Re: Alternating arrays as D-spec

Justin,

Looks like Barbara didn't notice that there's an unused byte (positions 15, 22, 29, etc) between the array elements in your I-specs.

But, you can just change the q1 in her example to overlay(elem: 6) to skip that empty position, or still a filler field in between the two.

Alternately, you could try something like this instead:

D elem_t ds qualified
D c1 7p 0
D 1a
D q1 2s 0

D rec ds
D 10a
D elem likeds(elem_t) dim(50)


Then you'd refer to stuff as rec.elem(1).q1, rec.elem(1).c1, etc. I kinda like that approach, for some reason...

-SK


On 10/30/2015 4:33 PM, Barbara Morris wrote:

On 10/30/2015 10:38 AM, Justin Taylor wrote:

I have a non externally-described PF that contains two 50 element
arrays, but they alternate.

Example:
I P 11 14 0C1(1)
I 16 17 0Q1(1)
I P 18 21 0C1(2)
I 23 24 0Q1(2)
I P 25 28 0C1(3)
I 30 31 0Q1(3)
I P 32 35 0C1(4)
I 37 38 0Q1(4)
I P 39 42 0C1(5)
I 44 45 0Q1(5)


I need to declare this PF in a procedure, but I'm not sure the best
way to do it. Any suggestions?

You can define a data structure like this, and then code this data
structure as the result field of your READ operation.

dcl-ds ds;
elem dim(50) pos(11);
c1 packed(7) overlay(elem:*next);
q1 char(2) overlay(elem:*next);
end-ds;

Here's the fixed-form version of the DS.

D ds ds
D elem overlay(ds : 11) dim(50)
D c1 7p 0 overlay(elem : *next)
D q1 2a overlay(elem : *next)