Question

我想基于代码列数据返回具有相同数据的多行。我可以使用UNION ALL来实现。没有UNION ALL，有没有办法做到这一点。我有超过10列（S1，S2，.... S10）。在样本数据中，我仅提供了3列（S1，S2，S3）。

SELECT ID , NAME , S1 SCODE 
  FROM TBL1
  WHERE S1 IS NOT NULL
UNION ALL
SELECT ID , NAME , S2 SCODE 
  FROM TBL1
  WHERE S2 IS NOT NULL

实际表格：

ID NAME S1 S2 S3
1  T1   AA BB CC
2  T2   AA    KK

预期结果：

ID  NAME SCODE
1   T1   AA
1   T1   BB
1   T1   CC
2   T2   AA
2   T2   KK

Answer 1

这里是一个选择：连接Sx列，然后进行相反的操作-将它们拆分为行。

SQL> with test (id, name, s1, s2, s3) as
  2    (select 1, 'T1', 'AA', 'BB', 'CC' from dual union all
  3     select 2, 'T2', 'AA', null, 'KK' from dual
  4    ),
  5  temp as
  6    (select id,
  7            name,
  8            s1 || decode(s1, null, null, ':') ||
  9            s2 || decode(s2, null, null, ':') ||
 10            s3 col
 11     from test
 12    )
 13  select id,
 14         name,
 15         regexp_substr(col, '[^:]+', 1, column_value) scode
 16  from temp join table(cast(multiset(select level from dual
 17                                     connect by level<= regexp_count(col, ':') + 1
 18                                    ) as sys.odcinumberlist)) on 1 = 1;

        ID NA SCODE
---------- -- --------------------------------
         1 T1 AA
         1 T1 BB
         1 T1 CC
         2 T2 AA
         2 T2 KK

SQL>

Answer 2

尝试使用UNPIVOT子句：

with
  test (id, name, s1, s2, s3) as (
    select 1, 'T1', 'AA', 'BB', 'CC' from dual union all
    select 2, 'T2', 'AA', null, 'KK' from dual
  )
select id, name, scode
from test
unpivot (scode for cols in (S1, S2, S3)) up;

输出：

| ID | NAME | SCODE |
+----+------+-------+
|  1 | T1   | AA    |
|  1 | T1   | BB    |
|  1 | T1   | CC    |
|  2 | T2   | AA    |
|  2 | T2   | KK    |

使用rextester.com在线进行测试。

根据多列生成行

2 个答案: