replicate the functionality from delta lake to use topic configuration `tansu.lake.partition` to partition data into iceberg