| Author |
Message
|
| zpat |
 Posted: Thu Jul 29, 2004 4:59 am Post subject: XML parsing of MQ messages on a mainframe |
 |
|
Jedi Council
Joined: 19 May 2001
Posts: 5866
Location: UK
|
I would be interested to know how other installations are handling the need to receive MQ messages in XML format on their IBM mainframes and invoke "legacy" CICS transactions that accept fixed offset parameters.
Do you parse the XML on the mainframe (and how)? Or are you parsing it in a message broker (eg WBIMB) off the mainframe?
Anyone using COBOL V3 to parse XML on z/OS? |
|
| Back to top |
|
 |
| PGoodhart |
| Posted: Thu Jul 29, 2004 5:47 am Post subject: |
|
|
Master
Joined: 17 Jun 2004
Posts: 278
Location: Harrisburg PA
|
We are parsing in the Broker. We are either inserting into a database from the broker (which isn't much fun) or parsing the message into a CWF (custom wire frame) fixed length message from an XML message. The problem that we have had in regards to XML beyond the fixed length is the variable number of nodes in the XML message, and how to resolve this into a fixed length message, hence the database inserts at the broker.
_________________
Patrick Goodhart
MQ Admin/Web Developer/Consultant
WebSphere Application Server Admin |
|
| Back to top |
|
|
| RogerLacroix |
| Posted: Thu Jul 29, 2004 8:17 am Post subject: |
|
|
Jedi Knight
Joined: 15 May 2001
Posts: 3264
Location: London, ON Canada
|
Hi,
At my previous client, they ran Mercator on the mainframe.
The Mercator maps reads an XML message from the queue (A), transformed it to a COBOL COPYBOOK or C structure and puts it to another queue(B). The backend applications read the non-XML message from the queue does its processing, and puts the reply message back to a queue ('C). Another Mercator map transforms the non XML message to XML and puts it to a queue (D).
The Mercator transformation engine works really well and has lots and lots of connectors to other components (i.e. Oracle, DB2, ftp, http, email, etc...).
Regards,
Roger Lacroix
Capitalware Inc.
_________________
Capitalware: Transforming tomorrow into today.
Connected to MQ!
Twitter |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Thu Jul 29, 2004 1:11 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
This is the classic case for installing a message transformation broker.
That said, there are redbooks for working with XML on Z/OS.
There are also a variety of message transformation brokers available. If the environment is really as simple as XML->COBOL->XML, and will stay that way, then you could probably away with an extremly light weight transformation engine rather than a full featured broker.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| JT |
| Posted: Thu Jul 29, 2004 1:22 pm Post subject: |
|
|
Padawan
Joined: 27 Mar 2003
Posts: 1564
Location: Hartford, CT.
|
Since it was installed a few months ago, we have a couple of new, IMS applications that are using the XML parsing feature that was made available in Enterprise COBOL for z/OS.
But, the introduction of the XML parsing feature will probably not lead to a large-scale proliferation of XML messages on our mainframe. For the most part our applications are IMS-based. That being the case, the IMS limitation of 64K for the passing of IMS messages and the overhead of the XML tags increasing the size of the payload, restricts which applications can readily use XML structured data.
Our existing CICS, IMS and batch applications continue to have XML messages reconstituted into "COBOL" format by WBIMB. |
|
| Back to top |
|
|
| zpat |
| Posted: Fri Jul 30, 2004 12:58 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5866
Location: UK
|
Do you have an idea of the MIPS (CPU) load incurred by using Enterprise COBOL for XML parsing?
How difficult it is to code a comprehensive XML parser in COBOL V3 that is easy to maintain without requiring source code updates? |
|
| Back to top |
|
|
| Michael000 |
| Posted: Sun Aug 01, 2004 10:32 pm Post subject: |
|
|
Newbie
Joined: 01 Aug 2004
Posts: 1
|
Hi
where I am, we receive the XML on the mainframe where we parse and transform for the mainframe environment (ADABAS/Natural) using a C++ program which can be called from Natural.
So far this approach has worked quite well, although it can be expensive in terms of CPU. XML code was largely found from open source initiatives with the largest amount of programming effort in transforming XML into standard Natural data structures and how to change schemas into datasets.
Michael |
|
| Back to top |
|
|
|
|
