DB2 for i series: passing table row as parameter - db2

Question

I've got a table with many columns:
CREATE TABLE TESTTABLE (ID INT GENERATED ALWAYS AS IDENTITY (START WITH 1 INCREMENT BY 1) PRIMARY KEY, F1 INT, F2 INT, F3 INT, F4 INT, F5 INT, F6 INT, F7 INT, F8 INT, F9 INT, F10 INT)
Then I've got a function that takes row-data and performs some type of operation. For the sake of this example, i just sum them up (not caring of null values).
CREATE OR REPLACE FUNCTION ROWFUNCTION(ID INT, F1 INT, F2 INT, F3 INT, F4 INT, F5 INT, F6 INT, F7 INT, F8 INT, F9 INT, F10 INT) RETURNS INT
LANGUAGE SQL
CONTAINS SQL
RETURN F1 + F2 + F3 + F4 + F5 + F6 + F7 + F8 + F9 + F10;
Then I've got some sort of function that calls this row-level function for each record.
CREATE OR REPLACE FUNCTION DOSOMETHING() RETURNS INT
LANGUAGE SQL
READS SQL DATA
BEGIN
DECLARE RES INT DEFAULT 0;
FOR C CURSOR FOR SELECT * FROM TESTTABLE DO
SET RES = RES + ROWFUNCTION(ID, F1, F2, F3, F4, F5, F6, F7, F8, F9, F10);
END FOR;
RETURN RES;
END
Now the question is this. Having a function with a parameter for each table column is cumbersome.
It would be nice to have a 'record type' that just correspond to the table columns. This way I could pass a single parameter to the function. Maybe I missed something, but I couldn't find any solution to this problem in DB2 for i series.
If this solution is not possible, are there any workarounds?
I considered passing the ID, that allows the function to identify and load the record:
CREATE OR REPLACE FUNCTION ROWFUNCTION(ID INT) RETURNS INT
LANGUAGE SQL
READS SQL DATA
FOR SELECT * FROM TESTTABLE WHERE ID=ROWFUNCTION.ID DO
RETURN F1 + F2 + F3 + F4 + F5 + F6 + F7 + F8 + F9 + F10;
END FOR;
and the call would be:
...
SET RES = RES + ROWFUNCTION(ID);
...
Despite being almost exactly what i was looking for, this solution has the major drawback of generating a second access to the table for each record instead of using the data already retrieved by the main loop. The efficiency loss maybe negligible in most cases but in case we want also modify the records, having a double access to the table seems to add a unnecessary level of complexity. This is why I would welcome other solutions to this problem.

Answer 1

You may try the ARRAY data type, if you operate with just a set of integers.
CREATE TYPE INTARR AS INT ARRAY[]
CREATE OR REPLACE FUNCTION TEST_ARR (P_ARR INTARR)
RETURNS INT
BEGIN
RETURN (SELECT SUM (ID) FROM UNNEST (P_ARR) AS T (ID));
END
CREATE TABLE JUST_4_RES (A INT)
BEGIN
DECLARE V_RES INT DEFAULT 0;
DECLARE V_ARR INTARR;
FOR V_C AS
SELECT F1, F2, F3
FROM
(
VALUES
(1, 1, 1)
, (1, 1, 1)
) T (F1, F2, F3)
DO
SET V_ARR = ARRAY [V_C.F1, V_C.F2, V_C.F3];
SET V_RES = V_RES + TEST_ARR (V_ARR);
END FOR;
INSERT INTO JUST_4_RES (A) VALUES (V_RES);
END
SELECT * FROM JUST_4_RES
A
6
fiddle