我需要找到表之间的差异,作为另一个表中不可用的数据(反之亦然)。 我能够找到差异如下:
select *
from (select input_name_id, count(1) as cnt
from Table1
group by input_name_id
) a join
(select input_name_id, count(1) as cnt
from Table2
group by input_name_id
) b
on (a.input_name_id = b.input_name_id)
where a.cnt <> b.cnt
我尝试了数字方式来提取数据,但我无法做到! 所以你的帮助非常感谢。谢谢
答案 0 :(得分:1)
两件事:(1)全外连接; (2)枚举具有相同值的行:
select *
from (select input_name_id, match_id, name,
row_number() over (partition by input_name_id, match_id, name order by name) as seqnum
from Table1
) a full join
(select input_name_id, match_id, name,
row_number() over (partition by input_name_id, match_id, name order by name) as seqnum
from Table2
) b
on a.input_name_id = b.input_name_id and
a.match_id = b.match_id and
a.name = b.name and
a.seqnum = b.seqnum
where a.seqnum is null or b.seqnum is null;
答案 1 :(得分:0)
使用除和结合所有
的简单解决方案select null as a_input_name_id, null as a_matchid, null as a_name, input_name_id as b_input_name_id, matchid as b_matchid, name as b_name
from
(select input_name_id, matchid, name, row_number() over (partition by input_name_id, matchid, name order by input_name_id) as rw_num
from t2
except
select input_name_id, matchid, name, row_number() over (partition by input_name_id, matchid, name order by input_name_id) as rw_num
from t1 ) a
union all
select input_name_id as a_input_name_id, matchid as a_matchid, name as a_name, null as b_input_name_id, null as b_matchid, null as b_name
from
(select input_name_id, matchid, name, row_number() over (partition by input_name_id, matchid, name order by input_name_id) as rw_num
from t1
except
select input_name_id, matchid, name, row_number() over (partition by input_name_id, matchid, name order by input_name_id) as rw_num
from t2) b