Leveraging Large Language Models for Building Interpretable Rule-Based Data-to-Text Systems