C4322LR@UMVMA.BITNET (Len Rugen) (02/20/90)
I haven't used the exit to do this. What we have done is to have a
callable module to make a change to TCTTE that remains in effect for the
duration of the session, or until reset. Just changing the UCTRAN bit
doesn't seem to do anything, the field that works is TCTUECTB.
This could be done from a simple transaction, maybe even a 'prefix' to
the transaction that needs mixed case. Here's a part of the code......
SETON EQU * 00000680
MVI TCTEUCTB,X'01' SET UCTRAN ON 00000690
B RETURN 00000700
SPACE 3 00000710
SETOFF EQU * 00000720
NI TCTEUCTB,X'00' SET UCTRAN OFF 00000730
B RETURN 00000740RJD@UNC.BITNET (919 Joel Dunn 966-5837) (02/21/90)
I will try once again to respond to this note! The mail system has now eaten my last two attempts. I don't have a direct answer to this question, but I think I can offer some relevant comments. I have written a XZCIN exit, and I found that the only way I could debug it effectively was to use WTO's. I suggest that you code a simple exit, put in some WTO's and write the TIOA out to the log and check on the case translation. You should be able to determine where XZCIN is invoked. The subject of UCTRAN under MRO is extremely timely. We have just gone to production using MRO (CICS 2.1.1 under MVS/XA) and had a few problems with this. We have many application programs that link to systems- supplied assembler routines to change TCTEUCTB to dynamically change the case translation in a VTAM environment. When these routines are invoked in an AOR, the change has no effect, since the setting of TCTEUCTB in the TOR determines the UCTRAN translation. To solve this, I modified our routines to start a transaction back in the TOR that does a DFHTC CTYPE=LOCATE to find the matching TCTTE and zap the bit to match the setting in the AOR. It seems to work very well! I am willing to share the code if anyone is interested. Joel Dunn Manager, Communications UNC-Chapel Hill Administrative Data Processing > Can someone tell me precisely when exit XZCIN is driven?? Is it before > or after UCTRAN occurs?? > > I am trying to install some IBM office automation software in an MRO > environment. All our existing applications require UCTRAN to be "on", but > the office automation stuff wants it "off". Among the replies I have > received via our SE (EQUAL items and replies) were two apparently conflicting > answers: One said to write a GLUE for XZCIN to flip the UCTRAN bit in the > TCTTE as required, and the other said that the XZCIN exit won't work because > translation occurs BEFORE xzcin is driven. We are on CICS 1.7.0.