def test_something(path=None, new_home=None):
# will refuse to work on dataset without a dataset
assert_raises(ValueError, ConfigManager, source='branch')
# now read the example config
cfg = ConfigManager(GitRepo(opj(path, 'ds'), create=True), source='branch')
assert_equal(len(cfg), 5)
assert_in('something.user', cfg)
# multi-value
assert_equal(len(cfg['something.user']), 2)
assert_equal(cfg['something.user'],
('name=Jane Doe', '[email protected]'))
assert_true(cfg.has_section('something'))
assert_false(cfg.has_section('somethingelse'))
assert_equal(sorted(cfg.sections()),
[u'onemore.complicated の beast with.dot', 'something'])
assert_true(cfg.has_option('something', 'user'))
assert_false(cfg.has_option('something', 'us?er'))
assert_false(cfg.has_option('some?thing', 'user'))
assert_equal(sorted(cfg.options('something')),
['empty', 'myint', 'novalue', 'user'])
assert_equal(cfg.options(u'onemore.complicated の beast with.dot'),
['findme'])
assert_equal(sorted(cfg.items()),
[(u'onemore.complicated の beast with.dot.findme', '5.0'),
('something.empty', ''), ('something.myint', '3'),
('something.novalue', None),
('something.user',
('name=Jane Doe', '[email protected]'))])
assert_equal(sorted(cfg.items('something')),
[('something.empty', ''), ('something.myint', '3'),
('something.novalue', None),
('something.user',
('name=Jane Doe', '[email protected]'))])
# by default get last value only
assert_equal(cfg.get('something.user'), '[email protected]')
# but can get all values
assert_equal(cfg.get('something.user', get_all=True),
('name=Jane Doe', '[email protected]'))
assert_raises(KeyError, cfg.__getitem__, 'somedthing.user')
assert_equal(
cfg.getfloat(u'onemore.complicated の beast with.dot', 'findme'), 5.0)
assert_equal(cfg.getint('something', 'myint'), 3)
assert_equal(cfg.getbool('something', 'myint'), True)
# git demands a key without value at all to be used as a flag, thus True
assert_equal(cfg.getbool('something', 'novalue'), True)
assert_equal(cfg.get('something.novalue'), None)
# empty value is False
assert_equal(cfg.getbool('something', 'empty'), False)
assert_equal(cfg.get('something.empty'), '')
assert_equal(cfg.getbool('doesnot', 'exist', default=True), True)
assert_raises(TypeError, cfg.getbool, 'something', 'user')
# gitpython-style access
assert_equal(cfg.get('something.myint'),
cfg.get_value('something', 'myint'))
assert_equal(cfg.get_value('doesnot', 'exist', default='oohaaa'), 'oohaaa')
# weird, but that is how it is
assert_raises(KeyError, cfg.get_value, 'doesnot', 'exist', default=None)
# modification follows
cfg.add('something.new', 'の')
assert_equal(cfg.get('something.new'), u'の')
# sections are added on demand
cfg.add('unheard.of', 'fame')
assert_true(cfg.has_section('unheard.of'))
comp = cfg.items('something')
cfg.rename_section('something', 'this')
assert_true(cfg.has_section('this'))
assert_false(cfg.has_section('something'))
# direct comparison would fail, because of section prefix
assert_equal(len(cfg.items('this')), len(comp))
# fail if no such section
with swallow_logs():
assert_raises(CommandError, cfg.rename_section, 'nothere',
'irrelevant')
assert_true(cfg.has_option('this', 'myint'))
cfg.unset('this.myint')
assert_false(cfg.has_option('this', 'myint'))
# batch a changes
cfg.add('mike.wants.to', 'know', reload=False)
assert_false('mike.wants.to' in cfg)
cfg.add('mike.wants.to', 'eat')
assert_true('mike.wants.to' in cfg)
assert_equal(len(cfg['mike.wants.to']), 2)
# set a new one:
cfg.set('mike.should.have', 'known')
assert_in('mike.should.have', cfg)
assert_equal(cfg['mike.should.have'], 'known')
# set an existing one:
cfg.set('mike.should.have', 'known better')
assert_equal(cfg['mike.should.have'], 'known better')
# set, while there are several matching ones already:
cfg.add('mike.should.have', 'a meal')
assert_equal(len(cfg['mike.should.have']), 2)
# raises with force=False
assert_raises(CommandError,
cfg.set,
'mike.should.have',
'a beer',
force=False)
assert_equal(len(cfg['mike.should.have']), 2)
# replaces all matching ones with force=True
cfg.set('mike.should.have', 'a beer', force=True)
assert_equal(cfg['mike.should.have'], 'a beer')
# test deprecated 'where' interface and old 'dataset' (not 'branch') value
# TODO: remove along with the removal of deprecated 'where'
cfg.set('mike.should.have', 'wasknown', where='dataset')
assert_equal(cfg['mike.should.have'], 'wasknown')
assert_equal(cfg.get_from_source('dataset', 'mike.should.have'),
'wasknown')
# fails unknown location
assert_raises(ValueError, cfg.add, 'somesuch', 'shit', scope='umpalumpa')
# very carefully test non-local config
# so carefully that even in case of bad weather Yarik doesn't find some
# lame datalad unittest sections in his precious ~/.gitconfig
# Note: An easier way to test this, would be to just set GIT_CONFIG_GLOBAL
# to point somewhere else. However, this is not supported by git before
# 2.32. Hence, stick with changed HOME in this test, but be sure to unset a
# possible GIT_CONFIG_GLOBAL in addition.
patched_env = os.environ.copy()
patched_env.pop('GIT_CONFIG_GLOBAL', None)
patched_env.update(get_home_envvars(new_home))
with patch.dict('os.environ',
dict(patched_env, DATALAD_SNEAKY_ADDITION='ignore'),
clear=True):
global_gitconfig = opj(new_home, '.gitconfig')
assert (not exists(global_gitconfig))
globalcfg = ConfigManager()
assert_not_in('datalad.unittest.youcan', globalcfg)
assert_in('datalad.sneaky.addition', globalcfg)
cfg.add('datalad.unittest.youcan', 'removeme', scope='global')
assert (exists(global_gitconfig))
# it did not go into the dataset's config!
assert_not_in('datalad.unittest.youcan', cfg)
# does not monitor additions!
globalcfg.reload(force=True)
assert_in('datalad.unittest.youcan', globalcfg)
with swallow_logs():
assert_raises(CommandError,
globalcfg.unset,
'datalad.unittest.youcan',
scope='local')
assert (globalcfg.has_section('datalad.unittest'))
globalcfg.unset('datalad.unittest.youcan', scope='global')
# but after we unset the only value -- that section is no longer listed
assert (not globalcfg.has_section('datalad.unittest'))
assert_not_in('datalad.unittest.youcan', globalcfg)
ok_file_has_content(global_gitconfig, "")
cfg = ConfigManager(Dataset(opj(path, 'ds')),
source='branch',
overrides={'datalad.godgiven': True})
assert_equal(cfg.get('datalad.godgiven'), True)
# setter has no effect
cfg.set('datalad.godgiven', 'false')
assert_equal(cfg.get('datalad.godgiven'), True)
def test_something(path, new_home):
# will refuse to work on dataset without a dataset
assert_raises(ValueError, ConfigManager, source='dataset')
# now read the example config
cfg = ConfigManager(Dataset(opj(path, 'ds')), source='dataset')
assert_equal(len(cfg), 5)
assert_in('something.user', cfg)
# multi-value
assert_equal(len(cfg['something.user']), 2)
assert_equal(cfg['something.user'],
('name=Jane Doe', '[email protected]'))
assert_true(cfg.has_section('something'))
assert_false(cfg.has_section('somethingelse'))
assert_equal(sorted(cfg.sections()),
[u'onemore.complicated の beast with.dot', 'something'])
assert_true(cfg.has_option('something', 'user'))
assert_false(cfg.has_option('something', 'us?er'))
assert_false(cfg.has_option('some?thing', 'user'))
assert_equal(sorted(cfg.options('something')),
['empty', 'myint', 'novalue', 'user'])
assert_equal(cfg.options(u'onemore.complicated の beast with.dot'),
['findme'])
assert_equal(sorted(cfg.items()),
[(u'onemore.complicated の beast with.dot.findme', '5.0'),
('something.empty', ''), ('something.myint', '3'),
('something.novalue', None),
('something.user',
('name=Jane Doe', '[email protected]'))])
assert_equal(sorted(cfg.items('something')),
[('something.empty', ''), ('something.myint', '3'),
('something.novalue', None),
('something.user',
('name=Jane Doe', '[email protected]'))])
# always get all values
assert_equal(cfg.get('something.user'),
('name=Jane Doe', '[email protected]'))
assert_raises(KeyError, cfg.__getitem__, 'somedthing.user')
assert_equal(
cfg.getfloat(u'onemore.complicated の beast with.dot', 'findme'), 5.0)
assert_equal(cfg.getint('something', 'myint'), 3)
assert_equal(cfg.getbool('something', 'myint'), True)
# git demands a key without value at all to be used as a flag, thus True
assert_equal(cfg.getbool('something', 'novalue'), True)
assert_equal(cfg.get('something.novalue'), None)
# empty value is False
assert_equal(cfg.getbool('something', 'empty'), False)
assert_equal(cfg.get('something.empty'), '')
assert_equal(cfg.getbool('doesnot', 'exist', default=True), True)
assert_raises(TypeError, cfg.getbool, 'something', 'user')
# gitpython-style access
assert_equal(cfg.get('something.myint'),
cfg.get_value('something', 'myint'))
assert_equal(cfg.get_value('doesnot', 'exist', default='oohaaa'), 'oohaaa')
# weired, but that is how it is
assert_raises(KeyError, cfg.get_value, 'doesnot', 'exist', default=None)
# modification follows
cfg.add('something.new', 'の')
assert_equal(cfg.get('something.new'), u'の')
# sections are added on demand
cfg.add('unheard.of', 'fame')
assert_true(cfg.has_section('unheard.of'))
comp = cfg.items('something')
cfg.rename_section('something', 'this')
assert_true(cfg.has_section('this'))
assert_false(cfg.has_section('something'))
# direct comparision would fail, because of section prefix
assert_equal(len(cfg.items('this')), len(comp))
# fail if no such section
with swallow_logs():
assert_raises(CommandError, cfg.rename_section, 'nothere',
'irrelevant')
assert_true(cfg.has_option('this', 'myint'))
cfg.unset('this.myint')
assert_false(cfg.has_option('this', 'myint'))
# batch a changes
cfg.add('mike.wants.to', 'know', reload=False)
assert_false('mike.wants.to' in cfg)
cfg.add('mike.wants.to', 'eat')
assert_true('mike.wants.to' in cfg)
assert_equal(len(cfg['mike.wants.to']), 2)
# set a new one:
cfg.set('mike.should.have', 'known')
assert_in('mike.should.have', cfg)
assert_equal(cfg['mike.should.have'], 'known')
# set an existing one:
cfg.set('mike.should.have', 'known better')
assert_equal(cfg['mike.should.have'], 'known better')
# set, while there are several matching ones already:
cfg.add('mike.should.have', 'a meal')
assert_equal(len(cfg['mike.should.have']), 2)
# raises with force=False
assert_raises(CommandError,
cfg.set,
'mike.should.have',
'a beer',
force=False)
assert_equal(len(cfg['mike.should.have']), 2)
# replaces all matching ones with force=True
cfg.set('mike.should.have', 'a beer', force=True)
assert_equal(cfg['mike.should.have'], 'a beer')
# fails unknown location
assert_raises(ValueError, cfg.add, 'somesuch', 'shit', where='umpalumpa')
# very carefully test non-local config
# so carefully that even in case of bad weather Yarik doesn't find some
# lame datalad unittest sections in his precious ~/.gitconfig
with patch.dict('os.environ', {
'HOME': new_home,
'DATALAD_SNEAKY_ADDITION': 'ignore'
}):
global_gitconfig = opj(new_home, '.gitconfig')
assert (not exists(global_gitconfig))
globalcfg = ConfigManager()
assert_not_in('datalad.unittest.youcan', globalcfg)
assert_in('datalad.sneaky.addition', globalcfg)
cfg.add('datalad.unittest.youcan', 'removeme', where='global')
assert (exists(global_gitconfig))
# it did not go into the dataset's config!
assert_not_in('datalad.unittest.youcan', cfg)
# does not monitor additions!
globalcfg.reload(force=True)
assert_in('datalad.unittest.youcan', globalcfg)
with swallow_logs():
assert_raises(CommandError,
globalcfg.unset,
'datalad.unittest.youcan',
where='local')
assert (globalcfg.has_section('datalad.unittest'))
globalcfg.unset('datalad.unittest.youcan', where='global')
# but after we unset the only value -- that section is no longer listed
assert (not globalcfg.has_section('datalad.unittest'))
assert_not_in('datalad.unittest.youcan', globalcfg)
if external_versions['cmd:git'] < '2.18':
# older versions leave empty section behind in the file
ok_file_has_content(global_gitconfig,
'[datalad "unittest"]',
strip=True)
# remove_section to clean it up entirely
globalcfg.remove_section('datalad.unittest', where='global')
ok_file_has_content(global_gitconfig, "")
cfg = ConfigManager(Dataset(opj(path, 'ds')),
source='dataset',
overrides={'datalad.godgiven': True})
assert_equal(cfg.get('datalad.godgiven'), True)
# setter has no effect
cfg.set('datalad.godgiven', 'false')
assert_equal(cfg.get('datalad.godgiven'), True)
def _install_subds_from_flexible_source(ds, sm, **kwargs):
"""Tries to obtain a given subdataset from several meaningful locations
Parameters
----------
ds : Dataset
Parent dataset of to-be-installed subdataset.
sm : dict
Submodule record as produced by `subdatasets()`.
**kwargs
Passed onto clone()
"""
sm_path = op.relpath(sm['path'], start=sm['parentds'])
# compose a list of candidate clone URLs
clone_urls = _get_flexible_source_candidates_for_submodule(ds, sm)
# prevent inevitable exception from `clone`
dest_path = op.join(ds.path, sm_path)
clone_urls_ = [src['url'] for src in clone_urls if src['url'] != dest_path]
if not clone_urls:
# yield error
yield get_status_dict(
action='install',
ds=ds,
status='error',
message=("Have got no candidates to install subdataset %s from.",
sm_path),
logger=lgr,
)
return
for res in clone_dataset(clone_urls_,
Dataset(dest_path),
cfg=ds.config,
checkout_gitsha=sm['gitshasum'],
**kwargs):
if res.get('action', None) == 'install' and \
res.get('status', None) == 'ok' and \
res.get('type', None) == 'dataset' and \
res.get('path', None) == dest_path:
_fixup_submodule_dotgit_setup(ds, sm_path)
section_name = 'submodule.{}'.format(sm['gitmodule_name'])
# register the submodule as "active" in the superdataset
ds.config.set(
'{}.active'.format(section_name),
'true',
reload=False,
force=True,
where='local',
)
ds.config.set(
'{}.url'.format(section_name),
# record the actual source URL of the successful clone
# and not a funky prediction based on the parent ds
# like ds.repo.update_submodule() would do (does not
# accept a URL)
res['source']['giturl'],
reload=True,
force=True,
where='local',
)
yield res
subds = Dataset(dest_path)
if not subds.is_installed():
lgr.debug('Desired subdataset %s did not materialize, stopping', subds)
return
# check whether clone URL generators were involved
cand_cfg = [rec for rec in clone_urls if rec.get('from_config', False)]
if cand_cfg:
# get a handle on the configuration that is specified in the
# dataset itself (local and dataset)
super_cfg = ConfigManager(dataset=ds, source='dataset-local')
need_reload = False
for rec in cand_cfg:
# check whether any of this configuration originated from the
# superdataset. if so, inherit the config in the new subdataset
# clone. if not, keep things clean in order to be able to move with
# any outside configuration change
for c in ('datalad.get.subdataset-source-candidate-{}{}'.format(
rec['cost'], rec['name']),
'datalad.get.subdataset-source-candidate-{}'.format(
rec['name'])):
if c in super_cfg.keys():
subds.config.set(c,
super_cfg.get(c),
where='local',
reload=False)
need_reload = True
break
if need_reload:
subds.config.reload(force=True)
def _install_subds_from_flexible_source(ds, sm, **kwargs):
"""Tries to obtain a given subdataset from several meaningful locations
Parameters
----------
ds : Dataset
Parent dataset of to-be-installed subdataset.
sm : dict
Submodule record as produced by `subdatasets()`.
**kwargs
Passed onto clone()
"""
sm_path = op.relpath(sm['path'], start=sm['parentds'])
# compose a list of candidate clone URLs
clone_urls = _get_flexible_source_candidates_for_submodule(ds, sm)
# prevent inevitable exception from `clone`
dest_path = op.join(ds.path, sm_path)
clone_urls_ = [src['url'] for src in clone_urls if src['url'] != dest_path]
if not clone_urls:
# yield error
yield get_status_dict(
action='install',
ds=ds,
status='error',
message=("Have got no candidates to install subdataset %s from.",
sm_path),
logger=lgr,
)
return
for res in clone_dataset(clone_urls_, Dataset(dest_path), **kwargs):
# make sure to fix a detached HEAD before yielding the install success
# result. The resetting of the branch would undo any change done
# to the repo by processing in response to the result
if res.get('action', None) == 'install' and \
res.get('status', None) == 'ok' and \
res.get('type', None) == 'dataset' and \
res.get('path', None) == dest_path:
_fixup_submodule_dotgit_setup(ds, sm_path)
# do fancy update
lgr.debug(
"Update cloned subdataset {0} in parent".format(dest_path))
ds.repo.update_submodule(sm_path, init=True)
yield res
subds = Dataset(dest_path)
if not subds.is_installed():
lgr.debug('Desired subdataset %s did not materialize, stopping', subds)
return
# check whether clone URL generators were involved
cand_cfg = [rec for rec in clone_urls if rec.get('from_config', False)]
if cand_cfg:
# get a handle on the configuration that is specified in the
# dataset itself (local and dataset)
super_cfg = ConfigManager(dataset=ds, source='dataset-local')
need_reload = False
for rec in cand_cfg:
# check whether any of this configuration originated from the
# superdataset. if so, inherit the config in the new subdataset
# clone. if not, keep things clean in order to be able to move with
# any outside configuration change
for c in ('datalad.get.subdataset-source-candidate-{}{}'.format(
rec['cost'], rec['name']),
'datalad.get.subdataset-source-candidate-{}'.format(
rec['name'])):
if c in super_cfg.keys():
subds.config.set(c,
super_cfg.get(c),
where='local',
reload=False)
need_reload = True
break
if need_reload:
subds.config.reload(force=True)
def test_something(path, new_home):
# read nothing, has nothing
cfg = ConfigManager(dataset_only=True)
assert_false(len(cfg))
# now read the example config
cfg = ConfigManager(Dataset(opj(path, 'ds')), dataset_only=True)
assert_equal(len(cfg), 3)
assert_in('something.user', cfg)
# multi-value
assert_equal(len(cfg['something.user']), 2)
assert_equal(cfg['something.user'], ('name=Jane Doe', '[email protected]'))
assert_true(cfg.has_section('something'))
assert_false(cfg.has_section('somethingelse'))
assert_equal(sorted(cfg.sections()),
[u'onemore.complicated の beast with.dot', 'something'])
assert_true(cfg.has_option('something', 'user'))
assert_false(cfg.has_option('something', 'us?er'))
assert_false(cfg.has_option('some?thing', 'user'))
assert_equal(sorted(cfg.options('something')), ['myint', 'user'])
assert_equal(cfg.options(u'onemore.complicated の beast with.dot'), ['findme'])
assert_equal(
sorted(cfg.items()),
[(u'onemore.complicated の beast with.dot.findme', '5.0'),
('something.myint', '3'),
('something.user', ('name=Jane Doe', '[email protected]'))])
assert_equal(
sorted(cfg.items('something')),
[('something.myint', '3'),
('something.user', ('name=Jane Doe', '[email protected]'))])
# always get all values
assert_equal(
cfg.get('something.user'),
('name=Jane Doe', '[email protected]'))
assert_raises(KeyError, cfg.__getitem__, 'somedthing.user')
assert_equal(cfg.getfloat(u'onemore.complicated の beast with.dot', 'findme'), 5.0)
assert_equal(cfg.getint('something', 'myint'), 3)
assert_equal(cfg.getbool('something', 'myint'), True)
assert_equal(cfg.getbool('doesnot', 'exist', default=True), True)
assert_raises(TypeError, cfg.getbool, 'something', 'user')
# gitpython-style access
assert_equal(cfg.get('something.myint'), cfg.get_value('something', 'myint'))
assert_equal(cfg.get_value('doesnot', 'exist', default='oohaaa'), 'oohaaa')
# weired, but that is how it is
assert_raises(KeyError, cfg.get_value, 'doesnot', 'exist', default=None)
# modification follows
cfg.add('something.new', 'の')
assert_equal(cfg.get('something.new'), u'の')
# sections are added on demand
cfg.add('unheard.of', 'fame')
assert_true(cfg.has_section('unheard.of'))
comp = cfg.items('something')
cfg.rename_section('something', 'this')
assert_true(cfg.has_section('this'))
assert_false(cfg.has_section('something'))
# direct comparision would fail, because of section prefix
assert_equal(len(cfg.items('this')), len(comp))
# fail if no such section
with swallow_logs():
assert_raises(CommandError, cfg.rename_section, 'nothere', 'irrelevant')
assert_true(cfg.has_option('this', 'myint'))
cfg.unset('this.myint')
assert_false(cfg.has_option('this', 'myint'))
# batch a changes
cfg.add('mike.wants.to', 'know', reload=False)
assert_false('mike.wants.to' in cfg)
cfg.add('mike.wants.to', 'eat')
assert_true('mike.wants.to' in cfg)
assert_equal(len(cfg['mike.wants.to']), 2)
# set a new one:
cfg.set('mike.should.have', 'known')
assert_in('mike.should.have', cfg)
assert_equal(cfg['mike.should.have'], 'known')
# set an existing one:
cfg.set('mike.should.have', 'known better')
assert_equal(cfg['mike.should.have'], 'known better')
# set, while there are several matching ones already:
cfg.add('mike.should.have', 'a meal')
assert_equal(len(cfg['mike.should.have']), 2)
# raises with force=False
assert_raises(CommandError,
cfg.set, 'mike.should.have', 'a beer', force=False)
assert_equal(len(cfg['mike.should.have']), 2)
# replaces all matching ones with force=True
cfg.set('mike.should.have', 'a beer', force=True)
assert_equal(cfg['mike.should.have'], 'a beer')
# fails unknown location
assert_raises(ValueError, cfg.add, 'somesuch', 'shit', where='umpalumpa')
# very carefully test non-local config
# so carefully that even in case of bad weather Yarik doesn't find some
# lame datalad unittest sections in his precious ~/.gitconfig
with patch.dict('os.environ',
{'HOME': new_home, 'DATALAD_SNEAKY_ADDITION': 'ignore'}):
global_gitconfig = opj(new_home, '.gitconfig')
assert(not exists(global_gitconfig))
globalcfg = ConfigManager(dataset_only=False)
assert_not_in('datalad.unittest.youcan', globalcfg)
assert_in('datalad.sneaky.addition', globalcfg)
cfg.add('datalad.unittest.youcan', 'removeme', where='global')
assert(exists(global_gitconfig))
# it did not go into the dataset's config!
assert_not_in('datalad.unittest.youcan', cfg)
# does not monitor additions!
globalcfg.reload(force=True)
assert_in('datalad.unittest.youcan', globalcfg)
with swallow_logs():
assert_raises(
CommandError,
globalcfg.unset,
'datalad.unittest.youcan',
where='local')
assert(globalcfg.has_section('datalad.unittest'))
globalcfg.unset('datalad.unittest.youcan', where='global')
# but after we unset the only value -- that section is no longer listed
assert (not globalcfg.has_section('datalad.unittest'))
assert_not_in('datalad.unittest.youcan', globalcfg)
if external_versions['cmd:git'] < '2.18':
# older versions leave empty section behind in the file
ok_file_has_content(global_gitconfig, '[datalad "unittest"]', strip=True)
# remove_section to clean it up entirely
globalcfg.remove_section('datalad.unittest', where='global')
ok_file_has_content(global_gitconfig, "")
cfg = ConfigManager(
Dataset(opj(path, 'ds')),
dataset_only=True,
overrides={'datalad.godgiven': True})
assert_equal(cfg.get('datalad.godgiven'), True)
# setter has no effect
cfg.set('datalad.godgiven', 'false')
assert_equal(cfg.get('datalad.godgiven'), True)
def _install_subds_from_flexible_source(ds, sm, **kwargs):
"""Tries to obtain a given subdataset from several meaningful locations
Parameters
----------
ds : Dataset
Parent dataset of to-be-installed subdataset.
sm : dict
Submodule record as produced by `subdatasets()`.
**kwargs
Passed onto clone()
"""
sm_path = op.relpath(sm['path'], start=sm['parentds'])
# compose a list of candidate clone URLs
clone_urls = _get_flexible_source_candidates_for_submodule(ds, sm)
# prevent inevitable exception from `clone`
dest_path = op.join(ds.path, sm_path)
clone_urls_ = [src['url'] for src in clone_urls if src['url'] != dest_path]
if not clone_urls:
# yield error
yield get_status_dict(
action='install',
ds=ds,
status='error',
message=("Have got no candidates to install subdataset %s from.",
sm_path),
logger=lgr,
)
return
for res in clone_dataset(clone_urls_,
Dataset(dest_path),
cfg=ds.config,
**kwargs):
# make sure to fix a detached HEAD before yielding the install success
# result. The resetting of the branch would undo any change done
# to the repo by processing in response to the result
if res.get('action', None) == 'install' and \
res.get('status', None) == 'ok' and \
res.get('type', None) == 'dataset' and \
res.get('path', None) == dest_path:
_fixup_submodule_dotgit_setup(ds, sm_path)
target_commit = sm['gitshasum']
lgr.debug(
"Update cloned subdataset {0} in parent".format(dest_path))
section_name = 'submodule.{}'.format(sm['gitmodule_name'])
# do not use `git-submodule update --init`, it would make calls
# to git-config which will not obey datalad inter-process locks for
# modifying .git/config
sub = GitRepo(res['path'])
# record what branch we were on right after the clone
# TODO instead of the active branch, this should first consider
# a configured branch in the submodule record of the superdataset
sub_orig_branch = sub.get_active_branch()
# if we are on a branch this hexsha will be the tip of that branch
sub_orig_hexsha = sub.get_hexsha()
if sub_orig_hexsha != target_commit:
# make sure we have the desired commit locally
# expensive and possibly error-prone fetch conditional on cheap
# local check
if not sub.commit_exists(target_commit):
try:
sub.fetch(remote='origin', refspec=target_commit)
except CommandError:
pass
# instead of inspecting the fetch results for possible ways
# with which it could failed to produced the desired result
# let's verify the presence of the commit directly, we are in
# expensive-land already anyways
if not sub.commit_exists(target_commit):
res.update(
status='error',
message=
('Target commit %s does not exist in the clone, and '
'a fetch that commit from origin failed',
target_commit[:8]),
)
yield res
# there is nothing we can do about this
# MIH thinks that removing the clone is not needed, as a likely
# next step will have to be a manual recovery intervention
# and not another blind attempt
continue
# checkout the desired commit
sub.call_git(['checkout', target_commit])
# did we detach?
# XXX: This is a less generic variant of a part of
# GitRepo.update_submodule(). It makes use of already available
# information and trusts the existence of the just cloned repo
# and avoids (redoing) some safety checks
if sub_orig_branch and not sub.get_active_branch():
# trace if current state is a predecessor of the branch_hexsha
lgr.debug(
"Detached HEAD after updating submodule %s "
"(original branch: %s)", sub, sub_orig_branch)
if sub.get_merge_base([sub_orig_hexsha,
target_commit]) == target_commit:
# TODO: config option?
# MIH: There is no real need here. IMHO this should all not
# happen, unless the submodule record has a branch
# configured. And Datalad should leave such a record, when
# a submodule is registered.
# we assume the target_commit to be from the same branch,
# because it is an ancestor -- update that original branch
# to point to the target_commit, and update HEAD to point to
# that location -- this readies the subdataset for
# further modification
lgr.info(
"Reset subdataset branch '%s' to %s (from %s) to "
"avoid a detached HEAD", sub_orig_branch,
target_commit[:8], sub_orig_hexsha[:8])
branch_ref = 'refs/heads/%s' % sub_orig_branch
sub.update_ref(branch_ref, target_commit)
sub.update_ref('HEAD', branch_ref, symbolic=True)
else:
lgr.warning(
"%s has a detached HEAD, because the recorded "
"subdataset state %s has no unique ancestor with "
"branch '%s'", sub, target_commit[:8],
sub_orig_branch)
# register the submodule as "active" in the superdataset
ds.config.set(
'{}.active'.format(section_name),
'true',
reload=False,
force=True,
where='local',
)
ds.config.set(
'{}.url'.format(section_name),
# record the actual source URL of the successful clone
# and not a funky prediction based on the parent ds
# like ds.repo.update_submodule() would do (does not
# accept a URL)
res['source']['giturl'],
reload=True,
force=True,
where='local',
)
yield res
subds = Dataset(dest_path)
if not subds.is_installed():
lgr.debug('Desired subdataset %s did not materialize, stopping', subds)
return
# check whether clone URL generators were involved
cand_cfg = [rec for rec in clone_urls if rec.get('from_config', False)]
if cand_cfg:
# get a handle on the configuration that is specified in the
# dataset itself (local and dataset)
super_cfg = ConfigManager(dataset=ds, source='dataset-local')
need_reload = False
for rec in cand_cfg:
# check whether any of this configuration originated from the
# superdataset. if so, inherit the config in the new subdataset
# clone. if not, keep things clean in order to be able to move with
# any outside configuration change
for c in ('datalad.get.subdataset-source-candidate-{}{}'.format(
rec['cost'], rec['name']),
'datalad.get.subdataset-source-candidate-{}'.format(
rec['name'])):
if c in super_cfg.keys():
subds.config.set(c,
super_cfg.get(c),
where='local',
reload=False)
need_reload = True
break
if need_reload:
subds.config.reload(force=True)
