Reference for ultralytics/utils/downloads.py#

def is_url(url: str | Path, check: bool = False) -> bool:
    """Validate if the given string is a URL and optionally check if the URL exists online.

    Args:
        url (str | Path): The string to be validated as a URL.
        check (bool, optional): If True, performs an additional check to see if the URL exists online.

    Returns:
        (bool): True if the string is a valid URL and, when 'check' is True, the URL is also reachable online.

    Examples:
        >>> valid = is_url("https://www.example.com")
        >>> valid_and_exists = is_url("https://www.example.com", check=True)
    """
    try:
        url = str(url)
        result = parse.urlparse(url)
        if not (result.scheme and result.netloc):
            return False
        if check:
            import requests  # scoped as slow import

            return requests.head(url, timeout=3, allow_redirects=True).ok
        return True
    except Exception:
        return False

def delete_dsstore(path: str | Path, files_to_delete: tuple[str, ...] = (".DS_Store", "__MACOSX")) -> None:
    """Delete all specified system files and directories in a directory.

    Args:
        path (str | Path): The directory path where the files should be deleted.
        files_to_delete (tuple[str, ...]): Names of files and directories to delete.

    Examples:
        >>> from ultralytics.utils.downloads import delete_dsstore
        >>> delete_dsstore("path/to/dir")

    Notes:
        ".DS_Store" files are created by the Apple operating system and contain metadata about folders and files. They
        are hidden system files and can cause issues when transferring files between different operating systems.
    """
    for file in files_to_delete:
        matches = sorted(Path(path).rglob(file), key=lambda x: len(x.parts), reverse=True)
        LOGGER.info(f"Deleting {file} files: {matches}")
        for f in matches:
            if f.is_dir() and not f.is_symlink():
                shutil.rmtree(f)
            else:
                f.unlink()

def zip_directory(
    directory: str | Path,
    compress: bool = True,
    exclude: tuple[str, ...] = (".DS_Store", "__MACOSX"),
    progress: bool = True,
) -> Path:
    """Zip the contents of a directory, excluding specified files.

    The resulting zip file is named after the directory and placed alongside it.

    Args:
        directory (str | Path): The path to the directory to be zipped.
        compress (bool): Whether to compress the files while zipping.
        exclude (tuple[str, ...], optional): A tuple of filename strings to be excluded.
        progress (bool, optional): Whether to display a progress bar.

    Returns:
        (Path): The path to the resulting zip file.

    Raises:
        FileNotFoundError: If the directory does not exist.

    Examples:
        >>> from ultralytics.utils.downloads import zip_directory
        >>> file = zip_directory("path/to/dir")
    """
    from zipfile import ZIP_DEFLATED, ZIP_STORED, ZipFile

    delete_dsstore(directory)
    directory = Path(directory)
    if not directory.is_dir():
        raise FileNotFoundError(f"Directory '{directory}' does not exist.")

    # Zip with progress bar
    files = [f for f in directory.rglob("*") if f.is_file() and all(x not in f.name for x in exclude)]  # files to zip
    zip_file = directory.with_suffix(".zip")
    compression = ZIP_DEFLATED if compress else ZIP_STORED
    with ZipFile(zip_file, "w", compression) as f:
        for file in TQDM(files, desc=f"Zipping {directory} to {zip_file}...", unit="files", disable=not progress):
            f.write(file, file.relative_to(directory))

    return zip_file  # return path to zip file

def unzip_file(
    file: str | Path,
    path: str | Path | None = None,
    exclude: tuple[str, ...] = (".DS_Store", "__MACOSX"),
    exist_ok: bool = False,
    progress: bool = True,
) -> Path:
    """Unzip a *.zip file to the specified path, excluding specified files.

    If the zipfile does not contain a single top-level directory, the function will create a new directory with the same
    name as the zipfile (without the extension) to extract its contents. If a path is not provided, the function will
    use the parent directory of the zipfile as the default path.

    Args:
        file (str | Path): The path to the zipfile to be extracted.
        path (str | Path, optional): The path to extract the zipfile to.
        exclude (tuple[str, ...], optional): A tuple of filename strings to be excluded.
        exist_ok (bool, optional): Whether to overwrite existing contents if they exist.
        progress (bool, optional): Whether to display a progress bar.

    Returns:
        (Path): The path to the directory where the zipfile was extracted.

    Raises:
        BadZipFile: If the provided file does not exist or is not a valid zipfile.

    Examples:
        >>> from ultralytics.utils.downloads import unzip_file
        >>> directory = unzip_file("path/to/file.zip")
    """
    from zipfile import BadZipFile, ZipFile, is_zipfile

    if not (Path(file).exists() and is_zipfile(file)):
        raise BadZipFile(f"File '{file}' does not exist or is a bad zip file.")
    if path is None:
        path = Path(file).parent  # default path

    # Unzip the file contents
    with ZipFile(file) as zipObj:
        files = [f for f in zipObj.namelist() if all(x not in f for x in exclude)]
        top_level_dirs = {Path(f).parts[0] for f in files}

        # Decide to unzip directly or unzip into a directory
        unzip_as_dir = len(top_level_dirs) == 1  # (len(files) > 1 and not files[0].endswith("/"))
        if unzip_as_dir:
            # Zip has 1 top-level directory
            extract_path = path  # i.e. ../datasets
            path = Path(path) / next(iter(top_level_dirs))  # i.e. extract coco8/ dir to ../datasets/
        else:
            # Zip has multiple files at top level
            path = extract_path = Path(path) / Path(file).stem  # i.e. extract multiple files to ../datasets/coco8/

        # Skip existing files or non-empty directories unless overwriting
        if path.exists() and (path.is_file() or any(path.iterdir())) and not exist_ok:
            LOGGER.warning(f"Skipping {file} unzip as destination path {path} already exists.")
            return path

        extract_path = Path(extract_path).resolve()
        for f in TQDM(files, desc=f"Unzipping {file} to {Path(path).resolve()}...", unit="files", disable=not progress):
            f_path = Path(f)
            target = (extract_path / f_path).resolve()
            if (
                f_path.is_absolute()
                or ".." in f_path.parts
                or target.parts[: len(extract_path.parts)] != extract_path.parts
            ):
                LOGGER.warning(f"Potentially insecure file path: {f}, skipping extraction.")
                continue
            zipObj.extract(f, extract_path)

    return path  # return unzip dir

def check_disk_space(
    file_bytes: int,
    path: str | Path | None = None,
    sf: float = 1.5,
    hard: bool = True,
) -> bool:
    """Check if there is sufficient disk space to download and store a file.

    Args:
        file_bytes (int): The file size in bytes.
        path (str | Path, optional): The path or drive to check the available free space on. Defaults to the current
            working directory.
        sf (float, optional): Safety factor, the multiplier for the required free space.
        hard (bool, optional): Whether to throw an error or not on insufficient disk space.

    Returns:
        (bool): True if there is sufficient disk space, False otherwise.

    Raises:
        MemoryError: If there is insufficient disk space and `hard` is True.
    """
    total, _used, free = shutil.disk_usage(path or Path.cwd())  # bytes
    # A filesystem that cannot report usage returns 0 total blocks; free == 0 against a valid total is genuinely
    # full and must still be caught, since `free` counts blocks available to an unprivileged process.
    if not total or file_bytes * sf < free:
        return True  # sufficient space

    def fmt_bytes(b):
        if b < (1 << 20):  # without a KB tier every value under 51 KB renders "0.0 MB", hiding how full the disk is
            return f"{b / (1 << 10):.1f} KB"
        return f"{b / (1 << 20):.1f} MB" if b < (1 << 30) else f"{b / (1 << 30):.3f} GB"

    # Insufficient space
    text = (
        f"Insufficient free disk space {fmt_bytes(free)} < {fmt_bytes(int(file_bytes * sf))} required, "
        f"Please free {fmt_bytes(int(file_bytes * sf - free))} additional disk space and try again."
    )
    if hard:
        raise MemoryError(text)
    LOGGER.warning(text)
    return False

def get_google_drive_file_info(link: str) -> tuple[str, str | None]:
    """Retrieve the direct download link and filename for a shareable Google Drive file link.

    Args:
        link (str): The shareable link of the Google Drive file.

    Returns:
        url (str): Direct download URL for the Google Drive file.
        filename (str | None): Original filename of the Google Drive file. If filename extraction fails, returns None.

    Raises:
        ConnectionError: If the Google Drive download quota for the file has been exceeded.

    Examples:
        >>> from ultralytics.utils.downloads import get_google_drive_file_info
        >>> link = "https://drive.google.com/file/d/1cqT-cJgANNrhIHCrEufUYhQ4RqiWG_lJ/view?usp=drive_link"
        >>> url, filename = get_google_drive_file_info(link)
    """
    import requests  # scoped as slow import

    file_id = link.split("/d/")[1].split("/view", 1)[0]
    drive_url = f"https://drive.google.com/uc?export=download&id={file_id}"
    filename = None

    # Start session
    with requests.Session() as session:
        response = session.get(drive_url, stream=True)
        if "quota exceeded" in str(response.content.lower()):
            raise ConnectionError(
                emojis(
                    f"❌  Google Drive file download quota exceeded. "
                    f"Please try again later or download this file manually at {link}."
                )
            )
        for k, v in response.cookies.items():
            if k.startswith("download_warning"):
                drive_url += f"&confirm={v}"  # v is token
        if cd := response.headers.get("content-disposition"):
            filename = re.findall('filename="(.+)"', cd)[0]
    return drive_url, filename

def safe_download(
    url: str | Path,
    file: str | Path | None = None,
    dir: str | Path | None = None,
    unzip: bool = True,
    delete: bool = False,
    curl: bool = False,
    retry: int = 3,
    min_bytes: float = 1e0,
    exist_ok: bool = False,
    progress: bool = True,
) -> Path:
    """Download a file from a URL with options for retrying, unzipping, and deleting the downloaded file.

    Partial downloads are detected using Content-Length validation and resumed with HTTP Range requests on retry. If
    `url` is an existing local file path, no download occurs and the file is optionally unzipped.

    Args:
        url (str | Path): The URL of the file to be downloaded, or a local file path.
        file (str | Path, optional): The filename of the downloaded file. If not provided, the file will be saved with
            the same name as the URL.
        dir (str | Path, optional): The directory to save the downloaded file. If not provided, the file will be saved
            in the current working directory.
        unzip (bool, optional): Whether to unzip the downloaded file.
        delete (bool, optional): Whether to delete the downloaded file after unzipping.
        curl (bool, optional): Whether to use curl command line tool for downloading.
        retry (int, optional): The number of times to retry the download in case of failure.
        min_bytes (float, optional): The minimum number of bytes that the downloaded file should have, to be considered
            a successful download.
        exist_ok (bool, optional): Whether to overwrite existing contents during unzipping.
        progress (bool, optional): Whether to display a progress bar during the download.

    Returns:
        (Path): The path to the downloaded file or extracted directory.

    Raises:
        ConnectionError: If the download fails after all retries or the environment is offline.
        MemoryError: If there is insufficient disk space for the download.

    Examples:
        >>> from ultralytics.utils.downloads import safe_download
        >>> link = "https://ultralytics.com/assets/bus.jpg"
        >>> path = safe_download(link)
    """
    url = str(url)
    if "://" not in url and Path(url).is_file():  # local file path ('://' check required in Windows Python<3.10)
        f = Path(url)
    else:
        import requests  # scoped as slow import

        gdrive = url.startswith("https://drive.google.com/")  # check if the URL is a Google Drive link
        if gdrive:
            url, file = get_google_drive_file_info(url)
        url = url.replace(" ", "%20")  # encode spaces for curl compatibility

        f = Path(dir or ".") / (file or url2file(url))  # URL converted to filename
        if not f.is_file():  # URL and file do not exist
            uri = (url if gdrive else clean_url(url)).replace(ASSETS_URL, "https://ultralytics.com/assets")  # clean
            desc = f"Downloading {uri} to '{f}'"
            f.parent.mkdir(parents=True, exist_ok=True)  # make directory if missing
            target = f
            f = target.with_name(f".{target.name}.{uuid4().hex}.part")  # publish only after size validation
            curl_installed = shutil.which("curl")
            expected_size = 0  # total bytes from Content-Length, kept across retries to validate them
            # Both transports save the body as sent, so Content-Length and Range describe the file even when the server
            # encodes it despite `Accept-Encoding: identity`, e.g. gzip objects on S3; it is decoded once complete
            encoding = ""
            for i in range(retry + 1):
                try:
                    resume = f.stat().st_size if f.exists() else 0  # partial bytes kept from a failed attempt
                    if curl and not resume and curl_installed:  # explicit curl download
                        s = "sS" * (not progress)  # silent
                        # Stall bounds (not a total-transfer cap): abort if <1 B/s for 300 s so a dead connection
                        # cannot block interpreter shutdown while a non-daemon plot thread waits on a font download
                        args = ["--retry", "4", "--connect-timeout", "30", "--speed-limit", "1", "--speed-time", "300"]
                        # -f is required: without it curl writes the server error page as the file and exits 0
                        r = subprocess.run(
                            ["curl", "-#", f"-{s}fL", url, "-o", f, "-D", "-", *args],
                            check=False,
                            stdout=subprocess.PIPE,
                        )
                        if r.returncode:
                            raise ConnectionError(f"Curl return value {r.returncode}")
                        # Final response, after any redirect, proxy or retry blocks and before any trailer block
                        final_headers = [h for h in r.stdout.split(b"\r\n\r\n") if h.startswith(b"HTTP/")][-1]
                        encoding = (
                            b"".join(re.findall(rb"(?im)^content-encoding:\s*(\S+)", final_headers)).decode().lower()
                        )
                    else:  # requests download; timeout bounds connect and per-chunk read gaps, not total transfer
                        headers = {"Accept-Encoding": "identity"}
                        if resume:
                            headers["Range"] = f"bytes={resume}-"
                        with requests.get(url, stream=True, headers=headers, timeout=(30, 300)) as response:
                            if response.status_code == 416:  # nothing left to resume, so the next retry restarts
                                f.unlink()
                            response.raise_for_status()
                            encoding = response.headers.get("Content-Encoding", "").lower()
                            if response.status_code != 206:  # Range ignored, e.g. transcoded GCS objects, so restart
                                resume = 0
                                expected_size = int(response.headers.get("Content-Length", 0)) or expected_size
                            elif not expected_size:  # partial left by curl, so take the total from 'bytes 5-9/10'
                                total = response.headers.get("Content-Range", "").rpartition("/")[2]
                                expected_size = int(total) if total.isdigit() else 0
                            if i == 0 and expected_size > 1048576:
                                check_disk_space(expected_size, path=f.parent)
                            buffer_size = max(8192, min(1048576, expected_size // 1000)) if expected_size else 8192
                            with TQDM(
                                total=expected_size,
                                desc=desc,
                                disable=not progress,
                                unit="B",
                                unit_scale=True,
                                unit_divisor=1024,
                                initial=resume,
                            ) as pbar, open(f, "ab" if resume else "wb") as f_opened:
                                for data in response.raw.stream(buffer_size, decode_content=False):
                                    f_opened.write(data)
                                    pbar.update(len(data))

                    if f.exists():
                        file_size = f.stat().st_size
                        if expected_size and file_size != expected_size:  # only if Content-Length is known
                            LOGGER.warning(
                                f"Partial download: {file_size}/{expected_size} bytes ({file_size / expected_size * 100:.1f}%)"
                            )
                        else:
                            if encoding not in {"", "identity"}:  # undo the transfer encoding of the complete body
                                decoded = f.with_name(f"{f.name}.decoded")
                                try:
                                    with open(f, "rb") as src, open(decoded, "wb") as dst:
                                        wbits = 47  # detects a gzip or zlib header
                                        try:
                                            zlib.decompressobj(wbits).decompress(src.read(1024))
                                        except zlib.error:  # some servers send 'deflate' as raw DEFLATE without one
                                            wbits = -15
                                        src.seek(0)
                                        d = zlib.decompressobj(wbits)
                                        for chunk in iter(lambda: src.read(1048576), b""):
                                            while chunk:
                                                if d.eof:  # next gzip member, after any zero padding
                                                    chunk = chunk.lstrip(b"\0")
                                                    if not chunk:
                                                        break
                                                    d = zlib.decompressobj(wbits)
                                                dst.write(d.decompress(chunk))
                                                chunk = d.unused_data
                                    if not d.eof:  # not an assert, which `python -O` removes
                                        raise ConnectionError("Encoded body ended before its end-of-stream marker")
                                    # A gzip encoding under a gzip name means the gzip is the file itself, e.g. a
                                    # .tar.gz object stored with a gzip Content-Encoding, so keep its verified bytes
                                    if encoding != "gzip" or target.suffix not in {".gz", ".tgz"}:
                                        decoded.replace(f)
                                finally:
                                    decoded.unlink(missing_ok=True)
                            if f.stat().st_size > min_bytes:
                                f.replace(target)
                                f = target
                                break  # success
                        f.unlink()  # remove partial downloads
                except MemoryError:
                    raise  # Re-raise immediately - no point retrying if insufficient disk space
                except Exception as e:
                    # Only on the terminal failure: retries resume the partial file via a Range request, but leaving
                    # one behind makes the `not f.is_file()` guard above serve it as a complete cache hit forever.
                    if i == 0 and not is_online():
                        f.unlink(missing_ok=True)
                        raise ConnectionError(
                            emojis(f"❌  Download failure for {uri}. Environment may be offline.")
                        ) from e
                    status = getattr(getattr(e, "response", None), "status_code", None)
                    # A 4xx other than timeout, range or rate-limit answers will not change on retry, so fail fast
                    if i >= retry or (status and status < 500 and status not in {408, 416, 429}):
                        f.unlink(missing_ok=True)
                        limit = "Retry limit reached. " * (i >= retry)
                        raise ConnectionError(emojis(f"❌  Download failure for {uri}. {limit}{e}")) from e
                    delay = 5 * 2**i  # 5, 10, 20 s rides out the ~20-40 s HTTP 5xx bursts GitHub Releases returns
                    LOGGER.warning(f"Download failure, retrying {i + 1}/{retry} in {delay}s {uri}... {e}")
                    time.sleep(delay)
            else:  # no attempt reached `break`, so every one failed size validation and unlinked its download
                raise ConnectionError(emojis(f"❌  Download failure for {uri}. Retry limit reached."))

    if unzip and f.exists() and f.suffix in {"", ".zip", ".tar", ".gz", ".tgz", ".xz", ".bz2", ".txz", ".tbz2"}:
        from zipfile import is_zipfile

        unzip_dir = Path(dir or f.parent).resolve()  # unzip to dir if provided else unzip in place
        if is_zipfile(f):
            unzip_dir = unzip_file(file=f, path=unzip_dir, exist_ok=exist_ok, progress=progress)  # unzip
        elif tarfile.is_tarfile(f):
            LOGGER.info(f"Unzipping {f} to {unzip_dir}...")
            top_level_dirs = set()
            with tarfile.open(f, "r:*") as tar:
                for m in tar:
                    if not (m.isfile() or m.isdir()) or m.issym() or m.islnk():
                        LOGGER.warning(f"Potentially insecure tar member: {m.name}, skipping extraction.")
                        continue
                    m_path = Path(m.name)
                    target = (unzip_dir / m_path).resolve()
                    if (
                        m_path.is_absolute()
                        or ".." in m_path.parts
                        or target.parts[: len(unzip_dir.parts)] != unzip_dir.parts
                    ):
                        LOGGER.warning(f"Potentially insecure file path: {m.name}, skipping extraction.")
                        continue
                    top_level_dirs.update(m_path.parts[:1])  # slice as './' root entries have no parts
                    if m.isdir():
                        target.mkdir(parents=True, exist_ok=True)
                    elif source := tar.extractfile(m):
                        target.parent.mkdir(parents=True, exist_ok=True)
                        with source, open(target, "wb") as out:  # 'f' is the archive path, deleted below
                            shutil.copyfileobj(source, out)
                        os.utime(target, (m.mtime, m.mtime))  # keep archive mtimes so re-extraction keeps caches valid
            if len(top_level_dirs) == 1:
                unzip_dir /= next(iter(top_level_dirs))  # return the single extracted file or directory
        else:
            unzip_dir = f  # not a zip or tar, i.e. an HTML error page or plain gzip, return the file
        if delete:
            f.unlink()  # remove archive
        return unzip_dir
    return f

def get_github_assets(
    repo: str = "ultralytics/assets",
    version: str = "latest",
    retry: bool = False,
) -> tuple[str, list[str]]:
    """Retrieve the specified version's tag and assets from a GitHub repository.

    If the version is not specified, the function fetches the latest release assets.

    Args:
        repo (str, optional): The GitHub repository in the format 'owner/repo'.
        version (str, optional): The release version to fetch assets from.
        retry (bool, optional): Flag to retry the request once in case of a failure.

    Returns:
        tag (str): The release tag, or an empty string if the request fails.
        assets (list[str]): A list of asset names, or an empty list if the request fails.

    Examples:
        >>> tag, assets = get_github_assets(repo="ultralytics/assets", version="latest")
    """
    import requests  # scoped as slow import

    if version != "latest":
        version = f"tags/{version}"  # i.e. tags/v6.2
    url = f"https://api.github.com/repos/{repo}/releases/{version}"
    attempts = 2 if retry else 1  # retry once on transient network errors or non-200 responses
    for attempt in range(attempts):
        try:
            r = requests.get(url, timeout=30)  # github api
        except requests.exceptions.RequestException as e:
            if attempt < attempts - 1:
                continue  # transient network error, try again
            LOGGER.warning(f"GitHub assets check failure for {url}: {e}")
            return "", []
        if r.status_code == 200 or r.reason == "rate limit exceeded":  # do not retry 403 rate limits
            break
    if r.status_code != 200:
        LOGGER.warning(f"GitHub assets check failure for {url}: {r.status_code} {r.reason}")
        return "", []
    data = r.json()
    return data["tag_name"], [x["name"] for x in data["assets"]]  # tag, assets i.e. ['yolo26n.pt', 'yolo11s.pt', ...]

def attempt_download_asset(
    file: str | Path,
    repo: str = "ultralytics/assets",
    release: str = "v8.4.0",
    **kwargs,
) -> str:
    """Attempt to download a file from GitHub release assets if it is not found locally.

    Args:
        file (str | Path): The filename or file path to be downloaded.
        repo (str, optional): The GitHub repository in the format 'owner/repo'.
        release (str, optional): The specific release version to be downloaded.
        **kwargs (Any): Additional keyword arguments passed to `safe_download`.

    Returns:
        (str): The path to the local or downloaded file.

    Examples:
        >>> file_path = attempt_download_asset("yolo26n.pt", repo="ultralytics/assets")
    """
    from ultralytics.utils import SETTINGS  # scoped for circular import

    # YOLOv3/5u updates
    file = Path(checks.check_yolov5u_filename(str(file).strip().replace("'", "")))
    if file.exists():
        return str(file)
    elif (SETTINGS["weights_dir"] / file).exists():
        return str(SETTINGS["weights_dir"] / file)
    else:
        # URL specified
        name = Path(parse.unquote(str(file))).name  # decode '%2F' to '/' etc.
        download_url = f"https://github.com/{repo}/releases/download"
        if str(file).startswith(("http:/", "https:/")):  # download
            url = str(file).replace(":/", "://")  # Pathlib turns :// -> :/
            file = url2file(name)  # parse authentication query strings
            if Path(file).is_file():
                LOGGER.info(f"Found {clean_url(url)} locally at {file}")  # file already exists
            else:
                safe_download(url=url, file=file, min_bytes=1e5, **kwargs)

        elif repo == GITHUB_ASSETS_REPO and name in GITHUB_ASSETS_NAMES:
            safe_download(url=f"{download_url}/{release}/{name}", file=file, min_bytes=1e5, **kwargs)

        else:
            tag, assets = get_github_assets(repo, release)
            if not assets:
                tag, assets = get_github_assets(repo)  # latest release
            if name in assets:
                safe_download(url=f"{download_url}/{tag}/{name}", file=file, min_bytes=1e5, **kwargs)

        return str(file)

def download(
    url: str | list[str] | Path,
    dir: str | Path | None = None,
    unzip: bool = True,
    delete: bool = False,
    curl: bool = False,
    threads: int = 1,
    retry: int = 3,
    exist_ok: bool = False,
) -> None:
    """Download files from specified URLs to a given directory.

    Supports concurrent downloads if multiple threads are specified.

    Args:
        url (str | list[str] | Path): The URL or list of URLs of the files to be downloaded.
        dir (str | Path, optional): The directory where the files will be saved. Defaults to the current working
            directory.
        unzip (bool, optional): Flag to unzip the files after downloading.
        delete (bool, optional): Flag to delete the zip files after extraction.
        curl (bool, optional): Flag to use curl for downloading.
        threads (int, optional): Number of threads to use for concurrent downloads.
        retry (int, optional): Number of retries in case of download failure.
        exist_ok (bool, optional): Whether to overwrite existing contents during unzipping.

    Examples:
        >>> download("https://github.com/ultralytics/assets/releases/download/v0.0.0/bus.jpg", dir="path/to/dir")
    """
    dir = Path(dir or Path.cwd())
    dir.mkdir(parents=True, exist_ok=True)  # make directory
    urls = [url] if isinstance(url, (str, Path)) else url
    if threads > 1:
        LOGGER.info(f"Downloading {len(urls)} file(s) with {threads} threads to {dir}...")
        with ThreadPool(threads) as pool:
            pool.map(
                lambda x: safe_download(
                    url=x[0],
                    dir=x[1],
                    unzip=unzip,
                    delete=delete,
                    curl=curl,
                    retry=retry,
                    exist_ok=exist_ok,
                    progress=True,
                ),
                zip(urls, repeat(dir)),
            )
            pool.close()
            pool.join()
    else:
        for u in urls:
            safe_download(url=u, dir=dir, unzip=unzip, delete=delete, curl=curl, retry=retry, exist_ok=exist_ok)