Re: Another evil of LVLCHK(*NO)

Rob,

Isn't data validated on writes with DDL, and on reads with DDS? Would that explain your observations, or am I just missing your point...

Crispin.

----- Original Message ----- From: <rob@xxxxxxxxx>
To: "Midrange Systems Technical Discussion" <midrange-l@xxxxxxxxxxxx>
Sent: Wednesday, September 23, 2009 10:22 AM
Subject: Re: Another evil of LVLCHK(*NO)

Tried a workaround and I'm a little confused by data management.

Adding a trigger to a file created with DDS can be used to verify the
numeric field.

Added a trigger that basically did this:
if focDesc='INSERT' or focDesc='UPDATE';
// Check for decimal data error from LVLCHK(*NO)
Monitor;
NumericTest=new.mynbr*1;
on-error;
new.mynbr=0;
ENDMON;
endif;

Worked great on a file created with DDS. However, the first program dies
without the trigger being called when the file is created with DDL. Really
switches around that data management, eh?

From DSPFD:

Trigger Description
Trigger name . . . . . . . . . . . . . . . : TRG QSYS_TRIG_ROB____
____LVL1_______000001
Trigger library . . . . . . . . . . . . . : ROB
Trigger state . . . . . . . . . . . . . . : STATE *ENABLED
Trigger status . . . . . . . . . . . . . : *OPERATIVE
Trigger event . . . . . . . . . . . . . . : TRGEVENT *INSERT
Trigger time . . . . . . . . . . . . . . : TRGTIME *BEFORE
Allow repeated change . . . . . . . . . . : ALWREPCHG *YES
Program Name . . . . . . . . . . . . . . : PGM LVL1TRIG
Library . . . . . . . . . . . . . . . . : ROB