Re: Sequence error in MR

[Hans Boldt <boldt@xxxxxxxxxx>]

Gary Kuznitz wrote:
> Hi Doug,
>
> Great catch.  But I don't know if this is the answer.  The file KAISXMITL1 does 
> not have a descending index or is case-insensitive.  The file KAISLISTL1 is 
> case-insensitive.  I ran this in debug and found no records have been read from 
> the file KAISLISTL1.  The record the program is looking at from the file 
> KAISXMITL1 is "ABEL E".  The next record in the file is "Abiah".  The dump 
> shows all blanks in the FirstNameL field coming from KAISLISTL1 which means it 
> didn't read a record yet because there are no records with a blank in 
> FirstNameL .  It's acting like the case-insensitive is on the wrong file.  Is 
> there a way I can have the LF treet the field as uppercase?

If I'm not mistaken, I believe the match fields logic compares match 
fields using a straight hex comparison. That is, "Abiah" compares 
less than "ABEL E" (EBCDIC), even thought data management serves up 
the latter key first.

The match fields logic does take the alternate collating sequence 
into account, and so if you want lower and upper case alphabetics to 
compare equal, you can code an altseq table. (But I'm sure there's a 
better solution using just DDS or SQL.)

Cheers! Hans