It would be useful for users to be able to split large results into many Parquet files. https://duckdb.org/docs/stable/sql/statements/copy