from spark_queries import utils

Q_NUM = 16


def q():
    query_str = f"""
    select
        p_brand,
        p_type,
        p_size,
        count(distinct ps_suppkey) as supplier_cnt
    from
        partsupp,
        part
    where
        p_partkey = ps_partkey
        and p_brand <> 'Brand#45'
        and p_type not like 'MEDIUM POLISHED%'
        and p_size in (49, 14, 23, 45, 19, 3, 36, 9)
        and ps_suppkey not in (
            select
                s_suppkey
            from
                supplier
            where
                s_comment like '%Customer%Complaints%'
        )
    group by
        p_brand,
        p_type,
        p_size
    order by
        supplier_cnt desc,
        p_brand,
        p_type,
        p_size
	"""

    utils.get_part_supp_ds()
    utils.get_part_ds()
    utils.get_supplier_ds()

    q_final = utils.get_or_create_spark().sql(query_str)

    utils.run_query(Q_NUM, q_final)


if __name__ == "__main__":
    q()