This is an ETL (extract / transform / load) question.  The DB2 IMPORT
and LOAD utilities handle the "L" portion of ETL, and as such are not
"smart" - they are written to be as fast as possible.

You need a tool (or a hand written program) to do this for you.  You
could certainly load your 2 files into 2 "staging" tables and use SQL
to do the transformation.

Hi PASCAL,

You can create a stored procedure which will help you getting the
desired result.

1.) First make a temp table with the same structure as your file.

Create table schemaname.temp(Name character(),Age
integer,jobdescription character())

2.) Then import your file to the temp table

3.) Create a stored procedure which will get you the desired result.
I am not good in that, but will give you a template from which you can
proceed or some one from the group can give you an enhancement.

Drop PROCEDURE schemaname.import@
CREATE PROCEDURE schemaname.import(IN procname CHARACTER(2))
   SPECIFIC schemaname.import
   DYNAMIC RESULT SETS 1
P1: BEGIN

   DECLARE SQLCODE               INTEGER;
   DECLARE SPV_ID                INTEGER;

---------------------
--- Declare cursors
----------------------
DECLARE C1 CURSOR WITH HOLD FOR
               SELECT * from schemaname.temp ;
----------------------
---OPEN CURSOR
-----------------------

OPEN C1;

-----------------------------
--FETCH THE VALUE OF CURSOR
------------------------------

FETCH C1 into
        SPV_name,
    SPV_age,
    SPV_jobdescription;

WHILE SQLCODE <> 100 DO

IF (SPV_job='Programmer')
THEN
    SET SPV_ID=1;
END IF;
IF (SPV_job='Manager')
THEN
    SET SPV_JOBID=2;
END IF;
IF (SPV_job='Secretary')
THEN
    SET SPV_ID=3;
END IF;

INSERT INTO schemaname.tableperson values(SPV_name,SPV_age,SPV_JOBID);
INSERT INTO schemaname.tablejob values(SPV_JOBID,SPV_jobdescription);

END WHILE;  

CLOSE C1;
RETURN;

END P1@

Regards
Tariq Rahiman