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
423
return versionedfile.add_lines_with_ghosts(
424
self._new_revision_id, parents, new_lines,
425
nostore_sha=nostore_sha, random_id=self.random_revid,
426
check_content=False)[0:2]
429
class RootCommitBuilder(CommitBuilder):
430
"""This commitbuilder actually records the root id"""
432
# the root entry gets versioned properly by this builder.
433
_versioned_root = True
435
def _check_root(self, ie, parent_invs, tree):
436
"""Helper for record_entry_contents.
438
:param ie: An entry being added.
439
:param parent_invs: The inventories of the parent revisions of the
441
:param tree: The tree that is being committed.
445
######################################################################
448
class Repository(object):
449
"""Repository holding history for one or more branches.
451
The repository holds and retrieves historical information including
452
revisions and file history. It's normally accessed only by the Branch,
453
which views a particular line of development through that history.
455
The Repository builds on top of Stores and a Transport, which respectively
456
describe the disk data format and the way of accessing the (possibly
460
# What class to use for a CommitBuilder. Often its simpler to change this
461
# in a Repository class subclass rather than to override
462
# get_commit_builder.
463
_commit_builder_class = CommitBuilder
464
# The search regex used by xml based repositories to determine what things
465
# where changed in a single commit.
466
_file_ids_altered_regex = lazy_regex.lazy_compile(
467
r'file_id="(?P<file_id>[^"]+)"'
468
r'.* revision="(?P<revision_id>[^"]+)"'
471
def abort_write_group(self):
472
"""Commit the contents accrued within the current write group.
474
:seealso: start_write_group.
476
if self._write_group is not self.get_transaction():
477
# has an unlock or relock occured ?
478
raise errors.BzrError('mismatched lock context and write group.')
479
self._abort_write_group()
480
self._write_group = None
482
def _abort_write_group(self):
483
"""Template method for per-repository write group cleanup.
485
This is called during abort before the write group is considered to be
486
finished and should cleanup any internal state accrued during the write
487
group. There is no requirement that data handed to the repository be
488
*not* made available - this is not a rollback - but neither should any
489
attempt be made to ensure that data added is fully commited. Abort is
490
invoked when an error has occured so futher disk or network operations
491
may not be possible or may error and if possible should not be
495
def add_inventory(self, revision_id, inv, parents):
496
"""Add the inventory inv to the repository as revision_id.
498
:param parents: The revision ids of the parents that revision_id
499
is known to have and are in the repository already.
501
:returns: The validator(which is a sha1 digest, though what is sha'd is
502
repository format specific) of the serialized inventory.
504
assert self.is_in_write_group()
505
_mod_revision.check_not_reserved_id(revision_id)
506
assert inv.revision_id is None or inv.revision_id == revision_id, \
507
"Mismatch between inventory revision" \
508
" id and insertion revid (%r, %r)" % (inv.revision_id, revision_id)
509
assert inv.root is not None
510
inv_lines = self._serialise_inventory_to_lines(inv)
511
inv_vf = self.get_inventory_weave()
512
return self._inventory_add_lines(inv_vf, revision_id, parents,
513
inv_lines, check_content=False)
515
def _inventory_add_lines(self, inv_vf, revision_id, parents, lines,
517
"""Store lines in inv_vf and return the sha1 of the inventory."""
519
for parent in parents:
521
final_parents.append(parent)
522
return inv_vf.add_lines(revision_id, final_parents, lines,
523
check_content=check_content)[0]
525
def add_revision(self, revision_id, rev, inv=None, config=None):
526
"""Add rev to the revision store as revision_id.
528
:param revision_id: the revision id to use.
529
:param rev: The revision object.
530
:param inv: The inventory for the revision. if None, it will be looked
531
up in the inventory storer
532
:param config: If None no digital signature will be created.
533
If supplied its signature_needed method will be used
534
to determine if a signature should be made.
536
# TODO: jam 20070210 Shouldn't we check rev.revision_id and
538
_mod_revision.check_not_reserved_id(revision_id)
539
if config is not None and config.signature_needed():
541
inv = self.get_inventory(revision_id)
542
plaintext = Testament(rev, inv).as_short_text()
543
self.store_revision_signature(
544
gpg.GPGStrategy(config), plaintext, revision_id)
545
inventory_vf = self.get_inventory_weave()
546
if not revision_id in inventory_vf:
548
raise errors.WeaveRevisionNotPresent(revision_id,
551
# yes, this is not suitable for adding with ghosts.
552
rev.inventory_sha1 = self.add_inventory(revision_id, inv,
555
rev.inventory_sha1 = inventory_vf.get_sha1s([revision_id])[0]
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
graph = self.get_graph()
599
parent_map = graph.get_parent_map(revision_ids)
600
# The old API returned a list, should this actually be a set?
601
return parent_map.keys()
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()
1116
def get_deltas_for_revisions(self, revisions):
1117
"""Produce a generator of revision deltas.
1119
Note that the input is a sequence of REVISIONS, not revision_ids.
1120
Trees will be held in memory until the generator exits.
1121
Each delta is relative to the revision's lefthand predecessor.
1123
required_trees = set()
1124
for revision in revisions:
1125
required_trees.add(revision.revision_id)
1126
required_trees.update(revision.parent_ids[:1])
1127
trees = dict((t.get_revision_id(), t) for
1128
t in self.revision_trees(required_trees))
1129
for revision in revisions:
1130
if not revision.parent_ids:
1131
old_tree = self.revision_tree(None)
1133
old_tree = trees[revision.parent_ids[0]]
1134
yield trees[revision.revision_id].changes_from(old_tree)
1137
def get_revision_delta(self, revision_id):
1138
"""Return the delta for one revision.
1140
The delta is relative to the left-hand predecessor of the
1143
r = self.get_revision(revision_id)
1144
return list(self.get_deltas_for_revisions([r]))[0]
1147
def store_revision_signature(self, gpg_strategy, plaintext, revision_id):
1148
signature = gpg_strategy.sign(plaintext)
1149
self.add_signature_text(revision_id, signature)
1152
def add_signature_text(self, revision_id, signature):
1153
self._revision_store.add_revision_signature_text(revision_id,
1155
self.get_transaction())
1157
def find_text_key_references(self):
1158
"""Find the text key references within the repository.
1160
:return: a dictionary mapping (file_id, revision_id) tuples to altered file-ids to an iterable of
1161
revision_ids. Each altered file-ids has the exact revision_ids that
1162
altered it listed explicitly.
1163
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
1164
to whether they were referred to by the inventory of the
1165
revision_id that they contain. The inventory texts from all present
1166
revision ids are assessed to generate this report.
1168
revision_ids = self.all_revision_ids()
1169
w = self.get_inventory_weave()
1170
pb = ui.ui_factory.nested_progress_bar()
1172
return self._find_text_key_references_from_xml_inventory_lines(
1173
w.iter_lines_added_or_present_in_versions(revision_ids, pb=pb))
1177
def _find_text_key_references_from_xml_inventory_lines(self,
1179
"""Core routine for extracting references to texts from inventories.
1181
This performs the translation of xml lines to revision ids.
1183
:param line_iterator: An iterator of lines, origin_version_id
1184
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
1185
to whether they were referred to by the inventory of the
1186
revision_id that they contain. Note that if that revision_id was
1187
not part of the line_iterator's output then False will be given -
1188
even though it may actually refer to that key.
1190
if not self._serializer.support_altered_by_hack:
1191
raise AssertionError(
1192
"_find_text_key_references_from_xml_inventory_lines only "
1193
"supported for branches which store inventory as unnested xml"
1194
", not on %r" % self)
1197
# this code needs to read every new line in every inventory for the
1198
# inventories [revision_ids]. Seeing a line twice is ok. Seeing a line
1199
# not present in one of those inventories is unnecessary but not
1200
# harmful because we are filtering by the revision id marker in the
1201
# inventory lines : we only select file ids altered in one of those
1202
# revisions. We don't need to see all lines in the inventory because
1203
# only those added in an inventory in rev X can contain a revision=X
1205
unescape_revid_cache = {}
1206
unescape_fileid_cache = {}
1208
# jam 20061218 In a big fetch, this handles hundreds of thousands
1209
# of lines, so it has had a lot of inlining and optimizing done.
1210
# Sorry that it is a little bit messy.
1211
# Move several functions to be local variables, since this is a long
1213
search = self._file_ids_altered_regex.search
1214
unescape = _unescape_xml
1215
setdefault = result.setdefault
1216
for line, version_id in line_iterator:
1217
match = search(line)
1220
# One call to match.group() returning multiple items is quite a
1221
# bit faster than 2 calls to match.group() each returning 1
1222
file_id, revision_id = match.group('file_id', 'revision_id')
1224
# Inlining the cache lookups helps a lot when you make 170,000
1225
# lines and 350k ids, versus 8.4 unique ids.
1226
# Using a cache helps in 2 ways:
1227
# 1) Avoids unnecessary decoding calls
1228
# 2) Re-uses cached strings, which helps in future set and
1230
# (2) is enough that removing encoding entirely along with
1231
# the cache (so we are using plain strings) results in no
1232
# performance improvement.
1234
revision_id = unescape_revid_cache[revision_id]
1236
unescaped = unescape(revision_id)
1237
unescape_revid_cache[revision_id] = unescaped
1238
revision_id = unescaped
1240
# Note that unconditionally unescaping means that we deserialise
1241
# every fileid, which for general 'pull' is not great, but we don't
1242
# really want to have some many fulltexts that this matters anyway.
1245
file_id = unescape_fileid_cache[file_id]
1247
unescaped = unescape(file_id)
1248
unescape_fileid_cache[file_id] = unescaped
1251
key = (file_id, revision_id)
1252
setdefault(key, False)
1253
if revision_id == version_id:
1257
def _find_file_ids_from_xml_inventory_lines(self, line_iterator,
1259
"""Helper routine for fileids_altered_by_revision_ids.
1261
This performs the translation of xml lines to revision ids.
1263
:param line_iterator: An iterator of lines, origin_version_id
1264
:param revision_ids: The revision ids to filter for. This should be a
1265
set or other type which supports efficient __contains__ lookups, as
1266
the revision id from each parsed line will be looked up in the
1267
revision_ids filter.
1268
:return: a dictionary mapping altered file-ids to an iterable of
1269
revision_ids. Each altered file-ids has the exact revision_ids that
1270
altered it listed explicitly.
1273
setdefault = result.setdefault
1274
for file_id, revision_id in \
1275
self._find_text_key_references_from_xml_inventory_lines(
1276
line_iterator).iterkeys():
1277
# once data is all ensured-consistent; then this is
1278
# if revision_id == version_id
1279
if revision_id in revision_ids:
1280
setdefault(file_id, set()).add(revision_id)
1283
def fileids_altered_by_revision_ids(self, revision_ids, _inv_weave=None):
1284
"""Find the file ids and versions affected by revisions.
1286
:param revisions: an iterable containing revision ids.
1287
:param _inv_weave: The inventory weave from this repository or None.
1288
If None, the inventory weave will be opened automatically.
1289
:return: a dictionary mapping altered file-ids to an iterable of
1290
revision_ids. Each altered file-ids has the exact revision_ids that
1291
altered it listed explicitly.
1293
selected_revision_ids = set(revision_ids)
1294
w = _inv_weave or self.get_inventory_weave()
1295
pb = ui.ui_factory.nested_progress_bar()
1297
return self._find_file_ids_from_xml_inventory_lines(
1298
w.iter_lines_added_or_present_in_versions(
1299
selected_revision_ids, pb=pb),
1300
selected_revision_ids)
1304
def iter_files_bytes(self, desired_files):
1305
"""Iterate through file versions.
1307
Files will not necessarily be returned in the order they occur in
1308
desired_files. No specific order is guaranteed.
1310
Yields pairs of identifier, bytes_iterator. identifier is an opaque
1311
value supplied by the caller as part of desired_files. It should
1312
uniquely identify the file version in the caller's context. (Examples:
1313
an index number or a TreeTransform trans_id.)
1315
bytes_iterator is an iterable of bytestrings for the file. The
1316
kind of iterable and length of the bytestrings are unspecified, but for
1317
this implementation, it is a list of lines produced by
1318
VersionedFile.get_lines().
1320
:param desired_files: a list of (file_id, revision_id, identifier)
1323
transaction = self.get_transaction()
1324
for file_id, revision_id, callable_data in desired_files:
1326
weave = self.weave_store.get_weave(file_id, transaction)
1327
except errors.NoSuchFile:
1328
raise errors.NoSuchIdInRepository(self, file_id)
1329
yield callable_data, weave.get_lines(revision_id)
1331
def _generate_text_key_index(self, text_key_references=None,
1333
"""Generate a new text key index for the repository.
1335
This is an expensive function that will take considerable time to run.
1337
:return: A dict mapping text keys ((file_id, revision_id) tuples) to a
1338
list of parents, also text keys. When a given key has no parents,
1339
the parents list will be [NULL_REVISION].
1341
# All revisions, to find inventory parents.
1342
if ancestors is None:
1343
graph = self.get_graph()
1344
ancestors = graph.get_parent_map(self.all_revision_ids())
1345
if text_key_references is None:
1346
text_key_references = self.find_text_key_references()
1347
pb = ui.ui_factory.nested_progress_bar()
1349
return self._do_generate_text_key_index(ancestors,
1350
text_key_references, pb)
1354
def _do_generate_text_key_index(self, ancestors, text_key_references, pb):
1355
"""Helper for _generate_text_key_index to avoid deep nesting."""
1356
revision_order = tsort.topo_sort(ancestors)
1357
invalid_keys = set()
1359
for revision_id in revision_order:
1360
revision_keys[revision_id] = set()
1361
text_count = len(text_key_references)
1362
# a cache of the text keys to allow reuse; costs a dict of all the
1363
# keys, but saves a 2-tuple for every child of a given key.
1365
for text_key, valid in text_key_references.iteritems():
1367
invalid_keys.add(text_key)
1369
revision_keys[text_key[1]].add(text_key)
1370
text_key_cache[text_key] = text_key
1371
del text_key_references
1373
text_graph = graph.Graph(graph.DictParentsProvider(text_index))
1374
NULL_REVISION = _mod_revision.NULL_REVISION
1375
# Set a cache with a size of 10 - this suffices for bzr.dev but may be
1376
# too small for large or very branchy trees. However, for 55K path
1377
# trees, it would be easy to use too much memory trivially. Ideally we
1378
# could gauge this by looking at available real memory etc, but this is
1379
# always a tricky proposition.
1380
inventory_cache = lru_cache.LRUCache(10)
1381
batch_size = 10 # should be ~150MB on a 55K path tree
1382
batch_count = len(revision_order) / batch_size + 1
1384
pb.update("Calculating text parents.", processed_texts, text_count)
1385
for offset in xrange(batch_count):
1386
to_query = revision_order[offset * batch_size:(offset + 1) *
1390
for rev_tree in self.revision_trees(to_query):
1391
revision_id = rev_tree.get_revision_id()
1392
parent_ids = ancestors[revision_id]
1393
for text_key in revision_keys[revision_id]:
1394
pb.update("Calculating text parents.", processed_texts)
1395
processed_texts += 1
1396
candidate_parents = []
1397
for parent_id in parent_ids:
1398
parent_text_key = (text_key[0], parent_id)
1400
check_parent = parent_text_key not in \
1401
revision_keys[parent_id]
1403
# the parent parent_id is a ghost:
1404
check_parent = False
1405
# truncate the derived graph against this ghost.
1406
parent_text_key = None
1408
# look at the parent commit details inventories to
1409
# determine possible candidates in the per file graph.
1412
inv = inventory_cache[parent_id]
1414
inv = self.revision_tree(parent_id).inventory
1415
inventory_cache[parent_id] = inv
1416
parent_entry = inv._byid.get(text_key[0], None)
1417
if parent_entry is not None:
1419
text_key[0], parent_entry.revision)
1421
parent_text_key = None
1422
if parent_text_key is not None:
1423
candidate_parents.append(
1424
text_key_cache[parent_text_key])
1425
parent_heads = text_graph.heads(candidate_parents)
1426
new_parents = list(parent_heads)
1427
new_parents.sort(key=lambda x:candidate_parents.index(x))
1428
if new_parents == []:
1429
new_parents = [NULL_REVISION]
1430
text_index[text_key] = new_parents
1432
for text_key in invalid_keys:
1433
text_index[text_key] = [NULL_REVISION]
1436
def item_keys_introduced_by(self, revision_ids, _files_pb=None):
1437
"""Get an iterable listing the keys of all the data introduced by a set
1440
The keys will be ordered so that the corresponding items can be safely
1441
fetched and inserted in that order.
1443
:returns: An iterable producing tuples of (knit-kind, file-id,
1444
versions). knit-kind is one of 'file', 'inventory', 'signatures',
1445
'revisions'. file-id is None unless knit-kind is 'file'.
1447
# XXX: it's a bit weird to control the inventory weave caching in this
1448
# generator. Ideally the caching would be done in fetch.py I think. Or
1449
# maybe this generator should explicitly have the contract that it
1450
# should not be iterated until the previously yielded item has been
1452
inv_w = self.get_inventory_weave()
1454
# file ids that changed
1455
file_ids = self.fileids_altered_by_revision_ids(revision_ids, inv_w)
1457
num_file_ids = len(file_ids)
1458
for file_id, altered_versions in file_ids.iteritems():
1459
if _files_pb is not None:
1460
_files_pb.update("fetch texts", count, num_file_ids)
1462
yield ("file", file_id, altered_versions)
1463
# We're done with the files_pb. Note that it finished by the caller,
1464
# just as it was created by the caller.
1468
yield ("inventory", None, revision_ids)
1471
revisions_with_signatures = set()
1472
for rev_id in revision_ids:
1474
self.get_signature_text(rev_id)
1475
except errors.NoSuchRevision:
1479
revisions_with_signatures.add(rev_id)
1480
yield ("signatures", None, revisions_with_signatures)
1483
yield ("revisions", None, revision_ids)
1486
def get_inventory_weave(self):
1487
return self.control_weaves.get_weave('inventory',
1488
self.get_transaction())
1491
def get_inventory(self, revision_id):
1492
"""Get Inventory object by revision id."""
1493
return self.iter_inventories([revision_id]).next()
1495
def iter_inventories(self, revision_ids):
1496
"""Get many inventories by revision_ids.
1498
This will buffer some or all of the texts used in constructing the
1499
inventories in memory, but will only parse a single inventory at a
1502
:return: An iterator of inventories.
1504
assert None not in revision_ids
1505
assert _mod_revision.NULL_REVISION not in revision_ids
1506
return self._iter_inventories(revision_ids)
1508
def _iter_inventories(self, revision_ids):
1509
"""single-document based inventory iteration."""
1510
texts = self.get_inventory_weave().get_texts(revision_ids)
1511
for text, revision_id in zip(texts, revision_ids):
1512
yield self.deserialise_inventory(revision_id, text)
1514
def deserialise_inventory(self, revision_id, xml):
1515
"""Transform the xml into an inventory object.
1517
:param revision_id: The expected revision id of the inventory.
1518
:param xml: A serialised inventory.
1520
result = self._serializer.read_inventory_from_string(xml, revision_id)
1521
if result.revision_id != revision_id:
1522
raise AssertionError('revision id mismatch %s != %s' % (
1523
result.revision_id, revision_id))
1526
def serialise_inventory(self, inv):
1527
return self._serializer.write_inventory_to_string(inv)
1529
def _serialise_inventory_to_lines(self, inv):
1530
return self._serializer.write_inventory_to_lines(inv)
1532
def get_serializer_format(self):
1533
return self._serializer.format_num
1536
def get_inventory_xml(self, revision_id):
1537
"""Get inventory XML as a file object."""
1539
assert isinstance(revision_id, str), type(revision_id)
1540
iw = self.get_inventory_weave()
1541
return iw.get_text(revision_id)
1543
raise errors.HistoryMissing(self, 'inventory', revision_id)
1546
def get_inventory_sha1(self, revision_id):
1547
"""Return the sha1 hash of the inventory entry
1549
return self.get_revision(revision_id).inventory_sha1
1552
@deprecated_method(symbol_versioning.one_four)
1553
def get_revision_graph(self, revision_id=None):
1554
"""Return a dictionary containing the revision graph.
1556
NB: This method should not be used as it accesses the entire graph all
1557
at once, which is much more data than most operations should require.
1559
:param revision_id: The revision_id to get a graph from. If None, then
1560
the entire revision graph is returned. This is a deprecated mode of
1561
operation and will be removed in the future.
1562
:return: a dictionary of revision_id->revision_parents_list.
1564
raise NotImplementedError(self.get_revision_graph)
1567
@deprecated_method(symbol_versioning.one_three)
1568
def get_revision_graph_with_ghosts(self, revision_ids=None):
1569
"""Return a graph of the revisions with ghosts marked as applicable.
1571
:param revision_ids: an iterable of revisions to graph or None for all.
1572
:return: a Graph object with the graph reachable from revision_ids.
1574
if 'evil' in debug.debug_flags:
1576
"get_revision_graph_with_ghosts scales with size of history.")
1577
result = deprecated_graph.Graph()
1578
if not revision_ids:
1579
pending = set(self.all_revision_ids())
1582
pending = set(revision_ids)
1583
# special case NULL_REVISION
1584
if _mod_revision.NULL_REVISION in pending:
1585
pending.remove(_mod_revision.NULL_REVISION)
1586
required = set(pending)
1589
revision_id = pending.pop()
1591
rev = self.get_revision(revision_id)
1592
except errors.NoSuchRevision:
1593
if revision_id in required:
1596
result.add_ghost(revision_id)
1598
for parent_id in rev.parent_ids:
1599
# is this queued or done ?
1600
if (parent_id not in pending and
1601
parent_id not in done):
1603
pending.add(parent_id)
1604
result.add_node(revision_id, rev.parent_ids)
1605
done.add(revision_id)
1608
def iter_reverse_revision_history(self, revision_id):
1609
"""Iterate backwards through revision ids in the lefthand history
1611
:param revision_id: The revision id to start with. All its lefthand
1612
ancestors will be traversed.
1614
graph = self.get_graph()
1615
next_id = revision_id
1617
if next_id in (None, _mod_revision.NULL_REVISION):
1620
# Note: The following line may raise KeyError in the event of
1621
# truncated history. We decided not to have a try:except:raise
1622
# RevisionNotPresent here until we see a use for it, because of the
1623
# cost in an inner loop that is by its very nature O(history).
1624
# Robert Collins 20080326
1625
parents = graph.get_parent_map([next_id])[next_id]
1626
if len(parents) == 0:
1629
next_id = parents[0]
1632
def get_revision_inventory(self, revision_id):
1633
"""Return inventory of a past revision."""
1634
# TODO: Unify this with get_inventory()
1635
# bzr 0.0.6 and later imposes the constraint that the inventory_id
1636
# must be the same as its revision, so this is trivial.
1637
if revision_id is None:
1638
# This does not make sense: if there is no revision,
1639
# then it is the current tree inventory surely ?!
1640
# and thus get_root_id() is something that looks at the last
1641
# commit on the branch, and the get_root_id is an inventory check.
1642
raise NotImplementedError
1643
# return Inventory(self.get_root_id())
1645
return self.get_inventory(revision_id)
1648
def is_shared(self):
1649
"""Return True if this repository is flagged as a shared repository."""
1650
raise NotImplementedError(self.is_shared)
1653
def reconcile(self, other=None, thorough=False):
1654
"""Reconcile this repository."""
1655
from bzrlib.reconcile import RepoReconciler
1656
reconciler = RepoReconciler(self, thorough=thorough)
1657
reconciler.reconcile()
1660
def _refresh_data(self):
1661
"""Helper called from lock_* to ensure coherency with disk.
1663
The default implementation does nothing; it is however possible
1664
for repositories to maintain loaded indices across multiple locks
1665
by checking inside their implementation of this method to see
1666
whether their indices are still valid. This depends of course on
1667
the disk format being validatable in this manner.
1671
def revision_tree(self, revision_id):
1672
"""Return Tree for a revision on this branch.
1674
`revision_id` may be None for the empty tree revision.
1676
# TODO: refactor this to use an existing revision object
1677
# so we don't need to read it in twice.
1678
if revision_id is None or revision_id == _mod_revision.NULL_REVISION:
1679
return RevisionTree(self, Inventory(root_id=None),
1680
_mod_revision.NULL_REVISION)
1682
inv = self.get_revision_inventory(revision_id)
1683
return RevisionTree(self, inv, revision_id)
1685
def revision_trees(self, revision_ids):
1686
"""Return Tree for a revision on this branch.
1688
`revision_id` may not be None or 'null:'"""
1689
inventories = self.iter_inventories(revision_ids)
1690
for inv in inventories:
1691
yield RevisionTree(self, inv, inv.revision_id)
1694
def get_ancestry(self, revision_id, topo_sorted=True):
1695
"""Return a list of revision-ids integrated by a revision.
1697
The first element of the list is always None, indicating the origin
1698
revision. This might change when we have history horizons, or
1699
perhaps we should have a new API.
1701
This is topologically sorted.
1703
if _mod_revision.is_null(revision_id):
1705
if not self.has_revision(revision_id):
1706
raise errors.NoSuchRevision(self, revision_id)
1707
w = self.get_inventory_weave()
1708
candidates = w.get_ancestry(revision_id, topo_sorted)
1709
return [None] + candidates # self._eliminate_revisions_not_present(candidates)
1712
"""Compress the data within the repository.
1714
This operation only makes sense for some repository types. For other
1715
types it should be a no-op that just returns.
1717
This stub method does not require a lock, but subclasses should use
1718
@needs_write_lock as this is a long running call its reasonable to
1719
implicitly lock for the user.
1723
def print_file(self, file, revision_id):
1724
"""Print `file` to stdout.
1726
FIXME RBC 20060125 as John Meinel points out this is a bad api
1727
- it writes to stdout, it assumes that that is valid etc. Fix
1728
by creating a new more flexible convenience function.
1730
tree = self.revision_tree(revision_id)
1731
# use inventory as it was in that revision
1732
file_id = tree.inventory.path2id(file)
1734
# TODO: jam 20060427 Write a test for this code path
1735
# it had a bug in it, and was raising the wrong
1737
raise errors.BzrError("%r is not present in revision %s" % (file, revision_id))
1738
tree.print_file(file_id)
1740
def get_transaction(self):
1741
return self.control_files.get_transaction()
1743
def revision_parents(self, revision_id):
1744
return self.get_inventory_weave().parent_names(revision_id)
1746
@deprecated_method(symbol_versioning.one_one)
1747
def get_parents(self, revision_ids):
1748
"""See StackedParentsProvider.get_parents"""
1749
parent_map = self.get_parent_map(revision_ids)
1750
return [parent_map.get(r, None) for r in revision_ids]
1752
def get_parent_map(self, keys):
1753
"""See graph._StackedParentsProvider.get_parent_map"""
1755
for revision_id in keys:
1756
if revision_id == _mod_revision.NULL_REVISION:
1757
parent_map[revision_id] = ()
1760
parent_id_list = self.get_revision(revision_id).parent_ids
1761
except errors.NoSuchRevision:
1764
if len(parent_id_list) == 0:
1765
parent_ids = (_mod_revision.NULL_REVISION,)
1767
parent_ids = tuple(parent_id_list)
1768
parent_map[revision_id] = parent_ids
1771
def _make_parents_provider(self):
1774
def get_graph(self, other_repository=None):
1775
"""Return the graph walker for this repository format"""
1776
parents_provider = self._make_parents_provider()
1777
if (other_repository is not None and
1778
not self.has_same_location(other_repository)):
1779
parents_provider = graph._StackedParentsProvider(
1780
[parents_provider, other_repository._make_parents_provider()])
1781
return graph.Graph(parents_provider)
1783
def _get_versioned_file_checker(self):
1784
"""Return an object suitable for checking versioned files."""
1785
return _VersionedFileChecker(self)
1787
def revision_ids_to_search_result(self, result_set):
1788
"""Convert a set of revision ids to a graph SearchResult."""
1789
result_parents = set()
1790
for parents in self.get_graph().get_parent_map(
1791
result_set).itervalues():
1792
result_parents.update(parents)
1793
included_keys = result_set.intersection(result_parents)
1794
start_keys = result_set.difference(included_keys)
1795
exclude_keys = result_parents.difference(result_set)
1796
result = graph.SearchResult(start_keys, exclude_keys,
1797
len(result_set), result_set)
1801
def set_make_working_trees(self, new_value):
1802
"""Set the policy flag for making working trees when creating branches.
1804
This only applies to branches that use this repository.
1806
The default is 'True'.
1807
:param new_value: True to restore the default, False to disable making
1810
raise NotImplementedError(self.set_make_working_trees)
1812
def make_working_trees(self):
1813
"""Returns the policy for making working trees on new branches."""
1814
raise NotImplementedError(self.make_working_trees)
1817
def sign_revision(self, revision_id, gpg_strategy):
1818
plaintext = Testament.from_revision(self, revision_id).as_short_text()
1819
self.store_revision_signature(gpg_strategy, plaintext, revision_id)
1822
def has_signature_for_revision_id(self, revision_id):
1823
"""Query for a revision signature for revision_id in the repository."""
1824
return self._revision_store.has_signature(revision_id,
1825
self.get_transaction())
1828
def get_signature_text(self, revision_id):
1829
"""Return the text for a signature."""
1830
return self._revision_store.get_signature_text(revision_id,
1831
self.get_transaction())
1834
def check(self, revision_ids=None):
1835
"""Check consistency of all history of given revision_ids.
1837
Different repository implementations should override _check().
1839
:param revision_ids: A non-empty list of revision_ids whose ancestry
1840
will be checked. Typically the last revision_id of a branch.
1842
return self._check(revision_ids)
1844
def _check(self, revision_ids):
1845
result = check.Check(self)
1849
def _warn_if_deprecated(self):
1850
global _deprecation_warning_done
1851
if _deprecation_warning_done:
1853
_deprecation_warning_done = True
1854
warning("Format %s for %s is deprecated - please use 'bzr upgrade' to get better performance"
1855
% (self._format, self.bzrdir.transport.base))
1857
def supports_rich_root(self):
1858
return self._format.rich_root_data
1860
def _check_ascii_revisionid(self, revision_id, method):
1861
"""Private helper for ascii-only repositories."""
1862
# weave repositories refuse to store revisionids that are non-ascii.
1863
if revision_id is not None:
1864
# weaves require ascii revision ids.
1865
if isinstance(revision_id, unicode):
1867
revision_id.encode('ascii')
1868
except UnicodeEncodeError:
1869
raise errors.NonAsciiRevisionId(method, self)
1872
revision_id.decode('ascii')
1873
except UnicodeDecodeError:
1874
raise errors.NonAsciiRevisionId(method, self)
1876
def revision_graph_can_have_wrong_parents(self):
1877
"""Is it possible for this repository to have a revision graph with
1880
If True, then this repository must also implement
1881
_find_inconsistent_revision_parents so that check and reconcile can
1882
check for inconsistencies before proceeding with other checks that may
1883
depend on the revision index being consistent.
1885
raise NotImplementedError(self.revision_graph_can_have_wrong_parents)
1888
# remove these delegates a while after bzr 0.15
1889
def __make_delegated(name, from_module):
1890
def _deprecated_repository_forwarder():
1891
symbol_versioning.warn('%s moved to %s in bzr 0.15'
1892
% (name, from_module),
1895
m = __import__(from_module, globals(), locals(), [name])
1897
return getattr(m, name)
1898
except AttributeError:
1899
raise AttributeError('module %s has no name %s'
1901
globals()[name] = _deprecated_repository_forwarder
1904
'AllInOneRepository',
1905
'WeaveMetaDirRepository',
1906
'PreSplitOutRepositoryFormat',
1907
'RepositoryFormat4',
1908
'RepositoryFormat5',
1909
'RepositoryFormat6',
1910
'RepositoryFormat7',
1912
__make_delegated(_name, 'bzrlib.repofmt.weaverepo')
1916
'RepositoryFormatKnit',
1917
'RepositoryFormatKnit1',
1919
__make_delegated(_name, 'bzrlib.repofmt.knitrepo')
1922
def install_revision(repository, rev, revision_tree):
1923
"""Install all revision data into a repository."""
1924
install_revisions(repository, [(rev, revision_tree, None)])
1927
def install_revisions(repository, iterable, num_revisions=None, pb=None):
1928
"""Install all revision data into a repository.
1930
Accepts an iterable of revision, tree, signature tuples. The signature
1933
repository.start_write_group()
1935
for n, (revision, revision_tree, signature) in enumerate(iterable):
1936
_install_revision(repository, revision, revision_tree, signature)
1938
pb.update('Transferring revisions', n + 1, num_revisions)
1940
repository.abort_write_group()
1943
repository.commit_write_group()
1946
def _install_revision(repository, rev, revision_tree, signature):
1947
"""Install all revision data into a repository."""
1948
present_parents = []
1950
for p_id in rev.parent_ids:
1951
if repository.has_revision(p_id):
1952
present_parents.append(p_id)
1953
parent_trees[p_id] = repository.revision_tree(p_id)
1955
parent_trees[p_id] = repository.revision_tree(None)
1957
inv = revision_tree.inventory
1958
entries = inv.iter_entries()
1959
# backwards compatibility hack: skip the root id.
1960
if not repository.supports_rich_root():
1961
path, root = entries.next()
1962
if root.revision != rev.revision_id:
1963
raise errors.IncompatibleRevision(repr(repository))
1964
# Add the texts that are not already present
1965
for path, ie in entries:
1966
w = repository.weave_store.get_weave_or_empty(ie.file_id,
1967
repository.get_transaction())
1968
if ie.revision not in w:
1970
# FIXME: TODO: The following loop *may* be overlapping/duplicate
1971
# with InventoryEntry.find_previous_heads(). if it is, then there
1972
# is a latent bug here where the parents may have ancestors of each
1974
for revision, tree in parent_trees.iteritems():
1975
if ie.file_id not in tree:
1977
parent_id = tree.inventory[ie.file_id].revision
1978
if parent_id in text_parents:
1980
text_parents.append(parent_id)
1982
vfile = repository.weave_store.get_weave_or_empty(ie.file_id,
1983
repository.get_transaction())
1984
lines = revision_tree.get_file(ie.file_id).readlines()
1985
vfile.add_lines(rev.revision_id, text_parents, lines)
1987
# install the inventory
1988
repository.add_inventory(rev.revision_id, inv, present_parents)
1989
except errors.RevisionAlreadyPresent:
1991
if signature is not None:
1992
repository.add_signature_text(rev.revision_id, signature)
1993
repository.add_revision(rev.revision_id, rev, inv)
1996
class MetaDirRepository(Repository):
1997
"""Repositories in the new meta-dir layout."""
1999
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
2000
super(MetaDirRepository, self).__init__(_format,
2006
dir_mode = self.control_files._dir_mode
2007
file_mode = self.control_files._file_mode
2010
def is_shared(self):
2011
"""Return True if this repository is flagged as a shared repository."""
2012
return self.control_files._transport.has('shared-storage')
2015
def set_make_working_trees(self, new_value):
2016
"""Set the policy flag for making working trees when creating branches.
2018
This only applies to branches that use this repository.
2020
The default is 'True'.
2021
:param new_value: True to restore the default, False to disable making
2026
self.control_files._transport.delete('no-working-trees')
2027
except errors.NoSuchFile:
2030
self.control_files.put_utf8('no-working-trees', '')
2032
def make_working_trees(self):
2033
"""Returns the policy for making working trees on new branches."""
2034
return not self.control_files._transport.has('no-working-trees')
2037
class MetaDirVersionedFileRepository(MetaDirRepository):
2038
"""Repositories in a meta-dir, that work via versioned file objects."""
2040
def __init__(self, _format, a_bzrdir, control_files, _revision_store, control_store, text_store):
2041
super(MetaDirVersionedFileRepository, self).__init__(_format, a_bzrdir,
2042
control_files, _revision_store, control_store, text_store)
2043
_revision_store.get_scope = self.get_transaction
2044
control_store.get_scope = self.get_transaction
2045
text_store.get_scope = self.get_transaction
2048
class RepositoryFormatRegistry(registry.Registry):
2049
"""Registry of RepositoryFormats."""
2051
def get(self, format_string):
2052
r = registry.Registry.get(self, format_string)
2058
format_registry = RepositoryFormatRegistry()
2059
"""Registry of formats, indexed by their identifying format string.
2061
This can contain either format instances themselves, or classes/factories that
2062
can be called to obtain one.
2066
#####################################################################
2067
# Repository Formats
2069
class RepositoryFormat(object):
2070
"""A repository format.
2072
Formats provide three things:
2073
* An initialization routine to construct repository data on disk.
2074
* a format string which is used when the BzrDir supports versioned
2076
* an open routine which returns a Repository instance.
2078
There is one and only one Format subclass for each on-disk format. But
2079
there can be one Repository subclass that is used for several different
2080
formats. The _format attribute on a Repository instance can be used to
2081
determine the disk format.
2083
Formats are placed in an dict by their format string for reference
2084
during opening. These should be subclasses of RepositoryFormat
2087
Once a format is deprecated, just deprecate the initialize and open
2088
methods on the format class. Do not deprecate the object, as the
2089
object will be created every system load.
2091
Common instance attributes:
2092
_matchingbzrdir - the bzrdir format that the repository format was
2093
originally written to work with. This can be used if manually
2094
constructing a bzrdir and repository, or more commonly for test suite
2098
# Set to True or False in derived classes. True indicates that the format
2099
# supports ghosts gracefully.
2100
supports_ghosts = None
2101
# Can this repository be given external locations to lookup additional
2102
# data. Set to True or False in derived classes.
2103
supports_external_lookups = None
2106
return "<%s>" % self.__class__.__name__
2108
def __eq__(self, other):
2109
# format objects are generally stateless
2110
return isinstance(other, self.__class__)
2112
def __ne__(self, other):
2113
return not self == other
2116
def find_format(klass, a_bzrdir):
2117
"""Return the format for the repository object in a_bzrdir.
2119
This is used by bzr native formats that have a "format" file in
2120
the repository. Other methods may be used by different types of
2124
transport = a_bzrdir.get_repository_transport(None)
2125
format_string = transport.get("format").read()
2126
return format_registry.get(format_string)
2127
except errors.NoSuchFile:
2128
raise errors.NoRepositoryPresent(a_bzrdir)
2130
raise errors.UnknownFormatError(format=format_string,
2134
def register_format(klass, format):
2135
format_registry.register(format.get_format_string(), format)
2138
def unregister_format(klass, format):
2139
format_registry.remove(format.get_format_string())
2142
def get_default_format(klass):
2143
"""Return the current default format."""
2144
from bzrlib import bzrdir
2145
return bzrdir.format_registry.make_bzrdir('default').repository_format
2147
def _get_control_store(self, repo_transport, control_files):
2148
"""Return the control store for this repository."""
2149
raise NotImplementedError(self._get_control_store)
2151
def get_format_string(self):
2152
"""Return the ASCII format string that identifies this format.
2154
Note that in pre format ?? repositories the format string is
2155
not permitted nor written to disk.
2157
raise NotImplementedError(self.get_format_string)
2159
def get_format_description(self):
2160
"""Return the short description for this format."""
2161
raise NotImplementedError(self.get_format_description)
2163
def _get_revision_store(self, repo_transport, control_files):
2164
"""Return the revision store object for this a_bzrdir."""
2165
raise NotImplementedError(self._get_revision_store)
2167
def _get_text_rev_store(self,
2174
"""Common logic for getting a revision store for a repository.
2176
see self._get_revision_store for the subclass-overridable method to
2177
get the store for a repository.
2179
from bzrlib.store.revision.text import TextRevisionStore
2180
dir_mode = control_files._dir_mode
2181
file_mode = control_files._file_mode
2182
text_store = TextStore(transport.clone(name),
2184
compressed=compressed,
2186
file_mode=file_mode)
2187
_revision_store = TextRevisionStore(text_store, serializer)
2188
return _revision_store
2190
# TODO: this shouldn't be in the base class, it's specific to things that
2191
# use weaves or knits -- mbp 20070207
2192
def _get_versioned_file_store(self,
2197
versionedfile_class=None,
2198
versionedfile_kwargs={},
2200
if versionedfile_class is None:
2201
versionedfile_class = self._versionedfile_class
2202
weave_transport = control_files._transport.clone(name)
2203
dir_mode = control_files._dir_mode
2204
file_mode = control_files._file_mode
2205
return VersionedFileStore(weave_transport, prefixed=prefixed,
2207
file_mode=file_mode,
2208
versionedfile_class=versionedfile_class,
2209
versionedfile_kwargs=versionedfile_kwargs,
2212
def initialize(self, a_bzrdir, shared=False):
2213
"""Initialize a repository of this format in a_bzrdir.
2215
:param a_bzrdir: The bzrdir to put the new repository in it.
2216
:param shared: The repository should be initialized as a sharable one.
2217
:returns: The new repository object.
2219
This may raise UninitializableFormat if shared repository are not
2220
compatible the a_bzrdir.
2222
raise NotImplementedError(self.initialize)
2224
def is_supported(self):
2225
"""Is this format supported?
2227
Supported formats must be initializable and openable.
2228
Unsupported formats may not support initialization or committing or
2229
some other features depending on the reason for not being supported.
2233
def check_conversion_target(self, target_format):
2234
raise NotImplementedError(self.check_conversion_target)
2236
def open(self, a_bzrdir, _found=False):
2237
"""Return an instance of this format for the bzrdir a_bzrdir.
2239
_found is a private parameter, do not use it.
2241
raise NotImplementedError(self.open)
2244
class MetaDirRepositoryFormat(RepositoryFormat):
2245
"""Common base class for the new repositories using the metadir layout."""
2247
rich_root_data = False
2248
supports_tree_reference = False
2249
supports_external_lookups = False
2250
_matchingbzrdir = bzrdir.BzrDirMetaFormat1()
2253
super(MetaDirRepositoryFormat, self).__init__()
2255
def _create_control_files(self, a_bzrdir):
2256
"""Create the required files and the initial control_files object."""
2257
# FIXME: RBC 20060125 don't peek under the covers
2258
# NB: no need to escape relative paths that are url safe.
2259
repository_transport = a_bzrdir.get_repository_transport(self)
2260
control_files = lockable_files.LockableFiles(repository_transport,
2261
'lock', lockdir.LockDir)
2262
control_files.create_lock()
2263
return control_files
2265
def _upload_blank_content(self, a_bzrdir, dirs, files, utf8_files, shared):
2266
"""Upload the initial blank content."""
2267
control_files = self._create_control_files(a_bzrdir)
2268
control_files.lock_write()
2270
control_files._transport.mkdir_multi(dirs,
2271
mode=control_files._dir_mode)
2272
for file, content in files:
2273
control_files.put(file, content)
2274
for file, content in utf8_files:
2275
control_files.put_utf8(file, content)
2277
control_files.put_utf8('shared-storage', '')
2279
control_files.unlock()
2282
# formats which have no format string are not discoverable
2283
# and not independently creatable, so are not registered. They're
2284
# all in bzrlib.repofmt.weaverepo now. When an instance of one of these is
2285
# needed, it's constructed directly by the BzrDir. Non-native formats where
2286
# the repository is not separately opened are similar.
2288
format_registry.register_lazy(
2289
'Bazaar-NG Repository format 7',
2290
'bzrlib.repofmt.weaverepo',
2294
format_registry.register_lazy(
2295
'Bazaar-NG Knit Repository Format 1',
2296
'bzrlib.repofmt.knitrepo',
2297
'RepositoryFormatKnit1',
2300
format_registry.register_lazy(
2301
'Bazaar Knit Repository Format 3 (bzr 0.15)\n',
2302
'bzrlib.repofmt.knitrepo',
2303
'RepositoryFormatKnit3',
2306
format_registry.register_lazy(
2307
'Bazaar Knit Repository Format 4 (bzr 1.0)\n',
2308
'bzrlib.repofmt.knitrepo',
2309
'RepositoryFormatKnit4',
2312
# Pack-based formats. There is one format for pre-subtrees, and one for
2313
# post-subtrees to allow ease of testing.
2314
# NOTE: These are experimental in 0.92. Stable in 1.0 and above
2315
format_registry.register_lazy(
2316
'Bazaar pack repository format 1 (needs bzr 0.92)\n',
2317
'bzrlib.repofmt.pack_repo',
2318
'RepositoryFormatKnitPack1',
2320
format_registry.register_lazy(
2321
'Bazaar pack repository format 1 with subtree support (needs bzr 0.92)\n',
2322
'bzrlib.repofmt.pack_repo',
2323
'RepositoryFormatKnitPack3',
2325
format_registry.register_lazy(
2326
'Bazaar pack repository format 1 with rich root (needs bzr 1.0)\n',
2327
'bzrlib.repofmt.pack_repo',
2328
'RepositoryFormatKnitPack4',
2330
# Development formats.
2332
# development 0 - stub to introduce development versioning scheme.
2333
format_registry.register_lazy(
2334
"Bazaar development format 0 (needs bzr.dev from before 1.3)\n",
2335
'bzrlib.repofmt.pack_repo',
2336
'RepositoryFormatPackDevelopment0',
2338
format_registry.register_lazy(
2339
("Bazaar development format 0 with subtree support "
2340
"(needs bzr.dev from before 1.3)\n"),
2341
'bzrlib.repofmt.pack_repo',
2342
'RepositoryFormatPackDevelopment0Subtree',
2344
# 1.3->1.4 go below here
2347
class InterRepository(InterObject):
2348
"""This class represents operations taking place between two repositories.
2350
Its instances have methods like copy_content and fetch, and contain
2351
references to the source and target repositories these operations can be
2354
Often we will provide convenience methods on 'repository' which carry out
2355
operations with another repository - they will always forward to
2356
InterRepository.get(other).method_name(parameters).
2360
"""The available optimised InterRepository types."""
2362
def copy_content(self, revision_id=None):
2363
raise NotImplementedError(self.copy_content)
2365
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2366
"""Fetch the content required to construct revision_id.
2368
The content is copied from self.source to self.target.
2370
:param revision_id: if None all content is copied, if NULL_REVISION no
2372
:param pb: optional progress bar to use for progress reports. If not
2373
provided a default one will be created.
2375
Returns the copied revision count and the failed revisions in a tuple:
2378
raise NotImplementedError(self.fetch)
2380
def _walk_to_common_revisions(self, revision_ids):
2381
"""Walk out from revision_ids in source to revisions target has.
2383
:param revision_ids: The start point for the search.
2384
:return: A set of revision ids.
2386
target_graph = self.target.get_graph()
2387
revision_ids = frozenset(revision_ids)
2388
if set(target_graph.get_parent_map(revision_ids)) == revision_ids:
2389
return graph.SearchResult(revision_ids, set(), 0, set())
2390
missing_revs = set()
2391
source_graph = self.source.get_graph()
2392
# ensure we don't pay silly lookup costs.
2393
searcher = source_graph._make_breadth_first_searcher(revision_ids)
2394
null_set = frozenset([_mod_revision.NULL_REVISION])
2397
next_revs, ghosts = searcher.next_with_ghosts()
2398
except StopIteration:
2400
if revision_ids.intersection(ghosts):
2401
absent_ids = set(revision_ids.intersection(ghosts))
2402
# If all absent_ids are present in target, no error is needed.
2403
absent_ids.difference_update(
2404
set(target_graph.get_parent_map(absent_ids)))
2406
raise errors.NoSuchRevision(self.source, absent_ids.pop())
2407
# we don't care about other ghosts as we can't fetch them and
2408
# haven't been asked to.
2409
next_revs = set(next_revs)
2410
# we always have NULL_REVISION present.
2411
have_revs = set(target_graph.get_parent_map(next_revs)).union(null_set)
2412
missing_revs.update(next_revs - have_revs)
2413
searcher.stop_searching_any(have_revs)
2414
return searcher.get_result()
2416
@deprecated_method(symbol_versioning.one_two)
2418
def missing_revision_ids(self, revision_id=None, find_ghosts=True):
2419
"""Return the revision ids that source has that target does not.
2421
These are returned in topological order.
2423
:param revision_id: only return revision ids included by this
2425
:param find_ghosts: If True find missing revisions in deep history
2426
rather than just finding the surface difference.
2428
return list(self.search_missing_revision_ids(
2429
revision_id, find_ghosts).get_keys())
2432
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2433
"""Return the revision ids that source has that target does not.
2435
:param revision_id: only return revision ids included by this
2437
:param find_ghosts: If True find missing revisions in deep history
2438
rather than just finding the surface difference.
2439
:return: A bzrlib.graph.SearchResult.
2441
# stop searching at found target revisions.
2442
if not find_ghosts and revision_id is not None:
2443
return self._walk_to_common_revisions([revision_id])
2444
# generic, possibly worst case, slow code path.
2445
target_ids = set(self.target.all_revision_ids())
2446
if revision_id is not None:
2447
source_ids = self.source.get_ancestry(revision_id)
2448
assert source_ids[0] is None
2451
source_ids = self.source.all_revision_ids()
2452
result_set = set(source_ids).difference(target_ids)
2453
return self.source.revision_ids_to_search_result(result_set)
2456
def _same_model(source, target):
2457
"""True if source and target have the same data representation."""
2458
if source.supports_rich_root() != target.supports_rich_root():
2460
if source._serializer != target._serializer:
2465
class InterSameDataRepository(InterRepository):
2466
"""Code for converting between repositories that represent the same data.
2468
Data format and model must match for this to work.
2472
def _get_repo_format_to_test(self):
2473
"""Repository format for testing with.
2475
InterSameData can pull from subtree to subtree and from non-subtree to
2476
non-subtree, so we test this with the richest repository format.
2478
from bzrlib.repofmt import knitrepo
2479
return knitrepo.RepositoryFormatKnit3()
2482
def is_compatible(source, target):
2483
return InterRepository._same_model(source, target)
2486
def copy_content(self, revision_id=None):
2487
"""Make a complete copy of the content in self into destination.
2489
This copies both the repository's revision data, and configuration information
2490
such as the make_working_trees setting.
2492
This is a destructive operation! Do not use it on existing
2495
:param revision_id: Only copy the content needed to construct
2496
revision_id and its parents.
2499
self.target.set_make_working_trees(self.source.make_working_trees())
2500
except NotImplementedError:
2502
# but don't bother fetching if we have the needed data now.
2503
if (revision_id not in (None, _mod_revision.NULL_REVISION) and
2504
self.target.has_revision(revision_id)):
2506
self.target.fetch(self.source, revision_id=revision_id)
2509
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2510
"""See InterRepository.fetch()."""
2511
from bzrlib.fetch import GenericRepoFetcher
2512
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2513
self.source, self.source._format, self.target,
2514
self.target._format)
2515
f = GenericRepoFetcher(to_repository=self.target,
2516
from_repository=self.source,
2517
last_revision=revision_id,
2518
pb=pb, find_ghosts=find_ghosts)
2519
return f.count_copied, f.failed_revisions
2522
class InterWeaveRepo(InterSameDataRepository):
2523
"""Optimised code paths between Weave based repositories.
2525
This should be in bzrlib/repofmt/weaverepo.py but we have not yet
2526
implemented lazy inter-object optimisation.
2530
def _get_repo_format_to_test(self):
2531
from bzrlib.repofmt import weaverepo
2532
return weaverepo.RepositoryFormat7()
2535
def is_compatible(source, target):
2536
"""Be compatible with known Weave formats.
2538
We don't test for the stores being of specific types because that
2539
could lead to confusing results, and there is no need to be
2542
from bzrlib.repofmt.weaverepo import (
2548
return (isinstance(source._format, (RepositoryFormat5,
2550
RepositoryFormat7)) and
2551
isinstance(target._format, (RepositoryFormat5,
2553
RepositoryFormat7)))
2554
except AttributeError:
2558
def copy_content(self, revision_id=None):
2559
"""See InterRepository.copy_content()."""
2560
# weave specific optimised path:
2562
self.target.set_make_working_trees(self.source.make_working_trees())
2563
except (errors.RepositoryUpgradeRequired, NotImplemented):
2565
# FIXME do not peek!
2566
if self.source.control_files._transport.listable():
2567
pb = ui.ui_factory.nested_progress_bar()
2569
self.target.weave_store.copy_all_ids(
2570
self.source.weave_store,
2572
from_transaction=self.source.get_transaction(),
2573
to_transaction=self.target.get_transaction())
2574
pb.update('copying inventory', 0, 1)
2575
self.target.control_weaves.copy_multi(
2576
self.source.control_weaves, ['inventory'],
2577
from_transaction=self.source.get_transaction(),
2578
to_transaction=self.target.get_transaction())
2579
self.target._revision_store.text_store.copy_all_ids(
2580
self.source._revision_store.text_store,
2585
self.target.fetch(self.source, revision_id=revision_id)
2588
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2589
"""See InterRepository.fetch()."""
2590
from bzrlib.fetch import GenericRepoFetcher
2591
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2592
self.source, self.source._format, self.target, self.target._format)
2593
f = GenericRepoFetcher(to_repository=self.target,
2594
from_repository=self.source,
2595
last_revision=revision_id,
2596
pb=pb, find_ghosts=find_ghosts)
2597
return f.count_copied, f.failed_revisions
2600
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2601
"""See InterRepository.missing_revision_ids()."""
2602
# we want all revisions to satisfy revision_id in source.
2603
# but we don't want to stat every file here and there.
2604
# we want then, all revisions other needs to satisfy revision_id
2605
# checked, but not those that we have locally.
2606
# so the first thing is to get a subset of the revisions to
2607
# satisfy revision_id in source, and then eliminate those that
2608
# we do already have.
2609
# this is slow on high latency connection to self, but as as this
2610
# disk format scales terribly for push anyway due to rewriting
2611
# inventory.weave, this is considered acceptable.
2613
if revision_id is not None:
2614
source_ids = self.source.get_ancestry(revision_id)
2615
assert source_ids[0] is None
2618
source_ids = self.source._all_possible_ids()
2619
source_ids_set = set(source_ids)
2620
# source_ids is the worst possible case we may need to pull.
2621
# now we want to filter source_ids against what we actually
2622
# have in target, but don't try to check for existence where we know
2623
# we do not have a revision as that would be pointless.
2624
target_ids = set(self.target._all_possible_ids())
2625
possibly_present_revisions = target_ids.intersection(source_ids_set)
2626
actually_present_revisions = set(
2627
self.target._eliminate_revisions_not_present(possibly_present_revisions))
2628
required_revisions = source_ids_set.difference(actually_present_revisions)
2629
if revision_id is not None:
2630
# we used get_ancestry to determine source_ids then we are assured all
2631
# revisions referenced are present as they are installed in topological order.
2632
# and the tip revision was validated by get_ancestry.
2633
result_set = required_revisions
2635
# if we just grabbed the possibly available ids, then
2636
# we only have an estimate of whats available and need to validate
2637
# that against the revision records.
2639
self.source._eliminate_revisions_not_present(required_revisions))
2640
return self.source.revision_ids_to_search_result(result_set)
2643
class InterKnitRepo(InterSameDataRepository):
2644
"""Optimised code paths between Knit based repositories."""
2647
def _get_repo_format_to_test(self):
2648
from bzrlib.repofmt import knitrepo
2649
return knitrepo.RepositoryFormatKnit1()
2652
def is_compatible(source, target):
2653
"""Be compatible with known Knit formats.
2655
We don't test for the stores being of specific types because that
2656
could lead to confusing results, and there is no need to be
2659
from bzrlib.repofmt.knitrepo import RepositoryFormatKnit
2661
are_knits = (isinstance(source._format, RepositoryFormatKnit) and
2662
isinstance(target._format, RepositoryFormatKnit))
2663
except AttributeError:
2665
return are_knits and InterRepository._same_model(source, target)
2668
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2669
"""See InterRepository.fetch()."""
2670
from bzrlib.fetch import KnitRepoFetcher
2671
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2672
self.source, self.source._format, self.target, self.target._format)
2673
f = KnitRepoFetcher(to_repository=self.target,
2674
from_repository=self.source,
2675
last_revision=revision_id,
2676
pb=pb, find_ghosts=find_ghosts)
2677
return f.count_copied, f.failed_revisions
2680
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2681
"""See InterRepository.missing_revision_ids()."""
2682
if revision_id is not None:
2683
source_ids = self.source.get_ancestry(revision_id)
2684
assert source_ids[0] is None
2687
source_ids = self.source.all_revision_ids()
2688
source_ids_set = set(source_ids)
2689
# source_ids is the worst possible case we may need to pull.
2690
# now we want to filter source_ids against what we actually
2691
# have in target, but don't try to check for existence where we know
2692
# we do not have a revision as that would be pointless.
2693
target_ids = set(self.target.all_revision_ids())
2694
possibly_present_revisions = target_ids.intersection(source_ids_set)
2695
actually_present_revisions = set(
2696
self.target._eliminate_revisions_not_present(possibly_present_revisions))
2697
required_revisions = source_ids_set.difference(actually_present_revisions)
2698
if revision_id is not None:
2699
# we used get_ancestry to determine source_ids then we are assured all
2700
# revisions referenced are present as they are installed in topological order.
2701
# and the tip revision was validated by get_ancestry.
2702
result_set = required_revisions
2704
# if we just grabbed the possibly available ids, then
2705
# we only have an estimate of whats available and need to validate
2706
# that against the revision records.
2708
self.source._eliminate_revisions_not_present(required_revisions))
2709
return self.source.revision_ids_to_search_result(result_set)
2712
class InterPackRepo(InterSameDataRepository):
2713
"""Optimised code paths between Pack based repositories."""
2716
def _get_repo_format_to_test(self):
2717
from bzrlib.repofmt import pack_repo
2718
return pack_repo.RepositoryFormatKnitPack1()
2721
def is_compatible(source, target):
2722
"""Be compatible with known Pack formats.
2724
We don't test for the stores being of specific types because that
2725
could lead to confusing results, and there is no need to be
2728
from bzrlib.repofmt.pack_repo import RepositoryFormatPack
2730
are_packs = (isinstance(source._format, RepositoryFormatPack) and
2731
isinstance(target._format, RepositoryFormatPack))
2732
except AttributeError:
2734
return are_packs and InterRepository._same_model(source, target)
2737
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2738
"""See InterRepository.fetch()."""
2739
from bzrlib.repofmt.pack_repo import Packer
2740
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2741
self.source, self.source._format, self.target, self.target._format)
2742
self.count_copied = 0
2743
if revision_id is None:
2745
# everything to do - use pack logic
2746
# to fetch from all packs to one without
2747
# inventory parsing etc, IFF nothing to be copied is in the target.
2749
revision_ids = self.source.all_revision_ids()
2750
revision_keys = [(revid,) for revid in revision_ids]
2751
index = self.target._pack_collection.revision_index.combined_index
2752
present_revision_ids = set(item[1][0] for item in
2753
index.iter_entries(revision_keys))
2754
revision_ids = set(revision_ids) - present_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
if len(revision_ids) == 0:
2773
packs = self.source._pack_collection.all_packs()
2774
pack = Packer(self.target._pack_collection, packs, '.fetch',
2775
revision_ids).pack()
2776
if pack is not None:
2777
self.target._pack_collection._save_pack_names()
2778
# Trigger an autopack. This may duplicate effort as we've just done
2779
# a pack creation, but for now it is simpler to think about as
2780
# 'upload data, then repack if needed'.
2781
self.target._pack_collection.autopack()
2782
return (pack.get_revision_count(), [])
2787
def search_missing_revision_ids(self, revision_id=None, find_ghosts=True):
2788
"""See InterRepository.missing_revision_ids().
2790
:param find_ghosts: Find ghosts throughout the ancestry of
2793
if not find_ghosts and revision_id is not None:
2794
return self._walk_to_common_revisions([revision_id])
2795
elif revision_id is not None:
2796
source_ids = self.source.get_ancestry(revision_id)
2797
assert source_ids[0] is None
2800
source_ids = self.source.all_revision_ids()
2801
# source_ids is the worst possible case we may need to pull.
2802
# now we want to filter source_ids against what we actually
2803
# have in target, but don't try to check for existence where we know
2804
# we do not have a revision as that would be pointless.
2805
target_ids = set(self.target.all_revision_ids())
2806
result_set = set(source_ids).difference(target_ids)
2807
return self.source.revision_ids_to_search_result(result_set)
2810
class InterModel1and2(InterRepository):
2813
def _get_repo_format_to_test(self):
2817
def is_compatible(source, target):
2818
if not source.supports_rich_root() and target.supports_rich_root():
2824
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2825
"""See InterRepository.fetch()."""
2826
from bzrlib.fetch import Model1toKnit2Fetcher
2827
f = Model1toKnit2Fetcher(to_repository=self.target,
2828
from_repository=self.source,
2829
last_revision=revision_id,
2830
pb=pb, find_ghosts=find_ghosts)
2831
return f.count_copied, f.failed_revisions
2834
def copy_content(self, revision_id=None):
2835
"""Make a complete copy of the content in self into destination.
2837
This is a destructive operation! Do not use it on existing
2840
:param revision_id: Only copy the content needed to construct
2841
revision_id and its parents.
2844
self.target.set_make_working_trees(self.source.make_working_trees())
2845
except NotImplementedError:
2847
# but don't bother fetching if we have the needed data now.
2848
if (revision_id not in (None, _mod_revision.NULL_REVISION) and
2849
self.target.has_revision(revision_id)):
2851
self.target.fetch(self.source, revision_id=revision_id)
2854
class InterKnit1and2(InterKnitRepo):
2857
def _get_repo_format_to_test(self):
2861
def is_compatible(source, target):
2862
"""Be compatible with Knit1 source and Knit3 target"""
2863
from bzrlib.repofmt.knitrepo import RepositoryFormatKnit3
2865
from bzrlib.repofmt.knitrepo import (RepositoryFormatKnit1,
2866
RepositoryFormatKnit3)
2867
from bzrlib.repofmt.pack_repo import (
2868
RepositoryFormatKnitPack1,
2869
RepositoryFormatKnitPack3,
2870
RepositoryFormatPackDevelopment0,
2871
RepositoryFormatPackDevelopment0Subtree,
2874
RepositoryFormatKnit1,
2875
RepositoryFormatKnitPack1,
2876
RepositoryFormatPackDevelopment0,
2879
RepositoryFormatKnit3,
2880
RepositoryFormatKnitPack3,
2881
RepositoryFormatPackDevelopment0Subtree,
2883
return (isinstance(source._format, nosubtrees) and
2884
isinstance(target._format, subtrees))
2885
except AttributeError:
2889
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2890
"""See InterRepository.fetch()."""
2891
from bzrlib.fetch import Knit1to2Fetcher
2892
mutter("Using fetch logic to copy between %s(%s) and %s(%s)",
2893
self.source, self.source._format, self.target,
2894
self.target._format)
2895
f = Knit1to2Fetcher(to_repository=self.target,
2896
from_repository=self.source,
2897
last_revision=revision_id,
2898
pb=pb, find_ghosts=find_ghosts)
2899
return f.count_copied, f.failed_revisions
2902
class InterDifferingSerializer(InterKnitRepo):
2905
def _get_repo_format_to_test(self):
2909
def is_compatible(source, target):
2910
"""Be compatible with Knit2 source and Knit3 target"""
2911
if source.supports_rich_root() != target.supports_rich_root():
2913
# Ideally, we'd support fetching if the source had no tree references
2914
# even if it supported them...
2915
if (getattr(source, '_format.supports_tree_reference', False) and
2916
not getattr(target, '_format.supports_tree_reference', False)):
2921
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2922
"""See InterRepository.fetch()."""
2923
revision_ids = self.target.search_missing_revision_ids(self.source,
2924
revision_id, find_ghosts=find_ghosts).get_keys()
2925
revision_ids = tsort.topo_sort(
2926
self.source.get_graph().get_parent_map(revision_ids))
2927
def revisions_iterator():
2928
for current_revision_id in revision_ids:
2929
revision = self.source.get_revision(current_revision_id)
2930
tree = self.source.revision_tree(current_revision_id)
2932
signature = self.source.get_signature_text(
2933
current_revision_id)
2934
except errors.NoSuchRevision:
2936
yield revision, tree, signature
2938
my_pb = ui.ui_factory.nested_progress_bar()
2943
install_revisions(self.target, revisions_iterator(),
2944
len(revision_ids), pb)
2946
if my_pb is not None:
2948
return len(revision_ids), 0
2951
class InterRemoteToOther(InterRepository):
2953
def __init__(self, source, target):
2954
InterRepository.__init__(self, source, target)
2955
self._real_inter = None
2958
def is_compatible(source, target):
2959
if not isinstance(source, remote.RemoteRepository):
2961
# Is source's model compatible with target's model?
2962
source._ensure_real()
2963
real_source = source._real_repository
2964
assert not isinstance(real_source, remote.RemoteRepository), (
2965
"We don't support remote repos backed by remote repos yet.")
2966
return InterRepository._same_model(real_source, target)
2969
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
2970
"""See InterRepository.fetch()."""
2971
from bzrlib.fetch import RemoteToOtherFetcher
2972
mutter("Using fetch logic to copy between %s(remote) and %s(%s)",
2973
self.source, self.target, self.target._format)
2974
# TODO: jam 20070210 This should be an assert, not a translate
2975
revision_id = osutils.safe_revision_id(revision_id)
2976
f = RemoteToOtherFetcher(to_repository=self.target,
2977
from_repository=self.source,
2978
last_revision=revision_id,
2979
pb=pb, find_ghosts=find_ghosts)
2980
return f.count_copied, f.failed_revisions
2983
def _get_repo_format_to_test(self):
2987
class InterOtherToRemote(InterRepository):
2989
def __init__(self, source, target):
2990
InterRepository.__init__(self, source, target)
2991
self._real_inter = None
2994
def is_compatible(source, target):
2995
if isinstance(target, remote.RemoteRepository):
2999
def _ensure_real_inter(self):
3000
if self._real_inter is None:
3001
self.target._ensure_real()
3002
real_target = self.target._real_repository
3003
self._real_inter = InterRepository.get(self.source, real_target)
3005
def copy_content(self, revision_id=None):
3006
self._ensure_real_inter()
3007
self._real_inter.copy_content(revision_id=revision_id)
3009
def fetch(self, revision_id=None, pb=None, find_ghosts=False):
3010
self._ensure_real_inter()
3011
self._real_inter.fetch(revision_id=revision_id, pb=pb,
3012
find_ghosts=find_ghosts)
3015
def _get_repo_format_to_test(self):
3019
InterRepository.register_optimiser(InterDifferingSerializer)
3020
InterRepository.register_optimiser(InterSameDataRepository)
3021
InterRepository.register_optimiser(InterWeaveRepo)
3022
InterRepository.register_optimiser(InterKnitRepo)
3023
InterRepository.register_optimiser(InterModel1and2)
3024
InterRepository.register_optimiser(InterKnit1and2)
3025
InterRepository.register_optimiser(InterPackRepo)
3026
InterRepository.register_optimiser(InterRemoteToOther)
3027
InterRepository.register_optimiser(InterOtherToRemote)
3030
class CopyConverter(object):
3031
"""A repository conversion tool which just performs a copy of the content.
3033
This is slow but quite reliable.
3036
def __init__(self, target_format):
3037
"""Create a CopyConverter.
3039
:param target_format: The format the resulting repository should be.
3041
self.target_format = target_format
3043
def convert(self, repo, pb):
3044
"""Perform the conversion of to_convert, giving feedback via pb.
3046
:param to_convert: The disk object to convert.
3047
:param pb: a progress bar to use for progress information.
3052
# this is only useful with metadir layouts - separated repo content.
3053
# trigger an assertion if not such
3054
repo._format.get_format_string()
3055
self.repo_dir = repo.bzrdir
3056
self.step('Moving repository to repository.backup')
3057
self.repo_dir.transport.move('repository', 'repository.backup')
3058
backup_transport = self.repo_dir.transport.clone('repository.backup')
3059
repo._format.check_conversion_target(self.target_format)
3060
self.source_repo = repo._format.open(self.repo_dir,
3062
_override_transport=backup_transport)
3063
self.step('Creating new repository')
3064
converted = self.target_format.initialize(self.repo_dir,
3065
self.source_repo.is_shared())
3066
converted.lock_write()
3068
self.step('Copying content into repository.')
3069
self.source_repo.copy_content_into(converted)
3072
self.step('Deleting old repository content.')
3073
self.repo_dir.transport.delete_tree('repository.backup')
3074
self.pb.note('repository converted')
3076
def step(self, message):
3077
"""Update the pb by a step."""
3079
self.pb.update(message, self.count, self.total)
3091
def _unescaper(match, _map=_unescape_map):
3092
code = match.group(1)
3096
if not code.startswith('#'):
3098
return unichr(int(code[1:])).encode('utf8')
3104
def _unescape_xml(data):
3105
"""Unescape predefined XML entities in a string of data."""
3107
if _unescape_re is None:
3108
_unescape_re = re.compile('\&([^;]*);')
3109
return _unescape_re.sub(_unescaper, data)
3112
class _VersionedFileChecker(object):
3114
def __init__(self, repository):
3115
self.repository = repository
3116
self.text_index = self.repository._generate_text_key_index()
3118
def calculate_file_version_parents(self, revision_id, file_id):
3119
"""Calculate the correct parents for a file version according to
3122
parent_keys = self.text_index[(file_id, revision_id)]
3123
if parent_keys == [_mod_revision.NULL_REVISION]:
3125
# strip the file_id, for the weave api
3126
return tuple([revision_id for file_id, revision_id in parent_keys])
3128
def check_file_version_parents(self, weave, file_id):
3129
"""Check the parents stored in a versioned file are correct.
3131
It also detects file versions that are not referenced by their
3132
corresponding revision's inventory.
3134
:returns: A tuple of (wrong_parents, dangling_file_versions).
3135
wrong_parents is a dict mapping {revision_id: (stored_parents,
3136
correct_parents)} for each revision_id where the stored parents
3137
are not correct. dangling_file_versions is a set of (file_id,
3138
revision_id) tuples for versions that are present in this versioned
3139
file, but not used by the corresponding inventory.
3142
unused_versions = set()
3143
versions = weave.versions()
3144
parent_map = weave.get_parent_map(versions)
3145
for num, revision_id in enumerate(versions):
3147
correct_parents = self.calculate_file_version_parents(
3148
revision_id, file_id)
3150
# The version is not part of the used keys.
3151
unused_versions.add(revision_id)
3154
knit_parents = tuple(parent_map[revision_id])
3155
except errors.RevisionNotPresent:
3157
if correct_parents != knit_parents:
3158
wrong_parents[revision_id] = (knit_parents, correct_parents)
3159
return wrong_parents, unused_versions
3162
def _old_get_graph(repository, revision_id):
3163
"""DO NOT USE. That is all. I'm serious."""
3164
graph = repository.get_graph()
3165
revision_graph = dict(((key, value) for key, value in
3166
graph.iter_ancestry([revision_id]) if value is not None))
3167
return _strip_NULL_ghosts(revision_graph)
3170
def _strip_NULL_ghosts(revision_graph):
3171
"""Also don't use this. more compatibility code for unmigrated clients."""
3172
# Filter ghosts, and null:
3173
if _mod_revision.NULL_REVISION in revision_graph:
3174
del revision_graph[_mod_revision.NULL_REVISION]
3175
for key, parents in revision_graph.items():
3176
revision_graph[key] = tuple(parent for parent in parents if parent
3178
return revision_graph
added
removed
bzrlib/repository.py
