How to use the pyflink.common.dependency_manager.DependencyManager function in pyflink

.. note::

            Please make sure the uploaded python environment matches the platform that the cluster
            is running on and that the python version must be 3.5 or higher.

        .. note::

            The python udf worker depends on Apache Beam (version == 2.15.0),
            Pip (version >= 7.1.0) and SetupTools (version >= 37.0.0).
            Please ensure that the specified environment meets the above requirements.

        :param python_exec: The path of python interpreter.
        :type python_exec: str
        """
        self.get_configuration().set_string(DependencyManager.PYTHON_EXEC, python_exec)

def __init__(self, j_tenv, is_blink_planner, serializer=PickleSerializer()):
        self._j_tenv = j_tenv
        self._is_blink_planner = is_blink_planner
        self._serializer = serializer
        self._dependency_manager = DependencyManager(self.get_config().get_configuration(),
                                                     self._get_j_env())
        self._dependency_manager.load_from_env(os.environ)

def get_python_executable(self):
        """
        Gets the path of the python interpreter which is used to execute the python udf workers.
        If no path is specified before, it will return a None value.

        :return: The path of the python interpreter which is used to execute the python udf workers.
        :rtype: str
        """
        return self.get_configuration().get_string(DependencyManager.PYTHON_EXEC, None)