def deserialize_item(data: dict) -> Any:
# Figures out what type of manifest is being decoded with some heuristics
# and returns a Lhotse manifest object rather than a raw dict.
from lhotse import Features, MonoCut, Recording, SupervisionSegment
from lhotse.array import deserialize_array
from lhotse.cut import MixedCut
if "shape" in data or "array" in data:
return deserialize_array(data)
if "sources" in data:
return Recording.from_dict(data)
if "num_features" in data:
return Features.from_dict(data)
if "type" not in data:
return SupervisionSegment.from_dict(data)
cut_type = data.pop("type")
if cut_type == "MonoCut":
return MonoCut.from_dict(data)
if cut_type == "Cut":
warnings.warn(
"Your manifest was created with Lhotse version earlier than v0.8, when MonoCut was called Cut. "
"Please re-generate it with Lhotse v0.8 as it might stop working in a future version "
"(using manifest.from_file() and then manifest.to_file() should be sufficient)."
)
return MonoCut.from_dict(data)
if cut_type == "MixedCut":
return MixedCut.from_dict(data)
raise ValueError(
f"Unexpected cut type during deserialization: '{cut_type}'")
def simple_mixed_cut():
return MixedCut(id='simple-mixed-cut',
tracks=[
MixTrack(cut=dummy_cut('cut0', duration=10.0)),
MixTrack(cut=dummy_cut('cut1', duration=10.0),
offset=5.0),
])
def cut_set_with_mixed_cut(cut1, cut2):
mixed_cut = MixedCut(
id="mixed-cut-id",
tracks=[MixTrack(cut=cut1),
MixTrack(cut=cut2, offset=1.0, snr=10)],
)
return CutSet({cut.id: cut for cut in [cut1, cut2, mixed_cut]})
def simple_mixed_cut():
return MixedCut(
id="simple-mixed-cut",
tracks=[
MixTrack(cut=dummy_cut(0, duration=10.0)),
MixTrack(cut=dummy_cut(1, duration=10.0), offset=5.0),
],
)
def deserialize_item(data: dict) -> Any:
# Figures out what type of manifest is being decoded with some heuristics
# and returns a Lhotse manifest object rather than a raw dict.
from lhotse import Cut, Features, Recording, SupervisionSegment
from lhotse.cut import MixedCut
data = arr2list_recursive(data)
if 'sources' in data:
return Recording.from_dict(data)
if 'num_features' in data:
return Features.from_dict(data)
if 'type' not in data:
return SupervisionSegment.from_dict(data)
cut_type = data.pop('type')
if cut_type == 'Cut':
return Cut.from_dict(data)
if cut_type == 'MixedCut':
return MixedCut.from_dict(data)
raise ValueError(f"Unexpected cut type during deserialization: '{cut_type}'")
def deserialize_item(data: dict) -> Any:
# Figures out what type of manifest is being decoded with some heuristics
# and returns a Lhotse manifest object rather than a raw dict.
from lhotse import MonoCut, Features, Recording, SupervisionSegment
from lhotse.cut import MixedCut
data = arr2list_recursive(data)
if 'sources' in data:
return Recording.from_dict(data)
if 'num_features' in data:
return Features.from_dict(data)
if 'type' not in data:
return SupervisionSegment.from_dict(data)
cut_type = data.pop('type')
if cut_type == 'MonoCut':
return MonoCut.from_dict(data)
if cut_type == 'Cut':
warnings.warn('Your manifest was created with Lhotse version earlier than v0.8, when MonoCut was called Cut. '
'Please re-generate it with Lhotse v0.8 as it might stop working in a future version '
'(using manifest.from_file() and then manifest.to_file() should be sufficient).')
return MonoCut.from_dict(data)
if cut_type == 'MixedCut':
return MixedCut.from_dict(data)
raise ValueError(f"Unexpected cut type during deserialization: '{cut_type}'")