Question

我有两张桌子 - 员工和经理。一位经理有很多员工。我想使用CONNECT BY和GROUP BY命令在父子关系中显示它。

EMPLOYEE表的结构：

CREATE TABLE employee (
  employee_id INTEGER,
  manager_id INTEGER,
  first_name VARCHAR2(10) NOT NULL,
  last_name VARCHAR2(10) NOT NULL
);

经理表的结构：

CREATE TABLE manager(
  manager_id INTEGER NOT NULL,
  manager_dept VARCHAR2(20) NOT NULL,
  first_name VARCHAR2(30) NOT NULL,
  last_name  VARCHAR2(30) NOT NULL
);

我写了这个查询：

SELECT E.EMPLOYEE_ID, M.MANAGER_ID, E.FIRST_NAME, E.LAST_NAME, CONNECT_BY_ISLEAF, CONNECT_BY_ISCYCLE
FROM EMPLOYEE E
LEFT OUTER JOIN MANAGER M
ON E.MANAGER_ID=M.MANAGER_ID
connect by NOCYCLE m.manager_id=prior e.employee_id
GROUP BY manager_id;

它给出了这个错误：

ORA-00918: column ambiguously defined
00918. 00000 -  "column ambiguously defined"
Error at Line: 6 Column: 10

这个group by子句有什么问题？

Answer 1

您将获得column ambiguously defined，因为oracle无法确定哪个表的manager_id应该用于组，因为您尚未指定它。正确的查询是：

SELECT M.MANAGER_ID
FROM EMPLOYEE E
LEFT OUTER JOIN MANAGER M
ON E.MANAGER_ID=M.MANAGER_ID
connect by NOCYCLE m.manager_id=prior e.employee_id
GROUP BY m.manager_id;

Answer 2

GROUP BY中的列manager_id是不明确的 - 它可以是employee表的manager_id，也可以是manager表的manager_id。

您还需要将剩余的列添加到GROUP BY。然后完整的查询变为：

SELECT 
  E.EMPLOYEE_ID, M.MANAGER_ID, E.FIRST_NAME, 
  E.LAST_NAME, CONNECT_BY_ISLEAF, CONNECT_BY_ISCYCLE
FROM EMPLOYEE E
LEFT OUTER JOIN MANAGER M
  ON E.MANAGER_ID=M.MANAGER_ID
connect by NOCYCLE m.manager_id=prior e.employee_id
GROUP BY 
  E.EMPLOYEE_ID, M.MANAGER_ID, E.FIRST_NAME, 
  E.LAST_NAME, CONNECT_BY_ISLEAF, CONNECT_BY_ISCYCLE;

不能一起使用CONNECT BY和GROUP BY

2 个答案: