Apply a select for every element of a column - sql

Question

I would like to update third with a select that uses column first
|first #|second #|third #|
|_______|________|_______|
|___1___|___1____|_null__|
|___5___|___2____|_null__|
|___3___|___6____|_null__|
|___2___|___4____|_null__|
In pseudo code:
for row in table:
row.third = result_of_a_select(row.first)
What is the equivalent on SQL?
My wrong attempt:
update example_table
set third=
(
SELECT MAX(CDARTI) FROM
(
SELECT A.CDARTI
FROM PGMR.UT_ART_CODALT T,
PGMR.MRP_ARCH_ARTICOLI A
WHERE
A.CDARTI = T.CDARTI AND
T.CDARTI = first
UNION
SELECT A.CDARTI
FROM PGMR.UT_ART_CODALT T,
PGMR.MRP_ARCH_ARTICOLI A
WHERE
A.CDARTI = T.CDARTI_A AND
T.CDARTI = first
UNION
SELECT A.CDARTI
FROM PGMR.UT_ART_CODALT T,
PGMR.MRP_ARCH_ARTICOLI A
WHERE
A.CDARTI = T.CDARTI AND
T.CDARTI_A = first
UNION
SELECT A.CDARTI
FROM PGMR.UT_ART_CODALT T,
PGMR.MRP_ARCH_ARTICOLI A
WHERE
A.CDARTI = T.CDARTI_A AND
T.CDARTI = (SELECT T2.CDARTI FROM PGMR.UT_ART_CODALT T2 WHERE T2.CDARTI_A = first)
)
);
commit;

Answer 1

Works for me.
update t1 set third = (select third from t2 where t2.first = t1.first);
Fiddle (switched to MySql DB since I can't make Oracle work in it, but tested against my local Oracle as well).
The problem is not in the approach, but in your query. To fix that, you'll need to provide a Minimal, Complete, and Verifiable example.

Answer 2

A simplified test case:
SQL> create table yourTable (first, second, third) as (
2 select 1, 1, cast (null as number) from dual union all
3 select 5, 2, cast (null as number) from dual union all
4 select 3, 6, cast (null as number) from dual union all
5 select 2, 4, cast (null as number) from dual
6 );
Table created.
SQL> update yourTable t
2 set third = (select t.first * 2 from dual);
4 rows updated.
SQL> select * from yourTable;
FIRST SECOND THIRD
---------- ---------- ----------
1 1 2
5 2 10
3 6 6
2 4 4
SQL>

Answer 3

To make it a slightly more interesting/illustrative example I'm updating the third column with the value from a mapping table and have included duplicate values.
You can use MERGE and match on the pseudocolumn ROWID:
Oracle Setup:
CREATE TABLE table_name ( first, second, third ) AS
SELECT 1, 1, CAST( NULL AS NUMBER ) FROM DUAL UNION ALL
SELECT 5, 2, NULL FROM DUAL UNION ALL
SELECT 3, 6, NULL FROM DUAL UNION ALL
SELECT 2, 4, NULL FROM DUAL UNION ALL
SELECT 3, 4, NULL FROM DUAL;
CREATE TABLE table_name_map ( first, value ) AS
SELECT 1, 9 FROM DUAL UNION ALL
SELECT 2, 8 FROM DUAL UNION ALL
SELECT 3, 7 FROM DUAL UNION ALL
SELECT 4, 6 FROM DUAL UNION ALL
SELECT 5, 5 FROM DUAL;
Update:
MERGE INTO table_name dst
USING ( SELECT t.ROWID AS ri,
m.value
FROM table_name t
INNER JOIN table_name_map m
ON ( t.first = m.first )
) src
ON ( src.ri = dst.ROWID )
WHEN MATCHED THEN
UPDATE SET third = src.value;
Result:
FIRST SECOND THIRD
----- ------ -----
1 1 9
5 2 5
3 6 7
2 4 8
3 4 7