def load_lyrics(fhandle: TextIO) -> annotations.LyricData:
"""Load an ikala lyrics annotation
Args:
fhandle (str or file-like): File-like object or path to lyric annotation file
Raises:
IOError: if lyrics_path does not exist
Returns:
LyricData: lyric annotation data
"""
# input: start time (ms), end time (ms), lyric, [pronunciation]
reader = csv.reader(fhandle, delimiter=" ")
start_times = []
end_times = []
lyrics = []
pronunciations = []
for line in reader:
start_times.append(float(line[0]) / 1000.0)
end_times.append(float(line[1]) / 1000.0)
lyrics.append(line[2])
if len(line) > 2:
pronunciation = " ".join(line[3:])
pronunciations.append(pronunciation)
else:
pronunciations.append("")
lyrics_data = annotations.LyricData(
np.array([start_times, end_times]).T,
lyrics,
pronunciations,
)
return lyrics_data
def load_annotations_granularity(annotations_path, granularity):
"""Load annotations at the specified level of granularity
Args:
annotations_path (str): path to a DALI annotation file
granularity (str): one of 'notes', 'words', 'lines', 'paragraphs'
Returns:
NoteData for granularity='notes' or LyricData otherwise
"""
try:
with gzip.open(annotations_path, "rb") as f:
output = pickle.load(f)
except Exception as e:
with gzip.open(annotations_path, "r") as f:
output = pickle.load(f)
text = []
notes = []
begs = []
ends = []
for annot in output.annotations["annot"][granularity]:
notes.append(round(annot["freq"][0], 3))
begs.append(round(annot["time"][0], 3))
ends.append(round(annot["time"][1], 3))
text.append(annot["text"])
if granularity == "notes":
annotation = annotations.NoteData(
np.array([begs, ends]).T, np.array(notes), None)
else:
annotation = annotations.LyricData(
np.array([begs, ends]).T, text, None)
return annotation
def test_lyric_data():
intervals = np.array([[1.0, 2.0], [1.5, 3.0], [2.0, 3.0]])
lyrics = ["E:m", "A", "G:7"]
pronunciations = ["a", "", "b"]
lyric_data = annotations.LyricData(intervals, lyrics, pronunciations)
assert np.allclose(lyric_data.intervals, intervals)
assert lyric_data.lyrics == lyrics
assert lyric_data.pronunciations == pronunciations
def load_phrases(fhandle: TextIO) -> annotations.LyricData:
"""Load lyric phrases annotation
Args:
fhandle (str or file-like): path or file-like object pointing to a lyric annotation file
Returns:
LyricData: lyric phrase annotation
"""
start_times = []
end_times = []
lyrics = []
reader = csv.reader(fhandle, delimiter="\t")
for line in reader:
start_times.append(float(line[0]))
end_times.append(float(line[1]))
lyrics.append(line[2] if line[2] != "sil" else "")
return annotations.LyricData(
np.array([start_times, end_times]).T,
lyrics,
)
def test_lyrics():
lyrics_data_1 = [
(
annotations.LyricData(
np.array([[0.027, 0.232], [0.227, 0.742]]).T,
["The", "Test"],
["", ""],
),
None,
)
]
lyrics_data_2 = [
(
annotations.LyricData(
np.array([[0.027, 0.232], [0.227, 0.742]]).T,
["The", "Test"],
["", ""],
),
"lyrics_1",
)
]
lyrics_data_3 = [
(
annotations.LyricData(
np.array([[0.027, 0.232], [0.227, 0.742]]).T,
["The", "Test"],
["", ""],
),
"lyrics_1",
),
(
annotations.LyricData(
np.array([[0.0, 0.232], [0.227, 0.742]]).T,
["is", "cool"],
["", ""],
),
"lyrics_2",
),
]
lyrics_data_4 = (
annotations.LyricData(
np.array([[0.027, 0.232], [0.227, 0.742]]).T,
["The", "Test"],
["", ""],
),
"lyrics_1",
)
lyrics_data_5 = [
(
annotations.LyricData(
np.array([[0.027, 0.232], [0.227, 0.742]]).T,
["The", "Test"],
["", ""],
),
"lyrics_1",
),
[
annotations.LyricData(
np.array([[0.0, 0.232], [0.227, 0.742]]).T,
["is", "cool"],
["", ""],
),
"lyrics_2",
],
]
lyrics_data_6 = [(None, None)]
lyrics_data_7 = [
(
annotations.EventData(
np.array([[0.2, 0.3], [0.3, 0.4]]).T,
["event A", "event B"],
),
None,
)
]
jam_1 = jams_utils.jams_converter(lyrics_data=lyrics_data_1)
jam_2 = jams_utils.jams_converter(lyrics_data=lyrics_data_2)
jam_3 = jams_utils.jams_converter(lyrics_data=lyrics_data_3)
jam_6 = jams_utils.jams_converter(lyrics_data=lyrics_data_6)
time, duration, value, confidence = get_jam_data(jam_1, "lyrics", 0)
assert time == [0.027, 0.232]
assert duration == [0.2, 0.51]
assert value == ["The", "Test"]
assert confidence == [None, None]
assert jam_2.annotations[0]["sandbox"]["name"] == "lyrics_1"
time, duration, value, confidence = get_jam_data(jam_3, "lyrics", 0)
assert time == [0.027, 0.232]
assert duration == [0.2, 0.51]
assert value == ["The", "Test"]
assert confidence == [None, None]
time, duration, value, confidence = get_jam_data(jam_3, "lyrics", 1)
assert time == [0.0, 0.232]
assert duration == [0.227, 0.51]
assert value == ["is", "cool"]
assert confidence == [None, None]
time, duration, value, confidence = get_jam_data(jam_6, "lyrics", 0)
assert time == []
assert duration == []
assert value == []
assert confidence == []
assert type(jam_1) == jams.JAMS
with pytest.raises(TypeError):
jams_utils.jams_converter(lyrics_data=lyrics_data_4)
with pytest.raises(TypeError):
jams_utils.jams_converter(lyrics_data=lyrics_data_5)
with pytest.raises(TypeError):
jams_utils.jams_converter(lyrics_data=lyrics_data_7)