Question

我需要编写一个加入大量查询的查询。它们通过ID连接。问题是在其中一个查询中，我必须选择的值是“单组组函数”，我必须评估必须使用第二个选定列的ON子句的条件，即我我无法添加。我会写一个比我的代码更简单的例子：

Select frst.FirstResult, scnd.FirstResult 
from(
    Select something AS FirstResult, ID_to_compare as SecondResult from table
    Where -- [...]
) frst

join(
    Select something AS FirstResult, ID_to_compare as SecondResult from table2
    Where -- [...] 
) scnd 
ON frst.SecondResult=scnd.SecondResult

join(
    Select something AS FirstResult, ID_to_compare as SecondResult from table3
    Where -- [...]
) trd 
ON scnd.SecondResult=trd.SecondResult

-- [...]

join(
    Select single_group_function(params) AS FirstResult, ID_to_compare as SecondResult from table3
--This select cannot be done because of the group function cannot be executed
    Where -- [...]
) trd 
ON svn.SecondResult=eit.SecondResult

我的问题是我需要比较每个选择的“SecondResult”，但是不允许成员组进行这样的查询。我试过“双”表，但它对我来说真的很乱，我不明白如何使用它。我也尝试过在宏查询中加入它们，但是每一个选择都足够大，真的很难写。我已经从this post中获取了很多想法。你有任何暗示要完成我的任务吗？

对于MT0的请愿，最小完整且可验证的问题是：

Select frst.FirstField, scnd.FirstField, frst.SecondField
from
(Select a1.DUMMY as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1) frst
  join
(Select a1.DUMMY as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1) scnd
  ON frst.SecondField = scnd.SecondField

  join
  (Select sum(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1) trd
  ON trd.SecondField = scnd.SecondField
  ;

我希望在这种情况下

frst.FirstField  scnd.FirstField  frst.SecondField
---------------- ---------------- --------------------
X                X                1

但我得到了

Error en la línea de comandos : 11 Columna : 40
Informe de error -
Error SQL: ORA-00937: la función de grupo no es de grupo único
00937. 00000 -  "not a single-group group function"
*Cause:    
*Action:

Answer 1

这是您的示例子查询：

Select sum(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
from dual a1 join dual a2 on 1=1

那么你期望一个结果行吗？这是没有GROUP BY的聚合。您可以获得所有记录的总和，但是在此结果记录中显示哪个a2.DUMMY？这就是DBMS所抱怨的。

您可能想要的是每a2.DUMMY的总和。因此，请按此列分组：

Select sum(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
from dual a1 join dual a2 on 1=1
GROUP BY a2.DUMMY

Answer 2

在您所做的简化示例中：

  join
  (Select sum(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1) trd
  ON trd.SecondField = scnd.SecondField

您可以进一步简化以查看问题：

Select sum(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
from dual a1 join dual a2 on 1=1;

ORA-00937: not a single-group group function

您必须在group-by子句中包含 not 聚合的任何列：

Select sum(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
from dual a1 join dual a2 on 1=1
group by a2.DUMMY;

现在将获得ORA-01722，因为dummy是一个字符串;使用max代替：

Select frst.FirstField, scnd.FirstField, frst.SecondField
from
(Select a1.DUMMY as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1) frst
  join
(Select a1.DUMMY as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1) scnd
  ON frst.SecondField = scnd.SecondField

  join
  (Select max(a1.DUMMY) as FirstField, a2.DUMMY as SecondField
    from dual a1 join dual a2 on 1=1 group by a2.DUMMY) trd
  ON trd.SecondField = scnd.SecondField
  ;

F F S
- - -
X X X

所以你真正的查询需要做类似的事情：

    Select single_group_function(params) AS FirstResult, ID_to_compare as SecondResult from table3
    Where -- [...]
    group by ID_to_compare

在子查询中

如何在具有“单组组功能”的SELECT子句中使用两个字段

2 个答案: