speechdat.py

# coding=utf-8
# Copyright 2021 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Speech Dat dataset"""

import datasets
import json
import os
from datasets.tasks import AutomaticSpeechRecognition
from pathlib import Path

_DESCRIPTION = """\
Speechdat dataset
"""

_HOMEPAGE = ""

_LICENSE = ""

class SpeechDat(datasets.GeneratorBasedBuilder):

    DEFAULT_WRITER_BATCH_SIZE = 1000

    VERSION = datasets.Version("1.1.0")
 
    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="audio", version=VERSION, description="SpeechDat dataset"),
    ]

    def _info(self):
        features = datasets.Features(
            {
                "path": datasets.Value("string"),
                "audio": datasets.Audio(sampling_rate=16_000),
                "sentence": datasets.Value("string"),
            }
        )

        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            supervised_keys=None,
            homepage=_HOMEPAGE,
            license=_LICENSE,
            task_templates=[
                AutomaticSpeechRecognition(audio_file_path_column="path", transcription_column="sentence")
            ],
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""

        manual_dir = os.path.abspath(os.path.expanduser(dl_manager.manual_dir))
        path_to_data = "/".join(["wav"])

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "data_dir": manual_dir
                },
            )
        ]

    def _generate_examples(self, data_dir):
        """Yields examples."""
        data_fields = list(self._info().features.keys())

        def get_single_line(path):
            lines = []
            with open(path, 'r', encoding="utf-8") as f:
                for line in f:
                    line = line.strip()
                    lines.append(line)
            if len(lines) == 1:
                return lines[0]
            elif len(lines) == 0:
                return None
            else:
                return " ".join(lines)

        data_path = Path(data_dir)
        for wav_file in data_path.glob("*.wav"):
            text_file = Path(str(wav_file).replace(".wav", ".svo"))
            if not text_file.is_file():
                continue
            text_line = get_single_line(text_file)
            if text_line is None or text_line == "":
                continue
            size = os.path.getsize(wav_file)
            if size > 1024:
                with open(wav_file, "rb") as wav_data:
                    yield str(wav_file), {
                        "path": str(wav_file),
                        "sentence": text_line,
                        "audio": {
                            "path": str(wav_file),
                            "bytes": wav_data.read()
                        }
                    }



def normalize(text):
    # remove ~


    return text