如何从表中选择oracle sql中带有group by子句的嵌套json对象？

假设我有以下陈述：

WITH t AS
(
  SELECT 'A' AS level_0, 'A1' AS level_1_1, 'object_1' AS level_1_2, 'A11' AS level_2_1, 'B11' AS level_2_2 FROM dual
  UNION ALL
  SELECT 'A' AS level_0, 'A1' AS level_1_1, 'object_1' AS level_1_2, 'A12' AS level_2_1, 'B12' AS level_2_2 FROM dual
  UNION ALL
  SELECT 'A' AS level_0, 'A1' AS level_1_1, 'object_1' AS level_1_2, 'A13' AS level_2_1, 'B13' AS level_2_2 FROM dual
  UNION ALL
  SELECT 'A' AS level_0, 'A1' AS level_1_1, 'object_1' AS level_1_2, 'A14' AS level_2_1, 'B14' AS level_2_2 FROM dual
  UNION ALL
  SELECT 'B' AS level_0, 'A1' AS level_1_1, 'object_2' AS level_1_2, 'A11' AS level_2_1, 'B15' AS level_2_2 FROM dual
  UNION ALL
  SELECT 'B' AS level_0, 'A1' AS level_1_1, 'object_2' AS level_1_2, 'A12' AS level_2_1, 'B16' AS level_2_2 FROM dual
  UNION ALL
  SELECT 'B' AS level_0, 'A1' AS level_1_1, 'object_2' AS level_1_2, 'A13' AS level_2_1, 'B17' AS level_2_2 FROM dual
)
SELECT * FROM t

我想要的输出如下所示：

+---------+---------------------------------------------------------------------------------+
| Level 0 |                                      JSON                                       |
+---------+---------------------------------------------------------------------------------+
| A       | {"level_1_1":"A1","object_1":{"A11":"B11","A12":"B12","A13":"B13","A14":"B14"}} |
| B       | {"level_1_1":"A1","object_2":{"A11":"B15","A12":"B16","A13":"B17"}}             |
+---------+---------------------------------------------------------------------------------+

如何使用select语句获得此输出

---

非常感谢你的帮助

您可以使用

分组依据

、

列表标记

和

连接

，如下所示：

SQL> with t as(
  2  select 'A' as level_0, 'A1' as level_1_1, 'object_1' as level_1_2, 'A11' as level_2_1, 'B11' as level_2_2 from dual union all
  3  select 'A' as level_0, 'A1' as level_1_1, 'object_1' as level_1_2, 'A12' as level_2_1, 'B12' as level_2_2  from dual union all
  4  select 'A' as level_0, 'A1' as level_1_1, 'object_1' as level_1_2, 'A13' as level_2_1, 'B13' as level_2_2  from dual union all
  5  select 'A' as level_0, 'A1' as level_1_1, 'object_1' as level_1_2, 'A14' as level_2_1, 'B14' as level_2_2  from dual union all
  6  select 'B' as level_0, 'A1' as level_1_1, 'object_2' as level_1_2, 'A11' as level_2_1, 'B15' as level_2_2  from dual union all
  7  select 'B' as level_0, 'A1' as level_1_1, 'object_2' as level_1_2, 'A12' as level_2_1, 'B16' as level_2_2  from dual union all
  8  select 'B' as level_0, 'A1' as level_1_1, 'object_2' as level_1_2, 'A13' as level_2_1, 'B17' as level_2_2  from dual
  9  )
 10  select LEVEL_0,
 11     '{"level_1_1":"'|| level_1_1 || '","'||level_1_2 ||'":{'
 12     || LISTAGG('"' || level_2_1 || '":"' || level_2_2 || '"', ',')
 13  WITHIN GROUP (ORDER BY level_2_1, level_2_2)
 14     || '}}' AS JSON
 15  from  t
 16  GROUP BY LEVEL_0, level_1_1, level_1_2;

LEVEL_0    JSON
---------- -------------------------------------------------------------------------------
A          {"level_1_1":"A1","object_1":{"A11":"B11","A12":"B12","A13":"B13","A14":"B14"}}
B          {"level_1_1":"A1","object_2":{"A11":"B15","A12":"B16","A13":"B17"}}

SQL>

JSON\u OBJECT（）

和

JSON\u OBJECTAGG（）

函数可以与适当的分组一起使用，例如

SELECT level_0,
       JSON_OBJECT(
                   'level_1_1' VALUE level_1_1,
                    level_1_2  VALUE JSON_OBJECTAGG(level_2_1 VALUE level_2_2)
        )
    AS "Result JSON"                           
  FROM t
 GROUP BY level_1_2, level_1_1, level_0

---

您可能非常感谢您的解决方案。我认为listagg函数中应该有一个分隔符参数，比如listagg（“'”| | level_2_1 | |“：“'| | level_2_2 | |“，”）