You could maybe set up a single empty dataset, say, SYS1.NULLFILE
or SYS1.DUMMY, that could be used by any job, and refer its DCB
to the original dataset that you are nullifying. For example:
// DD DSN=ORIGINAL.FILE,DISP=SHR
becomes
// DD DSN=SYS1.NULLFILE,DCB=ORIGINAL.FILE,DISP=SHR
so you just have to insert "SYS1.NULLFILE,DCB=" after "DSN=".
> Date: Tue, 8 Mar 2011 06:42:20 -0700
> From: st...@trainersfriend.com
> Subject: Re: QSAM concatenation where the first DD is DUMMY?
> To: IBM-MAIN@bama.ua.edu
>
> On 3/7/2011 10:36 PM, ನಾಗೇಶ್ ಸುಬ್ರಹ್ಮಣ್ಯ (Nagesh S) wrote:
> > Would the effect be the same with DSN=NULLFILE ?
>
> Yes.
>
> I wish I could test; but, I
> > do not have access to a z system these days. The manuals seem to convey
> > NULLFILE and DUMMY do not differ.
> >
> > Of course, one can actually allocate an empty dataset in ISPF 3.2 and then
> > provide this name in the override. ;-)
>
> Well, you can also allocate an empty dataset in a prior
> job step, no need to use ISPF. But, unless you are using
> SMS for this allocation, the file will not have an end
> of data indicator, so in the step where you try to
> reference this empty file for input, it will likely fail
> with an abend.
>
> So, it's a little more complicated. Of course you could
> create an empty file, edit it, including deleting all
> the lines, then exit. Then use that file in multiple
> jobs that use input files with the same DCB attributes.
>
>
> > <snippage>
>
> --
>
> Kind regards,
>
> -Steve Comstock
> The Trainer's Friend, Inc.
