def test_create(probe, path):
# only as a probe whether this FS is a crippled one
ar = AnnexRepo(probe, create=True)
ds = Dataset(path)
ds.create(
description="funny",
# custom git init option
initopts=dict(shared='world') if not ar.is_managed_branch() else None)
ok_(ds.is_installed())
assert_repo_status(ds.path, annex=True)
# check default backend
(ds.pathobj / "f1").write_text("1")
ds.save()
eq_(ds.repo.get_file_backend(["f1"]), ['MD5E'])
if not ar.is_managed_branch():
eq_(ds.config.get("core.sharedrepository"), '2')
# check description in `info`
cmlout = ds.repo.call_annex(['info'])
assert_in('funny [here]', cmlout)
# check dataset ID
eq_(ds.config.get_value('datalad.dataset', 'id'),
ds.id)
def test_surprise_subds(path):
# https://github.com/datalad/datalad/issues/3139
ds = create(path, force=True)
# a lonely repo without any commit
somerepo = AnnexRepo(path=op.join(path, 'd1', 'subrepo'), create=True)
# a proper subdataset
subds = create(op.join(path, 'd2', 'subds'), force=True)
# If subrepo is an adjusted branch, it would have a commit, making most of
# this test irrelevant because it is about the unborn branch edge case.
adjusted = somerepo.is_managed_branch()
# This edge case goes away with Git v2.22.0.
fixed_git = external_versions['cmd:git'] >= '2.22.0'
# save non-recursive
res = ds.save(recursive=False, on_failure='ignore')
if not adjusted and fixed_git:
# We get an appropriate error about no commit being checked out.
assert_in_results(res, action='add_submodule', status='error')
# the content of both subds and subrepo are not added to their
# respective parent as no --recursive was given
assert_repo_status(subds.path, untracked=['subfile'])
assert_repo_status(somerepo.path, untracked=['subfile'])
if adjusted or fixed_git:
if adjusted:
# adjusted branch: #datalad/3178 (that would have a commit)
modified = [subds.repo.pathobj, somerepo.pathobj]
untracked = []
else:
# Newer Git versions refuse to add a sub-repository with no commits
# checked out.
modified = [subds.repo.pathobj]
untracked = ['d1']
assert_repo_status(ds.path, modified=modified, untracked=untracked)
assert_not_in(ds.repo.pathobj / 'd1' / 'subrepo' / 'subfile',
ds.repo.get_content_info())
else:
# however, while the subdataset is added (and reported as modified
# because it content is still untracked) the subrepo
# cannot be added (it has no commit)
# worse: its untracked file add been added to the superdataset
assert_repo_status(ds.path, modified=['d2/subds'])
assert_in(ds.repo.pathobj / 'd1' / 'subrepo' / 'subfile',
ds.repo.get_content_info())
# with proper subdatasets, all evil is gone
assert_not_in(ds.repo.pathobj / 'd2' / 'subds' / 'subfile',
ds.repo.get_content_info())
def test_create(probe, path):
# only as a probe whether this FS is a crippled one
ar = AnnexRepo(probe, create=True)
ds = Dataset(path)
ds.create(
description="funny",
# custom git init option
initopts=dict(shared='world') if not ar.is_managed_branch() else None)
ok_(ds.is_installed())
assert_repo_status(ds.path, annex=True)
# check default backend
eq_(ds.config.get("annex.backends"), 'MD5E')
if not ar.is_managed_branch():
eq_(ds.config.get("core.sharedrepository"), '2')
runner = Runner()
# check description in `info`
cmd = ['git', 'annex', 'info']
cmlout = runner.run(cmd, cwd=path)
assert_in('funny [here]', cmlout[0])
# check datset ID
eq_(ds.config.get_value('datalad.dataset', 'id'),
ds.id)
def test_save_amend(dspath):
dspath = Path(dspath)
file_in_super = dspath / 'somefile'
file_in_sub = dspath / 'subds' / 'file_in_sub'
# test on a hierarchy including a plain git repo:
ds = Dataset(dspath).create(force=True, no_annex=True)
subds = ds.create('subds', force=True)
ds.save(recursive=True)
assert_repo_status(ds.repo)
# recursive and amend are mutually exclusive:
for d in (ds, subds):
assert_raises(ValueError, d.save, recursive=True, amend=True)
# in an annex repo the branch we are interested in might not be the active
# branch (adjusted):
sub_branch = subds.repo.get_corresponding_branch()
# amend in subdataset w/ new message; otherwise empty amendment:
last_sha = subds.repo.get_hexsha(sub_branch)
subds.save(message="new message in sub", amend=True)
# we did in fact commit something:
neq_(last_sha, subds.repo.get_hexsha(sub_branch))
# repo is clean:
assert_repo_status(subds.repo)
# message is correct:
eq_(
subds.repo.format_commit("%B", sub_branch).strip(),
"new message in sub")
# actually replaced the previous commit:
assert_not_in(last_sha, subds.repo.get_branch_commits_(sub_branch))
# amend modifications in subdataset w/o new message
if not subds.repo.is_managed_branch():
subds.unlock('file_in_sub')
file_in_sub.write_text("modified again")
last_sha = subds.repo.get_hexsha(sub_branch)
subds.save(amend=True)
neq_(last_sha, subds.repo.get_hexsha(sub_branch))
assert_repo_status(subds.repo)
# message unchanged:
eq_(
subds.repo.format_commit("%B", sub_branch).strip(),
"new message in sub")
# actually replaced the previous commit:
assert_not_in(last_sha, subds.repo.get_branch_commits_(sub_branch))
# save --amend with nothing to amend with:
res = subds.save(amend=True)
assert_result_count(res, 1)
assert_result_count(res, 1, status='notneeded', action='save')
# amend in superdataset w/ new message; otherwise empty amendment:
last_sha = ds.repo.get_hexsha()
ds.save(message="new message in super", amend=True)
neq_(last_sha, ds.repo.get_hexsha())
assert_repo_status(subds.repo)
eq_(ds.repo.format_commit("%B").strip(), "new message in super")
assert_not_in(last_sha, ds.repo.get_branch_commits_())
# amend modifications in superdataset w/o new message
file_in_super.write_text("changed content")
if not subds.repo.is_managed_branch():
subds.unlock('file_in_sub')
file_in_sub.write_text("modified once again")
last_sha = ds.repo.get_hexsha()
last_sha_sub = subds.repo.get_hexsha(sub_branch)
ds.save(amend=True)
neq_(last_sha, ds.repo.get_hexsha())
eq_(ds.repo.format_commit("%B").strip(), "new message in super")
assert_not_in(last_sha, ds.repo.get_branch_commits_())
# we didn't mess with the subds:
assert_repo_status(ds.repo, modified=["subds"])
eq_(last_sha_sub, subds.repo.get_hexsha(sub_branch))
eq_(
subds.repo.format_commit("%B", sub_branch).strip(),
"new message in sub")
# save --amend with nothing to amend with:
last_sha = ds.repo.get_hexsha()
res = ds.save(amend=True)
assert_result_count(res, 1)
assert_result_count(res, 1, status='notneeded', action='save')
eq_(last_sha, ds.repo.get_hexsha())
# we didn't mess with the subds:
assert_repo_status(ds.repo, modified=["subds"])
eq_(last_sha_sub, subds.repo.get_hexsha(sub_branch))
eq_(
subds.repo.format_commit("%B", sub_branch).strip(),
"new message in sub")
# amend with different identity:
orig_author = ds.repo.format_commit("%an")
orig_email = ds.repo.format_commit("%ae")
orig_date = ds.repo.format_commit("%ad")
orig_committer = ds.repo.format_commit("%cn")
orig_committer_mail = ds.repo.format_commit("%ce")
eq_(orig_author, orig_committer)
eq_(orig_email, orig_committer_mail)
with patch.dict(
'os.environ', {
'GIT_COMMITTER_NAME': 'Hopefully Different',
'GIT_COMMITTER_EMAIL': '*****@*****.**'
}):
ds.config.reload(force=True)
ds.save(amend=True, message="amend with hope")
# author was kept:
eq_(orig_author, ds.repo.format_commit("%an"))
eq_(orig_email, ds.repo.format_commit("%ae"))
eq_(orig_date, ds.repo.format_commit("%ad"))
# committer changed:
eq_(ds.repo.format_commit("%cn"), "Hopefully Different")
eq_(ds.repo.format_commit("%ce"), "*****@*****.**")
# corner case: amend empty commit with no parent:
rmtree(str(dspath))
# When adjusted branch is enforced by git-annex detecting a crippled FS,
# git-annex produces an empty commit before switching to adjusted branch:
# "commit before entering adjusted branch"
# The commit by `create` would be the second one already.
# Therefore go with plain annex repo and create an (empty) commit only when
# not on adjusted branch:
repo = AnnexRepo(dspath, create=True)
if not repo.is_managed_branch():
repo.commit(msg="initial", options=['--allow-empty'])
ds = Dataset(dspath)
branch = ds.repo.get_corresponding_branch() or ds.repo.get_active_branch()
# test pointless if we start with more than one commit
eq_(len(list(ds.repo.get_branch_commits_(branch))),
1,
msg="More than on commit '{}': {}".format(
branch, ds.repo.call_git(['log', branch])))
last_sha = ds.repo.get_hexsha(branch)
ds.save(message="new initial commit", amend=True)
assert_repo_status(ds.repo)
eq_(len(list(ds.repo.get_branch_commits_(branch))),
1,
msg="More than on commit '{}': {}".format(
branch, ds.repo.call_git(['log', branch])))
assert_not_in(last_sha, ds.repo.get_branch_commits_(branch))
eq_(ds.repo.format_commit("%B", branch).strip(), "new initial commit")
