def test_yaml_save_load_roundtrip(extension):
data = {'some': ['data']}
with NamedTemporaryFile() as f:
path = Path(f.name).with_suffix(extension)
save_to_yaml(data, path)
f.flush()
data_deserialized = load_yaml(path)
assert data == data_deserialized
def from_yaml(path: Pathlike) -> 'FeatureSet':
data = load_yaml(path)
return FeatureSet(
feature_extractor=FeatureExtractor.from_dict(
data['feature_extractor']),
features=[
Features(**feature_data) for feature_data in data['features']
],
)
def load_manifest(path: Pathlike) -> Manifest:
"""Generic utility for reading an arbitrary manifest."""
raw_data = load_yaml(path)
data_set = None
for manifest_type in [RecordingSet, SupervisionSet, FeatureSet, CutSet]:
try:
data_set = manifest_type.from_dicts(raw_data)
except Exception:
pass
if data_set is None:
raise ValueError(f'Unknown type of manifest: {path}')
return data_set
def from_yaml(path: Pathlike) -> 'CutSet':
raw_cuts = load_yaml(path)
def deserialize_one(raw_cut: dict) -> AnyCut:
cut_type = raw_cut.pop('type')
if cut_type == 'Cut':
return Cut.from_dict(raw_cut)
if cut_type == 'MixedCut':
return MixedCut.from_dict(raw_cut)
raise ValueError(
f"Unexpected cut type during deserialization: '{cut_type}'")
return CutSet.from_cuts(deserialize_one(cut) for cut in raw_cuts)
def from_yaml(cls, path: Pathlike) -> 'FeatureExtractor':
return cls.from_dict(load_yaml(path))
def from_yaml(path: Pathlike) -> 'RecordingSet':
raw_recordings = load_yaml(path)
return RecordingSet.from_recordings(
Recording.from_dict(raw_rec) for raw_rec in raw_recordings)
def from_yaml(path: Pathlike) -> 'SupervisionSet':
raw_segments = load_yaml(path)
return SupervisionSet.from_segments(
SupervisionSegment.from_dict(s) for s in raw_segments)
