从不直接相关的表的子查询中合并SQL中的列

Question

I have 4 Subqueries that I need to run as a unit and retrieve the distinct results in 1 list. My java application will call this query with Spring Data JPA repository and it should return a List.

## S_NAME based on TABLE_1.Id = 1
SELECT DISTINCT(S_NAME) FROM TABLE_1 WHERE S_ID = 1;

## I_NAME based on TABLE_1.Id = 1
SELECT DISTINCT(I_NAME) FROM I_TABLE WHERE I_ID
IN (SELECT I_ID FROM E_TABLE WHERE S_ID = 1);

## C_NAME based on TABLE_1.Id = 1
SELECT DISTINCT(C_NAME) FROM C_TABLE WHERE C_ID
IN (SELECT C_ID FROM C_TABLE WHERE E_ID
IN (SELECT E_ID FROM E_TABLE WHERE S_ID = 1));

## P_NAME based on TABLE_1.Id = 1
SELECT DISTINCT(P_NAME) FROM P_TABLE WHERE C_ID
IN (SELECT C_ID FROM C_TABLE WHERE E_ID
IN (SELECT E_ID FROM E_TABLE WHERE S_ID = 1));

The columns S_NAME, I_NAME, C_NAME, and P_NAME contain the same data. Simply different names so we know the source of that data. However, for this query I need distinct results from all these columns in 1 column.

Can this be done easily without making a stored procedure? I would, but I don't have the authorization for such things. Same constaint on making a view.

Answer 1

You can use union all to get the value in rows or put them in separate columns. Here is one way:

SELECT (SELECT DISTINCT(S_NAME) FROM TABLE_1 WHERE S_ID = 1
       ) as val1,
       (SELECT DISTINCT(I_NAME) FROM I_TABLE WHERE I_ID IN (SELECT I_ID FROM E_TABLE WHERE S_ID = 1
       ) as val2,
       (SELECT DISTINCT(C_NAME) FROM C_TABLE
        WHERE C_ID IN (SELECT C_ID FROM C_TABLE WHERE E_ID IN (SELECT E_ID FROM E_TABLE WHERE S_ID = 1))
       ) as val3,
       (SELECT DISTINCT(P_NAME) FROM P_TABLE
        WHERE C_ID IN (SELECT C_ID FROM C_TABLE WHERE E_ID IN (SELECT E_ID FROM E_TABLE WHERE S_ID = 1))
       ) as val4;

Although the queries are similar, I am not sure if it is worth trying to combine them.

Answer 2

Use UNION , which implies DISTINCT so you don't need to say that:

SELECT S_NAME FROM TABLE_1 WHERE S_ID = 1
UNION
SELECT I_NAME FROM I_TABLE WHERE I_ID IN (
   SELECT I_ID FROM E_TABLE WHERE S_ID = 1)
UNION
SELECT C_NAME FROM C_TABLE WHERE C_ID IN (
   SELECT C_ID FROM C_TABLE WHERE E_ID IN (
      SELECT E_ID FROM E_TABLE WHERE S_ID = 1))
UNION
SELECT P_NAME FROM P_TABLE WHERE C_ID IN (
   SELECT C_ID FROM C_TABLE WHERE E_ID IN (
      SELECT E_ID FROM E_TABLE WHERE S_ID = 1))
;

FYI: The third query is double-reading C_TABLE , which is a waste. I left it alone, assuming it's a simple mistake setting up the example.

Answer 3

SELECT (S_NAME) FROM TABLE_1 WHERE S_ID = 1;

UNION
SELECT (I_NAME) FROM I_TABLE WHERE I_ID
IN (SELECT I_ID FROM E_TABLE WHERE S_ID = 1);

UNION
SELECT (C_NAME) FROM C_TABLE WHERE C_ID
IN (SELECT C_ID FROM C_TABLE WHERE E_ID
IN (SELECT E_ID FROM E_TABLE WHERE S_ID = 1));
-- HERE UNION will only Take DISTINCT row from the each query no need of Distinct
UNION
SELECT (P_NAME) FROM P_TABLE WHERE C_ID
IN (SELECT C_ID FROM C_TABLE WHERE E_ID
IN (SELECT E_ID FROM E_TABLE WHERE S_ID = 1));