Question

is it possible to increment the field a and b of a table (A.a and A.b) using the value c and d of a different table (B.c B.d) for all the row of A where A.x == B.z?

I’m getting crazy with this query

Answer 1

DB2 and the SQL standard don’t have a FROM clause in an UPDATE statement.
So you have to clearly separate the steps to

1. identify the rows to be modified and to
2. compute the new value.

.

Here is an example:

UPDATE TABLE A
SET A.FLD_SUPV = ( SELECT B.FLD_SUPV
FROM TABLEA A, TABLEB B, TABLEC C,TABLED D
WHERE A.FLD1= B.FLD1
AND A.FLD_DT >= B.FLD_FM_DT
AND A.FLD_DT <= B.FLD_THRU_DT
AND A.FLD_DT > D.FLD_THRU_DT
AND A.FLD_DT < C.FLD_EFF_DT )
WHERE EXISTS ( SELECT B.FLD_SUPV
FROM TABLEA A, TABLEB B, TABLEC C,TABLED D
WHERE A.FLD1= B.FLD1
AND A.FLD_DT >= B.FLD_FM_DT
AND A.FLD_DT <= B.FLD_THRU_DT
AND A.FLD_DT > D.FLD_THRU_DT
AND A.FLD_DT < C.FLD_EFF_DT )

To update two fields you may use an example like this:

UPDATE table1 t1 
 SET (col1, col2) = (
  SELECT col3, col4 
  FROM  table2 t2 
  WHERE t1.col8=t2.col9
 )

The optimizer will see that the sub-queries in the SET and the FROM clause
are identical and it should merge them in the internal execution plan.