原来以为not exists 会比not in 效率高,但在实际运用过程中,发现两表数据差不多的情况下,用另一种方式效率会更好.
原设计语句:
select * from (select * from ex_22222 where exattribute1 = '13') K1 where not exists
(
select * from
(
select ex_22222.entityid from ex_22222,ut_users where exattribute1 = '13'
and ex_22222.entityid = ut_users."UID"
) K2
where K1.ENTITYID = K2.ENTITYID
)
新改进语句:
select K1.ENTITYID,K2.ENTITYID from (select * from ex_22222 where exattribute1 = '13') K1,
(
select * from
(
select ex_22222.entityid from ex_22222,ut_users where exattribute1 = '13'
and ex_22222.entityid = ut_users."UID"
)
) K2
where K1.ENTITYID = K2.ENTITYID(+)
and K2.ENTITYID is null