Rob Scott wrote:
>Try adding the following lines to your test exec after the ISFCALLS
statement :
>Rc=ISFQUERY("INIT")
>Say "Prefix="isfprefix " Owner="isfowner " Dest="isfdest " JES="isfjesname
But when I do so, those variables do not get set:
/* REXX */
arg j
rc = isfcalls('ON')
rc = isfquery('INIT')
say "RC from isfquery('INIT') was" rc
say "Prefix="isfprefix "Owner="isfowner "Dest="isfdest ,
"JES="isfjesname
isfprefix = '*'
address SDSF 'ISFEXEC ST' j
say 'RC from ISFEXEC ST' j 'was' rc';' jname.0 'jobs found'
output is:
********************************* TOP OF DATA
**********************************
READY
%ST VSH1
RC from isfquery('INIT') was 1
Prefix=ISFPREFIX Owner=ISFOWNER Dest=ISFDEST JES=ISFJESNAME
RC from ISFEXEC ST VSH1 was 0; 1 jobs found
READY
END
******************************** BOTTOM OF DATA
********************************
Hm, rc from ISFQUERY was 1. Which seems to be undocumented. Mind you,
Googling "ISFQUERY" gets a total of FOUR hits, which seems unlikely
(especially since none of those is a copy of the SDSF User's Guide on
IBM-there is one on manualzz). I hunted down the actual IBM SDSF User's
Guide, which still doesn't document ISFQUERY return codes, but I found this
code snippet in there:
rc=isfquery()
if rc<sym;>sym;0 then
do
Say "** SDSF environment does not exist, exec ending."
Exit 20
end
Yes, complete with <sym;>sym; -- wow. Anyway, that seems unlikely. What
am I missing this time? I get the same results whether I put the isfquery
before or after the isfcalls (yes, Rob, you said after, but the IBM sample
has it before, so I tried both).
Lionel: I'm thinking that if we get this working, it would be useful output
from your utility, since at a minimum it will save some "Why isn't this
working?" (may lead to "How do I override it?" but at last it won't be a
total mystery!)
Wait, just as I was about to hit send:
https://www.ibm.com/support/pages/apar/PH27154 sounds like "ISFQUERY is
flat-out busted in 2.4, we don't care, will fix eventually maybe". Am I
reading that right? Or is this talking about something else? It would be
nice if it documented what "will not complete successfully" meant-RC=1?
...phsiii
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
