假设你有两张table：

t1:

DB1_count   DB2_count
500         500

和

t2:

DB1_count   DB2_count
376         325

可以添加“表名”列和“匹配”列，并执行联合：

select 'table_1' as Table_name,
       DB1_count, DB2_count,
       case when DB1_count = DB2_count then Y else N end as Match
from t1
union all
select 'table_2' as Table_name,
       DB1_count, DB2_count,
       case when DB1_count = DB2_count then Y else N end as Match
from t2

要动态生成查询，可以尝试：

val names = Seq(("table_1", "t1"), ("table_2", "t2"))
val query = names.map(x => s"""
select '${x._1}' as Table_name,
       DB1_count, DB2_count,
       case when DB1_count = DB2_count then Y else N end as Match
from ${x._2}
""").mkString("\nunion_all\n")

val result = spark.sql(query)