By: newkid

newkid — Thu, 01 Jul 2010 21:43:46 +0000
If I change your data a little bit:
CREATE table t5 as
SELECT  1 as ID, 'A' as A, to_char(null) as B, 'C' as C , to_char(null) as D
FROM    dual
UNION ALL
SELECT  2, to_char(null), 'B', to_char(null) as C ,'D'
FROM    dual
UNION ALL
SELECT  3, 'A1', 'A' as B, to_char(null) as C, 'D1'
FROM    dual
UNION ALL
SELECT 4, to_char(null) as A ,'B1', 'C1', to_char(null) as D
FROM dual;

SELECT * FROM t5;

        ID A  B  C  D
---------- -- -- -- --
         1 A     C
         2    B     D
         3 A1 A     D1  -------- this row is different
         4    B1 C1
         

your result will be:

A  B  C  D
-- -- -- --
A  B  C  D
A1 B1 C1 D1
   B1 C1

Here's my revised version:
SELECT  MAX(A) AS A,
        MAX(B) AS B,
        MAX(C) AS C,
        MAX(D) AS D
FROM   (SELECT row_number( ) OVER (PARTITION BY LEV ORDER BY ID) AS rn,
                DECODE(lev,1,A) A,
                DECODE(lev,2,B) B,
                DECODE(lev,3,C) C,
                DECODE(lev,4,D) D 
        FROM    t5,(SELECT LEVEL lev FROM dual CONNECT BY LEVEL <=4)
        WHERE
                CASE
                        WHEN    Lev = 1
                        THEN    A
                        WHEN    Lev = 2
                        THEN    B
                        WHEN    Lev = 3
                        THEN    C
                        WHEN    Lev = 4
                        THEN    D
                END IS NOT NULL
        )
GROUP BY RN;

A  B  C  D
-- -- -- --
A  B  C  D
A1 A  C1 D1
   B1
Comments on: How to collapse all nulls for every table column in a SQL statement

By: newkid
