audiocaps.py

# coding=utf-8

"""AudioCaps dataset."""


import os
import gzip
import shutil
import joblib
import pathlib
import logging
import datasets
import typing as tp
import pandas as pd
import urllib.request
from pathlib import Path
from copy import deepcopy
from tqdm.auto import tqdm
from rich.logging import RichHandler

logger = logging.getLogger(__name__)
logger.addHandler(RichHandler())
logger.setLevel(logging.INFO)

VERSION = "0.0.1"

# Cache location
DEFAULT_XDG_CACHE_HOME = "~/.cache"
XDG_CACHE_HOME = os.getenv("XDG_CACHE_HOME", DEFAULT_XDG_CACHE_HOME)
DEFAULT_HF_CACHE_HOME = os.path.join(XDG_CACHE_HOME, "huggingface")
HF_CACHE_HOME = os.path.expanduser(os.getenv("HF_HOME", DEFAULT_HF_CACHE_HOME))
DEFAULT_HF_DATASETS_CACHE = os.path.join(HF_CACHE_HOME, "datasets")
HF_DATASETS_CACHE = Path(os.getenv("HF_DATASETS_CACHE", DEFAULT_HF_DATASETS_CACHE))


class AudioCapsConfig(datasets.BuilderConfig):
    """BuilderConfig for AudioCaps."""
    
    def __init__(self, features, **kwargs):
        super(AudioCapsConfig, self).__init__(version=datasets.Version(VERSION, ""), **kwargs)
        self.features = features


class AudioCaps(datasets.GeneratorBasedBuilder):

    BUILDER_CONFIGS = [
        AudioCapsConfig(
            features=datasets.Features(
                {
                    "file": datasets.Value("string"),
                    "audio": datasets.Audio(sampling_rate=None),
                    "captions": atasets.Sequence(datasets.Value("string")),
                }
            ),
            name="audiocaps", 
            description="",
        ), 
    ]

    DEFAULT_CONFIG_NAME = "audiocaps"

    def _info(self):
        return datasets.DatasetInfo(
            description="",
            features=self.config.features,
            supervised_keys=None,
            homepage="",
            citation="",
            task_templates=None,
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        extensions = ['.wav']
        
        # Development sets
        def _download_and_extract(filename):
            DEV_URL = f'https://huggingface.co/datasets/confit/audiocaps/resolve/main/data/{filename}'
            _dev_save_path = os.path.join(
                HF_DATASETS_CACHE, 'confit___audiocaps/audiocaps', VERSION
            )
            download_file(
                source=DEV_URL, 
                dest=os.path.join(_dev_save_path, filename), 
                unpack=True, 
                dest_unpack=os.path.join(_dev_save_path, 'extracted', 'train'), 
            )
            
        joblib.Parallel(n_jobs=8, verbose=10)(
            joblib.delayed(_download_and_extract)(
                filename=_filename
            ) for _filename in tqdm([f'train{i}.zip' for i in range(1, 12+1)])
        )
        train_audio_paths = []
        for _filename in [f'train{i}.zip' for i in range(1, 12+1)]:
            DEV_URL = f'https://huggingface.co/datasets/confit/audiocaps/resolve/main/data/{_filename}'
            _dev_save_path = os.path.join(
                HF_DATASETS_CACHE, 'confit___audiocaps/audiocaps', VERSION
            )
            train_archive_path = os.path.join(_dev_save_path, 'extracted', 'train')
            _, _walker = fast_scandir(train_archive_path, extensions, recursive=True)
            train_audio_paths.extend(_walker)
        
        # Validation set
        VAL_URL = 'https://huggingface.co/datasets/confit/audiocaps/resolve/main/data/val.zip'
        _val_save_path = os.path.join(
            HF_DATASETS_CACHE, 'confit___audiocaps/audiocaps', VERSION
        )
        _filename = 'val.zip'
        download_file(
            source=VAL_URL, 
            dest=os.path.join(_val_save_path, _filename), 
            unpack=True, 
            dest_unpack=os.path.join(_val_save_path, 'extracted', 'validation'), 
        )
        validation_archive_path = os.path.join(_val_save_path, 'extracted', 'validation')
        _, validation_audio_paths = fast_scandir(validation_archive_path, extensions, recursive=True)

        # Evaluation set
        EVAL_URL = 'https://huggingface.co/datasets/confit/audiocaps/resolve/main/data/test.zip'
        _eval_save_path = os.path.join(
            HF_DATASETS_CACHE, 'confit___audiocaps/audiocaps', VERSION
        )
        _filename = 'test.zip'
        download_file(
            source=EVAL_URL, 
            dest=os.path.join(_eval_save_path, _filename), 
            unpack=True, 
            dest_unpack=os.path.join(_eval_save_path, 'extracted', 'test'), 
        )
        test_archive_path = os.path.join(_eval_save_path, 'extracted', 'test')
        _, test_audio_paths = fast_scandir(test_archive_path, extensions, recursive=True)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN, gen_kwargs={"audio_paths": train_audio_paths, "split": "train"}
            ), 
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION, gen_kwargs={"audio_paths": validation_audio_paths, "split": "validation"}
            ), 
            datasets.SplitGenerator(
                name=datasets.Split.TEST, gen_kwargs={"audio_paths": test_audio_paths, "split": "test"}
            ), 
        ]

    def _generate_examples(self, audio_paths, split=None):
        if split == 'train':
            metadata_df = pd.read_csv('https://huggingface.co/datasets/confit/audiocaps/raw/main/metadata/train.csv')
        elif split == 'validation':
            metadata_df = pd.read_csv('https://huggingface.co/datasets/confit/audiocaps/raw/main/metadata/val.csv')
        elif split == 'test':
            metadata_df = pd.read_csv('https://huggingface.co/datasets/confit/audiocaps/raw/main/metadata/test.csv')

        fileid2caption = {}
        for idx, row in metadata_df.iterrows():
            fileid2caption[f"{row['audiocap_id']}.wav"] = row['caption'] # this filename doesn't have suffix

        for guid, audio_path in enumerate(audio_paths):
            fileid = Path(audio_path).name
            caption = fileid2caption.get(fileid)
            yield guid, {
                "id": str(guid),
                "file": audio_path, 
                "audio": audio_path, 
                "captions": [caption], 
            }


def fast_scandir(path: str, exts: tp.List[str], recursive: bool = False):
    # Scan files recursively faster than glob
    # From github.com/drscotthawley/aeiou/blob/main/aeiou/core.py
    subfolders, files = [], []

    try:  # hope to avoid 'permission denied' by this try
        for f in os.scandir(path):
            try:  # 'hope to avoid too many levels of symbolic links' error
                if f.is_dir():
                    subfolders.append(f.path)
                elif f.is_file():
                    if os.path.splitext(f.name)[1].lower() in exts:
                        files.append(f.path)
            except Exception:
                pass
    except Exception:
        pass

    if recursive:
        for path in list(subfolders):
            sf, f = fast_scandir(path, exts, recursive=recursive)
            subfolders.extend(sf)
            files.extend(f)  # type: ignore

    return subfolders, files


def download_file(
    source,
    dest,
    unpack=False,
    dest_unpack=None,
    replace_existing=False,
    write_permissions=False,
):
    """Downloads the file from the given source and saves it in the given
    destination path.
     Arguments
    ---------
    source : path or url
        Path of the source file. If the source is an URL, it downloads it from
        the web.
    dest : path
        Destination path.
    unpack : bool
        If True, it unpacks the data in the dest folder.
    dest_unpack: path
        Path where to store the unpacked dataset
    replace_existing : bool
        If True, replaces the existing files.
    write_permissions: bool
        When set to True, all the files in the dest_unpack directory will be granted write permissions.
        This option is active only when unpack=True.
    """
    class DownloadProgressBar(tqdm):
        """DownloadProgressBar class."""

        def update_to(self, b=1, bsize=1, tsize=None):
            """Needed to support multigpu training."""
            if tsize is not None:
                self.total = tsize
            self.update(b * bsize - self.n)

    # Create the destination directory if it doesn't exist
    dest_dir = pathlib.Path(dest).resolve().parent
    dest_dir.mkdir(parents=True, exist_ok=True)
    if "http" not in source:
        shutil.copyfile(source, dest)

    elif not os.path.isfile(dest) or (
        os.path.isfile(dest) and replace_existing
    ):
        logger.info(f"Downloading {source} to {dest}")
        with DownloadProgressBar(
            unit="B",
            unit_scale=True,
            miniters=1,
            desc=source.split("/")[-1],
        ) as t:
            urllib.request.urlretrieve(
                source, filename=dest, reporthook=t.update_to
            )
    else:
        logger.info(f"{dest} exists. Skipping download")

    # Unpack if necessary
    if unpack:
        if dest_unpack is None:
            dest_unpack = os.path.dirname(dest)
        if os.path.exists(dest_unpack):
            logger.info(f"{dest_unpack} already exists. Skipping extraction")
        else:
            logger.info(f"Extracting {dest} to {dest_unpack}")
            # shutil unpack_archive does not work with tar.gz files
            if (
                source.endswith(".tar.gz")
                or source.endswith(".tgz")
                or source.endswith(".gz")
            ):
                out = dest.replace(".gz", "")
                with gzip.open(dest, "rb") as f_in:
                    with open(out, "wb") as f_out:
                        shutil.copyfileobj(f_in, f_out)
            else:
                shutil.unpack_archive(dest, dest_unpack)
            if write_permissions:
                set_writing_permissions(dest_unpack)


def set_writing_permissions(folder_path):
    """
    This function sets user writing permissions to all the files in the given folder.
    Arguments
    ---------
    folder_path : folder
        Folder whose files will be granted write permissions.
    """
    for root, dirs, files in os.walk(folder_path):
        for file_name in files:
            file_path = os.path.join(root, file_name)
            # Set writing permissions (mode 0o666) to the file
            os.chmod(file_path, 0o666)