Chris Nault

Bruce, I appreciate the quick reply. I will take a look at WriteItems and see if it's something I can tackle. I am not super Java saavy so it's a bit daunting. Thanks, Chris

I don't know if this question belongs here on on the JRecord discussion. When using Data2Xml, Is there a way to create attributes on elements in the XML output? I was looking at the code and I see public static final String ATTRIBUTE_PREFIX = "". Is there a way to add a prefix to the fields in a copybook that will cause them to be added as attributes of an element rather than an element? Thanks, Chris

Simon, Thanks for the quick reply. FUNCTION SUBSTITUTE worked like a champ! Chris

Hello all, I am unable to get this code working. 01 WS-TEMP-STRING PIC X(50). ... INITIALIZE WS-TEMP-STRING MOVE 'THIS IS A TEST ñ' TO WS-TEMP-STRING DISPLAY 'TEMP STRING BEFORE: ' WS-TEMP-STRING INSPECT WS-TEMP-STRING REPLACING ALL 'ñ' BY 'n' DISPLAY 'TEMP STRING AFTER: ' WS-TEMP-STRING When it runs, I get: TEMP STRING BEFORE: THIS IS A TEST ñ' TEMP STRING AFTER: THIS IS A TEST ñ' The ñ character doesn't get replaced. I suspect its something wrong with my environment. I am running GNU Cobol 2.2...

I verified that MF doesn't give any warnings for my example above. GC is handling it correctly (ie producing and error). Here are the MF compiler options. * Accepted - verbose * Accepted - nolist * Accepted - DIR(cob_parms) * Accepted - NOFOLDCALLNAME * Accepted - FOLDCOPYNAME'LOWER' * Accepted - XREF * Accepted - SIGN'EBCDIC' * Accepted - NLS * Accepted - DATAMAP * Accepted - COPYLIST * Accepted - opt"3" * Compiling pgm.cob * Total Messages: 0 Here are the precompile/compile commands (I removed...

I agree that the out of bounds error is valid. It concerns me that the MF compiler doesn't complain about it. I have not had time to go back and investigate what compiler flags we could be using on the MF side to let the go unalerted.

I forced an example of the out of bounds error. Below is an exerpt. 05 HOLD-FLBHX-TEXT PIC X(500) VALUE SPACES. then PROCEDURE DIVISION MOVE 'RANDOM TEXT' TO HOLD-FLBHX-TEXT(499:500) gives the following error at compile pgm.cbl: in section '0000-MAINLINE': pgm.cbl: 1375: error: length of 'HOLD-FLBHX-TEXT' out of bounds: 500 MF doesn't complain about it enough to break our build. Maybe it's only issuing a warning? I will investigate.

Simon, Thanks for the info. I gave you some wrong info about the need for the parser. The bit about area_a was needed for a different reason. In our last iteration we started using this unit test framework on the AIX side. It is very picky about where things are. The parser was already in place when we started this migration so we added the GC specific stuff to it. I did try messing around with -fnot-reserved but I couldn't get it to work in all cases. For instance, COPY CPYBOOK REPLACING 'XX' WITH...