How to use the mirdata.utils.LyricData function in mirdata

def test_load_lines():
    data_path = (
        'tests/resources/mir_datasets/DALI/annotations/'
        + '4b196e6c99574dd49ad00d56e132712b.gz'
    )
    line_data = dali.load_annotations_granularity(data_path, 'lines')

    assert type(line_data) == utils.LyricData
    assert type(line_data.start_times) == np.ndarray
    assert type(line_data.end_times) == np.ndarray
    assert type(line_data.lyrics) == np.ndarray

    print(line_data.start_times)
    print(line_data.end_times)
    print(line_data.lyrics)

    assert np.array_equal(line_data.start_times, np.array([24.125, 24.42]))
    assert np.array_equal(line_data.end_times, np.array([24.42, 24.568]))
    assert np.array_equal(line_data.lyrics, np.array(['why do', 'they']))

    # load a file which doesn't exist
    line_none = dali.load_annotations_granularity('fake/file/path', 'lines')
    assert line_none is None

def test_lyrics():
    lyrics_data_1 = [
        (
            utils.LyricData(
                np.array([0.027, 0.232]),
                np.array([0.227, 0.742]),
                np.array(['The', 'Test']),
                np.array([None, None]),
            ),
            None,
        )
    ]
    lyrics_data_2 = [
        (
            utils.LyricData(
                np.array([0.027, 0.232]),
                np.array([0.227, 0.742]),
                np.array(['The', 'Test']),
                np.array([None, None]),
            ),

),
            'lyrics_2',
        ),
    ]
    lyrics_data_4 = (
        utils.LyricData(
            np.array([0.027, 0.232]),
            np.array([0.227, 0.742]),
            np.array(['The', 'Test']),
            np.array([None, None]),
        ),
        'lyrics_1',
    )
    lyrics_data_5 = [
        (
            utils.LyricData(
                np.array([0.027, 0.232]),
                np.array([0.227, 0.742]),
                np.array(['The', 'Test']),
                np.array([None, None]),
            ),
            'lyrics_1',
        ),
        [
            utils.LyricData(
                np.array([0.0, 0.232]),
                np.array([0.227, 0.742]),
                np.array(['is', 'cool']),
                np.array([None, None]),
            ),
            'lyrics_2',
        ],

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 = utils.NoteData(np.array([begs, ends]).T, np.array(notes), None)
    else:
        annotation = utils.LyricData(
            np.array(begs), np.array(ends), np.array(text), None
        )
    return annotation

Convert lyrics annotations into jams format.

    Parameters
    ----------
    lyrics: tuple
        A tuple in the format (LyricData, str), where str describes the annotation
        and LyricData is the lyric mirdata annotation format.

    Returns
    -------
    jannot_lyric: JAM lyric annotation object.
    '''
    jannot_lyric = jams.Annotation(namespace='lyrics')
    jannot_lyric.annotation_metadata = jams.AnnotationMetadata(data_source='mirdata')
    if lyrics[0] is not None:
        if type(lyrics[0]) != utils.LyricData:
            raise TypeError('Type should be LyricData.')
        for beg, end, lyric in zip(
            lyrics[0].start_times, lyrics[0].end_times, lyrics[0].lyrics
        ):
            jannot_lyric.append(time=beg, duration=end - beg, value=lyric)
    if lyrics[1] is not None:
        jannot_lyric.sandbox = jams.Sandbox(name=lyrics[1])
    return jannot_lyric

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 if pronunciation != '' else None)
            else:
                pronunciations.append(None)

    lyrics_data = utils.LyricData(
        np.array(start_times),
        np.array(end_times),
        np.array(lyrics),
        np.array(pronunciations),
    )
    return lyrics_data