See the discussion in https://github.com/apache/iceberg-rust/issues/250 . We need a framework to set up its api so that metadata catalogs such as sql, hive can go one.