def test_index_file_base(self):
# read from file
index = IndexFile(self.rorepo, fixture_path("index"))
assert index.entries
assert index.version > 0
# test entry
entry = next(iter(index.entries.values()))
for attr in ("path", "ctime", "mtime", "dev", "inode", "mode", "uid",
"gid", "size", "binsha", "hexsha", "stage"):
getattr(entry, attr)
# END for each method
# test update
entries = index.entries
assert isinstance(index.update(), IndexFile)
assert entries is not index.entries
# test stage
index_merge = IndexFile(self.rorepo, fixture_path("index_merge"))
self.assertEqual(len(index_merge.entries), 106)
assert len(list(e for e in index_merge.entries.values() if e.stage != 0))
# write the data - it must match the original
tmpfile = tempfile.mktemp()
index_merge.write(tmpfile)
with open(tmpfile, 'rb') as fp:
self.assertEqual(fp.read(), fixture("index_merge"))
os.remove(tmpfile)
def test_index_merge_tree(self, rw_repo):
# A bit out of place, but we need a different repo for this:
self.assertNotEqual(self.rorepo, rw_repo)
self.assertEqual(len(set((self.rorepo, self.rorepo, rw_repo, rw_repo))), 2)
# SINGLE TREE MERGE
# current index is at the (virtual) cur_commit
next_commit = "4c39f9da792792d4e73fc3a5effde66576ae128c"
parent_commit = rw_repo.head.commit.parents[0]
manifest_key = IndexFile.entry_key('MANIFEST.in', 0)
manifest_entry = rw_repo.index.entries[manifest_key]
rw_repo.index.merge_tree(next_commit)
# only one change should be recorded
assert manifest_entry.binsha != rw_repo.index.entries[manifest_key].binsha
rw_repo.index.reset(rw_repo.head)
self.assertEqual(rw_repo.index.entries[manifest_key].binsha, manifest_entry.binsha)
# FAKE MERGE
#############
# Add a change with a NULL sha that should conflict with next_commit. We
# pretend there was a change, but we do not even bother adding a proper
# sha for it ( which makes things faster of course )
manifest_fake_entry = BaseIndexEntry((manifest_entry[0], b"\0" * 20, 0, manifest_entry[3]))
# try write flag
self._assert_entries(rw_repo.index.add([manifest_fake_entry], write=False))
# add actually resolves the null-hex-sha for us as a feature, but we can
# edit the index manually
assert rw_repo.index.entries[manifest_key].binsha != Object.NULL_BIN_SHA
# must operate on the same index for this ! Its a bit problematic as
# it might confuse people
index = rw_repo.index
index.entries[manifest_key] = IndexEntry.from_base(manifest_fake_entry)
index.write()
self.assertEqual(rw_repo.index.entries[manifest_key].hexsha, Diff.NULL_HEX_SHA)
# write an unchanged index ( just for the fun of it )
rw_repo.index.write()
# a three way merge would result in a conflict and fails as the command will
# not overwrite any entries in our index and hence leave them unmerged. This is
# mainly a protection feature as the current index is not yet in a tree
self.failUnlessRaises(GitCommandError, index.merge_tree, next_commit, base=parent_commit)
# the only way to get the merged entries is to safe the current index away into a tree,
# which is like a temporary commit for us. This fails as well as the NULL sha deos not
# have a corresponding object
# NOTE: missing_ok is not a kwarg anymore, missing_ok is always true
# self.failUnlessRaises(GitCommandError, index.write_tree)
# if missing objects are okay, this would work though ( they are always okay now )
# As we can't read back the tree with NULL_SHA, we rather set it to something else
index.entries[manifest_key] = IndexEntry(manifest_entry[:1] + (hex_to_bin('f' * 40),) + manifest_entry[2:])
tree = index.write_tree()
# now make a proper three way merge with unmerged entries
unmerged_tree = IndexFile.from_tree(rw_repo, parent_commit, tree, next_commit)
unmerged_blobs = unmerged_tree.unmerged_blobs()
self.assertEqual(len(unmerged_blobs), 1)
self.assertEqual(list(unmerged_blobs.keys())[0], manifest_key[0])
def test_index_merge_tree(self, rw_repo):
# A bit out of place, but we need a different repo for this:
self.assertNotEqual(self.rorepo, rw_repo)
self.assertEqual(len(set((self.rorepo, self.rorepo, rw_repo, rw_repo))), 2)
# SINGLE TREE MERGE
# current index is at the (virtual) cur_commit
next_commit = "4c39f9da792792d4e73fc3a5effde66576ae128c"
parent_commit = rw_repo.head.commit.parents[0]
manifest_key = IndexFile.entry_key('MANIFEST.in', 0)
manifest_entry = rw_repo.index.entries[manifest_key]
rw_repo.index.merge_tree(next_commit)
# only one change should be recorded
assert manifest_entry.binsha != rw_repo.index.entries[manifest_key].binsha
rw_repo.index.reset(rw_repo.head)
self.assertEqual(rw_repo.index.entries[manifest_key].binsha, manifest_entry.binsha)
# FAKE MERGE
#############
# Add a change with a NULL sha that should conflict with next_commit. We
# pretend there was a change, but we do not even bother adding a proper
# sha for it ( which makes things faster of course )
manifest_fake_entry = BaseIndexEntry((manifest_entry[0], b"\0" * 20, 0, manifest_entry[3]))
# try write flag
self._assert_entries(rw_repo.index.add([manifest_fake_entry], write=False))
# add actually resolves the null-hex-sha for us as a feature, but we can
# edit the index manually
assert rw_repo.index.entries[manifest_key].binsha != Object.NULL_BIN_SHA
# must operate on the same index for this ! Its a bit problematic as
# it might confuse people
index = rw_repo.index
index.entries[manifest_key] = IndexEntry.from_base(manifest_fake_entry)
index.write()
self.assertEqual(rw_repo.index.entries[manifest_key].hexsha, Diff.NULL_HEX_SHA)
# write an unchanged index ( just for the fun of it )
rw_repo.index.write()
# a three way merge would result in a conflict and fails as the command will
# not overwrite any entries in our index and hence leave them unmerged. This is
# mainly a protection feature as the current index is not yet in a tree
self.failUnlessRaises(GitCommandError, index.merge_tree, next_commit, base=parent_commit)
# the only way to get the merged entries is to safe the current index away into a tree,
# which is like a temporary commit for us. This fails as well as the NULL sha deos not
# have a corresponding object
# NOTE: missing_ok is not a kwarg anymore, missing_ok is always true
# self.failUnlessRaises(GitCommandError, index.write_tree)
# if missing objects are okay, this would work though ( they are always okay now )
# As we can't read back the tree with NULL_SHA, we rather set it to something else
index.entries[manifest_key] = IndexEntry(manifest_entry[:1] + (hex_to_bin('f' * 40),) + manifest_entry[2:])
tree = index.write_tree()
# now make a proper three way merge with unmerged entries
unmerged_tree = IndexFile.from_tree(rw_repo, parent_commit, tree, next_commit)
unmerged_blobs = unmerged_tree.unmerged_blobs()
self.assertEqual(len(unmerged_blobs), 1)
self.assertEqual(list(unmerged_blobs.keys())[0], manifest_key[0])
def _git_rel_path(path):
try:
repo = Repo(path, search_parent_directories=True)
index = IndexFile(repo)
path = Path(index._to_relative_path(path))
return path.as_posix()
except Exception:
return ''
def test_index_file_from_tree(self, rw_repo):
if sys.version_info < (2, 7):
## Skipped, not `assertRaisesRegexp` in py2.6
return
common_ancestor_sha = "5117c9c8a4d3af19a9958677e45cda9269de1541"
cur_sha = "4b43ca7ff72d5f535134241e7c797ddc9c7a3573"
other_sha = "39f85c4358b7346fee22169da9cad93901ea9eb9"
# simple index from tree
base_index = IndexFile.from_tree(rw_repo, common_ancestor_sha)
assert base_index.entries
self._cmp_tree_index(common_ancestor_sha, base_index)
# merge two trees - its like a fast-forward
two_way_index = IndexFile.from_tree(rw_repo, common_ancestor_sha,
cur_sha)
assert two_way_index.entries
self._cmp_tree_index(cur_sha, two_way_index)
# merge three trees - here we have a merge conflict
three_way_index = IndexFile.from_tree(rw_repo, common_ancestor_sha,
cur_sha, other_sha)
assert len(
list(e for e in three_way_index.entries.values() if e.stage != 0))
# ITERATE BLOBS
merge_required = lambda t: t[0] != 0
merge_blobs = list(three_way_index.iter_blobs(merge_required))
assert merge_blobs
assert merge_blobs[0][0] in (1, 2, 3)
assert isinstance(merge_blobs[0][1], Blob)
# test BlobFilter
prefix = 'lib/git'
for stage, blob in base_index.iter_blobs(BlobFilter(
[prefix])): # @UnusedVariable
assert blob.path.startswith(prefix)
# writing a tree should fail with an unmerged index
self.failUnlessRaises(UnmergedEntriesError, three_way_index.write_tree)
# removed unmerged entries
unmerged_blob_map = three_way_index.unmerged_blobs()
assert unmerged_blob_map
# pick the first blob at the first stage we find and use it as resolved version
three_way_index.resolve_blobs(l[0][1]
for l in unmerged_blob_map.values())
tree = three_way_index.write_tree()
assert isinstance(tree, Tree)
num_blobs = 0
for blob in tree.traverse(
predicate=lambda item, d: item.type == "blob"):
assert (blob.path, 0) in three_way_index.entries
num_blobs += 1
# END for each blob
self.assertEqual(num_blobs, len(three_way_index.entries))
def test_index_file_from_tree(self, rw_repo):
if sys.version_info < (2, 7):
## Skipped, not `assertRaisesRegexp` in py2.6
return
common_ancestor_sha = "5117c9c8a4d3af19a9958677e45cda9269de1541"
cur_sha = "4b43ca7ff72d5f535134241e7c797ddc9c7a3573"
other_sha = "39f85c4358b7346fee22169da9cad93901ea9eb9"
# simple index from tree
base_index = IndexFile.from_tree(rw_repo, common_ancestor_sha)
assert base_index.entries
self._cmp_tree_index(common_ancestor_sha, base_index)
# merge two trees - its like a fast-forward
two_way_index = IndexFile.from_tree(rw_repo, common_ancestor_sha, cur_sha)
assert two_way_index.entries
self._cmp_tree_index(cur_sha, two_way_index)
# merge three trees - here we have a merge conflict
three_way_index = IndexFile.from_tree(rw_repo, common_ancestor_sha, cur_sha, other_sha)
assert len(list(e for e in three_way_index.entries.values() if e.stage != 0))
# ITERATE BLOBS
merge_required = lambda t: t[0] != 0
merge_blobs = list(three_way_index.iter_blobs(merge_required))
assert merge_blobs
assert merge_blobs[0][0] in (1, 2, 3)
assert isinstance(merge_blobs[0][1], Blob)
# test BlobFilter
prefix = 'lib/git'
for stage, blob in base_index.iter_blobs(BlobFilter([prefix])): # @UnusedVariable
assert blob.path.startswith(prefix)
# writing a tree should fail with an unmerged index
self.failUnlessRaises(UnmergedEntriesError, three_way_index.write_tree)
# removed unmerged entries
unmerged_blob_map = three_way_index.unmerged_blobs()
assert unmerged_blob_map
# pick the first blob at the first stage we find and use it as resolved version
three_way_index.resolve_blobs(l[0][1] for l in unmerged_blob_map.values())
tree = three_way_index.write_tree()
assert isinstance(tree, Tree)
num_blobs = 0
for blob in tree.traverse(predicate=lambda item, d: item.type == "blob"):
assert (blob.path, 0) in three_way_index.entries
num_blobs += 1
# END for each blob
self.assertEqual(num_blobs, len(three_way_index.entries))
def test__to_relative_path_at_root(self):
root = osp.abspath(os.sep)
class Mocked(object):
bare = False
git_dir = root
working_tree_dir = root
repo = Mocked()
path = os.path.join(root, 'file')
index = IndexFile(repo)
rel = index._to_relative_path(path)
self.assertEqual(rel, os.path.relpath(path, root))
def test__to_relative_path_at_root(self):
root = osp.abspath(os.sep)
class Mocked(object):
bare = False
git_dir = root
working_tree_dir = root
repo = Mocked()
path = os.path.join(root, 'file')
index = IndexFile(repo)
rel = index._to_relative_path(path)
self.assertEqual(rel, os.path.relpath(path, root))
def dependencies(self, revision='HEAD', paths=None):
"""Return dependencies from a revision or paths."""
result = []
if paths:
paths = (self.normalize_path(path) for path in paths)
else:
if revision == 'HEAD':
index = self.client.repo.index
else:
from git import IndexFile
index = IndexFile.from_tree(self.client.repo, revision)
paths = (path for path, _ in index.entries.keys())
for path in paths:
try:
result.append(
Usage.from_revision(
self.client,
path=path,
revision=revision,
))
except KeyError:
continue
return result
def test_index_new(self):
B = self.rorepo.tree("6d9b1f4f9fa8c9f030e3207e7deacc5d5f8bba4e")
H = self.rorepo.tree("25dca42bac17d511b7e2ebdd9d1d679e7626db5f")
M = self.rorepo.tree("e746f96bcc29238b79118123028ca170adc4ff0f")
for args in ((B, ), (B, H), (B, H, M)):
index = IndexFile.new(self.rorepo, *args)
assert isinstance(index, IndexFile)
def test_index_new(self):
B = self.rorepo.tree("6d9b1f4f9fa8c9f030e3207e7deacc5d5f8bba4e")
H = self.rorepo.tree("25dca42bac17d511b7e2ebdd9d1d679e7626db5f")
M = self.rorepo.tree("e746f96bcc29238b79118123028ca170adc4ff0f")
for args in ((B,), (B, H), (B, H, M)):
index = IndexFile.new(self.rorepo, *args)
assert isinstance(index, IndexFile)
def test_index_file_base(self):
# read from file
index = IndexFile(self.rorepo, fixture_path("index"))
assert index.entries
assert index.version > 0
# test entry
entry = next(iter(index.entries.values()))
for attr in ("path", "ctime", "mtime", "dev", "inode", "mode", "uid",
"gid", "size", "binsha", "hexsha", "stage"):
getattr(entry, attr)
# END for each method
# test update
entries = index.entries
assert isinstance(index.update(), IndexFile)
assert entries is not index.entries
# test stage
index_merge = IndexFile(self.rorepo, fixture_path("index_merge"))
self.assertEqual(len(index_merge.entries), 106)
assert len([e for e in index_merge.entries.values() if e.stage != 0])
# write the data - it must match the original
tmpfile = tempfile.mktemp()
index_merge.write(tmpfile)
with open(tmpfile, 'rb') as fp:
self.assertEqual(fp.read(), fixture("index_merge"))
os.remove(tmpfile)
def add_data(self, index: git.IndexFile, path: str, data: bytes):
"""
Adds physical files to the database.
WARNING: this function touches physical files in the Git Repo which can result in a race
condition to modify a file while it is also being pushed. ONLY CALL THIS FUNCTION INSIDE
A COMMIT_LOCK.
@param index:
@param path:
@param data:
@return:
"""
fullpath = os.path.join(os.path.dirname(index.repo.git_dir), path)
pathlib.Path(fullpath).parent.mkdir(parents=True, exist_ok=True)
with open(fullpath, 'wb') as fp:
fp.write(data)
index.add([fullpath])
def time_warp(self, c_o):
"""
History rewriting occurs here. We read everything from the original
commit, reformat the python, and checkin, mirroring the original
commit history.
:param c_o: Commit object representing "before"
:return: None
"""
log.info('warping: {} | {} | {:f} MB | {}s'.format(
time_convert(c_o.authored_date, c_o.author_tz_offset), c_o.summary,
resource.getrusage(resource.RUSAGE_SELF).ru_maxrss / 1000,
time.clock()))
items = self.handle_commit(c_o)
parent_commits = tuple(self.converted[v] for v in c_o.parents)
# for the singular case of init / root / the genesis
if len(parent_commits) == 0:
parent_commits = {c_o}
self.repo.head.reference = c_o
self.repo.head.reset(index=True, working_tree=True)
idx = IndexFile(self.repo)
idx.add(items)
idx.write_tree()
com_msg = [c_o.message]
com_msg.extend('\n'.join(self.convert_errors))
self.convert_errors = [] # todo: rearchitect - too easy to forget
com_msg.append('\n[gitreformat yapf-ify (github/ghtdak) on {}]'.format(
time.strftime('%c')))
com_msg.append('\n[from commit: {}]'.format(c_o.hexsha))
c_n = idx.commit(''.join(com_msg),
parent_commits=parent_commits,
author=c_o.author,
author_date=time_convert(c_o.authored_date,
c_o.author_tz_offset),
committer=c_o.committer,
commit_date=time_convert(c_o.committed_date,
c_o.committer_tz_offset))
self.repo.head.reference = c_n
self.repo.head.reset(index=True, working_tree=True)
self.verify_paths(c_o.tree, c_n.tree)
self.converted[c_o] = c_n
return c_n
def git_commit(index: IndexFile, message: str) -> Commit:
"""
Commit all staged files in the current repo.
:param IndexFile index:
the current index of the current branch of the current repo
:param message:
the commit message to include with the files to be committed
:return:
a record of the commit (:py:class:`~git.objects.commit.Commit`), itself
"""
return index.commit(message, author=get_author())
def git_add(index: IndexFile, files) -> List[BaseIndexEntry]:
"""
`git add` all of the given files within the current repo
:param IndexFile index:
the current index of the current branch of the current repo
:param files:
the files to be `git-add`'ed
:return:
the entries just added
"""
return index.add(files)
def time_warp(self, c_o):
"""
History rewriting occurs here. We read everything from the original
commit, reformat the python, and checkin, mirroring the original
commit history.
:param c_o: Commit object representing "before"
:return: None
"""
log.info('warping: {} | {} | {:f} MB | {}s'.format(
time_convert(c_o.authored_date, c_o.author_tz_offset),
c_o.summary,
resource.getrusage(resource.RUSAGE_SELF).ru_maxrss / 1000,
time.clock()))
items = self.handle_commit(c_o)
parent_commits = tuple(self.converted[v] for v in c_o.parents)
# for the singular case of init / root / the genesis
if len(parent_commits) == 0:
parent_commits = {c_o}
self.repo.head.reference = c_o
self.repo.head.reset(index=True, working_tree=True)
idx = IndexFile(self.repo)
idx.add(items)
idx.write_tree()
com_msg = [c_o.message]
com_msg.extend('\n'.join(self.convert_errors))
self.convert_errors = [] # todo: rearchitect - too easy to forget
com_msg.append(
'\n[gitreformat yapf-ify (github/ghtdak) on {}]'.format(
time.strftime('%c')))
com_msg.append('\n[from commit: {}]'.format(c_o.hexsha))
c_n = idx.commit(
''.join(com_msg),
parent_commits=parent_commits,
author=c_o.author,
author_date=time_convert(c_o.authored_date,
c_o.author_tz_offset),
committer=c_o.committer,
commit_date=time_convert(c_o.committed_date,
c_o.committer_tz_offset))
self.repo.head.reference = c_n
self.repo.head.reset(index=True, working_tree=True)
self.verify_paths(c_o.tree, c_n.tree)
self.converted[c_o] = c_n
return c_n
def build_status(self, revision='HEAD'):
"""Return files from the revision grouped by their status."""
index = self.client.git.index if revision == 'HEAD' \
else IndexFile.from_tree(self.client.git, revision)
current_files = set()
for filepath, _ in index.entries.keys():
if not filepath.startswith('.renku') and \
filepath not in {'.gitignore', '.gitattributes'}:
self.add_file(filepath, revision=revision)
current_files.add(filepath)
# Prepare status info for each file.
self._need_update()
graph_files = sorted(
((commit, filepath)
for (commit, filepath) in self.G if filepath in current_files),
key=itemgetter(1))
status = {'up-to-date': {}, 'outdated': {}, 'multiple-versions': {}}
for filepath, keys in groupby(graph_files, itemgetter(1)):
keys = list(keys)
if len(keys) > 1:
status['multiple-versions'][filepath] = keys
nodes = [self.G.nodes[key] for key in keys]
# Any latest version of a file needs an update.
is_outdated = any(
len(node['_need_update']) > 1 for node in nodes
if node['latest'] is None)
if is_outdated:
updates = list(self.G.nodes[key]['_need_update']
for key in keys)
status['outdated'][filepath] = updates
else:
status['up-to-date'][filepath] = keys[0][0]
return status
def dependencies(self, revision='HEAD', paths=None):
"""Return dependencies from a revision or paths."""
if paths:
return {
Usage.from_revision(self.client, path=path, revision=revision)
for path in paths
}
if revision == 'HEAD':
index = self.client.git.index
else:
from git import IndexFile
index = IndexFile.from_tree(self.client.git, revision)
return {
Usage.from_revision(
client=self.client,
path=path,
revision=revision,
)
for path, _ in index.entries.keys()
}
def add_data(index: git.IndexFile, path: str, data: bytes):
fullpath = os.path.join(os.path.dirname(index.repo.git_dir), path)
pathlib.Path(fullpath).parent.mkdir(parents=True, exist_ok=True)
with open(fullpath, 'wb') as fp:
fp.write(data)
index.add([fullpath])
def squash_blame(preserve_committers):
"""Squash history while preserving blame"""
repo = Repo()
head_commit = repo.head.commit
mutated_blobs = []
author_blob_lists = defaultdict(list)
with click.progressbar(head_commit.tree.traverse()) as git_objects:
for blob in git_objects:
if isinstance(blob, Blob):
blob_bytes = blob.data_stream.read()
mutated_blob_bytes = blob_bytes.replace(b"\n", b"\r\n")
mutated_blob = make_blob(repo, mutated_blob_bytes, blob.mode,
blob.path)
mutated_blobs.append(mutated_blob)
author_lines = defaultdict(set)
for blame in repo.blame_incremental(head_commit, blob.path):
author_lines[blame.commit.author].update(blame.linenos)
for author, authors_line_numbers in author_lines.items():
authors_fixed_blob_bytes_lines = []
for line_number, line in enumerate(
mutated_blob_bytes.splitlines(keepends=True),
start=1):
if line_number in authors_line_numbers:
line = line.replace(b"\r\n", b"\n")
authors_fixed_blob_bytes_lines.append(line)
authors_fixed_blob_bytes = b"".join(
authors_fixed_blob_bytes_lines)
authors_fixed_blob = make_blob(repo,
authors_fixed_blob_bytes,
blob.mode, blob.path)
author_blob_lists[author].append(authors_fixed_blob)
base_index = IndexFile.from_tree(repo, head_commit)
base_index.add(mutated_blobs, write=False)
new_root = base_index.commit(message="Initial commit",
parent_commits=[],
head=False)
author_commits = []
for author, author_blobs in author_blob_lists.items():
index = IndexFile.from_tree(repo, new_root)
index.add(author_blobs, write=False)
author_commits.append(
index.commit(
message=f"Squash while preserving git-blame for {author.name}",
parent_commits=[new_root],
head=False,
author=author,
committer=author if preserve_committers else None,
))
final_index = IndexFile.from_tree(repo, head_commit)
final_commit = final_index.commit(
message="Merge commits that preserve git-blame",
parent_commits=author_commits,
head=False,
)
print(final_commit.hexsha)
def remove_data(self, index: git.IndexFile, path: str):
fullpath = os.path.join(os.path.dirname(index.repo.git_dir), path)
pathlib.Path(fullpath).parent.mkdir(parents=True, exist_ok=True)
index.remove([fullpath], working_tree=True)
def test_index_file_diffing(self, rw_repo):
# default Index instance points to our index
index = IndexFile(rw_repo)
assert index.path is not None
assert len(index.entries)
# write the file back
index.write()
# could sha it, or check stats
# test diff
# resetting the head will leave the index in a different state, and the
# diff will yield a few changes
cur_head_commit = rw_repo.head.reference.commit
rw_repo.head.reset('HEAD~6', index=True, working_tree=False)
# diff against same index is 0
diff = index.diff()
self.assertEqual(len(diff), 0)
# against HEAD as string, must be the same as it matches index
diff = index.diff('HEAD')
self.assertEqual(len(diff), 0)
# against previous head, there must be a difference
diff = index.diff(cur_head_commit)
assert len(diff)
# we reverse the result
adiff = index.diff(str(cur_head_commit), R=True)
odiff = index.diff(cur_head_commit,
R=False) # now its not reversed anymore
assert adiff != odiff
self.assertEqual(odiff, diff) # both unreversed diffs against HEAD
# against working copy - its still at cur_commit
wdiff = index.diff(None)
assert wdiff != adiff
assert wdiff != odiff
# against something unusual
self.failUnlessRaises(ValueError, index.diff, int)
# adjust the index to match an old revision
cur_branch = rw_repo.active_branch
cur_commit = cur_branch.commit
rev_head_parent = 'HEAD~1'
assert index.reset(rev_head_parent) is index
self.assertEqual(cur_branch, rw_repo.active_branch)
self.assertEqual(cur_commit, rw_repo.head.commit)
# there must be differences towards the working tree which is in the 'future'
assert index.diff(None)
# reset the working copy as well to current head,to pull 'back' as well
new_data = b"will be reverted"
file_path = osp.join(rw_repo.working_tree_dir, "CHANGES")
with open(file_path, "wb") as fp:
fp.write(new_data)
index.reset(rev_head_parent, working_tree=True)
assert not index.diff(None)
self.assertEqual(cur_branch, rw_repo.active_branch)
self.assertEqual(cur_commit, rw_repo.head.commit)
with open(file_path, 'rb') as fp:
assert fp.read() != new_data
# test full checkout
test_file = osp.join(rw_repo.working_tree_dir, "CHANGES")
with open(test_file, 'ab') as fd:
fd.write(b"some data")
rval = index.checkout(None, force=True, fprogress=self._fprogress)
assert 'CHANGES' in list(rval)
self._assert_fprogress([None])
assert osp.isfile(test_file)
os.remove(test_file)
rval = index.checkout(None, force=False, fprogress=self._fprogress)
assert 'CHANGES' in list(rval)
self._assert_fprogress([None])
assert osp.isfile(test_file)
# individual file
os.remove(test_file)
rval = index.checkout(test_file, fprogress=self._fprogress)
self.assertEqual(list(rval)[0], 'CHANGES')
self._assert_fprogress([test_file])
assert osp.exists(test_file)
# checking out non-existing file throws
self.failUnlessRaises(CheckoutError, index.checkout,
"doesnt_exist_ever.txt.that")
self.failUnlessRaises(CheckoutError,
index.checkout,
paths=["doesnt/exist"])
# checkout file with modifications
append_data = b"hello"
with open(test_file, "ab") as fp:
fp.write(append_data)
try:
index.checkout(test_file)
except CheckoutError as e:
self.assertEqual(len(e.failed_files), 1)
self.assertEqual(e.failed_files[0], osp.basename(test_file))
self.assertEqual(len(e.failed_files), len(e.failed_reasons))
self.assertIsInstance(e.failed_reasons[0], str)
self.assertEqual(len(e.valid_files), 0)
with open(test_file, 'rb') as fd:
s = fd.read()
self.assertTrue(s.endswith(append_data), s)
else:
raise AssertionError("Exception CheckoutError not thrown")
# if we force it it should work
index.checkout(test_file, force=True)
assert not open(test_file, 'rb').read().endswith(append_data)
# checkout directory
rmtree(osp.join(rw_repo.working_tree_dir, "lib"))
rval = index.checkout('lib')
assert len(list(rval)) > 1
def test_references_and_objects(self, rw_dir):
# [1-test_references_and_objects]
import git
repo = git.Repo.clone_from(self._small_repo_url(), os.path.join(rw_dir, 'repo'), branch='master')
heads = repo.heads
master = heads.master # lists can be accessed by name for convenience
master.commit # the commit pointed to by head called master
master.rename('new_name') # rename heads
master.rename('master')
# ![1-test_references_and_objects]
# [2-test_references_and_objects]
tags = repo.tags
tagref = tags[0]
tagref.tag # tags may have tag objects carrying additional information
tagref.commit # but they always point to commits
repo.delete_tag(tagref) # delete or
repo.create_tag("my_tag") # create tags using the repo for convenience
# ![2-test_references_and_objects]
# [3-test_references_and_objects]
head = repo.head # the head points to the active branch/ref
master = head.reference # retrieve the reference the head points to
master.commit # from here you use it as any other reference
# ![3-test_references_and_objects]
# [4-test_references_and_objects]
log = master.log()
log[0] # first (i.e. oldest) reflog entry
log[-1] # last (i.e. most recent) reflog entry
# ![4-test_references_and_objects]
# [5-test_references_and_objects]
new_branch = repo.create_head('new') # create a new one
new_branch.commit = 'HEAD~10' # set branch to another commit without changing index or working trees
repo.delete_head(new_branch) # delete an existing head - only works if it is not checked out
# ![5-test_references_and_objects]
# [6-test_references_and_objects]
new_tag = repo.create_tag('my_new_tag', message='my message')
# You cannot change the commit a tag points to. Tags need to be re-created
self.failUnlessRaises(AttributeError, setattr, new_tag, 'commit', repo.commit('HEAD~1'))
repo.delete_tag(new_tag)
# ![6-test_references_and_objects]
# [7-test_references_and_objects]
new_branch = repo.create_head('another-branch')
repo.head.reference = new_branch
# ![7-test_references_and_objects]
# [8-test_references_and_objects]
hc = repo.head.commit
hct = hc.tree
hc != hct
hc != repo.tags[0]
hc == repo.head.reference.commit
# ![8-test_references_and_objects]
# [9-test_references_and_objects]
assert hct.type == 'tree' # preset string type, being a class attribute
assert hct.size > 0 # size in bytes
assert len(hct.hexsha) == 40
assert len(hct.binsha) == 20
# ![9-test_references_and_objects]
# [10-test_references_and_objects]
assert hct.path == '' # root tree has no path
assert hct.trees[0].path != '' # the first contained item has one though
assert hct.mode == 0o40000 # trees have the mode of a linux directory
assert hct.blobs[0].mode == 0o100644 # blobs have a specific mode though comparable to a standard linux fs
# ![10-test_references_and_objects]
# [11-test_references_and_objects]
hct.blobs[0].data_stream.read() # stream object to read data from
hct.blobs[0].stream_data(open(os.path.join(rw_dir, 'blob_data'), 'wb')) # write data to given stream
# ![11-test_references_and_objects]
# [12-test_references_and_objects]
repo.commit('master')
repo.commit('v0.8.1')
repo.commit('HEAD~10')
# ![12-test_references_and_objects]
# [13-test_references_and_objects]
fifty_first_commits = list(repo.iter_commits('master', max_count=50))
assert len(fifty_first_commits) == 50
# this will return commits 21-30 from the commit list as traversed backwards master
ten_commits_past_twenty = list(repo.iter_commits('master', max_count=10, skip=20))
assert len(ten_commits_past_twenty) == 10
assert fifty_first_commits[20:30] == ten_commits_past_twenty
# ![13-test_references_and_objects]
# [14-test_references_and_objects]
headcommit = repo.head.commit
assert len(headcommit.hexsha) == 40
assert len(headcommit.parents) > 0
assert headcommit.tree.type == 'tree'
assert headcommit.author.name == 'Sebastian Thiel'
assert isinstance(headcommit.authored_date, int)
assert headcommit.committer.name == 'Sebastian Thiel'
assert isinstance(headcommit.committed_date, int)
assert headcommit.message != ''
# ![14-test_references_and_objects]
# [15-test_references_and_objects]
import time
time.asctime(time.gmtime(headcommit.committed_date))
time.strftime("%a, %d %b %Y %H:%M", time.gmtime(headcommit.committed_date))
# ![15-test_references_and_objects]
# [16-test_references_and_objects]
assert headcommit.parents[0].parents[0].parents[0] == repo.commit('master^^^')
# ![16-test_references_and_objects]
# [17-test_references_and_objects]
tree = repo.heads.master.commit.tree
assert len(tree.hexsha) == 40
# ![17-test_references_and_objects]
# [18-test_references_and_objects]
assert len(tree.trees) > 0 # trees are subdirectories
assert len(tree.blobs) > 0 # blobs are files
assert len(tree.blobs) + len(tree.trees) == len(tree)
# ![18-test_references_and_objects]
# [19-test_references_and_objects]
assert tree['smmap'] == tree / 'smmap' # access by index and by sub-path
for entry in tree: # intuitive iteration of tree members
print(entry)
blob = tree.trees[0].blobs[0] # let's get a blob in a sub-tree
assert blob.name
assert len(blob.path) < len(blob.abspath)
assert tree.trees[0].name + '/' + blob.name == blob.path # this is how the relative blob path is generated
assert tree[blob.path] == blob # you can use paths like 'dir/file' in tree[...]
# ![19-test_references_and_objects]
# [20-test_references_and_objects]
assert tree / 'smmap' == tree['smmap']
assert tree / blob.path == tree[blob.path]
# ![20-test_references_and_objects]
# [21-test_references_and_objects]
# This example shows the various types of allowed ref-specs
assert repo.tree() == repo.head.commit.tree
past = repo.commit('HEAD~5')
assert repo.tree(past) == repo.tree(past.hexsha)
assert repo.tree('v0.8.1').type == 'tree' # yes, you can provide any refspec - works everywhere
# ![21-test_references_and_objects]
# [22-test_references_and_objects]
assert len(tree) < len(list(tree.traverse()))
# ![22-test_references_and_objects]
# [23-test_references_and_objects]
index = repo.index
# The index contains all blobs in a flat list
assert len(list(index.iter_blobs())) == len([o for o in repo.head.commit.tree.traverse() if o.type == 'blob'])
# Access blob objects
for (path, stage), entry in index.entries.items():
pass
new_file_path = os.path.join(repo.working_tree_dir, 'new-file-name')
open(new_file_path, 'w').close()
index.add([new_file_path]) # add a new file to the index
index.remove(['LICENSE']) # remove an existing one
assert os.path.isfile(os.path.join(repo.working_tree_dir, 'LICENSE')) # working tree is untouched
assert index.commit("my commit message").type == 'commit' # commit changed index
repo.active_branch.commit = repo.commit('HEAD~1') # forget last commit
from git import Actor
author = Actor("An author", "*****@*****.**")
committer = Actor("A committer", "*****@*****.**")
# commit by commit message and author and committer
index.commit("my commit message", author=author, committer=committer)
# ![23-test_references_and_objects]
# [24-test_references_and_objects]
from git import IndexFile
# loads a tree into a temporary index, which exists just in memory
IndexFile.from_tree(repo, 'HEAD~1')
# merge two trees three-way into memory
merge_index = IndexFile.from_tree(repo, 'HEAD~10', 'HEAD', repo.merge_base('HEAD~10', 'HEAD'))
# and persist it
merge_index.write(os.path.join(rw_dir, 'merged_index'))
# ![24-test_references_and_objects]
# [25-test_references_and_objects]
empty_repo = git.Repo.init(os.path.join(rw_dir, 'empty'))
origin = empty_repo.create_remote('origin', repo.remotes.origin.url)
assert origin.exists()
assert origin == empty_repo.remotes.origin == empty_repo.remotes['origin']
origin.fetch() # assure we actually have data. fetch() returns useful information
# Setup a local tracking branch of a remote branch
empty_repo.create_head('master', origin.refs.master).set_tracking_branch(origin.refs.master)
origin.rename('new_origin') # rename remotes
# push and pull behaves similarly to `git push|pull`
origin.pull()
origin.push()
# assert not empty_repo.delete_remote(origin).exists() # create and delete remotes
# ![25-test_references_and_objects]
# [26-test_references_and_objects]
assert origin.url == repo.remotes.origin.url
cw = origin.config_writer
cw.set("pushurl", "other_url")
cw.release()
# Please note that in python 2, writing origin.config_writer.set(...) is totally safe.
# In py3 __del__ calls can be delayed, thus not writing changes in time.
# ![26-test_references_and_objects]
# [27-test_references_and_objects]
hcommit = repo.head.commit
hcommit.diff() # diff tree against index
hcommit.diff('HEAD~1') # diff tree against previous tree
hcommit.diff(None) # diff tree against working tree
index = repo.index
index.diff() # diff index against itself yielding empty diff
index.diff(None) # diff index against working copy
index.diff('HEAD') # diff index against current HEAD tree
# ![27-test_references_and_objects]
# [28-test_references_and_objects]
# Traverse added Diff objects only
for diff_added in hcommit.diff('HEAD~1').iter_change_type('A'):
print(diff_added)
# ![28-test_references_and_objects]
# [29-test_references_and_objects]
# Reset our working tree 10 commits into the past
past_branch = repo.create_head('past_branch', 'HEAD~10')
repo.head.reference = past_branch
assert not repo.head.is_detached
# reset the index and working tree to match the pointed-to commit
repo.head.reset(index=True, working_tree=True)
# To detach your head, you have to point to a commit directy
repo.head.reference = repo.commit('HEAD~5')
assert repo.head.is_detached
# now our head points 15 commits into the past, whereas the working tree
# and index are 10 commits in the past
# ![29-test_references_and_objects]
# [30-test_references_and_objects]
# checkout the branch using git-checkout. It will fail as the working tree appears dirty
self.failUnlessRaises(git.GitCommandError, repo.heads.master.checkout)
repo.heads.past_branch.checkout()
# ![30-test_references_and_objects]
# [31-test_references_and_objects]
git = repo.git
git.checkout('HEAD', b="my_new_branch") # create a new branch
git.branch('another-new-one')
git.branch('-D', 'another-new-one') # pass strings for full control over argument order
git.for_each_ref() # '-' becomes '_' when calling it
# ![31-test_references_and_objects]
# [32-test_references_and_objects]
ssh_executable = os.path.join(rw_dir, 'my_ssh_executable.sh')
with repo.git.custom_environment(GIT_SSH=ssh_executable):
# Note that we don't actually make the call here, as our test-setup doesn't permit it to
# succeed.
# It will in your case :)
repo.remotes.origin.fetch
def test_index_file_diffing(self, rw_repo):
# default Index instance points to our index
index = IndexFile(rw_repo)
assert index.path is not None
assert len(index.entries)
# write the file back
index.write()
# could sha it, or check stats
# test diff
# resetting the head will leave the index in a different state, and the
# diff will yield a few changes
cur_head_commit = rw_repo.head.reference.commit
rw_repo.head.reset('HEAD~6', index=True, working_tree=False)
# diff against same index is 0
diff = index.diff()
self.assertEqual(len(diff), 0)
# against HEAD as string, must be the same as it matches index
diff = index.diff('HEAD')
self.assertEqual(len(diff), 0)
# against previous head, there must be a difference
diff = index.diff(cur_head_commit)
assert len(diff)
# we reverse the result
adiff = index.diff(str(cur_head_commit), R=True)
odiff = index.diff(cur_head_commit, R=False) # now its not reversed anymore
assert adiff != odiff
self.assertEqual(odiff, diff) # both unreversed diffs against HEAD
# against working copy - its still at cur_commit
wdiff = index.diff(None)
assert wdiff != adiff
assert wdiff != odiff
# against something unusual
self.failUnlessRaises(ValueError, index.diff, int)
# adjust the index to match an old revision
cur_branch = rw_repo.active_branch
cur_commit = cur_branch.commit
rev_head_parent = 'HEAD~1'
assert index.reset(rev_head_parent) is index
self.assertEqual(cur_branch, rw_repo.active_branch)
self.assertEqual(cur_commit, rw_repo.head.commit)
# there must be differences towards the working tree which is in the 'future'
assert index.diff(None)
# reset the working copy as well to current head,to pull 'back' as well
new_data = b"will be reverted"
file_path = osp.join(rw_repo.working_tree_dir, "CHANGES")
with open(file_path, "wb") as fp:
fp.write(new_data)
index.reset(rev_head_parent, working_tree=True)
assert not index.diff(None)
self.assertEqual(cur_branch, rw_repo.active_branch)
self.assertEqual(cur_commit, rw_repo.head.commit)
with open(file_path, 'rb') as fp:
assert fp.read() != new_data
# test full checkout
test_file = osp.join(rw_repo.working_tree_dir, "CHANGES")
with open(test_file, 'ab') as fd:
fd.write(b"some data")
rval = index.checkout(None, force=True, fprogress=self._fprogress)
assert 'CHANGES' in list(rval)
self._assert_fprogress([None])
assert osp.isfile(test_file)
os.remove(test_file)
rval = index.checkout(None, force=False, fprogress=self._fprogress)
assert 'CHANGES' in list(rval)
self._assert_fprogress([None])
assert osp.isfile(test_file)
# individual file
os.remove(test_file)
rval = index.checkout(test_file, fprogress=self._fprogress)
self.assertEqual(list(rval)[0], 'CHANGES')
self._assert_fprogress([test_file])
assert osp.exists(test_file)
# checking out non-existing file throws
self.failUnlessRaises(CheckoutError, index.checkout, "doesnt_exist_ever.txt.that")
self.failUnlessRaises(CheckoutError, index.checkout, paths=["doesnt/exist"])
# checkout file with modifications
append_data = b"hello"
with open(test_file, "ab") as fp:
fp.write(append_data)
try:
index.checkout(test_file)
except CheckoutError as e:
self.assertEqual(len(e.failed_files), 1)
self.assertEqual(e.failed_files[0], osp.basename(test_file))
self.assertEqual(len(e.failed_files), len(e.failed_reasons))
self.assertIsInstance(e.failed_reasons[0], string_types)
self.assertEqual(len(e.valid_files), 0)
with open(test_file, 'rb') as fd:
s = fd.read()
self.assertTrue(s.endswith(append_data), s)
else:
raise AssertionError("Exception CheckoutError not thrown")
# if we force it it should work
index.checkout(test_file, force=True)
assert not open(test_file, 'rb').read().endswith(append_data)
# checkout directory
rmtree(osp.join(rw_repo.working_tree_dir, "lib"))
rval = index.checkout('lib')
assert len(list(rval)) > 1
