Coming out of starmsg, originally I had
...
'|Spec 17-* strip n',
...
Fine, until I decided I wanted to retain leading blanks.  Trying to be
clever, I coded
...
'|Spec a: 17-* . set #0:=Strip(a,"T"," ") print #0 n',
...
Well that produced a surprise:
PIPINX410E ABEND 000000C5 at 01EF7024; PSW 03EC1000 81EF7024 00040005.
PIPINX411I ... In FPLSPSRL; offset 0000089C in FPLSPS 01/23/11 12.44.
PIPINX412I ... GPR0: 01956400 01956FE8 01F4D4C0 00000003.
PIPINX412I ... GPR4: 03E9A1E0 00000001 00000020 01F4AC14.
PIPINX412I ... GPR8: 01F4CD40 01F4A110 01F4CD20 019562E0.
PIPINX412I ... GPRC: 81EF6FFC 01956F98 81EF700C 00000000.
PIPINX413I ... Store 01EF7018: 54208004 4142A000 BFFF4000 4720C030
000001C4.
PIPINX1198I Stage is active.
PIPMSG004I ... Issued from stage 2 of pipeline 3 name "GetSMsgResp".
PIPMSG001I ... Running "Spec a: 17-* . set #0:=Strip(a,"T"," ") print #0
n".
DMSABE141T Addressing exception occurred at 81EF7024 in routine PIPE
CMS
Aha! I thought, I need to explicitly indicate we're dealing with a string -
been there, done that.
So I changed to
...
'|Spec a: 17-* . set #0:=String(Strip(a,"T"," ")) print #0 n',
...
Same result, so for good measure reversed the function nesting: set
#0:=Strip(String(a),"T"," ") and still got the abend.

Curiously, when I use the same constructions in a simple "pipe literal
asdfads        |Spec a: etc" they behave as expected.
This is all with plastic pipes, I haven't tried with the product version.

Settling on expediency, I eventually just said <expletive> the trailing
blanks and changed to a "not chop 16".  My issue resolved.
--
Mike Harding
z/VM System Support

[email protected]
[email protected]
(925) 926-3179 (w)
(925) 323-2070 (c)
/sp

Reply via email to