Privacy Preserving In-Context-Learning Framework for Large Language Models