter source_ids against what we actually
2696
# have in target, but don't try to check for existence where we know
2697
# we do not have a revision as that would be pointless.
2698
target_ids = set(self.target.all_revision_ids())
2699
possibly_present_revisions = target_ids.intersection(source_ids_set)
2700
actually_present_revisions = set(
2701
self.target._eliminate_revisions_not_present(possibly_present_revisions))
2702
required_revisions = source_ids_set.difference(actually_present_revisions)
2703
if revision_id is not None:
2704
# we used get_ancestry to determine source_ids then we are assured all
2705
# revisions referenced are present as they are installed in topological order.
2706
# and the tip revision was validated by get_ancestry.
2707
result_set = required_revisions
2709
# if we just grabbed the possibly available ids, then
2710
# we only have an estimate of whats available and need to validate
2711
# that against the revision records.
2713
self.source._eliminate_revisions_not_present(required_revisions))
2714
return self.source.revision_ids_to_search_result(result_set)
2717
class InterPackRepo(InterSameDataRepository):
2718
"""Optimised code paths between Pack based repositories."""
2721
def _get_repo_format_to_test(self):
2722
from bzrlib.repofmt import pack_repo
2723
return pack_repo.RepositoryFormatKnitPack1()
2726
def is_compatible(source, target):
2727
"""Be compatible with known Pack formats.
2729
We don't test for the stores being of specific types because that
2730
could lead to confusing results, and there is no need to be
2733
from bzrlib.repofmt.pack_repo import RepositoryFormatPack
2735
are_packs = (isinstance(source._format, RepositoryFormatPack) and
2736
isinstance(target._format, RepositoryFormatPack))
2737
except AttributeError:
2739
return are_packs and InterRepository._same_model(source, target)
2742
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2743
"""See InterRepository.fetch()."""
2744
from bzrlib.repofmt.pack_repo import Packer
2745
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2746
self.source, self.source._format, self.target, self.target._format)
2747
self.count_copied = 0
2748
if revision_id is None:
2750
# everything to do - use pack logic
2751
# to fetch from all packs to one without
2752
# inventory parsing etc, IFF nothing to be copied is in the target.
2754
revision_ids = self.source.all_revision_ids()
2755
# implementing the TODO will involve:
2756
# - detecting when all of a pack is selected
2757
# - avoiding as much as possible pre-selection, so the
2758
# more-core routines such as create_pack_from_packs can filter in
2759
# a just-in-time fashion. (though having a HEADS list on a
2760
# repository might make this a lot easier, because we could
2761
# sensibly detect 'new revisions' without doing a full index scan.
2762
elif _mod_revision.is_null(revision_id):
2767
revision_ids = self.search_missing_revision_ids(revision_id,
2768
find_ghosts=find_ghosts).get_keys()
2769
except errors.NoSuchRevision:
2770
raise errors.InstallFailed([revision_id])
2771
packs = self.source._pack_collection.all_packs()
2772
pack = Packer(self.target._pack_collection, packs, '.fetch',
2773
revision_ids).pack()
2774
if pack is not None:
2775
self.target._pack_collection._save_pack_names()
2776
# Trigger an autopack. This may duplicate effort as we've just done
2777
# a pack creation, but for now it is simpler to think about as
2778
# 'upload data, then repack if needed'.
2779
self.target._pack_collection.autopack()
2780
return (pack.get_revision_count(), [])
2785
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2786
"""See InterRepository.missing_revision_ids().
2788
:param find_ghosts: Find ghosts throughout the ancestry of
2791
if not find_ghosts and revision_id is not None:
2792
return self._walk_to_common_revisions([revision_id])
2793
elif revision_id is not None:
2794
source_ids = self.source.get_ancestry(revision_id)
2795
assert source_ids[0] is None
2798
source_ids = self.source.all_revision_ids()
2799
# source_ids is the worst possible case we may need to pull.
2800
# now we want to filter source_ids against what we actually
2801
# have in target, but don't try to check for existence where we know
2802
# we do not have a revision as that would be pointless.
2803
target_ids = set(self.target.all_revision_ids())
2804
result_set = set(source_ids).difference(target_ids)
2805
return self.source.revision_ids_to_search_result(result_set)
2808
class InterModel1and2(InterRepository):
2811
def _get_repo_format_to_test(self):
2815
def is_compatible(source, target):
2816
if not source.supports_rich_root() and target.supports_rich_root():
2822
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2823
"""See InterRepository.fetch()."""
2824
from bzrlib.fetch import Model1toKnit2Fetcher
2825
f = Model1toKnit2Fetcher(to_repository=self.target,
2826
from_repository=self.source,
2827
last_revision=revision_id,
2828
pb=pb, find_ghosts=find_ghosts)
2829
return f.count_copied, f.failed_revisions
2832
def copy_content(self, revision_id=None):
2833
"""Make a complete copy of the content in self into destination.
2835
This is a destructive operation! Do not use it on existing
2838
:param revision_id: Only copy the content needed to construct
2839
revision_id and its parents.
2842
self.target.set_make_working_trees(self.source.make_working_trees())
2843
except NotImplementedError:
2845
# but don't bother fetching if we have the needed data now.
2846
if (revision_id not in (None, _mod_revision.NULL_REVISION) and
2847
self.target.has_revision(revision_id)):
2849
self.target.fetch(self.source, revision_id=revision_id)
2852
class InterKnit1and2(InterKnitRepo):
2855
def _get_repo_format_to_test(self):
2859
def is_compatible(source, target):
2860
"""Be compatible with Knit1 source and Knit3 target"""
2861
from bzrlib.repofmt.knitrepo import RepositoryFormatKnit3
2863
from bzrlib.repofmt.knitrepo import (RepositoryFormatKnit1,
2864
RepositoryFormatKnit3)
2865
from bzrlib.repofmt.pack_repo import (
2866
RepositoryFormatKnitPack1,
2867
RepositoryFormatKnitPack3,
2868
RepositoryFormatPackDevelopment0,
2869
RepositoryFormatPackDevelopment0Subtree,
2872
RepositoryFormatKnit1,
2873
RepositoryFormatKnitPack1,
2874
RepositoryFormatPackDevelopment0,
2877
RepositoryFormatKnit3,
2878
RepositoryFormatKnitPack3,
2879
RepositoryFormatPackDevelopment0Subtree,
2881
return (isinstance(source._format, nosubtrees) and
2882
isinstance(target._format, subtrees))
2883
except AttributeError:
2887
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2888
"""See InterRepository.fetch()."""
2889
from bzrlib.fetch import Knit1to2Fetcher
2890
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2891
self.source, self.source._format, self.target,
2892
self.target._format)
2893
f = Knit1to2Fetcher(to_repository=self.target,
2894
from_repository=self.source,
2895
last_revision=revision_id,
2896
pb=pb, find_ghosts=find_ghosts)
2897
return f.count_copied, f.failed_revisions
2900
class InterDifferingSerializer(InterKnitRepo):
2903
def _get_repo_format_to_test(self):
2907
def is_compatible(source, target):
2908
"""Be compatible with Knit2 source and Knit3 target"""
2909
if source.supports_rich_root() != target.supports_rich_root():
2911
# Ideally, we'd support fetching if the source had no tree references
2912
# even if it supported them...
2913
if (getattr(source, '_format.supports_tree_reference', False) and
2914
not getattr(target, '_format.supports_tree_reference', False)):
2919
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2920
"""See InterRepository.fetch()."""
2921
revision_ids = self.target.search_missing_revision_ids(self.source,
2922
revision_id, find_ghosts=find_ghosts).get_keys()
2923
revision_ids = tsort.topo_sort(
2924
self.source.get_graph().get_parent_map(revision_ids))
2925
def revisions_iterator():
2926
for current_revision_id in revision_ids:
2927
revision = self.source.get_revision(current_revision_id)
2928
tree = self.source.revision_tree(current_revision_id)
2930
signature = self.source.get_signature_text(
2931
current_revision_id)
2932
except errors.NoSuchRevision:
2934
yield revision, tree, signature
2936
my_pb = ui.ui_factory.nested_progress_bar()
2941
install_revisions(self.target, revisions_iterator(),
2942
len(revision_ids), pb)
2944
if my_pb is not None:
2946
return len(revision_ids), 0
2949
class InterRemoteToOther(InterRepository):
2951
def __init__(self, source, target):
2952
InterRepository.__init__(self, source, target)
2953
self._real_inter = None
2956
def is_compatible(source, target):
2957
if not isinstance(source, remote.RemoteRepository):
2959
# Is source's model compatible with target's model?
2960
source._ensure_real()
2961
real_source = source._real_repository
2962
assert not isinstance(real_source, remote.RemoteRepository), (
2963
"We don't support remote repos backed by remote repos yet.")
2964
return InterRepository._same_model(real_source, target)
2967
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2968
"""See InterRepository.fetch()."""
2969
from bzrlib.fetch import RemoteToOtherFetcher
2970
mutter("Using fetch logic to copy between %s(remote) and %s(%s)",
2971
self.source, self.target, self.target._format)
2972
# TODO: jam 20070210 This should be an assert, not a translate
2973
revision_id = osutils.safe_revision_id(revision_id)
2974
f = RemoteToOtherFetcher(to_repository=self.target,
2975
from_repository=self.source,
2976
last_revision=revision_id,
2977
pb=pb, find_ghosts=find_ghosts)
2978
return f.count_copied, f.failed_revisions
2981
def _get_repo_format_to_test(self):
2985
class InterOtherToRemote(InterRepository):
2987
def __init__(self, source, target):
2988
InterRepository.__init__(self, source, target)
2989
self._real_inter = None
2992
def is_compatible(source, target):
2993
if isinstance(target, remote.RemoteRepository):
2997
def _ensure_real_inter(self):
2998
if self._real_inter is None:
2999
self.target._ensure_real()
3000
real_target = self.target._real_repository
3001
self._real_inter = InterRepository.get(self.source, real_target)
3003
def copy_content(self, revision_id=None):
3004
self._ensure_real_inter()
3005
self._real_inter.copy_content(revision_id=revision_id)
3007
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
3008
self._ensure_real_inter()
3009
self._real_inter.fetch(revision_id=revision_id, pb=pb,
3010
find_ghosts=find_ghosts)
3013
def _get_repo_format_to_test(self):
3017
InterRepository.register_optimiser(InterDifferingSerializer)
3018
InterRepository.register_optimiser(InterSameDataRepository)
3019
InterRepository.register_optimiser(InterWeaveRepo)
3020
InterRepository.register_optimiser(InterKnitRepo)
3021
InterRepository.register_optimiser(InterModel1and2)
3022
InterRepository.register_optimiser(InterKnit1and2)
3023
InterRepository.register_optimiser(InterPackRepo)
3024
InterRepository.register_optimiser(InterRemoteToOther)
3025
InterRepository.register_optimiser(InterOtherToRemote)
3028
class CopyConverter(object):
3029
"""A repository conversion tool which just performs a copy of the content.
3031
This is slow but quite reliable.
3034
def __init__(self, target_format):
3035
"""Create a CopyConverter.
3037
:param target_format: The format the resulting repository should be.
3039
self.target_format = target_format
3041
def convert(self, repo, pb):
3042
"""Perform the conversion of to_convert, giving feedback via pb.
3044
:param to_convert: The disk object to convert.
3045
:param pb: a progress bar to use for progress information.
3050
# this is only useful with metadir layouts - separated repo content.
3051
# trigger an assertion if not such
3052
repo._format.get_format_string()
3053
self.repo_dir = repo.bzrdir
3054
self.step('Moving repository to repository.backup')
3055
self.repo_dir.transport.move('repository', 'repository.backup')
3056
backup_transport = self.repo_dir.transport.clone('repository.backup')
3057
repo._format.check_conversion_target(self.target_format)
3058
self.source_repo = repo._format.open(self.repo_dir,
3060
_override_transport=backup_transport)
3061
self.step('Creating new repository')
3062
converted = self.target_format.initialize(self.repo_dir,
3063
self.source_repo.is_shared())
3064
converted.lock_write()
3066
self.step('Copying content into repository.')
3067
self.source_repo.copy_content_into(converted)
3070
self.step('Deleting old repository content.')
3071
self.repo_dir.transport.delete_tree('repository.backup')
3072
self.pb.note('repository converted')
3074
def step(self, message):
3075
"""Update the pb by a step."""
3077
self.pb.update(message, self.count, self.total)
3089
def _unescaper(match, _map=_unescape_map):
3090
code = match.group(1)
3094
if not code.startswith('#'):
3096
return unichr(int(code[1:])).encode('utf8')
3102
def _unescape_xml(data):
3103
"""Unescape predefined XML entities in a string of data."""
3105
if _unescape_re is None:
3106
_unescape_re = re.compile('\&([^;]*);')
3107
return _unescape_re.sub(_unescaper, data)
3110
class _VersionedFileChecker(object):
3112
def __init__(self, repository):
3113
self.repository = repository
3114
self.text_index = self.repository._generate_text_key_index()
3116
def calculate_file_version_parents(self, revision_id, file_id):
3117
"""Calculate the correct parents for a file version according to
3120
parent_keys = self.text_index[(file_id, revision_id)]
3121
if parent_keys == [_mod_revision.NULL_REVISION]:
3123
# strip the file_id, for the weave api
3124
return tuple([revision_id for file_id, revision_id in parent_keys])
3126
def check_file_version_parents(self, weave, file_id):
3127
"""Check the parents stored in a versioned file are correct.
3129
It also detects file versions that are not referenced by their
3130
corresponding revision's inventory.
3132
:returns: A tuple of (wrong_parents, dangling_file_versions).
3133
wrong_parents is a dict mapping {revision_id: (stored_parents,
3134
correct_parents)} for each revision_id where the stored parents
3135
are not correct. dangling_file_versions is a set of (file_id,
3136
revision_id) tuples for versions that are present in this versioned
3137
file, but not used by the corresponding inventory.
3140
unused_versions = set()
3141
versions = weave.versions()
3142
parent_map = weave.get_parent_map(versions)
3143
for num, revision_id in enumerate(versions):
3145
correct_parents = self.calculate_file_version_parents(
3146
revision_id, file_id)
3148
# The version is not part of the used keys.
3149
unused_versions.add(revision_id)
3152
knit_parents = tuple(parent_map[revision_id])
3153
except errors.RevisionNotPresent:
3155
if correct_parents != knit_parents:
3156
wrong_parents[revision_id] = (knit_parents, correct_parents)
3157
return wrong_parents, unused_versions
3160
def _old_get_graph(repository, revision_id):
3161
"""DO NOT USE. That is all. I'm serious."""
3162
graph = repository.get_graph()
3163
revision_graph = dict(((key, value) for key, value in
3164
graph.iter_ancestry([revision_id]) if value is not None))
3165
return _strip_NULL_ghosts(revision_graph)
3168
def _strip_NULL_ghosts(revision_graph):
3169
"""Also don't use this. more compatibility code for unmigrated clients."""
3170
# Filter ghosts, and null:
3171
if _mod_revision.NULL_REVISION in revision_graph:
3172
del revision_graph[_mod_revision.NULL_REVISION]
3173
for key, parents in revision_graph.items():
3174
revision_graph[key] = tuple(parent for parent in parents if parent
3176
return revision_graph
added
removed
bzrlib/repository.py
