Oracle SQL：子查询在多个表中以不同条件两次查询同一列

Question

This question kind of links to my previous question.

I'm trying to break it down a little to just get the last two columns organized and without duplicates.

I'm attempting to do something similar to the top rated solution on this site.

I'm using excels query builder to try and do this as I am not that familiar beyond simple basics of sql and databases.

The statement is currently:

SELECT WORK_CENTER.EQNO,
(
SELECT
UD_DATA.CUSER AS U1
FROM UD_DATA
FULL OUTER JOIN 
WORK_CENTER 
ON UD_DATA.PARENT_ID = WORK_CENTER.ID 
LEFT JOIN 
UD_COLS 
ON UD_DATA.UD_COLS_ID = UD_COLS.ID
WHERE UD_COLS.ID='16'
),
(
SELECT
UD_DATA.CUSER AS U2
FROM UD_DATA
FULL OUTER JOIN WORK_CENTER ON UD_DATA.PARENT_ID = WORK_CENTER.ID 
LEFT JOIN UD_COLS ON UD_DATA.UD_COLS_ID = UD_COLS.ID
WHERE UD_COLS.ID='17'
)

FROM   
IQMS.UD_DATA UD_DATA
FULL OUTER JOIN 
IQMS.WORK_CENTER WORK_CENTER 
ON WORK_CENTER.ID=UD_DATA.PARENT_ID
LEFT OUTER JOIN 
IQMS.UD_COLS UD_COLS 
ON 
UD_DATA.UD_COLS_ID = UD_COLS.ID 
ORDER BY WORK_CENTER.EQNO

Desired Result:

EQNO | U1   | U2
-----+------+------
001  | 1000 | test1
002  | 2000 | test2
003  | 3000 | test3
004  | 4000 | test4

But I'm currently getting

ORA-01427: single-row subquery returns more than one row Thanks to Pattrick Bacon. Pointed out my overlooking having LEFT JOIN JOIN

In the linked question it was suggested that the cause of the wrong output was due to the two tables being cross joined, I tried to get an understanding of it online and I was unable to really see a solution around it except for trying this.

I haven't changed too much of the subqueries themselves but I'm obviously doing something wrong. would anyone be able to help?

Someone asked for more examples

From my linked question:

This is the desired result(not showing blank columns and apologies for badly formatted table):

EQNO | CNTR_TYPE | CNTR_DESC | MFGCELL | MFG_TYPE | CUSER | U2

001 | 110T-40MM | DEMAG SYSTEM 110-430 | MOLDING | INJECTION | 1000 | test1

002 | 150T-25MM | DEMAG SYSTEM 150-320 | MOLDING | INJECTION | 2000 | test2

003 | 150T-45MM | DEMAG SYSTEM 150-610 | MOLDING | INJECTION | 3000 | test3

005 | 150T-45MM | DEMAG EXTRA  500-610 | MOLDING | INJECTION | 4000 | test4

But what I'm getting is:

EQNO | CNTR_TYPE | CNTR_DESC | MFGCELL | MFG_TYPE | CUSER | U2 | E2

001 | 110T-40MM | DEMAG SYSTEM 110-430 | MOLDING | INJECTION | 1000 | test1 | 001

001 | 110T-40MM | DEMAG SYSTEM 110-430 | MOLDING | INJECTION | 1000 | test2 | 002

001 | 110T-40MM | DEMAG SYSTEM 110-430 | MOLDING | INJECTION | 1000 | test3 | 003

001 | 110T-40MM | DEMAG SYSTEM 110-430 | MOLDING | INJECTION | 1000 | test4 | 005

002 | 150T-25MM | DEMAG SYSTEM 150-320 | MOLDING | INJECTION | 2000 | test1 | 001

002 | 150T-25MM | DEMAG SYSTEM 150-320 | MOLDING | INJECTION | 2000 | test2 | 002

002 | 150T-25MM | DEMAG SYSTEM 150-320 | MOLDING | INJECTION | 2000 | test3 | 003

002 | 150T-25MM | DEMAG SYSTEM 150-320 | MOLDING | INJECTION | 2000 | test4 | 005

003 | 150T-45MM | DEMAG SYSTEM 150-610 | MOLDING | INJECTION | 3000 | test1 | 001

003 | 150T-45MM | DEMAG SYSTEM 150-610 | MOLDING | INJECTION | 3000 | test2 | 002

003 | 150T-45MM | DEMAG SYSTEM 150-610 | MOLDING | INJECTION | 3000 | test3 | 003

003 | 150T-45MM | DEMAG SYSTEM 150-610 | MOLDING | INJECTION | 3000 | test4 | 005

005 | 150T-45MM | DEMAG EXTRA  500-610 | MOLDING | INJECTION | 4000 | test1 | 001 

005 | 150T-45MM | DEMAG EXTRA  500-610 | MOLDING | INJECTION | 4000 | test2 | 002

005 | 150T-45MM | DEMAG EXTRA  500-610 | MOLDING | INJECTION | 4000 | test3 | 003

005 | 150T-45MM | DEMAG EXTRA  500-610 | MOLDING | INJECTION | 4000 | test4 | 005

Union doesn't really work as it just puts the two columns "CUSER" and "CUSER AS U2" together since they are the same column really.

Adding more sample data

Here's a bit from the UD_DATA table

ID     | UD_COLS_ID|PARENT_ID|CUSER
-------+-----------+---------+------
12312  | 16        | 5210    | 1000
12313  | 17        | 5210    | test1
12314  | 16        | 5212    | 2000
12315  | 17        | 5212    | test2
12316  | 16        | 5213    | 3000
12317  | 17        | 5213    | test4
12318  | 16        | 5214    | 4000
12319  | 17        | 5214    | test4

My main issue is just displaying the date from the CUSER column as

      |COL1 | COL2|
      +-----+-----+
      | 1000|test1|
      | 2000|test2|
      | 3000|test3|
      | 4000|test4|

But from the linked question I got something like:

      |COL1 | COL2|
      +-----+-----+
      | 1000|test1|
      | 1000|test2|
      | 1000|test3|
      | 1000|test4|
      | 2000|test1|
      | 2000|test2|
      | 2000|test3|
      | 2000|test4|
      | 3000|test1|
      | 3000|test2|
      | 3000|test3|
      | 3000|test4|

This is what I've been struggling with trying to fix

Answer 1

You're actually after a PIVOT; no need for the joins, etc.

Eg:

WITH ud_data AS (SELECT 12312 id, 16 ud_cols_id, 5210 parent_id, '1000' cuser FROM dual UNION ALL
                 SELECT 12313 id, 17 ud_cols_id, 5210 parent_id, 'test1' cuser FROM dual UNION ALL
                 SELECT 12314 id, 16 ud_cols_id, 5212 parent_id, '2000' cuser FROM dual UNION ALL
                 SELECT 12315 id, 17 ud_cols_id, 5212 parent_id, 'test2' cuser FROM dual UNION ALL
                 SELECT 12316 id, 16 ud_cols_id, 5213 parent_id, '3000' cuser FROM dual UNION ALL
                 SELECT 12317 id, 17 ud_cols_id, 5213 parent_id, 'test3' cuser FROM dual UNION ALL
                 SELECT 12318 id, 16 ud_cols_id, 5214 parent_id, '4000' cuser FROM dual UNION ALL
                 SELECT 12319 id, 17 ud_cols_id, 5214 parent_id, 'test4' cuser FROM dual)
SELECT parent_id,
       col1,
       col2
FROM   (SELECT parent_id,
               ud_cols_id,
               cuser
        FROM   ud_data)
PIVOT (max(cuser) FOR (ud_cols_id) IN (16 AS col1, 17 AS col2))
ORDER BY parent_id;

 PARENT_ID COL1  COL2
---------- ----- -----
      5210 1000  test1
      5212 2000  test2
      5213 3000  test3
      5214 4000  test4