1
# Copyright (C) 2008 Jelmer Vernooij <jelmer@samba.org>
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
from cStringIO import (
21
from dulwich.client import (
22
SimpleFetchGraphWalker,
24
from dulwich.objects import (
35
from bzrlib.errors import (
39
from bzrlib.inventory import (
45
from bzrlib.lru_cache import (
48
from bzrlib.repository import (
51
from bzrlib.revision import (
54
from bzrlib.tsort import (
58
from bzrlib.plugins.git.converter import (
61
from bzrlib.plugins.git.repository import (
66
from bzrlib.plugins.git.remote import (
71
class BzrFetchGraphWalker(object):
72
"""GraphWalker implementation that uses a Bazaar repository."""
74
def __init__(self, repository, mapping):
75
self.repository = repository
76
self.mapping = mapping
78
self.heads = set(repository.all_revision_ids())
82
return iter(self.next, None)
85
revid = self.mapping.revision_id_foreign_to_bzr(sha)
88
def remove(self, revid):
90
if revid in self.heads:
91
self.heads.remove(revid)
92
if revid in self.parents:
93
for p in self.parents[revid]:
98
ret = self.heads.pop()
99
ps = self.repository.get_parent_map([ret])[ret]
100
self.parents[ret] = ps
101
self.heads.update([p for p in ps if not p in self.done])
104
return self.mapping.revision_id_bzr_to_foreign(ret)[0]
105
except InvalidRevisionId:
110
def import_git_blob(texts, mapping, path, blob, base_inv, parent_id,
111
revision_id, parent_invs, shagitmap, executable):
112
"""Import a git blob object into a bzr repository.
114
:param texts: VersionedFiles to add to
115
:param path: Path in the tree
116
:param blob: A git blob
117
:return: Inventory delta for this file
119
file_id = mapping.generate_file_id(path)
120
# We just have to hope this is indeed utf-8:
121
ie = InventoryFile(file_id, urlutils.basename(path).decode("utf-8"),
123
ie.text_size = len(blob.data)
124
ie.text_sha1 = osutils.sha_string(blob.data)
125
ie.executable = executable
126
# If there were no changes compared to the base inventory, there's no need
128
if (file_id in base_inv and
129
base_inv[file_id].parent_id == ie.parent_id and
130
base_inv[file_id].text_sha1 == ie.text_sha1 and
131
base_inv[file_id].executable == ie.executable):
133
# Check what revision we should store
135
for pinv in parent_invs:
136
if not file_id in pinv:
138
if pinv[file_id].text_sha1 == ie.text_sha1:
139
# found a revision in one of the parents to use
140
ie.revision = pinv[file_id].revision
142
parent_keys.append((file_id, pinv[file_id].revision))
143
if ie.revision is None:
144
# Need to store a new revision
145
ie.revision = revision_id
146
assert file_id is not None
147
assert ie.revision is not None
148
texts.add_lines((file_id, ie.revision), parent_keys,
149
osutils.split_lines(blob.data))
150
shagitmap.add_entry(blob.sha().hexdigest(), "blob",
151
(ie.file_id, ie.revision))
152
if file_id in base_inv:
153
old_path = base_inv.id2path(file_id)
156
return [(old_path, path, file_id, ie)]
159
def import_git_tree(texts, mapping, path, tree, base_inv, parent_id,
160
revision_id, parent_invs, shagitmap, lookup_object):
161
"""Import a git tree object into a bzr repository.
163
:param texts: VersionedFiles object to add to
164
:param path: Path in the tree
165
:param tree: A git tree object
166
:param base_inv: Base inventory against which to return inventory delta
167
:return: Inventory delta for this subtree
170
file_id = mapping.generate_file_id(path)
171
# We just have to hope this is indeed utf-8:
172
ie = InventoryDirectory(file_id, urlutils.basename(path.decode("utf-8")),
174
if not file_id in base_inv:
175
# Newly appeared here
176
ie.revision = revision_id
177
texts.add_lines((file_id, ie.revision), [], [])
178
ret.append((None, path, file_id, ie))
180
# See if this has changed at all
182
base_sha = shagitmap.lookup_tree(path, base_inv.revision_id)
186
if base_sha == tree.id:
187
# If nothing has changed since the base revision, we're done
189
# Remember for next time
190
existing_children = set()
191
shagitmap.add_entry(tree.id, "tree", (file_id, revision_id))
192
for mode, name, hexsha in tree.entries():
193
entry_kind = (mode & 0700000) / 0100000
194
basename = name.decode("utf-8")
195
existing_children.add(basename)
199
child_path = urlutils.join(path, name)
200
obj = lookup_object(hexsha)
202
ret.extend(import_git_tree(texts, mapping, child_path, obj, base_inv,
203
file_id, revision_id, parent_invs, shagitmap, lookup_object))
204
elif entry_kind == 1:
205
fs_mode = mode & 0777
206
ret.extend(import_git_blob(texts, mapping, child_path, obj, base_inv,
207
file_id, revision_id, parent_invs, shagitmap,
208
bool(fs_mode & 0111)))
210
raise AssertionError("Unknown blob kind, perms=%r." % (mode,))
211
# Remove any children that have disappeared
212
if file_id in base_inv:
213
deletable = [v for k,v in base_inv[file_id].children.iteritems() if k not in existing_children]
216
ret.append((base_inv.id2path(ie.file_id), None, ie.file_id, None))
217
if ie.kind == "directory":
218
deletable.extend(ie.children.values())
222
def import_git_objects(repo, mapping, object_iter, target_git_object_retriever,
224
"""Import a set of git objects into a bzr repository.
226
:param repo: Bazaar repository
227
:param mapping: Mapping to use
228
:param object_iter: Iterator over Git objects.
230
# TODO: a more (memory-)efficient implementation of this
236
parent_invs_cache = LRUCache(50)
237
# Find and convert commit objects
240
pb.update("finding revisions to fetch", len(graph), None)
242
assert isinstance(head, str)
243
o = object_iter[head]
244
if isinstance(o, Commit):
245
rev = mapping.import_commit(o)
246
if repo.has_revision(rev.revision_id):
248
root_trees[rev.revision_id] = o.tree
249
revisions[rev.revision_id] = rev
250
graph.append((rev.revision_id, rev.parent_ids))
251
target_git_object_retriever._idmap.add_entry(o.sha().hexdigest(),
252
"commit", (rev.revision_id, o._tree))
253
heads.extend([p for p in o.parents if p not in checked])
254
elif isinstance(o, Tag):
255
heads.append(o.object[1])
257
trace.warning("Unable to import head object %r" % o)
259
# Order the revisions
260
# Create the inventory objects
261
for i, revid in enumerate(topo_sort(graph)):
263
pb.update("fetching revisions", i, len(graph))
264
root_tree = object_iter[root_trees[revid]]
265
rev = revisions[revid]
266
# We have to do this here, since we have to walk the tree and
267
# we need to make sure to import the blobs / trees with the right
268
# path; this may involve adding them more than once.
269
def lookup_object(sha):
271
return object_iter[sha]
273
return target_git_object_retriever[sha]
275
for parent_id in rev.parent_ids:
277
parent_invs.append(parent_invs_cache[parent_id])
279
parent_inv = repo.get_inventory(parent_id)
280
parent_invs.append(parent_inv)
281
parent_invs_cache[parent_id] = parent_inv
282
if parent_invs == []:
283
base_inv = Inventory(root_id=None)
285
base_inv = parent_invs[0]
286
inv_delta = import_git_tree(repo.texts, mapping, "", root_tree,
287
base_inv, None, revid, parent_invs,
288
target_git_object_retriever._idmap, lookup_object)
290
basis_id = rev.parent_ids[0]
292
basis_id = NULL_REVISION
293
rev.inventory_sha1, inv = repo.add_inventory_by_delta(basis_id,
294
inv_delta, rev.revision_id, rev.parent_ids)
295
parent_invs_cache[rev.revision_id] = inv
296
repo.add_revision(rev.revision_id, rev)
297
target_git_object_retriever._idmap.commit()
300
class InterGitNonGitRepository(InterRepository):
301
"""InterRepository that copies revisions from a Git into a non-Git
304
_matching_repo_format = GitRepositoryFormat()
307
def _get_repo_format_to_test():
310
def copy_content(self, revision_id=None, pb=None):
311
"""See InterRepository.copy_content."""
312
self.fetch(revision_id, pb, find_ghosts=False)
314
def fetch_objects(self, determine_wants, mapping, pb=None):
316
pb.update("git: %s" % text.rstrip("\r\n"), 0, 0)
317
graph_walker = BzrFetchGraphWalker(self.target, mapping)
320
create_pb = pb = ui.ui_factory.nested_progress_bar()
321
target_git_object_retriever = GitObjectConverter(self.target, mapping)
324
def record_determine_wants(heads):
325
wants = determine_wants(heads)
326
recorded_wants.extend(wants)
330
self.target.lock_write()
332
self.target.start_write_group()
334
objects_iter = self.source.fetch_objects(
335
record_determine_wants,
337
target_git_object_retriever.__getitem__,
339
import_git_objects(self.target, mapping, objects_iter,
340
target_git_object_retriever, recorded_wants, pb)
342
self.target.commit_write_group()
349
def fetch(self, revision_id=None, pb=None, find_ghosts=False, mapping=None,
351
self.fetch_refs(revision_id=revision_id, pb=pb, find_ghosts=find_ghosts,
352
mapping=mapping, fetch_spec=fetch_spec)
354
def fetch_refs(self, revision_id=None, pb=None, find_ghosts=False,
355
mapping=None, fetch_spec=None):
357
mapping = self.source.get_mapping()
358
if revision_id is not None:
359
interesting_heads = [revision_id]
360
elif fetch_spec is not None:
361
interesting_heads = fetch_spec.heads
363
interesting_heads = None
365
def determine_wants(refs):
367
if interesting_heads is None:
368
ret = [sha for (ref, sha) in refs.iteritems() if not ref.endswith("^{}")]
370
ret = [mapping.revision_id_bzr_to_foreign(revid)[0] for revid in interesting_heads]
371
return [rev for rev in ret if not self.target.has_revision(mapping.revision_id_foreign_to_bzr(rev))]
372
self.fetch_objects(determine_wants, mapping, pb)
376
def is_compatible(source, target):
377
"""Be compatible with GitRepository."""
378
# FIXME: Also check target uses VersionedFile
379
return (isinstance(source, GitRepository) and
380
target.supports_rich_root() and
381
not isinstance(target, GitRepository))
384
class InterGitRepository(InterRepository):
385
"""InterRepository that copies between Git repositories."""
387
_matching_repo_format = GitRepositoryFormat()
390
def _get_repo_format_to_test():
393
def copy_content(self, revision_id=None, pb=None):
394
"""See InterRepository.copy_content."""
395
self.fetch(revision_id, pb, find_ghosts=False)
397
def fetch(self, revision_id=None, pb=None, find_ghosts=False,
398
mapping=None, fetch_spec=None):
400
mapping = self.source.get_mapping()
402
trace.info("git: %s", text)
404
if revision_id is not None:
405
args = [mapping.revision_id_bzr_to_foreign(revision_id)[0]]
406
elif fetch_spec is not None:
407
args = [mapping.revision_id_bzr_to_foreign(revid)[0] for revid in fetch_spec.heads]
408
if fetch_spec is None and revision_id is None:
409
determine_wants = r.object_store.determine_wants_all
411
determine_wants = lambda x: [y for y in args if not y in r.object_store]
413
graphwalker = SimpleFetchGraphWalker(r.heads().values(), r.get_parents)
414
f, commit = r.object_store.add_pack()
416
self.source._git.fetch_pack(path, determine_wants, graphwalker, f.write, progress)
424
def is_compatible(source, target):
425
"""Be compatible with GitRepository."""
426
return (isinstance(source, GitRepository) and
427
isinstance(target, GitRepository))
added
removed
fetch.py
