Re: RDi 9.5.1.2 with V7R3

Buck,

I got a parsing error on this H-spec:


H debug option(*srcstmt:*nodebugio) indent('|')

Error pop-up: "(" expected instead of this input
<Puzzled-Look>

Which is what I reported to IBM with my PMR 44369, 442, 000

Yup, old punch-card format H-spec, instead of the newer (over 3-years now, Feb 2014 **FREE) ctl-opt [control option statement].


Unknown to me, there were more parsing errors, and the "actual" error on this COMPILED-PRODUCTION code was on this statement:

Select;
When not %ERROR and MIERRC = '00';
...
Endsl;

Error pop-ups:
"ENDSL" insert to complete scope
"WHEN(Opcode" expected instead of this input

Friendly tech support, Mike told me that I could change the %error to %error() to fix it. Which does work! Thanks Mike!

But, I was tracing down error on production with existing compile production code in browse mode.

This program had Free-Format Calculation specs, but not 100% **free with the H, F D old last century Punch-card specs....

This program was last updated in 2007 to the "/free" format by me! <Big Smile> After I was authorized to modernize code finally!

So it appears on the Select/When with the %error, it expects the %Error() for the outline parse to work... WTF?

So, when we Modify code, we "general" update the code if time allows it. But there is no way we can modernize all our code overnight....

Your corporate irony meter goes to 11, that CRACKED ME UP! Thanks Buck! Love that! <Big Smile>
[previous email thread]

PS. I was doing analytical work on this program, examining existing logic. Not check-out, nor changing it...

Should RDI work with existing Compiled Code?
Do I expect too much out of my favorite Development tool, RDI???? <shock>



-Ken Killian-


-----Original Message-----
From: WDSCI-L [mailto:wdsci-l-bounces@xxxxxxxxxxxx] On Behalf Of Buck Calabro
Sent: Friday, April 28, 2017 10:30 AM
To: wdsci-l@xxxxxxxxxxxx
Subject: Re: [WDSCI-L] RDi 9.5.1.2 with V7R3

On 4/28/2017 10:24 AM, Ken Killian wrote:

It does not like %error, because it thinks it should be like Java and be "%error()" with brackets....

The 9k service program I just pulled up has this:

if sqlState = normal;
test(DE) *mdy bday;
if not %error;
birthday = %date(bday:*mdy);
endif;
endif;

Parses fine here.

--
--buck