1
# Copyright (C) 2005-2011 Canonical Ltd
1
# Copyright (C) 2005 Canonical Ltd
3
3
# This program is free software; you can redistribute it and/or modify
4
4
# it under the terms of the GNU General Public License as published by
5
5
# the Free Software Foundation; either version 2 of the License, or
6
6
# (at your option) any later version.
8
8
# This program is distributed in the hope that it will be useful,
9
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
11
# GNU General Public License for more details.
13
13
# You should have received a copy of the GNU General Public License
14
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19
from .lazy_import import lazy_import
20
lazy_import(globals(), """
24
branch as _mod_branch,
25
conflicts as _mod_conflicts,
30
revision as _mod_revision,
38
from breezy.bzr import (
42
from breezy.i18n import gettext
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
24
import bzrlib.revision
25
from bzrlib.merge_core import merge_flex, ApplyMerge3, BackupBeforeChange
26
from bzrlib.merge_core import WeaveMerge
27
from bzrlib.changeset import generate_changeset, ExceptionConflictHandler
28
from bzrlib.changeset import Inventory, Diff3Merge, ReplaceContents
29
from bzrlib.branch import Branch
30
from bzrlib.errors import (BzrCommandError,
34
WorkingTreeNotRevision,
38
from bzrlib.delta import compare_trees
39
from bzrlib.trace import mutter, warning, note
40
from bzrlib.fetch import greedy_fetch, fetch
41
from bzrlib.revision import is_ancestor, NULL_REVISION
42
from bzrlib.osutils import rename
43
from bzrlib.revision import common_ancestor, MultipleRevisionSources
44
from bzrlib.errors import NoSuchRevision
51
46
# TODO: Report back as changes are merged in
54
def transform_tree(from_tree, to_tree, interesting_files=None):
55
with from_tree.lock_tree_write():
56
merge_inner(from_tree.branch, to_tree, from_tree,
57
ignore_zero=True, this_tree=from_tree,
58
interesting_files=interesting_files)
61
class MergeHooks(hooks.Hooks):
64
hooks.Hooks.__init__(self, "breezy.merge", "Merger.hooks")
65
self.add_hook('merge_file_content',
66
"Called with a breezy.merge.Merger object to create a per file "
67
"merge object when starting a merge. "
68
"Should return either None or a subclass of "
69
"``breezy.merge.AbstractPerFileMerger``. "
70
"Such objects will then be called per file "
71
"that needs to be merged (including when one "
72
"side has deleted the file and the other has changed it). "
73
"See the AbstractPerFileMerger API docs for details on how it is "
76
self.add_hook('pre_merge',
77
'Called before a merge. '
78
'Receives a Merger object as the single argument.',
80
self.add_hook('post_merge',
81
'Called after a merge. '
82
'Receives a Merger object as the single argument. '
83
'The return value is ignored.',
87
class AbstractPerFileMerger(object):
88
"""PerFileMerger objects are used by plugins extending merge for breezy.
90
See ``breezy.plugins.news_merge.news_merge`` for an example concrete class.
92
:ivar merger: The Merge3Merger performing the merge.
95
def __init__(self, merger):
96
"""Create a PerFileMerger for use with merger."""
99
def merge_contents(self, merge_params):
100
"""Attempt to merge the contents of a single file.
102
:param merge_params: A breezy.merge.MergeFileHookParams
103
:return: A tuple of (status, chunks), where status is one of
104
'not_applicable', 'success', 'conflicted', or 'delete'. If status
105
is 'success' or 'conflicted', then chunks should be an iterable of
106
strings for the new file contents.
108
return ('not applicable', None)
111
class PerFileMerger(AbstractPerFileMerger):
112
"""Merge individual files when self.file_matches returns True.
114
This class is intended to be subclassed. The file_matches and
115
merge_matching methods should be overridden with concrete implementations.
118
def file_matches(self, params):
119
"""Return True if merge_matching should be called on this file.
121
Only called with merges of plain files with no clear winner.
123
Subclasses must override this.
125
raise NotImplementedError(self.file_matches)
127
def merge_contents(self, params):
128
"""Merge the contents of a single file."""
129
# Check whether this custom merge logic should be used.
131
# OTHER is a straight winner, rely on default merge.
132
params.winner == 'other' or
133
# THIS and OTHER aren't both files.
134
not params.is_file_merge() or
135
# The filename doesn't match
136
not self.file_matches(params)):
137
return 'not_applicable', None
138
return self.merge_matching(params)
140
def merge_matching(self, params):
141
"""Merge the contents of a single file that has matched the criteria
142
in PerFileMerger.merge_contents (is a conflict, is a file,
143
self.file_matches is True).
145
Subclasses must override this.
147
raise NotImplementedError(self.merge_matching)
150
class ConfigurableFileMerger(PerFileMerger):
151
"""Merge individual files when configured via a .conf file.
153
This is a base class for concrete custom file merging logic. Concrete
154
classes should implement ``merge_text``.
156
See ``breezy.plugins.news_merge.news_merge`` for an example concrete class.
158
:ivar affected_files: The configured file paths to merge.
160
:cvar name_prefix: The prefix to use when looking up configuration
161
details. <name_prefix>_merge_files describes the files targeted by the
164
:cvar default_files: The default file paths to merge when no configuration
171
def __init__(self, merger):
172
super(ConfigurableFileMerger, self).__init__(merger)
173
self.affected_files = None
174
self.default_files = self.__class__.default_files or []
175
self.name_prefix = self.__class__.name_prefix
176
if self.name_prefix is None:
177
raise ValueError("name_prefix must be set.")
179
def file_matches(self, params):
180
"""Check whether the file should call the merge hook.
182
<name_prefix>_merge_files configuration variable is a list of files
183
that should use the hook.
185
affected_files = self.affected_files
186
if affected_files is None:
187
config = self.merger.this_branch.get_config()
188
# Until bzr provides a better policy for caching the config, we
189
# just add the part we're interested in to the params to avoid
190
# reading the config files repeatedly (breezy.conf, location.conf,
192
config_key = self.name_prefix + '_merge_files'
193
affected_files = config.get_user_option_as_list(config_key)
194
if affected_files is None:
195
# If nothing was specified in the config, use the default.
196
affected_files = self.default_files
197
self.affected_files = affected_files
199
filepath = params.this_path
200
if filepath in affected_files:
204
def merge_matching(self, params):
205
return self.merge_text(params)
207
def merge_text(self, params):
208
"""Merge the byte contents of a single file.
210
This is called after checking that the merge should be performed in
211
merge_contents, and it should behave as per
212
``breezy.merge.AbstractPerFileMerger.merge_contents``.
214
raise NotImplementedError(self.merge_text)
217
class MergeFileHookParams(object):
218
"""Object holding parameters passed to merge_file_content hooks.
220
There are some fields hooks can access:
222
:ivar base_path: Path in base tree
223
:ivar other_path: Path in other tree
224
:ivar this_path: Path in this tree
225
:ivar trans_id: the transform ID for the merge of this file
226
:ivar this_kind: kind of file in 'this' tree
227
:ivar other_kind: kind of file in 'other' tree
228
:ivar winner: one of 'this', 'other', 'conflict'
231
def __init__(self, merger, paths, trans_id, this_kind, other_kind,
233
self._merger = merger
235
self.base_path, self.other_path, self.this_path = paths
236
self.trans_id = trans_id
237
self.this_kind = this_kind
238
self.other_kind = other_kind
241
def is_file_merge(self):
242
"""True if this_kind and other_kind are both 'file'."""
243
return self.this_kind == 'file' and self.other_kind == 'file'
245
@decorators.cachedproperty
246
def base_lines(self):
247
"""The lines of the 'base' version of the file."""
248
return self._merger.get_lines(self._merger.base_tree, self.base_path)
250
@decorators.cachedproperty
251
def this_lines(self):
252
"""The lines of the 'this' version of the file."""
253
return self._merger.get_lines(self._merger.this_tree, self.this_path)
255
@decorators.cachedproperty
256
def other_lines(self):
257
"""The lines of the 'other' version of the file."""
258
return self._merger.get_lines(self._merger.other_tree, self.other_path)
48
# TODO: build_working_dir can be built on something simpler than merge()
50
# FIXME: merge() parameters seem oriented towards the command line
51
# NOTABUG: merge is a helper for commandline functions. merge_inner is the
52
# the core functionality.
54
# comments from abentley on irc: merge happens in two stages, each
55
# of which generates a changeset object
57
# stage 1: generate OLD->OTHER,
58
# stage 2: use MINE and OLD->OTHER to generate MINE -> RESULT
60
class MergeConflictHandler(ExceptionConflictHandler):
61
"""Handle conflicts encountered while merging.
63
This subclasses ExceptionConflictHandler, so that any types of
64
conflict that are not explicitly handled cause an exception and
67
def __init__(self, this_tree, base_tree, other_tree, ignore_zero=False):
68
ExceptionConflictHandler.__init__(self)
70
self.ignore_zero = ignore_zero
71
self.this_tree = this_tree
72
self.base_tree = base_tree
73
self.other_tree = other_tree
75
def copy(self, source, dest):
76
"""Copy the text and mode of a file
77
:param source: The path of the file to copy
78
:param dest: The distination file to create
80
s_file = file(source, "rb")
81
d_file = file(dest, "wb")
84
os.chmod(dest, 0777 & os.stat(source).st_mode)
86
def dump(self, lines, dest):
87
"""Copy the text and mode of a file
88
:param source: The path of the file to copy
89
:param dest: The distination file to create
91
d_file = file(dest, "wb")
95
def add_suffix(self, name, suffix, last_new_name=None, fix_inventory=True):
96
"""Rename a file to append a suffix. If the new name exists, the
97
suffix is added repeatedly until a non-existant name is found
99
:param name: The path of the file
100
:param suffix: The suffix to append
101
:param last_new_name: (used for recursive calls) the last name tried
103
if last_new_name is None:
105
new_name = last_new_name+suffix
107
rename(name, new_name)
108
if fix_inventory is True:
110
relpath = self.this_tree.relpath(name)
111
except NotBranchError:
113
if relpath is not None:
114
file_id = self.this_tree.path2id(relpath)
115
if file_id is not None:
116
new_path = self.this_tree.relpath(new_name)
117
rename(new_name, name)
118
self.this_tree.rename_one(relpath, new_path)
119
assert self.this_tree.id2path(file_id) == new_path
121
if e.errno != errno.EEXIST and e.errno != errno.ENOTEMPTY:
123
return self.add_suffix(name, suffix, last_new_name=new_name,
124
fix_inventory=fix_inventory)
127
def conflict(self, text):
132
def merge_conflict(self, new_file, this_path, base_lines, other_lines):
134
Handle diff3 conflicts by producing a .THIS, .BASE and .OTHER. The
135
main file will be a version with diff3 conflicts.
136
:param new_file: Path to the output file with diff3 markers
137
:param this_path: Path to the file text for the THIS tree
138
:param base_path: Path to the file text for the BASE tree
139
:param other_path: Path to the file text for the OTHER tree
141
self.add_suffix(this_path, ".THIS", fix_inventory=False)
142
self.dump(base_lines, this_path+".BASE")
143
self.dump(other_lines, this_path+".OTHER")
144
rename(new_file, this_path)
145
self.conflict("Diff3 conflict encountered in %s" % this_path)
147
def weave_merge_conflict(self, filename, weave, other_i, out_file):
149
Handle weave conflicts by producing a .THIS, and .OTHER. The
150
main file will be a version with diff3-style conflicts.
152
self.add_suffix(filename, ".THIS", fix_inventory=False)
154
self.dump(weave.get_iter(other_i), filename+".OTHER")
155
self.conflict("Text conflict encountered in %s" % filename)
157
def new_contents_conflict(self, filename, other_contents):
158
"""Conflicting contents for newly added file."""
159
other_contents(filename + ".OTHER", self, False)
160
self.conflict("Conflict in newly added file %s" % filename)
163
def target_exists(self, entry, target, old_path):
164
"""Handle the case when the target file or dir exists"""
165
moved_path = self.add_suffix(target, ".moved")
166
self.conflict("Moved existing %s to %s" % (target, moved_path))
168
def rmdir_non_empty(self, filename):
169
"""Handle the case where the dir to be removed still has contents"""
170
self.conflict("Directory %s not removed because it is not empty"\
174
def rem_contents_conflict(self, filename, this_contents, base_contents):
175
base_contents(filename+".BASE", self, False)
176
this_contents(filename+".THIS", self, False)
177
return ReplaceContents(this_contents, None)
179
def rem_contents_conflict(self, filename, this_contents, base_contents):
180
base_contents(filename+".BASE", self, False)
181
this_contents(filename+".THIS", self, False)
182
self.conflict("Other branch deleted locally modified file %s" %
184
return ReplaceContents(this_contents, None)
186
def abs_this_path(self, file_id):
187
"""Return the absolute path for a file_id in the this tree."""
188
return self.this_tree.id2abspath(file_id)
190
def add_missing_parents(self, file_id, tree):
191
"""If some of the parents for file_id are missing, add them."""
192
entry = tree.inventory[file_id]
193
if entry.parent_id not in self.this_tree:
194
return self.create_all_missing(entry.parent_id, tree)
196
return self.abs_this_path(entry.parent_id)
198
def create_all_missing(self, file_id, tree):
199
"""Add contents for a file_id and all its parents to a tree."""
200
entry = tree.inventory[file_id]
201
if entry.parent_id is not None and entry.parent_id not in self.this_tree:
202
abspath = self.create_all_missing(entry.parent_id, tree)
204
abspath = self.abs_this_path(entry.parent_id)
205
entry_path = os.path.join(abspath, entry.name)
206
if not os.path.isdir(entry_path):
207
self.create(file_id, entry_path, tree)
210
def create(self, file_id, path, tree, reverse=False):
211
"""Uses tree data to create a filesystem object for the file_id"""
212
from changeset import get_contents
213
get_contents(tree, file_id)(path, self, reverse)
215
def missing_for_merge(self, file_id, other_path):
216
"""The file_id doesn't exist in THIS, but does in OTHER and BASE"""
217
self.conflict("Other branch modified locally deleted file %s" %
219
parent_dir = self.add_missing_parents(file_id, self.other_tree)
220
stem = os.path.join(parent_dir, os.path.basename(other_path))
221
self.create(file_id, stem+".OTHER", self.other_tree)
222
self.create(file_id, stem+".BASE", self.base_tree)
224
def threeway_contents_conflict(filename, this_contents, base_contents,
226
self.conflict("Three-way conflict merging %s" % filename)
229
if self.conflicts == 0:
230
if not self.ignore_zero:
231
note("All changes applied successfully.")
233
note("%d conflicts encountered." % self.conflicts)
235
def get_tree(treespec, local_branch=None):
236
location, revno = treespec
237
branch = Branch.open_containing(location)[0]
241
revision = branch.last_revision()
243
revision = branch.get_rev_id(revno)
245
revision = NULL_REVISION
246
return branch, get_revid_tree(branch, revision, local_branch)
248
def get_revid_tree(branch, revision, local_branch):
250
base_tree = branch.working_tree()
252
if local_branch is not None:
253
greedy_fetch(local_branch, branch, revision)
254
base_tree = local_branch.revision_tree(revision)
256
base_tree = branch.revision_tree(revision)
260
def file_exists(tree, file_id):
261
return tree.has_filename(tree.id2path(file_id))
264
def build_working_dir(to_dir):
265
"""Build a working directory in an empty directory.
267
to_dir is a directory containing branch metadata but no working files,
268
typically constructed by cloning an existing branch.
270
This is split out as a special idiomatic case of merge. It could
271
eventually be done by just building the tree directly calling into
272
lower-level code (e.g. constructing a changeset).
274
# RBC 20051019 is this not just 'export' ?
275
# AB Well, export doesn't take care of inventory...
276
this_branch = Branch.open_containing(to_dir)[0]
277
transform_tree(this_branch.working_tree(), this_branch.basis_tree())
280
def transform_tree(from_tree, to_tree, interesting_ids=None):
281
merge_inner(from_tree.branch, to_tree, from_tree, ignore_zero=True,
282
interesting_ids=interesting_ids)
285
def merge(other_revision, base_revision,
286
check_clean=True, ignore_zero=False,
287
this_dir=None, backup_files=False, merge_type=ApplyMerge3,
288
file_list=None, show_base=False, reprocess=False):
289
"""Merge changes into a tree.
292
list(path, revno) Base for three-way merge.
293
If [None, None] then a base will be automatically determined.
295
list(path, revno) Other revision for three-way merge.
297
Directory to merge changes into; '.' by default.
299
If true, this_dir must have no uncommitted changes before the
301
ignore_zero - If true, suppress the "zero conflicts" message when
302
there are no conflicts; should be set when doing something we expect
303
to complete perfectly.
304
file_list - If supplied, merge only changes to selected files.
306
All available ancestors of other_revision and base_revision are
307
automatically pulled into the branch.
309
The revno may be -1 to indicate the last revision on the branch, which is
312
This function is intended for use from the command line; programmatic
313
clients might prefer to call merge_inner(), which has less magic behavior.
317
this_branch = Branch.open_containing(this_dir)[0]
318
if show_base and not merge_type is ApplyMerge3:
319
raise BzrCommandError("Show-base is not supported for this merge"
320
" type. %s" % merge_type)
321
if reprocess and not merge_type is ApplyMerge3:
322
raise BzrCommandError("Reprocess is not supported for this merge"
323
" type. %s" % merge_type)
324
if reprocess and show_base:
325
raise BzrCommandError("Cannot reprocess and show base.")
326
merger = Merger(this_branch)
327
merger.check_basis(check_clean)
328
merger.set_other(other_revision)
329
merger.set_base(base_revision)
330
if merger.base_rev_id == merger.other_rev_id:
331
note('Nothing to do.')
333
merger.backup_files = backup_files
334
merger.merge_type = merge_type
335
merger.set_interesting_files(file_list)
336
merger.show_base = show_base
337
merger.reprocess = reprocess
338
merger.conflict_handler = MergeConflictHandler(merger.this_tree,
341
ignore_zero=ignore_zero)
342
conflicts = merger.do_merge()
346
def merge_inner(this_branch, other_tree, base_tree, ignore_zero=False,
348
merge_type=ApplyMerge3,
349
interesting_ids=None,
353
interesting_files=None):
354
"""Primary interface for merging.
356
typical use is probably
357
'merge_inner(branch, branch.get_revision_tree(other_revision),
358
branch.get_revision_tree(base_revision))'
360
merger = Merger(this_branch, other_tree, base_tree)
361
merger.backup_files = backup_files
362
merger.merge_type = merge_type
363
merger.interesting_ids = interesting_ids
364
if interesting_files:
365
assert not interesting_ids, ('Only supply interesting_ids'
366
' or interesting_files')
367
merger._set_interesting_files(interesting_files)
368
merger.show_base = show_base
369
merger.reprocess = reprocess
370
merger.conflict_handler = MergeConflictHandler(merger.this_tree, base_tree,
372
ignore_zero=ignore_zero)
373
merger.other_rev_id = other_rev_id
374
merger.other_basis = other_rev_id
375
return merger.do_merge()
261
378
class Merger(object):
265
def __init__(self, this_branch, other_tree=None, base_tree=None,
266
this_tree=None, change_reporter=None,
267
recurse='down', revision_graph=None):
379
def __init__(self, this_branch, other_tree=None, base_tree=None):
268
380
object.__init__(self)
269
381
self.this_branch = this_branch
270
self.this_basis = _mod_revision.ensure_null(
271
this_branch.last_revision())
382
self.this_basis = this_branch.last_revision()
272
383
self.this_rev_id = None
273
self.this_tree = this_tree
384
self.this_tree = this_branch.working_tree()
274
385
self.this_revision_tree = None
275
386
self.this_basis_tree = None
276
387
self.other_tree = other_tree
277
self.other_branch = None
278
388
self.base_tree = base_tree
279
389
self.ignore_zero = False
280
390
self.backup_files = False
281
self.interesting_files = None
391
self.interesting_ids = None
282
392
self.show_base = False
283
393
self.reprocess = False
285
self.recurse = recurse
286
self.change_reporter = change_reporter
287
self._cached_trees = {}
288
self._revision_graph = revision_graph
289
self._base_is_ancestor = None
290
self._base_is_other_ancestor = None
291
self._is_criss_cross = None
292
self._lca_trees = None
294
def cache_trees_with_revision_ids(self, trees):
295
"""Cache any tree in trees if it has a revision_id."""
296
for maybe_tree in trees:
297
if maybe_tree is None:
300
rev_id = maybe_tree.get_revision_id()
301
except AttributeError:
303
self._cached_trees[rev_id] = maybe_tree
306
def revision_graph(self):
307
if self._revision_graph is None:
308
self._revision_graph = self.this_branch.repository.get_graph()
309
return self._revision_graph
311
def _set_base_is_ancestor(self, value):
312
self._base_is_ancestor = value
314
def _get_base_is_ancestor(self):
315
if self._base_is_ancestor is None:
316
self._base_is_ancestor = self.revision_graph.is_ancestor(
317
self.base_rev_id, self.this_basis)
318
return self._base_is_ancestor
320
base_is_ancestor = property(_get_base_is_ancestor, _set_base_is_ancestor)
322
def _set_base_is_other_ancestor(self, value):
323
self._base_is_other_ancestor = value
325
def _get_base_is_other_ancestor(self):
326
if self._base_is_other_ancestor is None:
327
if self.other_basis is None:
329
self._base_is_other_ancestor = self.revision_graph.is_ancestor(
330
self.base_rev_id, self.other_basis)
331
return self._base_is_other_ancestor
333
base_is_other_ancestor = property(_get_base_is_other_ancestor,
334
_set_base_is_other_ancestor)
337
def from_uncommitted(tree, other_tree, base_tree=None):
338
"""Return a Merger for uncommitted changes in other_tree.
340
:param tree: The tree to merge into
341
:param other_tree: The tree to get uncommitted changes from
342
:param base_tree: The basis to use for the merge. If unspecified,
343
other_tree.basis_tree() will be used.
345
if base_tree is None:
346
base_tree = other_tree.basis_tree()
347
merger = Merger(tree.branch, other_tree, base_tree, tree)
348
merger.base_rev_id = merger.base_tree.get_revision_id()
349
merger.other_rev_id = None
350
merger.other_basis = merger.base_rev_id
354
def from_mergeable(klass, tree, mergeable):
355
"""Return a Merger for a bundle or merge directive.
357
:param tree: The tree to merge changes into
358
:param mergeable: A merge directive or bundle
360
mergeable.install_revisions(tree.branch.repository)
361
base_revision_id, other_revision_id, verified =\
362
mergeable.get_merge_request(tree.branch.repository)
363
revision_graph = tree.branch.repository.get_graph()
364
if base_revision_id is not None:
365
if (base_revision_id != _mod_revision.NULL_REVISION and
366
revision_graph.is_ancestor(
367
base_revision_id, tree.branch.last_revision())):
368
base_revision_id = None
394
self.conflict_handler = MergeConflictHandler(self.this_tree, base_tree,
397
def revision_tree(self, revision_id):
398
return self.this_branch.revision_tree(revision_id)
400
def ensure_revision_trees(self):
401
if self.this_revision_tree is None:
402
self.this_basis_tree = self.this_branch.revision_tree(
404
if self.this_basis == self.this_rev_id:
405
self.this_revision_tree = self.this_basis_tree
408
if self.other_rev_id is None:
409
other_basis_tree = self.revision_tree(self.other_basis)
410
changes = compare_trees(self.other_tree, other_basis_tree)
411
if changes.has_changed():
412
raise WorkingTreeNotRevision(self.this_tree)
413
other_rev_id = other_basis
414
self.other_tree = other_basis_tree
417
def file_revisions(self, file_id):
418
self.ensure_revision_trees()
419
def get_id(tree, file_id):
420
revision_id = tree.inventory[file_id].revision
421
assert revision_id is not None
423
if self.this_rev_id is None:
424
if self.this_basis_tree.get_file_sha1(file_id) != \
425
self.this_tree.get_file_sha1(file_id):
426
raise WorkingTreeNotRevision(self.this_tree)
428
trees = (self.this_basis_tree, self.other_tree)
429
return [get_id(tree, file_id) for tree in trees]
432
def merge_factory(self, file_id, base, other):
433
if self.merge_type.history_based:
434
if self.show_base is True:
435
raise BzrError("Cannot show base for hisory-based merges")
436
if self.reprocess is True:
437
raise BzrError("Cannot reprocess history-based merges")
439
t_revid, o_revid = self.file_revisions(file_id)
440
weave = self.this_basis_tree.get_weave(file_id)
441
contents_change = self.merge_type(weave, t_revid, o_revid)
443
if self.show_base is True or self.reprocess is True:
444
contents_change = self.merge_type(file_id, base, other,
445
show_base=self.show_base,
446
reprocess=self.reprocess)
370
trace.warning('Performing cherrypick')
371
merger = klass.from_revision_ids(tree, other_revision_id,
372
base_revision_id, revision_graph=revision_graph)
373
return merger, verified
376
def from_revision_ids(tree, other, base=None, other_branch=None,
377
base_branch=None, revision_graph=None,
379
"""Return a Merger for revision-ids.
381
:param tree: The tree to merge changes into
382
:param other: The revision-id to use as OTHER
383
:param base: The revision-id to use as BASE. If not specified, will
385
:param other_branch: A branch containing the other revision-id. If
386
not supplied, tree.branch is used.
387
:param base_branch: A branch containing the base revision-id. If
388
not supplied, other_branch or tree.branch will be used.
389
:param revision_graph: If you have a revision_graph precomputed, pass
390
it in, otherwise it will be created for you.
391
:param tree_branch: The branch associated with tree. If not supplied,
392
tree.branch will be used.
394
if tree_branch is None:
395
tree_branch = tree.branch
396
merger = Merger(tree_branch, this_tree=tree,
397
revision_graph=revision_graph)
398
if other_branch is None:
399
other_branch = tree.branch
400
merger.set_other_revision(other, other_branch)
404
if base_branch is None:
405
base_branch = other_branch
406
merger.set_base_revision(base, base_branch)
409
def revision_tree(self, revision_id, branch=None):
410
if revision_id not in self._cached_trees:
412
branch = self.this_branch
414
tree = self.this_tree.revision_tree(revision_id)
415
except errors.NoSuchRevisionInTree:
416
tree = branch.repository.revision_tree(revision_id)
417
self._cached_trees[revision_id] = tree
418
return self._cached_trees[revision_id]
420
def _get_tree(self, treespec, possible_transports=None):
421
location, revno = treespec
423
tree = workingtree.WorkingTree.open_containing(location)[0]
424
return tree.branch, tree
425
branch = _mod_branch.Branch.open_containing(
426
location, possible_transports)[0]
428
revision_id = branch.last_revision()
430
revision_id = branch.get_rev_id(revno)
431
revision_id = _mod_revision.ensure_null(revision_id)
432
return branch, self.revision_tree(revision_id, branch)
448
contents_change = self.merge_type(file_id, base, other)
449
if self.backup_files:
450
contents_change = BackupBeforeChange(contents_change)
451
return contents_change
453
def check_basis(self, check_clean):
454
if self.this_basis is None:
455
raise BzrCommandError("This branch has no commits")
458
if self.this_basis != self.this_rev_id:
459
raise BzrCommandError("Working tree has uncommitted changes.")
461
def compare_basis(self):
462
changes = compare_trees(self.this_branch.working_tree(),
463
self.this_branch.basis_tree(), False)
464
if not changes.has_changed():
465
self.this_rev_id = self.this_basis
434
467
def set_interesting_files(self, file_list):
435
self.interesting_files = file_list
469
self._set_interesting_files(file_list)
470
except NotVersionedError, e:
471
raise BzrCommandError("%s is not a source file in any"
474
def _set_interesting_files(self, file_list):
475
"""Set the list of interesting ids from a list of files."""
476
if file_list is None:
477
self.interesting_ids = None
480
interesting_ids = set()
481
for fname in file_list:
482
path = self.this_tree.relpath(fname)
484
for tree in (self.this_tree, self.base_tree, self.other_tree):
485
file_id = tree.inventory.path2id(path)
486
if file_id is not None:
487
interesting_ids.add(file_id)
490
raise NotVersionedError(path=fname)
491
self.interesting_ids = interesting_ids
437
493
def set_pending(self):
438
if (not self.base_is_ancestor or not self.base_is_other_ancestor
439
or self.other_rev_id is None):
443
def _add_parent(self):
444
new_parents = self.this_tree.get_parent_ids() + [self.other_rev_id]
445
new_parent_trees = []
446
with contextlib.ExitStack() as stack:
447
for revision_id in new_parents:
449
tree = self.revision_tree(revision_id)
450
except errors.NoSuchRevision:
453
stack.enter_context(tree.lock_read())
454
new_parent_trees.append((revision_id, tree))
455
self.this_tree.set_parent_trees(new_parent_trees, allow_leftmost_as_ghost=True)
457
def set_other(self, other_revision, possible_transports=None):
458
"""Set the revision and tree to merge from.
460
This sets the other_tree, other_rev_id, other_basis attributes.
462
:param other_revision: The [path, revision] list to merge from.
464
self.other_branch, self.other_tree = self._get_tree(other_revision,
494
if not self.base_is_ancestor:
496
if self.other_rev_id is None:
498
if self.other_rev_id in self.this_branch.get_ancestry(self.this_basis):
500
self.this_branch.working_tree().add_pending_merge(self.other_rev_id)
502
def set_other(self, other_revision):
503
other_branch, self.other_tree = get_tree(other_revision,
466
505
if other_revision[1] == -1:
467
self.other_rev_id = _mod_revision.ensure_null(
468
self.other_branch.last_revision())
469
if _mod_revision.is_null(self.other_rev_id):
470
raise errors.NoCommits(self.other_branch)
506
self.other_rev_id = other_branch.last_revision()
507
if self.other_rev_id is None:
508
raise NoCommits(other_branch)
471
509
self.other_basis = self.other_rev_id
472
510
elif other_revision[1] is not None:
473
self.other_rev_id = self.other_branch.get_rev_id(other_revision[1])
511
self.other_rev_id = other_branch.get_rev_id(other_revision[1])
474
512
self.other_basis = self.other_rev_id
476
514
self.other_rev_id = None
477
self.other_basis = self.other_branch.last_revision()
515
self.other_basis = other_branch.last_revision()
478
516
if self.other_basis is None:
479
raise errors.NoCommits(self.other_branch)
480
if self.other_rev_id is not None:
481
self._cached_trees[self.other_rev_id] = self.other_tree
482
self._maybe_fetch(self.other_branch,
483
self.this_branch, self.other_basis)
485
def set_other_revision(self, revision_id, other_branch):
486
"""Set 'other' based on a branch and revision id
488
:param revision_id: The revision to use for a tree
489
:param other_branch: The branch containing this tree
491
self.other_rev_id = revision_id
492
self.other_branch = other_branch
493
self._maybe_fetch(other_branch, self.this_branch, self.other_rev_id)
494
self.other_tree = self.revision_tree(revision_id)
495
self.other_basis = revision_id
497
def set_base_revision(self, revision_id, branch):
498
"""Set 'base' based on a branch and revision id
500
:param revision_id: The revision to use for a tree
501
:param branch: The branch containing this tree
503
self.base_rev_id = revision_id
504
self.base_branch = branch
505
self._maybe_fetch(branch, self.this_branch, revision_id)
506
self.base_tree = self.revision_tree(revision_id)
508
def _maybe_fetch(self, source, target, revision_id):
509
if not source.repository.has_same_location(target.repository):
510
target.fetch(source, revision_id)
513
revisions = [_mod_revision.ensure_null(self.this_basis),
514
_mod_revision.ensure_null(self.other_basis)]
515
if _mod_revision.NULL_REVISION in revisions:
516
self.base_rev_id = _mod_revision.NULL_REVISION
517
self.base_tree = self.revision_tree(self.base_rev_id)
518
self._is_criss_cross = False
520
lcas = self.revision_graph.find_lca(revisions[0], revisions[1])
521
self._is_criss_cross = False
523
self.base_rev_id = _mod_revision.NULL_REVISION
525
self.base_rev_id = list(lcas)[0]
526
else: # len(lcas) > 1
527
self._is_criss_cross = True
529
# find_unique_lca can only handle 2 nodes, so we have to
530
# start back at the beginning. It is a shame to traverse
531
# the graph again, but better than re-implementing
533
self.base_rev_id = self.revision_graph.find_unique_lca(
534
revisions[0], revisions[1])
536
self.base_rev_id = self.revision_graph.find_unique_lca(
538
sorted_lca_keys = self.revision_graph.find_merge_order(
540
if self.base_rev_id == _mod_revision.NULL_REVISION:
541
self.base_rev_id = sorted_lca_keys[0]
543
if self.base_rev_id == _mod_revision.NULL_REVISION:
544
raise errors.UnrelatedBranches()
545
if self._is_criss_cross:
546
trace.warning('Warning: criss-cross merge encountered. See bzr'
547
' help criss-cross.')
548
trace.mutter('Criss-cross lcas: %r' % lcas)
549
if self.base_rev_id in lcas:
550
trace.mutter('Unable to find unique lca. '
551
'Fallback %r as best option.'
553
interesting_revision_ids = set(lcas)
554
interesting_revision_ids.add(self.base_rev_id)
555
interesting_trees = dict((t.get_revision_id(), t)
556
for t in self.this_branch.repository.revision_trees(
557
interesting_revision_ids))
558
self._cached_trees.update(interesting_trees)
559
if self.base_rev_id in lcas:
560
self.base_tree = interesting_trees[self.base_rev_id]
562
self.base_tree = interesting_trees.pop(self.base_rev_id)
563
self._lca_trees = [interesting_trees[key]
564
for key in sorted_lca_keys]
566
self.base_tree = self.revision_tree(self.base_rev_id)
567
self.base_is_ancestor = True
568
self.base_is_other_ancestor = True
569
trace.mutter('Base revid: %r' % self.base_rev_id)
517
raise NoCommits(other_branch)
518
fetch(from_branch=other_branch, to_branch=self.this_branch,
519
last_revision=self.other_basis)
571
521
def set_base(self, base_revision):
572
"""Set the base revision to use for the merge.
574
:param base_revision: A 2-list containing a path and revision number.
576
trace.mutter("doing merge() with no base_revision specified")
522
mutter("doing merge() with no base_revision specified")
577
523
if base_revision == [None, None]:
525
self.base_rev_id = common_ancestor(self.this_basis,
528
except NoCommonAncestor:
529
raise UnrelatedBranches()
530
self.base_tree = get_revid_tree(self.this_branch, self.base_rev_id,
532
self.base_is_ancestor = True
580
base_branch, self.base_tree = self._get_tree(base_revision)
534
base_branch, self.base_tree = get_tree(base_revision)
581
535
if base_revision[1] == -1:
582
536
self.base_rev_id = base_branch.last_revision()
583
537
elif base_revision[1] is None:
584
self.base_rev_id = _mod_revision.NULL_REVISION
586
self.base_rev_id = _mod_revision.ensure_null(
587
base_branch.get_rev_id(base_revision[1]))
588
self._maybe_fetch(base_branch, self.this_branch, self.base_rev_id)
590
def make_merger(self):
591
kwargs = {'working_tree': self.this_tree, 'this_tree': self.this_tree,
592
'other_tree': self.other_tree,
593
'interesting_files': self.interesting_files,
594
'this_branch': self.this_branch,
595
'other_branch': self.other_branch,
597
if self.merge_type.requires_base:
598
kwargs['base_tree'] = self.base_tree
599
if self.merge_type.supports_reprocess:
600
kwargs['reprocess'] = self.reprocess
602
raise errors.BzrError(
603
"Conflict reduction is not supported for merge"
604
" type %s." % self.merge_type)
605
if self.merge_type.supports_show_base:
606
kwargs['show_base'] = self.show_base
608
raise errors.BzrError("Showing base is not supported for this"
609
" merge type. %s" % self.merge_type)
610
if (not getattr(self.merge_type, 'supports_reverse_cherrypick', True)
611
and not self.base_is_other_ancestor):
612
raise errors.CannotReverseCherrypick()
613
if self.merge_type.supports_cherrypick:
614
kwargs['cherrypick'] = (not self.base_is_ancestor or
615
not self.base_is_other_ancestor)
616
if self._is_criss_cross and getattr(self.merge_type,
617
'supports_lca_trees', False):
618
kwargs['lca_trees'] = self._lca_trees
619
return self.merge_type(change_reporter=self.change_reporter,
622
def _do_merge_to(self):
623
merge = self.make_merger()
624
if self.other_branch is not None:
625
self.other_branch.update_references(self.this_branch)
626
for hook in Merger.hooks['pre_merge']:
629
for hook in Merger.hooks['post_merge']:
631
if self.recurse == 'down':
632
for relpath in self.this_tree.iter_references():
633
sub_tree = self.this_tree.get_nested_tree(relpath)
634
other_revision = self.other_tree.get_reference_revision(
636
if other_revision == sub_tree.last_revision():
638
sub_merge = Merger(sub_tree.branch, this_tree=sub_tree)
639
sub_merge.merge_type = self.merge_type
640
other_branch = self.other_tree.reference_parent(relpath)
641
sub_merge.set_other_revision(other_revision, other_branch)
642
base_tree_path = _mod_tree.find_previous_path(
643
self.this_tree, self.base_tree, relpath)
644
base_revision = self.base_tree.get_reference_revision(
646
sub_merge.base_tree = \
647
sub_tree.branch.repository.revision_tree(base_revision)
648
sub_merge.base_rev_id = base_revision
653
with contextlib.ExitStack() as stack:
654
stack.enter_context(self.this_tree.lock_tree_write())
655
if self.base_tree is not None:
656
stack.enter_context(self.base_tree.lock_read())
657
if self.other_tree is not None:
658
stack.enter_context(self.other_tree.lock_read())
659
merge = self._do_merge_to()
660
if len(merge.cooked_conflicts) == 0:
661
if not self.ignore_zero and not trace.is_quiet():
662
trace.note(gettext("All changes applied successfully."))
664
trace.note(gettext("%d conflicts encountered.")
665
% len(merge.cooked_conflicts))
667
return len(merge.cooked_conflicts)
670
class _InventoryNoneEntry(object):
671
"""This represents an inventory entry which *isn't there*.
673
It simplifies the merging logic if we always have an InventoryEntry, even
674
if it isn't actually present
681
symlink_target = None
684
def is_unmodified(self, other):
688
_none_entry = _InventoryNoneEntry()
691
class Merge3Merger(object):
692
"""Three-way merger that uses the merge3 text merger"""
694
supports_reprocess = True
695
supports_show_base = True
696
history_based = False
697
supports_cherrypick = True
698
supports_reverse_cherrypick = True
699
winner_idx = {"this": 2, "other": 1, "conflict": 1}
700
supports_lca_trees = True
701
requires_file_merge_plan = False
703
def __init__(self, working_tree, this_tree, base_tree, other_tree,
704
reprocess=False, show_base=False,
705
change_reporter=None, interesting_files=None, do_merge=True,
706
cherrypick=False, lca_trees=None, this_branch=None,
708
"""Initialize the merger object and perform the merge.
710
:param working_tree: The working tree to apply the merge to
711
:param this_tree: The local tree in the merge operation
712
:param base_tree: The common tree in the merge operation
713
:param other_tree: The other tree to merge changes from
714
:param this_branch: The branch associated with this_tree. Defaults to
715
this_tree.branch if not supplied.
716
:param other_branch: The branch associated with other_tree, if any.
717
:param: reprocess If True, perform conflict-reduction processing.
718
:param show_base: If True, show the base revision in text conflicts.
719
(incompatible with reprocess)
720
:param change_reporter: An object that should report changes made
721
:param interesting_files: The tree-relative paths of files that should
722
participate in the merge. If these paths refer to directories,
723
the contents of those directories will also be included. If not
724
specified, all files may participate in the
726
:param lca_trees: Can be set to a dictionary of {revision_id:rev_tree}
727
if the ancestry was found to include a criss-cross merge.
728
Otherwise should be None.
730
object.__init__(self)
731
if this_branch is None:
732
this_branch = this_tree.branch
733
self.interesting_files = interesting_files
734
self.working_tree = working_tree
735
self.this_tree = this_tree
736
self.base_tree = base_tree
737
self.other_tree = other_tree
738
self.this_branch = this_branch
739
self.other_branch = other_branch
740
self._raw_conflicts = []
741
self.cooked_conflicts = []
742
self.reprocess = reprocess
743
self.show_base = show_base
744
self._lca_trees = lca_trees
745
# Uncommenting this will change the default algorithm to always use
746
# _entries_lca. This can be useful for running the test suite and
747
# making sure we haven't missed any corner cases.
748
# if lca_trees is None:
749
# self._lca_trees = [self.base_tree]
750
self.change_reporter = change_reporter
751
self.cherrypick = cherrypick
756
with contextlib.ExitStack() as stack:
757
stack.enter_context(self.working_tree.lock_tree_write())
758
stack.enter_context(self.this_tree.lock_read())
759
stack.enter_context(self.base_tree.lock_read())
760
stack.enter_context(self.other_tree.lock_read())
761
self.tt = self.working_tree.transform()
762
stack.enter_context(self.tt)
763
self._compute_transform()
764
results = self.tt.apply(no_conflicts=True)
765
self.write_modified(results)
767
self.working_tree.add_conflicts(self.cooked_conflicts)
768
except errors.UnsupportedOperation:
771
def make_preview_transform(self):
772
with self.base_tree.lock_read(), self.other_tree.lock_read():
773
self.tt = self.working_tree.preview_transform()
774
self._compute_transform()
777
def _compute_transform(self):
778
if self._lca_trees is None:
779
entries = list(self._entries3())
780
resolver = self._three_way
782
entries = list(self._entries_lca())
783
resolver = self._lca_multi_way
784
# Prepare merge hooks
785
factories = Merger.hooks['merge_file_content']
786
# One hook for each registered one plus our default merger
787
hooks = [factory(self) for factory in factories] + [self]
788
self.active_hooks = [hook for hook in hooks if hook is not None]
789
with ui.ui_factory.nested_progress_bar() as child_pb:
790
for num, (file_id, changed, paths3, parents3, names3,
791
executable3) in enumerate(entries):
792
trans_id = self.tt.trans_id_file_id(file_id)
793
# Try merging each entry
794
child_pb.update(gettext('Preparing file merge'),
796
self._merge_names(trans_id, file_id, paths3, parents3,
797
names3, resolver=resolver)
799
file_status = self._do_merge_contents(paths3, trans_id, file_id)
801
file_status = 'unmodified'
802
self._merge_executable(paths3, trans_id, executable3,
803
file_status, resolver=resolver)
804
self.tt.fixup_new_roots()
805
self._finish_computing_transform()
807
def _finish_computing_transform(self):
808
"""Finalize the transform and report the changes.
810
This is the second half of _compute_transform.
812
with ui.ui_factory.nested_progress_bar() as child_pb:
813
fs_conflicts = transform.resolve_conflicts(
815
lambda t, c: transform.conflict_pass(t, c, self.other_tree))
816
if self.change_reporter is not None:
817
from breezy import delta
818
delta.report_changes(
819
self.tt.iter_changes(), self.change_reporter)
820
self.cook_conflicts(fs_conflicts)
821
for conflict in self.cooked_conflicts:
822
trace.warning('%s', conflict.describe())
825
"""Gather data about files modified between three trees.
827
Return a list of tuples of file_id, changed, parents3, names3,
828
executable3. changed is a boolean indicating whether the file contents
829
or kind were changed. parents3 is a tuple of parent ids for base,
830
other and this. names3 is a tuple of names for base, other and this.
831
executable3 is a tuple of execute-bit values for base, other and this.
833
iterator = self.other_tree.iter_changes(self.base_tree,
834
specific_files=self.interesting_files,
835
extra_trees=[self.this_tree])
836
this_interesting_files = self.this_tree.find_related_paths_across_trees(
837
self.interesting_files, trees=[self.other_tree])
838
this_entries = dict(self.this_tree.iter_entries_by_dir(
839
specific_files=this_interesting_files))
840
for change in iterator:
841
if change.path[0] is not None:
842
this_path = _mod_tree.find_previous_path(
843
self.base_tree, self.this_tree, change.path[0])
845
this_path = _mod_tree.find_previous_path(
846
self.other_tree, self.this_tree, change.path[1])
847
this_entry = this_entries.get(this_path)
848
if this_entry is not None:
849
this_name = this_entry.name
850
this_parent = this_entry.parent_id
851
this_executable = this_entry.executable
855
this_executable = None
856
parents3 = change.parent_id + (this_parent,)
857
names3 = change.name + (this_name,)
858
paths3 = change.path + (this_path, )
859
executable3 = change.executable + (this_executable,)
861
(change.file_id, change.changed_content, paths3,
862
parents3, names3, executable3))
864
def _entries_lca(self):
865
"""Gather data about files modified between multiple trees.
867
This compares OTHER versus all LCA trees, and for interesting entries,
868
it then compares with THIS and BASE.
870
For the multi-valued entries, the format will be (BASE, [lca1, lca2])
872
:return: [(file_id, changed, paths, parents, names, executable)], where:
874
* file_id: Simple file_id of the entry
875
* changed: Boolean, True if the kind or contents changed else False
876
* paths: ((base, [path, in, lcas]), path_other, path_this)
877
* parents: ((base, [parent_id, in, lcas]), parent_id_other,
879
* names: ((base, [name, in, lcas]), name_in_other, name_in_this)
880
* executable: ((base, [exec, in, lcas]), exec_in_other,
883
if self.interesting_files is not None:
884
lookup_trees = [self.this_tree, self.base_tree]
885
lookup_trees.extend(self._lca_trees)
886
# I think we should include the lca trees as well
887
interesting_files = self.other_tree.find_related_paths_across_trees(
888
self.interesting_files, lookup_trees)
890
interesting_files = None
891
from .multiwalker import MultiWalker
892
walker = MultiWalker(self.other_tree, self._lca_trees)
894
for other_path, file_id, other_ie, lca_values in walker.iter_all():
895
# Is this modified at all from any of the other trees?
897
other_ie = _none_entry
899
if interesting_files is not None and other_path not in interesting_files:
902
# If other_revision is found in any of the lcas, that means this
903
# node is uninteresting. This is because when merging, if there are
904
# multiple heads(), we have to create a new node. So if we didn't,
905
# we know that the ancestry is linear, and that OTHER did not
907
# See doc/developers/lca_merge_resolution.txt for details
908
# We can't use this shortcut when other_revision is None,
909
# because it may be None because things are WorkingTrees, and
910
# not because it is *actually* None.
911
is_unmodified = False
912
for lca_path, ie in lca_values:
913
if ie is not None and other_ie.is_unmodified(ie):
921
for lca_path, lca_ie in lca_values:
923
lca_entries.append(_none_entry)
924
lca_paths.append(None)
926
lca_entries.append(lca_ie)
927
lca_paths.append(lca_path)
930
base_path = self.base_tree.id2path(file_id)
931
except errors.NoSuchId:
933
base_ie = _none_entry
935
base_ie = next(self.base_tree.iter_entries_by_dir(specific_files=[base_path]))[1]
938
this_path = self.this_tree.id2path(file_id)
939
except errors.NoSuchId:
940
this_ie = _none_entry
943
this_ie = next(self.this_tree.iter_entries_by_dir(specific_files=[this_path]))[1]
949
for lca_ie in lca_entries:
950
lca_kinds.append(lca_ie.kind)
951
lca_parent_ids.append(lca_ie.parent_id)
952
lca_names.append(lca_ie.name)
953
lca_executable.append(lca_ie.executable)
955
kind_winner = self._lca_multi_way(
956
(base_ie.kind, lca_kinds),
957
other_ie.kind, this_ie.kind)
958
parent_id_winner = self._lca_multi_way(
959
(base_ie.parent_id, lca_parent_ids),
960
other_ie.parent_id, this_ie.parent_id)
961
name_winner = self._lca_multi_way(
962
(base_ie.name, lca_names),
963
other_ie.name, this_ie.name)
965
content_changed = True
966
if kind_winner == 'this':
967
# No kind change in OTHER, see if there are *any* changes
968
if other_ie.kind == 'directory':
969
if parent_id_winner == 'this' and name_winner == 'this':
970
# No change for this directory in OTHER, skip
972
content_changed = False
973
elif other_ie.kind is None or other_ie.kind == 'file':
974
def get_sha1(tree, path):
978
return tree.get_file_sha1(path)
979
except errors.NoSuchFile:
981
base_sha1 = get_sha1(self.base_tree, base_path)
982
lca_sha1s = [get_sha1(tree, lca_path)
984
in zip(self._lca_trees, lca_paths)]
985
this_sha1 = get_sha1(self.this_tree, this_path)
986
other_sha1 = get_sha1(self.other_tree, other_path)
987
sha1_winner = self._lca_multi_way(
988
(base_sha1, lca_sha1s), other_sha1, this_sha1,
989
allow_overriding_lca=False)
990
exec_winner = self._lca_multi_way(
991
(base_ie.executable, lca_executable),
992
other_ie.executable, this_ie.executable)
993
if (parent_id_winner == 'this' and name_winner == 'this'
994
and sha1_winner == 'this' and exec_winner == 'this'):
995
# No kind, parent, name, exec, or content change for
996
# OTHER, so this node is not considered interesting
998
if sha1_winner == 'this':
999
content_changed = False
1000
elif other_ie.kind == 'symlink':
1001
def get_target(ie, tree, path):
1002
if ie.kind != 'symlink':
1004
return tree.get_symlink_target(path)
1005
base_target = get_target(base_ie, self.base_tree, base_path)
1006
lca_targets = [get_target(ie, tree, lca_path) for ie, tree, lca_path
1007
in zip(lca_entries, self._lca_trees, lca_paths)]
1008
this_target = get_target(
1009
this_ie, self.this_tree, this_path)
1010
other_target = get_target(
1011
other_ie, self.other_tree, other_path)
1012
target_winner = self._lca_multi_way(
1013
(base_target, lca_targets),
1014
other_target, this_target)
1015
if (parent_id_winner == 'this' and name_winner == 'this'
1016
and target_winner == 'this'):
1017
# No kind, parent, name, or symlink target change
1020
if target_winner == 'this':
1021
content_changed = False
1022
elif other_ie.kind == 'tree-reference':
1023
# The 'changed' information seems to be handled at a higher
1024
# level. At least, _entries3 returns False for content
1025
# changed, even when at a new revision_id.
1026
content_changed = False
1027
if (parent_id_winner == 'this' and name_winner == 'this'):
1028
# Nothing interesting
1031
raise AssertionError('unhandled kind: %s' % other_ie.kind)
1033
# If we have gotten this far, that means something has changed
1034
yield (file_id, content_changed,
1035
((base_path, lca_paths),
1036
other_path, this_path),
1037
((base_ie.parent_id, lca_parent_ids),
1038
other_ie.parent_id, this_ie.parent_id),
1039
((base_ie.name, lca_names),
1040
other_ie.name, this_ie.name),
1041
((base_ie.executable, lca_executable),
1042
other_ie.executable, this_ie.executable)
1045
def write_modified(self, results):
1046
if not self.working_tree.supports_merge_modified():
1048
modified_hashes = {}
1049
for path in results.modified_paths:
1050
wt_relpath = self.working_tree.relpath(path)
1051
if not self.working_tree.is_versioned(wt_relpath):
1053
hash = self.working_tree.get_file_sha1(wt_relpath)
1056
modified_hashes[wt_relpath] = hash
1057
self.working_tree.set_merge_modified(modified_hashes)
1061
"""Determine the parent for a file_id (used as a key method)"""
1064
return entry.parent_id
1068
"""Determine the name for a file_id (used as a key method)"""
1074
def contents_sha1(tree, path):
1075
"""Determine the sha1 of the file contents (used as a key method)."""
1077
return tree.get_file_sha1(path)
1078
except errors.NoSuchFile:
1082
def executable(tree, path):
1083
"""Determine the executability of a file-id (used as a key method)."""
1085
if tree.kind(path) != "file":
1087
except errors.NoSuchFile:
1089
return tree.is_executable(path)
1092
def kind(tree, path):
1093
"""Determine the kind of a file-id (used as a key method)."""
1095
return tree.kind(path)
1096
except errors.NoSuchFile:
1100
def _three_way(base, other, this):
1102
# if 'base == other', either they all agree, or only 'this' has
1105
elif this not in (base, other):
1106
# 'this' is neither 'base' nor 'other', so both sides changed
1109
# "Ambiguous clean merge" -- both sides have made the same change.
1112
# this == base: only other has changed.
1116
def _lca_multi_way(bases, other, this, allow_overriding_lca=True):
1117
"""Consider LCAs when determining whether a change has occurred.
1119
If LCAS are all identical, this is the same as a _three_way comparison.
1121
:param bases: value in (BASE, [LCAS])
1122
:param other: value in OTHER
1123
:param this: value in THIS
1124
:param allow_overriding_lca: If there is more than one unique lca
1125
value, allow OTHER to override THIS if it has a new value, and
1126
THIS only has an lca value, or vice versa. This is appropriate for
1127
truly scalar values, not as much for non-scalars.
1128
:return: 'this', 'other', or 'conflict' depending on whether an entry
1131
# See doc/developers/lca_tree_merging.txt for details about this
1134
# Either Ambiguously clean, or nothing was actually changed. We
1137
base_val, lca_vals = bases
1138
# Remove 'base_val' from the lca_vals, because it is not interesting
1139
filtered_lca_vals = [lca_val for lca_val in lca_vals
1140
if lca_val != base_val]
1141
if len(filtered_lca_vals) == 0:
1142
return Merge3Merger._three_way(base_val, other, this)
1144
unique_lca_vals = set(filtered_lca_vals)
1145
if len(unique_lca_vals) == 1:
1146
return Merge3Merger._three_way(unique_lca_vals.pop(), other, this)
1148
if allow_overriding_lca:
1149
if other in unique_lca_vals:
1150
if this in unique_lca_vals:
1151
# Each side picked a different lca, conflict
1154
# This has a value which supersedes both lca values, and
1155
# other only has an lca value
1157
elif this in unique_lca_vals:
1158
# OTHER has a value which supersedes both lca values, and this
1159
# only has an lca value
1162
# At this point, the lcas disagree, and the tip disagree
1165
def _merge_names(self, trans_id, file_id, paths, parents, names, resolver):
1166
"""Perform a merge on file names and parents"""
1167
base_name, other_name, this_name = names
1168
base_parent, other_parent, this_parent = parents
1169
unused_base_path, other_path, this_path = paths
1171
name_winner = resolver(*names)
1173
parent_id_winner = resolver(*parents)
1174
if this_name is None:
1175
if name_winner == "this":
1176
name_winner = "other"
1177
if parent_id_winner == "this":
1178
parent_id_winner = "other"
1179
if name_winner == "this" and parent_id_winner == "this":
1181
if name_winner == 'conflict' or parent_id_winner == 'conflict':
1182
# Creating helpers (.OTHER or .THIS) here cause problems down the
1183
# road if a ContentConflict needs to be created so we should not do
1185
self._raw_conflicts.append(('path conflict', trans_id, file_id,
1186
this_parent, this_name,
1187
other_parent, other_name))
1188
if other_path is None:
1189
# it doesn't matter whether the result was 'other' or
1190
# 'conflict'-- if it has no file id, we leave it alone.
1192
parent_id = parents[self.winner_idx[parent_id_winner]]
1193
name = names[self.winner_idx[name_winner]]
1194
if parent_id is not None or name is not None:
1195
# if we get here, name_winner and parent_winner are set to safe
1197
if parent_id is None and name is not None:
1198
# if parent_id is None and name is non-None, current file is
1200
if names[self.winner_idx[parent_id_winner]] != '':
1201
raise AssertionError(
1202
'File looks like a root, but named %s' %
1203
names[self.winner_idx[parent_id_winner]])
1204
parent_trans_id = transform.ROOT_PARENT
1206
parent_trans_id = self.tt.trans_id_file_id(parent_id)
1207
self.tt.adjust_path(name, parent_trans_id, trans_id)
1209
def _do_merge_contents(self, paths, trans_id, file_id):
1210
"""Performs a merge on file_id contents."""
1211
def contents_pair(tree, path):
1215
kind = tree.kind(path)
1216
except errors.NoSuchFile:
1219
contents = tree.get_file_sha1(path)
1220
elif kind == "symlink":
1221
contents = tree.get_symlink_target(path)
1224
return kind, contents
1226
base_path, other_path, this_path = paths
1227
# See SPOT run. run, SPOT, run.
1228
# So we're not QUITE repeating ourselves; we do tricky things with
1230
other_pair = contents_pair(self.other_tree, other_path)
1231
this_pair = contents_pair(self.this_tree, this_path)
1233
(base_path, lca_paths) = base_path
1234
base_pair = contents_pair(self.base_tree, base_path)
1235
lca_pairs = [contents_pair(tree, path)
1236
for tree, path in zip(self._lca_trees, lca_paths)]
1237
winner = self._lca_multi_way((base_pair, lca_pairs), other_pair,
1238
this_pair, allow_overriding_lca=False)
1240
base_pair = contents_pair(self.base_tree, base_path)
1241
if base_pair == other_pair:
1244
# We delayed evaluating this_pair as long as we can to avoid
1245
# unnecessary sha1 calculation
1246
this_pair = contents_pair(self.this_tree, this_path)
1247
winner = self._three_way(base_pair, other_pair, this_pair)
1248
if winner == 'this':
1249
# No interesting changes introduced by OTHER
1251
# We have a hypothetical conflict, but if we have files, then we
1252
# can try to merge the content
1253
params = MergeFileHookParams(
1254
self, (base_path, other_path, this_path), trans_id, this_pair[0],
1255
other_pair[0], winner)
1256
hooks = self.active_hooks
1257
hook_status = 'not_applicable'
1259
hook_status, lines = hook.merge_contents(params)
1260
if hook_status != 'not_applicable':
1261
# Don't try any more hooks, this one applies.
1263
# If the merge ends up replacing the content of the file, we get rid of
1264
# it at the end of this method (this variable is used to track the
1265
# exceptions to this rule).
1268
if hook_status == 'not_applicable':
1269
# No merge hook was able to resolve the situation. Two cases exist:
1270
# a content conflict or a duplicate one.
1272
name = self.tt.final_name(trans_id)
1273
parent_id = self.tt.final_parent(trans_id)
1274
inhibit_content_conflict = False
1275
if params.this_kind is None: # file_id is not in THIS
1276
# Is the name used for a different file_id ?
1277
if self.this_tree.is_versioned(other_path):
1278
# Two entries for the same path
1280
# versioning the merged file will trigger a duplicate
1282
self.tt.version_file(trans_id, file_id=file_id)
1283
transform.create_from_tree(
1284
self.tt, trans_id, self.other_tree,
1286
filter_tree_path=self._get_filter_tree_path(other_path))
1287
inhibit_content_conflict = True
1288
elif params.other_kind is None: # file_id is not in OTHER
1289
# Is the name used for a different file_id ?
1290
if self.other_tree.is_versioned(this_path):
1291
# Two entries for the same path again, but here, the other
1292
# entry will also be merged. We simply inhibit the
1293
# 'content' conflict creation because we know OTHER will
1294
# create (or has already created depending on ordering) an
1295
# entry at the same path. This will trigger a 'duplicate'
1298
inhibit_content_conflict = True
1299
if not inhibit_content_conflict:
1300
if params.this_kind is not None:
1301
self.tt.unversion_file(trans_id)
1302
# This is a contents conflict, because none of the available
1303
# functions could merge it.
1304
file_group = self._dump_conflicts(
1305
name, (base_path, other_path, this_path), parent_id,
1306
file_id, set_version=True)
1307
self._raw_conflicts.append(('contents conflict', file_group))
1308
elif hook_status == 'success':
1309
self.tt.create_file(lines, trans_id)
1310
elif hook_status == 'conflicted':
1311
# XXX: perhaps the hook should be able to provide
1312
# the BASE/THIS/OTHER files?
1313
self.tt.create_file(lines, trans_id)
1314
self._raw_conflicts.append(('text conflict', trans_id))
1315
name = self.tt.final_name(trans_id)
1316
parent_id = self.tt.final_parent(trans_id)
1317
self._dump_conflicts(
1318
name, (base_path, other_path, this_path), parent_id, file_id)
1319
elif hook_status == 'delete':
1320
self.tt.unversion_file(trans_id)
1322
elif hook_status == 'done':
1323
# The hook function did whatever it needs to do directly, no
1324
# further action needed here.
1327
raise AssertionError('unknown hook_status: %r' % (hook_status,))
1328
if not this_path and result == "modified":
1329
self.tt.version_file(trans_id, file_id=file_id)
1331
# The merge has been performed and produced a new content, so the
1332
# old contents should not be retained.
1333
self.tt.delete_contents(trans_id)
1336
def _default_other_winner_merge(self, merge_hook_params):
1337
"""Replace this contents with other."""
1338
trans_id = merge_hook_params.trans_id
1339
if merge_hook_params.other_path is not None:
1340
# OTHER changed the file
1341
transform.create_from_tree(
1342
self.tt, trans_id, self.other_tree,
1343
merge_hook_params.other_path,
1344
filter_tree_path=self._get_filter_tree_path(merge_hook_params.other_path))
1346
elif merge_hook_params.this_path is not None:
1347
# OTHER deleted the file
1348
return 'delete', None
1350
raise AssertionError(
1351
'winner is OTHER, but file %r not in THIS or OTHER tree'
1352
% (merge_hook_params.base_path,))
1354
def merge_contents(self, merge_hook_params):
1355
"""Fallback merge logic after user installed hooks."""
1356
# This function is used in merge hooks as the fallback instance.
1357
# Perhaps making this function and the functions it calls be a
1358
# a separate class would be better.
1359
if merge_hook_params.winner == 'other':
1360
# OTHER is a straight winner, so replace this contents with other
1361
return self._default_other_winner_merge(merge_hook_params)
1362
elif merge_hook_params.is_file_merge():
1363
# THIS and OTHER are both files, so text merge. Either
1364
# BASE is a file, or both converted to files, so at least we
1365
# have agreement that output should be a file.
1367
self.text_merge(merge_hook_params.trans_id,
1368
merge_hook_params.paths)
1369
except errors.BinaryFile:
1370
return 'not_applicable', None
1373
return 'not_applicable', None
1375
def get_lines(self, tree, path):
1376
"""Return the lines in a file, or an empty list."""
1380
kind = tree.kind(path)
1381
except errors.NoSuchFile:
1386
return tree.get_file_lines(path)
1388
def text_merge(self, trans_id, paths):
1389
"""Perform a three-way text merge on a file"""
1390
# it's possible that we got here with base as a different type.
1391
# if so, we just want two-way text conflicts.
1392
base_path, other_path, this_path = paths
1393
base_lines = self.get_lines(self.base_tree, base_path)
1394
other_lines = self.get_lines(self.other_tree, other_path)
1395
this_lines = self.get_lines(self.this_tree, this_path)
1396
m3 = merge3.Merge3(base_lines, this_lines, other_lines,
1397
is_cherrypick=self.cherrypick)
1398
start_marker = b"!START OF MERGE CONFLICT!" + b"I HOPE THIS IS UNIQUE"
1399
if self.show_base is True:
1400
base_marker = b'|' * 7
1404
def iter_merge3(retval):
1405
retval["text_conflicts"] = False
1406
for line in m3.merge_lines(name_a=b"TREE",
1407
name_b=b"MERGE-SOURCE",
1408
name_base=b"BASE-REVISION",
1409
start_marker=start_marker,
1410
base_marker=base_marker,
1411
reprocess=self.reprocess):
1412
if line.startswith(start_marker):
1413
retval["text_conflicts"] = True
1414
yield line.replace(start_marker, b'<' * 7)
1418
merge3_iterator = iter_merge3(retval)
1419
self.tt.create_file(merge3_iterator, trans_id)
1420
if retval["text_conflicts"] is True:
1421
self._raw_conflicts.append(('text conflict', trans_id))
1422
name = self.tt.final_name(trans_id)
1423
parent_id = self.tt.final_parent(trans_id)
1424
file_id = self.tt.final_file_id(trans_id)
1425
file_group = self._dump_conflicts(name, paths, parent_id, file_id,
1426
this_lines, base_lines,
1428
file_group.append(trans_id)
1430
def _get_filter_tree_path(self, path):
1431
if self.this_tree.supports_content_filtering():
1432
# We get the path from the working tree if it exists.
1433
# That fails though when OTHER is adding a file, so
1434
# we fall back to the other tree to find the path if
1435
# it doesn't exist locally.
1436
filter_path = _mod_tree.find_previous_path(
1437
self.other_tree, self.working_tree, path)
1438
if filter_path is None:
1441
# Skip the lookup for older formats
1444
def _dump_conflicts(self, name, paths, parent_id, file_id, this_lines=None,
1445
base_lines=None, other_lines=None, set_version=False,
1447
"""Emit conflict files.
1448
If this_lines, base_lines, or other_lines are omitted, they will be
1449
determined automatically. If set_version is true, the .OTHER, .THIS
1450
or .BASE (in that order) will be created as versioned files.
1452
base_path, other_path, this_path = paths
1453
data = [('OTHER', self.other_tree, other_path, other_lines),
1454
('THIS', self.this_tree, this_path, this_lines)]
1456
data.append(('BASE', self.base_tree, base_path, base_lines))
1458
# We need to use the actual path in the working tree of the file here,
1459
if self.this_tree.supports_content_filtering():
1460
filter_tree_path = this_path
1462
# Skip the id2path lookup for older formats
1463
filter_tree_path = None
1467
for suffix, tree, path, lines in data:
1468
if path is not None:
1469
trans_id = self._conflict_file(
1470
name, parent_id, path, tree, suffix, lines,
1472
file_group.append(trans_id)
1473
if set_version and not versioned:
1474
self.tt.version_file(trans_id, file_id=file_id)
1478
def _conflict_file(self, name, parent_id, path, tree, suffix,
1479
lines=None, filter_tree_path=None):
1480
"""Emit a single conflict file."""
1481
name = name + '.' + suffix
1482
trans_id = self.tt.create_path(name, parent_id)
1483
transform.create_from_tree(
1484
self.tt, trans_id, tree, path,
1486
filter_tree_path=filter_tree_path)
1489
def _merge_executable(self, paths, trans_id, executable, file_status,
1491
"""Perform a merge on the execute bit."""
1492
base_executable, other_executable, this_executable = executable
1493
base_path, other_path, this_path = paths
1494
if file_status == "deleted":
1496
winner = resolver(*executable)
1497
if winner == "conflict":
1498
# There must be a None in here, if we have a conflict, but we
1499
# need executability since file status was not deleted.
1500
if other_path is None:
1504
if winner == 'this' and file_status != "modified":
1506
if self.tt.final_kind(trans_id) != "file":
1508
if winner == "this":
1509
executability = this_executable
1511
if other_path is not None:
1512
executability = other_executable
1513
elif this_path is not None:
1514
executability = this_executable
1515
elif base_path is not None:
1516
executability = base_executable
1517
if executability is not None:
1518
self.tt.set_executability(executability, trans_id)
1520
def cook_conflicts(self, fs_conflicts):
1521
"""Convert all conflicts into a form that doesn't depend on trans_id"""
1522
content_conflict_file_ids = set()
1523
cooked_conflicts = transform.cook_conflicts(fs_conflicts, self.tt)
1524
fp = transform.FinalPaths(self.tt)
1525
for conflict in self._raw_conflicts:
1526
conflict_type = conflict[0]
1527
if conflict_type == 'path conflict':
1529
this_parent, this_name,
1530
other_parent, other_name) = conflict[1:]
1531
if this_parent is None or this_name is None:
1532
this_path = '<deleted>'
1534
parent_path = fp.get_path(
1535
self.tt.trans_id_file_id(this_parent))
1536
this_path = osutils.pathjoin(parent_path, this_name)
1537
if other_parent is None or other_name is None:
1538
other_path = '<deleted>'
1540
if other_parent == self.other_tree.path2id(''):
1541
# The tree transform doesn't know about the other root,
1542
# so we special case here to avoid a NoFinalPath
1546
parent_path = fp.get_path(
1547
self.tt.trans_id_file_id(other_parent))
1548
other_path = osutils.pathjoin(parent_path, other_name)
1549
c = _mod_conflicts.Conflict.factory(
1550
'path conflict', path=this_path,
1551
conflict_path=other_path,
1553
elif conflict_type == 'contents conflict':
1554
for trans_id in conflict[1]:
1555
file_id = self.tt.final_file_id(trans_id)
1556
if file_id is not None:
1557
# Ok we found the relevant file-id
1559
path = fp.get_path(trans_id)
1560
for suffix in ('.BASE', '.THIS', '.OTHER'):
1561
if path.endswith(suffix):
1562
# Here is the raw path
1563
path = path[:-len(suffix)]
1565
c = _mod_conflicts.Conflict.factory(conflict_type,
1566
path=path, file_id=file_id)
1567
content_conflict_file_ids.add(file_id)
1568
elif conflict_type == 'text conflict':
1569
trans_id = conflict[1]
1570
path = fp.get_path(trans_id)
1571
file_id = self.tt.final_file_id(trans_id)
1572
c = _mod_conflicts.Conflict.factory(conflict_type,
1573
path=path, file_id=file_id)
1575
raise AssertionError('bad conflict type: %r' % (conflict,))
1576
cooked_conflicts.append(c)
1578
self.cooked_conflicts = []
1579
# We want to get rid of path conflicts when a corresponding contents
1580
# conflict exists. This can occur when one branch deletes a file while
1581
# the other renames *and* modifies it. In this case, the content
1582
# conflict is enough.
1583
for c in cooked_conflicts:
1584
if (c.typestring == 'path conflict'
1585
and c.file_id in content_conflict_file_ids):
1587
self.cooked_conflicts.append(c)
1588
self.cooked_conflicts.sort(key=_mod_conflicts.Conflict.sort_key)
1591
class WeaveMerger(Merge3Merger):
1592
"""Three-way tree merger, text weave merger."""
1593
supports_reprocess = True
1594
supports_show_base = False
1595
supports_reverse_cherrypick = False
1596
history_based = True
1597
requires_file_merge_plan = True
1599
def _generate_merge_plan(self, this_path, base):
1600
return self.this_tree.plan_file_merge(this_path, self.other_tree,
1603
def _merged_lines(self, this_path):
1604
"""Generate the merged lines.
1605
There is no distinction between lines that are meant to contain <<<<<<<
1609
base = self.base_tree
1612
plan = self._generate_merge_plan(this_path, base)
1613
if 'merge' in debug.debug_flags:
1615
trans_id = self.tt.trans_id_file_id(file_id)
1616
name = self.tt.final_name(trans_id) + '.plan'
1617
contents = (b'%11s|%s' % l for l in plan)
1618
self.tt.new_file(name, self.tt.final_parent(trans_id), contents)
1619
textmerge = versionedfile.PlanWeaveMerge(plan, b'<<<<<<< TREE\n',
1620
b'>>>>>>> MERGE-SOURCE\n')
1621
lines, conflicts = textmerge.merge_lines(self.reprocess)
1623
base_lines = textmerge.base_from_plan()
1626
return lines, base_lines
1628
def text_merge(self, trans_id, paths):
1629
"""Perform a (weave) text merge for a given file and file-id.
1630
If conflicts are encountered, .THIS and .OTHER files will be emitted,
1631
and a conflict will be noted.
1633
base_path, other_path, this_path = paths
1634
lines, base_lines = self._merged_lines(this_path)
1636
# Note we're checking whether the OUTPUT is binary in this case,
1637
# because we don't want to get into weave merge guts.
1638
textfile.check_text_lines(lines)
1639
self.tt.create_file(lines, trans_id)
1640
if base_lines is not None:
1642
self._raw_conflicts.append(('text conflict', trans_id))
1643
name = self.tt.final_name(trans_id)
1644
parent_id = self.tt.final_parent(trans_id)
1645
file_id = self.tt.final_file_id(trans_id)
1646
file_group = self._dump_conflicts(name, paths, parent_id, file_id,
1648
base_lines=base_lines)
1649
file_group.append(trans_id)
1652
class LCAMerger(WeaveMerger):
1654
requires_file_merge_plan = True
1656
def _generate_merge_plan(self, this_path, base):
1657
return self.this_tree.plan_file_lca_merge(this_path, self.other_tree,
1661
class Diff3Merger(Merge3Merger):
1662
"""Three-way merger using external diff3 for text merging"""
1664
requires_file_merge_plan = False
1666
def dump_file(self, temp_dir, name, tree, path):
1667
out_path = osutils.pathjoin(temp_dir, name)
1668
with open(out_path, "wb") as out_file:
1669
in_file = tree.get_file(path)
1670
for line in in_file:
1671
out_file.write(line)
1674
def text_merge(self, trans_id, paths):
1675
"""Perform a diff3 merge using a specified file-id and trans-id.
1676
If conflicts are encountered, .BASE, .THIS. and .OTHER conflict files
1677
will be dumped, and a will be conflict noted.
1680
base_path, other_path, this_path = paths
1681
temp_dir = osutils.mkdtemp(prefix="bzr-")
1683
new_file = osutils.pathjoin(temp_dir, "new")
1684
this = self.dump_file(
1685
temp_dir, "this", self.this_tree, this_path)
1686
base = self.dump_file(
1687
temp_dir, "base", self.base_tree, base_path)
1688
other = self.dump_file(
1689
temp_dir, "other", self.other_tree, other_path)
1690
status = breezy.patch.diff3(new_file, this, base, other)
1691
if status not in (0, 1):
1692
raise errors.BzrError("Unhandled diff3 exit code")
1693
with open(new_file, 'rb') as f:
1694
self.tt.create_file(f, trans_id)
1696
name = self.tt.final_name(trans_id)
1697
parent_id = self.tt.final_parent(trans_id)
1698
file_id = self.tt.final_file_id(trans_id)
1699
self._dump_conflicts(name, paths, parent_id, file_id)
1700
self._raw_conflicts.append(('text conflict', trans_id))
1702
osutils.rmtree(temp_dir)
1705
class PathNotInTree(errors.BzrError):
1707
_fmt = """Merge-into failed because %(tree)s does not contain %(path)s."""
1709
def __init__(self, path, tree):
1710
errors.BzrError.__init__(self, path=path, tree=tree)
1713
class MergeIntoMerger(Merger):
1714
"""Merger that understands other_tree will be merged into a subdir.
1716
This also changes the Merger api so that it uses real Branch, revision_id,
1717
and RevisonTree objects, rather than using revision specs.
1720
def __init__(self, this_tree, other_branch, other_tree, target_subdir,
1721
source_subpath, other_rev_id=None):
1722
"""Create a new MergeIntoMerger object.
1724
source_subpath in other_tree will be effectively copied to
1725
target_subdir in this_tree.
1727
:param this_tree: The tree that we will be merging into.
1728
:param other_branch: The Branch we will be merging from.
1729
:param other_tree: The RevisionTree object we want to merge.
1730
:param target_subdir: The relative path where we want to merge
1731
other_tree into this_tree
1732
:param source_subpath: The relative path specifying the subtree of
1733
other_tree to merge into this_tree.
1735
# It is assumed that we are merging a tree that is not in our current
1736
# ancestry, which means we are using the "EmptyTree" as our basis.
1737
null_ancestor_tree = this_tree.branch.repository.revision_tree(
1738
_mod_revision.NULL_REVISION)
1739
super(MergeIntoMerger, self).__init__(
1740
this_branch=this_tree.branch,
1741
this_tree=this_tree,
1742
other_tree=other_tree,
1743
base_tree=null_ancestor_tree,
1745
self._target_subdir = target_subdir
1746
self._source_subpath = source_subpath
1747
self.other_branch = other_branch
1748
if other_rev_id is None:
1749
other_rev_id = other_tree.get_revision_id()
1750
self.other_rev_id = self.other_basis = other_rev_id
1751
self.base_is_ancestor = True
1752
self.backup_files = True
1753
self.merge_type = Merge3Merger
1754
self.show_base = False
1755
self.reprocess = False
1756
self.interesting_files = None
1757
self.merge_type = _MergeTypeParameterizer(MergeIntoMergeType,
1758
target_subdir=self._target_subdir,
1759
source_subpath=self._source_subpath)
1760
if self._source_subpath != '':
1761
# If this isn't a partial merge make sure the revisions will be
1763
self._maybe_fetch(self.other_branch, self.this_branch,
1766
def set_pending(self):
1767
if self._source_subpath != '':
1769
Merger.set_pending(self)
1772
class _MergeTypeParameterizer(object):
1773
"""Wrap a merge-type class to provide extra parameters.
1775
This is hack used by MergeIntoMerger to pass some extra parameters to its
1776
merge_type. Merger.do_merge() sets up its own set of parameters to pass to
1777
the 'merge_type' member. It is difficult override do_merge without
1778
re-writing the whole thing, so instead we create a wrapper which will pass
1779
the extra parameters.
1782
def __init__(self, merge_type, **kwargs):
1783
self._extra_kwargs = kwargs
1784
self._merge_type = merge_type
1786
def __call__(self, *args, **kwargs):
1787
kwargs.update(self._extra_kwargs)
1788
return self._merge_type(*args, **kwargs)
1790
def __getattr__(self, name):
1791
return getattr(self._merge_type, name)
1794
class MergeIntoMergeType(Merge3Merger):
1795
"""Merger that incorporates a tree (or part of a tree) into another."""
1797
def __init__(self, *args, **kwargs):
1798
"""Initialize the merger object.
1800
:param args: See Merge3Merger.__init__'s args.
1801
:param kwargs: See Merge3Merger.__init__'s keyword args, except for
1802
source_subpath and target_subdir.
1803
:keyword source_subpath: The relative path specifying the subtree of
1804
other_tree to merge into this_tree.
1805
:keyword target_subdir: The relative path where we want to merge
1806
other_tree into this_tree
1808
# All of the interesting work happens during Merge3Merger.__init__(),
1809
# so we have have to hack in to get our extra parameters set.
1810
self._source_subpath = kwargs.pop('source_subpath')
1811
self._target_subdir = kwargs.pop('target_subdir')
1812
super(MergeIntoMergeType, self).__init__(*args, **kwargs)
1814
def _compute_transform(self):
1815
with ui.ui_factory.nested_progress_bar() as child_pb:
1816
entries = self._entries_to_incorporate()
1817
entries = list(entries)
1818
for num, (entry, parent_id, relpath) in enumerate(entries):
1819
child_pb.update(gettext('Preparing file merge'),
1821
parent_trans_id = self.tt.trans_id_file_id(parent_id)
1822
path = osutils.pathjoin(self._source_subpath, relpath)
1823
trans_id = transform.new_by_entry(path, self.tt, entry,
1824
parent_trans_id, self.other_tree)
1825
self._finish_computing_transform()
1827
def _entries_to_incorporate(self):
1828
"""Yields pairs of (inventory_entry, new_parent)."""
1829
subdir_id = self.other_tree.path2id(self._source_subpath)
1830
if subdir_id is None:
1831
# XXX: The error would be clearer if it gave the URL of the source
1832
# branch, but we don't have a reference to that here.
1833
raise PathNotInTree(self._source_subpath, "Source tree")
1834
subdir = next(self.other_tree.iter_entries_by_dir(
1835
specific_files=[self._source_subpath]))[1]
1836
parent_in_target = osutils.dirname(self._target_subdir)
1837
target_id = self.this_tree.path2id(parent_in_target)
1838
if target_id is None:
1839
raise PathNotInTree(self._target_subdir, "Target tree")
1840
name_in_target = osutils.basename(self._target_subdir)
1841
merge_into_root = subdir.copy()
1842
merge_into_root.name = name_in_target
1844
self.this_tree.id2path(merge_into_root.file_id)
1845
except errors.NoSuchId:
1848
# Give the root a new file-id.
1849
# This can happen fairly easily if the directory we are
1850
# incorporating is the root, and both trees have 'TREE_ROOT' as
1851
# their root_id. Users will expect this to Just Work, so we
1852
# change the file-id here.
1853
# Non-root file-ids could potentially conflict too. That's really
1854
# an edge case, so we don't do anything special for those. We let
1855
# them cause conflicts.
1856
merge_into_root.file_id = generate_ids.gen_file_id(name_in_target)
1857
yield (merge_into_root, target_id, '')
1858
if subdir.kind != 'directory':
1859
# No children, so we are done.
1861
for path, entry in self.other_tree.root_inventory.iter_entries_by_dir(subdir_id):
1862
parent_id = entry.parent_id
1863
if parent_id == subdir.file_id:
1864
# The root's parent ID has changed, so make sure children of
1865
# the root refer to the new ID.
1866
parent_id = merge_into_root.file_id
1867
yield (entry, parent_id, path)
1870
def merge_inner(this_branch, other_tree, base_tree, ignore_zero=False,
1872
merge_type=Merge3Merger,
1876
interesting_files=None,
1878
change_reporter=None):
1879
"""Primary interface for merging.
1881
Typical use is probably::
1883
merge_inner(branch, branch.get_revision_tree(other_revision),
1884
branch.get_revision_tree(base_revision))
1886
if this_tree is None:
1887
raise errors.BzrError("breezy.merge.merge_inner requires a this_tree "
1889
merger = Merger(this_branch, other_tree, base_tree, this_tree=this_tree,
1890
change_reporter=change_reporter)
1891
merger.backup_files = backup_files
1892
merger.merge_type = merge_type
1893
merger.ignore_zero = ignore_zero
1894
merger.interesting_files = interesting_files
1895
merger.show_base = show_base
1896
merger.reprocess = reprocess
1897
merger.other_rev_id = other_rev_id
1898
merger.other_basis = other_rev_id
1899
get_revision_id = getattr(base_tree, 'get_revision_id', None)
1900
if get_revision_id is None:
1901
get_revision_id = base_tree.last_revision
1902
merger.cache_trees_with_revision_ids([other_tree, base_tree, this_tree])
1903
merger.set_base_revision(get_revision_id(), this_branch)
1904
return merger.do_merge()
1907
merge_type_registry = registry.Registry()
1908
merge_type_registry.register('diff3', Diff3Merger,
1909
"Merge using external diff3.")
1910
merge_type_registry.register('lca', LCAMerger,
1911
"LCA-newness merge.")
1912
merge_type_registry.register('merge3', Merge3Merger,
1913
"Native diff3-style merge.")
1914
merge_type_registry.register('weave', WeaveMerger,
1915
"Weave-based merge.")
1918
def get_merge_type_registry():
1919
"""Merge type registry was previously in breezy.option
1921
This method provides a backwards compatible way to retrieve it.
1923
return merge_type_registry
1926
def _plan_annotate_merge(annotated_a, annotated_b, ancestors_a, ancestors_b):
1927
def status_a(revision, text):
1928
if revision in ancestors_b:
1929
return 'killed-b', text
1931
return 'new-a', text
1933
def status_b(revision, text):
1934
if revision in ancestors_a:
1935
return 'killed-a', text
1937
return 'new-b', text
1939
plain_a = [t for (a, t) in annotated_a]
1940
plain_b = [t for (a, t) in annotated_b]
1941
matcher = patiencediff.PatienceSequenceMatcher(None, plain_a, plain_b)
1942
blocks = matcher.get_matching_blocks()
1945
for ai, bi, l in blocks:
1946
# process all mismatched sections
1947
# (last mismatched section is handled because blocks always
1948
# includes a 0-length last block)
1949
for revision, text in annotated_a[a_cur:ai]:
1950
yield status_a(revision, text)
1951
for revision, text in annotated_b[b_cur:bi]:
1952
yield status_b(revision, text)
1953
# and now the matched section
1956
for text_a in plain_a[ai:a_cur]:
1957
yield "unchanged", text_a
1960
class _PlanMergeBase(object):
1962
def __init__(self, a_rev, b_rev, vf, key_prefix):
1965
:param a_rev: Revision-id of one revision to merge
1966
:param b_rev: Revision-id of the other revision to merge
1967
:param vf: A VersionedFiles containing both revisions
1968
:param key_prefix: A prefix for accessing keys in vf, typically
1974
self._last_lines = None
1975
self._last_lines_revision_id = None
1976
self._cached_matching_blocks = {}
1977
self._key_prefix = key_prefix
1978
self._precache_tip_lines()
1980
def _precache_tip_lines(self):
1981
lines = self.get_lines([self.a_rev, self.b_rev])
1982
self.lines_a = lines[self.a_rev]
1983
self.lines_b = lines[self.b_rev]
1985
def get_lines(self, revisions):
1986
"""Get lines for revisions from the backing VersionedFiles.
1988
:raises RevisionNotPresent: on absent texts.
1990
keys = [(self._key_prefix + (rev,)) for rev in revisions]
1992
for record in self.vf.get_record_stream(keys, 'unordered', True):
1993
if record.storage_kind == 'absent':
1994
raise errors.RevisionNotPresent(record.key, self.vf)
1995
result[record.key[-1]] = record.get_bytes_as('lines')
1998
def plan_merge(self):
1999
"""Generate a 'plan' for merging the two revisions.
2001
This involves comparing their texts and determining the cause of
2002
differences. If text A has a line and text B does not, then either the
2003
line was added to text A, or it was deleted from B. Once the causes
2004
are combined, they are written out in the format described in
2005
VersionedFile.plan_merge
2007
blocks = self._get_matching_blocks(self.a_rev, self.b_rev)
2008
unique_a, unique_b = self._unique_lines(blocks)
2009
new_a, killed_b = self._determine_status(self.a_rev, unique_a)
2010
new_b, killed_a = self._determine_status(self.b_rev, unique_b)
2011
return self._iter_plan(blocks, new_a, killed_b, new_b, killed_a)
2013
def _iter_plan(self, blocks, new_a, killed_b, new_b, killed_a):
2016
for i, j, n in blocks:
2017
for a_index in range(last_i, i):
2018
if a_index in new_a:
2019
if a_index in killed_b:
2020
yield 'conflicted-a', self.lines_a[a_index]
2022
yield 'new-a', self.lines_a[a_index]
2024
yield 'killed-b', self.lines_a[a_index]
2025
for b_index in range(last_j, j):
2026
if b_index in new_b:
2027
if b_index in killed_a:
2028
yield 'conflicted-b', self.lines_b[b_index]
2030
yield 'new-b', self.lines_b[b_index]
2032
yield 'killed-a', self.lines_b[b_index]
2033
# handle common lines
2034
for a_index in range(i, i + n):
2035
yield 'unchanged', self.lines_a[a_index]
2039
def _get_matching_blocks(self, left_revision, right_revision):
2040
"""Return a description of which sections of two revisions match.
2042
See SequenceMatcher.get_matching_blocks
2044
cached = self._cached_matching_blocks.get((left_revision,
2046
if cached is not None:
2048
if self._last_lines_revision_id == left_revision:
2049
left_lines = self._last_lines
2050
right_lines = self.get_lines([right_revision])[right_revision]
2052
lines = self.get_lines([left_revision, right_revision])
2053
left_lines = lines[left_revision]
2054
right_lines = lines[right_revision]
2055
self._last_lines = right_lines
2056
self._last_lines_revision_id = right_revision
2057
matcher = patiencediff.PatienceSequenceMatcher(None, left_lines,
2059
return matcher.get_matching_blocks()
2061
def _unique_lines(self, matching_blocks):
2062
"""Analyse matching_blocks to determine which lines are unique
2064
:return: a tuple of (unique_left, unique_right), where the values are
2065
sets of line numbers of unique lines.
2071
for i, j, n in matching_blocks:
2072
unique_left.extend(range(last_i, i))
2073
unique_right.extend(range(last_j, j))
2076
return unique_left, unique_right
2079
def _subtract_plans(old_plan, new_plan):
2080
"""Remove changes from new_plan that came from old_plan.
2082
It is assumed that the difference between the old_plan and new_plan
2083
is their choice of 'b' text.
2085
All lines from new_plan that differ from old_plan are emitted
2086
verbatim. All lines from new_plan that match old_plan but are
2087
not about the 'b' revision are emitted verbatim.
2089
Lines that match and are about the 'b' revision are the lines we
2090
don't want, so we convert 'killed-b' -> 'unchanged', and 'new-b'
2091
is skipped entirely.
2093
matcher = patiencediff.PatienceSequenceMatcher(None, old_plan,
2096
for i, j, n in matcher.get_matching_blocks():
2097
for jj in range(last_j, j):
2099
for jj in range(j, j + n):
2100
plan_line = new_plan[jj]
2101
if plan_line[0] == 'new-b':
2103
elif plan_line[0] == 'killed-b':
2104
yield 'unchanged', plan_line[1]
2110
class _PlanMerge(_PlanMergeBase):
2111
"""Plan an annotate merge using on-the-fly annotation"""
2113
def __init__(self, a_rev, b_rev, vf, key_prefix):
2114
super(_PlanMerge, self).__init__(a_rev, b_rev, vf, key_prefix)
2115
self.a_key = self._key_prefix + (self.a_rev,)
2116
self.b_key = self._key_prefix + (self.b_rev,)
2117
self.graph = _mod_graph.Graph(self.vf)
2118
heads = self.graph.heads((self.a_key, self.b_key))
2120
# one side dominates, so we can just return its values, yay for
2122
# Ideally we would know that before we get this far
2123
self._head_key = heads.pop()
2124
if self._head_key == self.a_key:
2128
trace.mutter('found dominating revision for %s\n%s > %s', self.vf,
2129
self._head_key[-1], other)
2132
self._head_key = None
2135
def _precache_tip_lines(self):
2136
# Turn this into a no-op, because we will do this later
2139
def _find_recursive_lcas(self):
2140
"""Find all the ancestors back to a unique lca"""
2141
cur_ancestors = (self.a_key, self.b_key)
2142
# graph.find_lca(uncommon, keys) now returns plain NULL_REVISION,
2143
# rather than a key tuple. We will just map that directly to no common
2147
next_lcas = self.graph.find_lca(*cur_ancestors)
2148
# Map a plain NULL_REVISION to a simple no-ancestors
2149
if next_lcas == {_mod_revision.NULL_REVISION}:
2151
# Order the lca's based on when they were merged into the tip
2152
# While the actual merge portion of weave merge uses a set() of
2153
# active revisions, the order of insertion *does* effect the
2154
# implicit ordering of the texts.
2155
for rev_key in cur_ancestors:
2156
ordered_parents = tuple(self.graph.find_merge_order(rev_key,
2158
parent_map[rev_key] = ordered_parents
2159
if len(next_lcas) == 0:
2161
elif len(next_lcas) == 1:
2162
parent_map[list(next_lcas)[0]] = ()
2164
elif len(next_lcas) > 2:
2165
# More than 2 lca's, fall back to grabbing all nodes between
2166
# this and the unique lca.
2167
trace.mutter('More than 2 LCAs, falling back to all nodes for:'
2169
self.a_key, self.b_key, cur_ancestors)
2170
cur_lcas = next_lcas
2171
while len(cur_lcas) > 1:
2172
cur_lcas = self.graph.find_lca(*cur_lcas)
2173
if len(cur_lcas) == 0:
2174
# No common base to find, use the full ancestry
2177
unique_lca = list(cur_lcas)[0]
2178
if unique_lca == _mod_revision.NULL_REVISION:
2179
# find_lca will return a plain 'NULL_REVISION' rather
2180
# than a key tuple when there is no common ancestor, we
2181
# prefer to just use None, because it doesn't confuse
2182
# _get_interesting_texts()
2184
parent_map.update(self._find_unique_parents(next_lcas,
2187
cur_ancestors = next_lcas
2190
def _find_unique_parents(self, tip_keys, base_key):
2191
"""Find ancestors of tip that aren't ancestors of base.
2193
:param tip_keys: Nodes that are interesting
2194
:param base_key: Cull all ancestors of this node
2195
:return: The parent map for all revisions between tip_keys and
2196
base_key. base_key will be included. References to nodes outside of
2197
the ancestor set will also be removed.
2199
# TODO: this would be simpler if find_unique_ancestors took a list
2200
# instead of a single tip, internally it supports it, but it
2201
# isn't a "backwards compatible" api change.
2202
if base_key is None:
2203
parent_map = dict(self.graph.iter_ancestry(tip_keys))
2204
# We remove NULL_REVISION because it isn't a proper tuple key, and
2205
# thus confuses things like _get_interesting_texts, and our logic
2206
# to add the texts into the memory weave.
2207
if _mod_revision.NULL_REVISION in parent_map:
2208
parent_map.pop(_mod_revision.NULL_REVISION)
2211
for tip in tip_keys:
2213
self.graph.find_unique_ancestors(tip, [base_key]))
2214
parent_map = self.graph.get_parent_map(interesting)
2215
parent_map[base_key] = ()
2216
culled_parent_map, child_map, tails = self._remove_external_references(
2218
# Remove all the tails but base_key
2219
if base_key is not None:
2220
tails.remove(base_key)
2221
self._prune_tails(culled_parent_map, child_map, tails)
2222
# Now remove all the uninteresting 'linear' regions
2223
simple_map = _mod_graph.collapse_linear_regions(culled_parent_map)
2227
def _remove_external_references(parent_map):
2228
"""Remove references that go outside of the parent map.
2230
:param parent_map: Something returned from Graph.get_parent_map(keys)
2231
:return: (filtered_parent_map, child_map, tails)
2232
filtered_parent_map is parent_map without external references
2233
child_map is the {parent_key: [child_keys]} mapping
2234
tails is a list of nodes that do not have any parents in the map
2236
# TODO: The basic effect of this function seems more generic than
2237
# _PlanMerge. But the specific details of building a child_map,
2238
# and computing tails seems very specific to _PlanMerge.
2239
# Still, should this be in Graph land?
2240
filtered_parent_map = {}
2243
for key, parent_keys in parent_map.items():
2244
culled_parent_keys = [p for p in parent_keys if p in parent_map]
2245
if not culled_parent_keys:
2247
for parent_key in culled_parent_keys:
2248
child_map.setdefault(parent_key, []).append(key)
2249
# TODO: Do we want to do this, it adds overhead for every node,
2250
# just to say that the node has no children
2251
child_map.setdefault(key, [])
2252
filtered_parent_map[key] = culled_parent_keys
2253
return filtered_parent_map, child_map, tails
2256
def _prune_tails(parent_map, child_map, tails_to_remove):
2257
"""Remove tails from the parent map.
2259
This will remove the supplied revisions until no more children have 0
2262
:param parent_map: A dict of {child: [parents]}, this dictionary will
2263
be modified in place.
2264
:param tails_to_remove: A list of tips that should be removed,
2265
this list will be consumed
2266
:param child_map: The reverse dict of parent_map ({parent: [children]})
2267
this dict will be modified
2268
:return: None, parent_map will be modified in place.
2270
while tails_to_remove:
2271
next = tails_to_remove.pop()
2272
parent_map.pop(next)
2273
children = child_map.pop(next)
2274
for child in children:
2275
child_parents = parent_map[child]
2276
child_parents.remove(next)
2277
if len(child_parents) == 0:
2278
tails_to_remove.append(child)
2280
def _get_interesting_texts(self, parent_map):
2281
"""Return a dict of texts we are interested in.
2283
Note that the input is in key tuples, but the output is in plain
2286
:param parent_map: The output from _find_recursive_lcas
2287
:return: A dict of {'revision_id':lines} as returned by
2288
_PlanMergeBase.get_lines()
2290
all_revision_keys = set(parent_map)
2291
all_revision_keys.add(self.a_key)
2292
all_revision_keys.add(self.b_key)
2294
# Everything else is in 'keys' but get_lines is in 'revision_ids'
2295
all_texts = self.get_lines([k[-1] for k in all_revision_keys])
2298
def _build_weave(self):
2299
from .bzr import weave
2300
self._weave = weave.Weave(weave_name='in_memory_weave',
2301
allow_reserved=True)
2302
parent_map = self._find_recursive_lcas()
2304
all_texts = self._get_interesting_texts(parent_map)
2306
# Note: Unfortunately, the order given by topo_sort will effect the
2307
# ordering resolution in the output. Specifically, if you add A then B,
2308
# then in the output text A lines will show up before B lines. And, of
2309
# course, topo_sort doesn't guarantee any real ordering.
2310
# So we use merge_sort, and add a fake node on the tip.
2311
# This ensures that left-hand parents will always be inserted into the
2312
# weave before right-hand parents.
2313
tip_key = self._key_prefix + (_mod_revision.CURRENT_REVISION,)
2314
parent_map[tip_key] = (self.a_key, self.b_key)
2316
for seq_num, key, depth, eom in reversed(tsort.merge_sort(parent_map,
2320
# for key in tsort.topo_sort(parent_map):
2321
parent_keys = parent_map[key]
2322
revision_id = key[-1]
2323
parent_ids = [k[-1] for k in parent_keys]
2324
self._weave.add_lines(revision_id, parent_ids,
2325
all_texts[revision_id])
2327
def plan_merge(self):
2328
"""Generate a 'plan' for merging the two revisions.
2330
This involves comparing their texts and determining the cause of
2331
differences. If text A has a line and text B does not, then either the
2332
line was added to text A, or it was deleted from B. Once the causes
2333
are combined, they are written out in the format described in
2334
VersionedFile.plan_merge
2336
if self._head_key is not None: # There was a single head
2337
if self._head_key == self.a_key:
2340
if self._head_key != self.b_key:
2341
raise AssertionError('There was an invalid head: %s != %s'
2342
% (self.b_key, self._head_key))
2344
head_rev = self._head_key[-1]
2345
lines = self.get_lines([head_rev])[head_rev]
2346
return ((plan, line) for line in lines)
2347
return self._weave.plan_merge(self.a_rev, self.b_rev)
2350
class _PlanLCAMerge(_PlanMergeBase):
2352
This merge algorithm differs from _PlanMerge in that:
2354
1. comparisons are done against LCAs only
2355
2. cases where a contested line is new versus one LCA but old versus
2356
another are marked as conflicts, by emitting the line as conflicted-a
2359
This is faster, and hopefully produces more useful output.
2362
def __init__(self, a_rev, b_rev, vf, key_prefix, graph):
2363
_PlanMergeBase.__init__(self, a_rev, b_rev, vf, key_prefix)
2364
lcas = graph.find_lca(key_prefix + (a_rev,), key_prefix + (b_rev,))
2367
if lca == _mod_revision.NULL_REVISION:
2370
self.lcas.add(lca[-1])
2371
for lca in self.lcas:
2372
if _mod_revision.is_null(lca):
2375
lca_lines = self.get_lines([lca])[lca]
2376
matcher = patiencediff.PatienceSequenceMatcher(None, self.lines_a,
2378
blocks = list(matcher.get_matching_blocks())
2379
self._cached_matching_blocks[(a_rev, lca)] = blocks
2380
matcher = patiencediff.PatienceSequenceMatcher(None, self.lines_b,
2382
blocks = list(matcher.get_matching_blocks())
2383
self._cached_matching_blocks[(b_rev, lca)] = blocks
2385
def _determine_status(self, revision_id, unique_line_numbers):
2386
"""Determines the status unique lines versus all lcas.
2388
Basically, determines why the line is unique to this revision.
2390
A line may be determined new, killed, or both.
2392
If a line is determined new, that means it was not present in at least
2393
one LCA, and is not present in the other merge revision.
2395
If a line is determined killed, that means the line was present in
2398
If a line is killed and new, this indicates that the two merge
2399
revisions contain differing conflict resolutions.
2401
:param revision_id: The id of the revision in which the lines are
2403
:param unique_line_numbers: The line numbers of unique lines.
2404
:return: a tuple of (new_this, killed_other)
2408
unique_line_numbers = set(unique_line_numbers)
2409
for lca in self.lcas:
2410
blocks = self._get_matching_blocks(revision_id, lca)
2411
unique_vs_lca, _ignored = self._unique_lines(blocks)
2412
new.update(unique_line_numbers.intersection(unique_vs_lca))
2413
killed.update(unique_line_numbers.difference(unique_vs_lca))
538
self.base_rev_id = None
540
self.base_rev_id = base_branch.get_rev_id(base_revision[1])
541
fetch(from_branch=base_branch, to_branch=self.this_branch)
542
self.base_is_ancestor = is_ancestor(self.this_basis,
547
def get_inventory(tree):
548
return tree.inventory
550
inv_changes = merge_flex(self.this_tree, self.base_tree,
552
generate_changeset, get_inventory,
553
self.conflict_handler,
554
merge_factory=self.merge_factory,
555
interesting_ids=self.interesting_ids)
558
for id, path in inv_changes.iteritems():
563
assert path.startswith('.' + '/') or path.startswith('.' + '\\'), "path is %s" % path
565
adjust_ids.append((path, id))
566
if len(adjust_ids) > 0:
567
self.this_branch.working_tree().set_inventory(self.regen_inventory(adjust_ids))
568
conflicts = self.conflict_handler.conflicts
569
self.conflict_handler.finalize()
572
def regen_inventory(self, new_entries):
573
old_entries = self.this_branch.working_tree().read_working_inventory()
577
for path, file_id in new_entries:
580
new_entries_map[file_id] = path
582
def id2path(file_id):
583
path = new_entries_map.get(file_id)
586
entry = old_entries[file_id]
587
if entry.parent_id is None:
589
return os.path.join(id2path(entry.parent_id), entry.name)
591
for file_id in old_entries:
592
entry = old_entries[file_id]
593
path = id2path(file_id)
594
new_inventory[file_id] = (path, file_id, entry.parent_id,
596
by_path[path] = file_id
601
for path, file_id in new_entries:
603
del new_inventory[file_id]
606
new_path_list.append((path, file_id))
607
if file_id not in old_entries:
609
# Ensure no file is added before its parent
611
for path, file_id in new_path_list:
615
parent = by_path[os.path.dirname(path)]
616
abspath = os.path.join(self.this_tree.basedir, path)
617
kind = bzrlib.osutils.file_kind(abspath)
618
new_inventory[file_id] = (path, file_id, parent, kind)
619
by_path[path] = file_id
621
# Get a list in insertion order
622
new_inventory_list = new_inventory.values()
623
mutter ("""Inventory regeneration:
624
old length: %i insertions: %i deletions: %i new_length: %i"""\
625
% (len(old_entries), insertions, deletions,
626
len(new_inventory_list)))
627
assert len(new_inventory_list) == len(old_entries) + insertions\
629
new_inventory_list.sort()
630
return new_inventory_list
632
merge_types = { "merge3": (ApplyMerge3, "Native diff3-style merge"),
633
"diff3": (Diff3Merge, "Merge using external diff3"),
634
'weave': (WeaveMerge, "Weave-based merge")