IBM Mainframe Forum

Hi,

I have a below JCL. My input file has 5 fields,
Field1: Position 1 to 3(Primary Key)
Field2: Position 3 to 6
Field3: Position 7 to 9
Field4: Position 10 to 17
Field5: Position 18 to 20
I am trying to Sum two fields on the input file based on the Primary Key.



My expectation is when Value for a particular field is empty, take the value from next record. Can this be achieved using SORT?

Thank You.

Hello,

You're getting the output as below, because *SORT retains the first record as the base, and sums the remaining records into this record on the keys specified in the 'SORT FIELDS' keyword:



Before we move to the solution, can you please let us know the expected output in case the input looks as below.
Note that I've shuffled the order of records slightly; but it still remains in-line with your original requirement.

@Aki88, Thanks for the response.
Expected result:
CDE20020020170831ABC
CDE20020020170831CDE

Hello,

Umm, the expected output in this case might be incorrect if we go by the original post's expectation.
Can you please explain the logic on how this output was deduced; once you do the maths, you'll see the rationale behind my first statement.

Once we have this part figured out, then in solution, ICETOOL's extended operators such as FIRST, LAST etc. can be employed; or a logic can be built in DFSORT to achieve what you want.