How to use the qsharp.get_available_operations_by_namespace function in qsharp

# To check the full name, we ask the client rather than going through
        # the public API for the qsharp package, so that we can check if the
        # client is currently busy. This can happen if anything below us in
        # meta_path needs to handle an import during an execute; this is the
        # case when ZeroMQ needs to import additional functionality from a
        # Cython module to handle a message.
        # See https://github.com/Microsoft/QuantumLibraries/issues/69 for an
        # example of this failure modality.

        # If the client is busy, we'll want to forego this request to find a
        # module and return None early.
        if qsharp.client.busy:
            return None

        # At this point, we should be safe to rely on the public API again.
        ops = qsharp.get_available_operations_by_namespace()

        if full_name not in ops:
            # We may have been given part of the qualified name of a namespace.
            # E.g., if we try to import Microsoft.Quantum.Intrinsic, we'll
            # see calls with "Microsoft" and "Microsoft.Quantum" first.
            if not any(
                ns_name.startswith(full_name + ".")
                for ns_name in ops
            ):
                return None

        return QSharpModuleLoader()

def __getattr__(self, name):
        ops = qsharp.get_available_operations_by_namespace()
        if name in ops[self._qs_name]:
            op_cls = new_class(name, (QSharpCallable, ))

            # Copy over metadata from the operation's header.
            metadata = qsharp.client.get_operation_metadata(f"{self._qs_name}.{name}")
            op_cls.__doc__ = metadata.get('documentation', '')
            op_cls.__file__ = metadata.get('source', None)
            return op_cls(f"{self._qs_name}.{name}", "workspace")
        raise AttributeError(f"Q# namespace {self._qs_name} does not contain a callable {name}.")