Function _unwrap_protocol

python/ray/data/datasource/path_util.py:435–466

Slice off any protocol prefixes on path.

(path)

def _unwrap_protocol(path):
    """
    Slice off any protocol prefixes on path.
    """
    if sys.platform == "win32" and _is_local_windows_path(path):
        # Represent as posix path such that downstream functions properly handle it.
        # This is executed when 'file://' is NOT included in the path.
        return pathlib.Path(path).as_posix()

    parsed = urlparse(path, allow_fragments=False)  # support '#' in path
    params = ";" + parsed.params if parsed.params else ""  # support ';' in path
    query = "?" + parsed.query if parsed.query else ""  # support '?' in path
    netloc = parsed.netloc
    if parsed.scheme == "s3" and "@" in parsed.netloc:
        # If the path contains an @, it is assumed to be an anonymous
        # credentialed path, and we need to strip off the credentials.
        netloc = parsed.netloc.split("@")[-1]

    parsed_path = parsed.path
    # urlparse prepends the path with a '/'. This does not work on Windows
    # so if this is the case strip the leading slash.
    if (
        sys.platform == "win32"
        and not netloc
        and len(parsed_path) >= 3
        and parsed_path[0] == "/"  # The problematic leading slash
        and parsed_path[1].isalpha()  # Ensure it is a drive letter.
        and parsed_path[2:4] in (":", ":/")
    ):
        parsed_path = parsed_path[1:]

    return netloc + parsed_path + params + query


def _is_http_url(path) -> bool: