Unicode mismatch RFC client and repository

[Former Member]

Hello,

we are trying to connect an old RFC client application (written in CPP with CRfcLib / RfcLib) to a new SAP ECC 6.0 Unicode system via SAP XI (because customer moved from SAP 4.6 to 6.0 and introduces XI). The client calls an RFC, and even gets a successful return code, however XI receives wrong data from the RFC client, looks like some bytes are shifted. I then noted the following statement in the NW 7.0 help:

"The RFC metadata repository must be compatible with the sender and receivers systems with respect to Unicode.

A Unicode sender or receiver system must have a Unicode RFC metadata repository.

A non-Unicode sender or receiver system must have a non-Unicode RFC metadata repository."

So I assume that XI reads the Unicode metadata and so parses wrong field lengths.

As our old application is non unicode, we have now a mismatch. We would like to avoid to rework the whole CPP source to make it Unicode compatible. Is there some way to get around that above restriction ? We do not have any Non-Unicode SAP system which we could use for metadata lookup. Bypassing XI (unfortunately) is not an option.

PI 7.0 SP18. SAP ECC 6.0 Unicode. CPP code from 2002

CSY

[former_member200962]

I suppose that you are using RFC destination, in there you need to mention Unicode/ Non-Unicode accordingly. Just make a check in SM59 ---> TCP/IP Connection ---> RFC Destination for XI connectivity and check what communication type is enabled. May be this is causing some error.

Regards,

Abhishek.

[Former Member]

No, we are using in XI and RFC sender adapter, and there set Non-Unicode. There we also have to give the RFC metadata repository.