PL/SQL ORACLE query as Result - oracle

Question

How to write SQL Query for below condition
Result must be in PL/SQL Query:
MY_TABLE and data is Like bellow:
| sl. no | col1 | col2 | col3 |col4 | col5 ---col30| col41|col42|....col50
+------ +------+------ +------ +------+
| 1001 | 50 | 101 | 12 | 40 |
| 1002 | 30 | 250 | 80 | |
| 1003 | 40 | 150 | 90 | |
| 1004 | 50 | 250 | 20 | |
| 1005 | 70 | 300 | 30 | 50 |
| 1006 | 80 | 400 |
col1, col2,col3,...col30
I want to retrieve col1..to col30 data(value) against sl.no (where condition is sl.no) if col data is available. Oracle 9I
as row wise
only in PL/SQL oracle 9i
result like:
1001 | 50
1001 | 101
1001 | 12
1001 | 40
1002 | 30
1002 |250
1002 | 80
1003 | 40
1003 | 150
1003 | 90
1004 | 50
1004 | 250
1004 | 20
1005 |70
1005 |300
1005 |30
1005 |50
1006 |80
1006 |400

Answer 1

select *
from (
select no, decode( rn, 1, col1, 2, col2, 3, col3, 4, col4, 5, col5, ...., 30, col30) col
from table_name, (select rownum rn from all_objects where rownum <= 30 )
)
where col is not null
or
select *
from (
select no, decode( rn, 1, col1, 2, col2, 3, col3, 4, col4, 5, col5, ...., 30, col30) col
from table_name, (select rownum rn from dual connect by rownum <= 30 )
)
where col is not null
the second one is better but I don't remember will it works in 9i or not

Answer 2

To unpivot the pre-11g way, you would need to do something like:
WITH my_table AS (SELECT 1001 sl_no, 50 col1, 101 col2, 12 col3, 40 col4 FROM dual UNION ALL
SELECT 1002 sl_no, 30 col1, 250 col2, 80 col3, NULL col4 FROM dual UNION ALL
SELECT 1003 sl_no, 40 col1, 150 col2, 90 col3, NULL col4 FROM dual UNION ALL
SELECT 1004 sl_no, 50 col1, 250 col2, 20 col3, NULL col4 FROM dual UNION ALL
SELECT 1005 sl_no, 70 col1, 300 col2, 30 col3, 50 col4 FROM dual UNION ALL
SELECT 1006 sl_no, 80 col1, 400 col2, NULL col3, NULL col4 FROM dual),
dummy AS (SELECT LEVEL lvl
FROM dual
CONNECT BY LEVEL <= 4 -- number of columns to unpivot
),
results AS (SELECT mt.sl_no,
d.lvl,
CASE WHEN d.lvl = 1 THEN 'COL1'
WHEN d.lvl = 2 THEN 'COL2'
WHEN d.lvl = 3 THEN 'COL3'
WHEN d.lvl = 4 THEN 'COL4'
END col_name,
CASE WHEN d.lvl = 1 THEN col1
WHEN d.lvl = 2 THEN col2
WHEN d.lvl = 3 THEN col3
WHEN d.lvl = 4 THEN col4
END col_value
FROM my_table mt
CROSS JOIN dummy d)
SELECT sl_no,
col_name,
col_value
FROM results
WHERE col_value IS NOT NULL
ORDER BY sl_no, lvl;
SL_NO COL_NAME COL_VALUE
---------- -------- ----------
1001 COL1 50
1001 COL2 101
1001 COL3 12
1001 COL4 40
1002 COL1 30
1002 COL2 250
1002 COL3 80
1003 COL1 40
1003 COL2 150
1003 COL3 90
1004 COL1 50
1004 COL2 250
1004 COL3 20
1005 COL1 70
1005 COL2 300
1005 COL3 30
1005 COL4 50
1006 COL1 80
1006 COL2 400
The dummy subquery is used to generate a set of rows that is the same number as the number of columns to be unpivoted. In your example, it's 4.
We then cross join that to the table - this means each row in the table is duplicated 4 times - one for each column to be unpivoted.
Once we have that, we say "display the first column to be unpivoted on the first row, the second column on the second row, the third column on the third row, etc..."
I've also included a column which lists the column name that the corresponding value came from; I know you didn't ask for this info in your question, but it's usually required, and easy enough to produce.
Once we have unpivoted the columns, we then remove any values that are null.