Special Characters not showing correctly in XI from JMS Adapter.

[Former Member]

Hi All,

Scenario: IBM Websphere MQ->XI->CRM (RFC Adapter)

We have a flat file being picked up by JMS adapter and content conversion is done. The scenario works fine except when specific swedish alphabets appear in the flat file. When i checked in XI, these characters appear as rubbish characters.

I understand that we need to use CCSID and put the value 00819 in the CCSID of Destination field for the Comm Channel. However i still did not have any luck after changing this field. Do i need to flush my queues and restart MQ?

Does anyone have any idea what i am missing?

Message was edited by:

lookman

[Former Member · Accepted solution]

Hi Lookman,

I am assuming that your data is not Unicode.

Is the CCSID set properly (this can be done at Queue manger level or at message level)?

To find out what is the current value, read the queue using a utility like Rfhutil and look for header info.

JMS adapter uses the MQ message header to understand the encoding of the data.

Make sure that MQMD.format is set to String if your data is not unicode.

cheers

danus

[Former Member]

Hi All,

The problem was pretty simple really. We tried to do some exotic things like changing the encoding of the text file before it entered the JMS queue. When we sent the file to the queue as-is, the JMS queue did all the hard work, as we have set the codeset for the queue and all ended well. Thanks for all the help!

[Former Member]

Hi,

After some time working on the problem this is what i have:

1) Legacy system has EBCDIC (Sweden/Finland) as codepage. It generates flat files and a program picks these flat files, changes the codepage to UTF-8, and passes it to MQ. At the same time the message is put onto the queue, we provided the CCSID to the MQ (1208 for UTF-8)

2) In the JMS Sender Comm. Channel, we provided the value 1208 in the field Character Code Set ID (CCSID) of destination.

3)We still have problems in XI and CRM. e.g. ÖREBRO becomes EBRO.

Does anyone know why this is happening even when we've changed the flat file contents from the source system to UTF-8?

[Former Member]

Hi Lookman,

We need to make sure that the conversion to UFT-8 is done correctly, I see that as a potential cause. So first, stop the communication channel in XI, send a test file and download it from the Queue manger where XI is supposed to read the message from (using a tool like RFHUtil).

Then use a code page detector tool (http://cpdetector.sourceforge.net/) to make sure that the data is indeed in UTF-8.

By the way, the destination code page field in a Sender communication channel has no effect(http://help.sap.com/saphelp_nw04s/helpdata/en/c1/739c4186c2a409e10000000a155106/content.htm).

cheers

danus

[Former Member]

Hi Danus,

Thanks for the reply.

I have get my MQ guy to check the state of data in the queue and the encoding and it looks correct. I think the problem could likely be at the content conversion. We did not include "Transform.ContentType" parameter for the content conversion. I'm going to include that, test and see if it would solve my problem.

[agasthuri_doss]

Hi,

>>Do i need to flush my queues and restart MQ?

This is not the right way..Still try it out

Regards

Agasthuri Doss