Question

Suppose I have these two tables :

TABLEA          TABLEB
-----------     -----------
ID  | NAME      ID | TABLEA_ID  | NAME
1   |  ...      1  | 1          | ...
2   |           2  | 2          | ...
3   |           3  | 2          |
4   |           4  | 2          |
                5  | 3          |
                6  | 3          |
                7  | 4          | 
                8  | 2          | 

I want an SQL SELECT statement that can generate such result when TABLEA.ID = TABLEB.TABLEA_ID, you can note here I don’t care about grouping or ordering, I just want to generate a incremented value for each line of the same TABLEB.TABLEA_ID.

ID | TABLEA_ID | ORDER_INDEX | NAME 
1  | 1         | 0           | ...  
2  | 2         | 0           | ...  
3  | 2         | 1           |      
4  | 2         | 2           |      
5  | 3         | 0           |      
6  | 3         | 1           |      
7  | 4         | 0           |      
8  | 2         | 3           |      

I tried without success to use rownum in several combination of sub-selects to generate the ORDER_INDEX depending on the value in TABLEA_ID.

• Do you have hint to do that in plain SQL, is it even possible with plain SQL.
• Is it possible via a PL/SQL ? And how if possible ?

Thank you very much in advance.

Answer 1

I believe that this is what you want:

SELECT B.ID, B.TABLEA_ID, 
       ROW_NUMBER() OVER(PARTITION BY B.TABLEA_ID ORDER BY B.ID) - 1 ORDER_INDEX,
       B.NAME -- OR A.NAME, its not clear on your question
FROM TABLEB B
LEFT JOIN TABLEA A
ON B.TABLEA_ID = A.ID