如果您可以将查询更改为输出
1 1 [Row1]
1 2 [Row2]
1 3 [Row3]
2 1 [Row4]
2 2 [Row5]
2 3 [Row6]
到一个临时表中tmp
,然后你可以使用类似的东西来查询它
select col1, col3 from tmp into tmp1 where col2 = 1;
select col1, col3 from tmp into tmp2 where col2 = 2;
select col1, col3 from tmp into tmp3 where col2 = 3;
select tmp1.col3, tmp2.col3, tmp3.col3 from tmp1, tmp2, tmp3 where tmp1.col1 = tmp2.col1 and tmp1.col1 = tmp3.col1;
您可以生成col1
和col2
使用rownum
,但它是非标准的,它需要对原始查询的输出进行正确排序。
编辑:
如果您不能使用临时表,我假设您可以使用子查询:
select tmp1.col3, tmp2.col3, tmp3.col3 from
(select col1, col3 from (ORIGINAL_QUERY) where col2 = 1) as tmp1,
(select col1, col3 from (ORIGINAL_QUERY) where col2 = 2) as tmp2,
(select col1, col3 from (ORIGINAL_QUERY) where col2 = 3) as tmp3
where tmp1.col1 = tmp2.col1 and tmp1.col1 = tmp3.col1;
并希望优化器很聪明。