Re: AW: Trigger weirdness with L-Date fields

On 8/24/2011 10:35 AM, James Lampert wrote:

Birgitta Hauser wrote:

Are the L-Fields formatted, i.e. is a date format specified in the DDS?
What Date Format is specified in the H-Specs of your trigger program.
I assume both formats must be identical.

The trigger program does not process (or even take notice of) the L-date
fields in any way. As far as it's concerned, they're just completely
unreferenced bytes in a buffer.

We copy the record image from the trigger buffer into an externally
defined data structure, so that we can access the fields, a typical
trigger program action.

But the same thing occurs whether we copy the DS back into the trigger
buffer, or we comment out all code that puts anything back into the
trigger buffer.

Which makes no sense. Unless, of course, you aren't really running the trigger program you think you are running. Which is why you should check the message, check the job stack, etc. Look for programs with QRPLOBJ as the library name.

Start trimming down the possibilities. Does it happen with a completely different file? Create a file with dates, write a no-op trigger program and check the result. If it does not fail, find out what's different. Does the trigger program call another program? Does it pass the data structure? Is the called program corrupting the data buffer?

If you think it is indeed just the fact that there is a date in the file, then you ought to be able to recreate this with a five trigger line program.