I am looking at doing that right now, but since I generate up to 1mill
records for some runs of this, it now means creating two of those files,
even though one is only temporary...I will see what it does to my
performance.
Thanks for the idea...looks like we were thinking the same thing!
Thank you and best regards,
Billy Ashton
------ Original Message ------
From "S.Karthik Premnath"
<[email protected]>
To [email protected]
Date 10/3/2025 3:56:45 PM
Subject Re: Generate nines-complement in DFSORT
Just guessing, If on the fly not possible, why not output the data into a
temp file and make another pass where the seqnum can be treated as any
normal integer ?
Karthik Premnath.
On Fri, 3 Oct, 2025, 21:14 Billy Ashton, <
[email protected]> wrote:
After further testing, it looks like I can do all sort of arithmetic on
other numeric fields, but can't seem to do anything on the SEQNUM field
on the fly.
So, this will not work with SEQNUM:
BUILD=(001🙁((SEQNUM,9,ZD,START=K1,INCR=I1),
ADD,-999999999),MUL,-1), . . .
But it will work on other numeric fields like a single record with a
date field on the input.
Any ideas?
Thank you and best regards,
Billy Ashton
------ Original Message ------
From "Billy Ashton" <[email protected]>
To "IBM Mainframe Discussion List" <[email protected]>
Date 10/3/2025 12:16:13 PM
Subject Generate nines-complement in DFSORT
>Hi all! I have a DFSORT that is generating a test file for me, and I
>currently pass in a starting numeric value and increment value to
>generate my first 9-byte field. This has been working fine and has been
>useful for lots of things.
>
>Now, though I want to take that first field, and make it a 9-s
>complement instead of the SEQNUM value. Can anyone tell me if there is
>a way to translate from 0123456789 to 9876543210, or to subtract the
>SEQNUM from 999999999? DFSORT does not allow incrementing by negative
>value for SEQNUM.
>
>Here is the relevant part of the SYSIN:
>OPTION COPY
>OUTFIL REPEAT=CNT,
>BUILD=(001:SEQNUM,9,ZD,START=K1,INCR=I1,
>. . .
>
>So if I start with K1 = 100000001, and with I1 = 4, my first few
>records start:
>100000001...
>100000005...
>100000009...
>
>and I want them to be
>899999998...
>899999994...
>899999991...
>
>Thank you and best regards,
>Billy Ashton
>
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
