RFC Problems with VB and Unicode.

[Former Member]

Hi all,

i've a VB application wich connects via RFC to an ERP 5.0 Systems and calls BAPI_DOCUMENT_CHANGE2 to checkin a document to a content server.

For non unicode this works fine.

For a unicode system there will be the following errors:

• ERROR file opened at , SAP-REL 710,0,0 RFC-VER 3 880476 MT-SL

T:2368 Error in program 'doc2bapi': *>RfcInstallUnicodeStrucure: field 11 (ROW, t:8, l:4) offset 502 too small

(at least 504 expected in sizeofR(SAP_UC) == 1 environment)

• ERROR file opened at 20070723 084619 Westeuropäisch, SAP-REL 640,0,113 RFC-VER 3 810862 MT-SL

T:1104 <* RfcDispatch : returns 1:RFC_FAILURE

• ERROR file opened at 20070723 084622 Westeuropäisch, SAP-REL 640,0,113 RFC-VER 3 810862 MT-SL

T:2372 <* RfcDispatch : returns 1:RFC_FAILURE

What is wrong and what is to do ?

Regards

Reinhold

[Former Member · Accepted solution]

Hi Reinhold,

how are you doing ?

-

---

• ERROR file opened at , SAP-REL 710,0,0 RFC-VER 3 880476 MT-SL

T:2368 Error in program 'doc2bapi': *>RfcInstallUnicodeStrucure: field 11 (ROW, t:8, l:4) offset 502 too small

(at least 504 expected in sizeofR(SAP_UC) == 1 environment)

-

---

the unicode system expects a greater offset in the input structure

please generate separate proxies for the unicode system ( in the case that unicode characters expand the input / output structures)

and consume them separately

with respect,

amit

[Former Member · Accepted solution]

Hi amit,

fine thanks.

Its not a proxy for an XI its a standalone VB Program.

I expect that there may be a parameter to tell the unicode Netweaver that this is an nonunicode Connect?

Regards

Reinhold

[Former Member · Accepted solution]

Hi Reinhold,

how are you doing ?

thank you for the helpful answer reward points.

what version of VB and connector are you using to connect ?

with NCo ( SAP .NET Connector) every rfc has a different proxy, when the system changes from unicode to non unicode, the proxy in only certain cases needs to be regeneerated ( when unicode chars are part of the i/p output structures)

if you are using the DCOM connector or connecting with your own connector, you will need to follow the same methodology, which in this case will be to point out the unicode character structure, and accomodate it's length.

alternatively you could write a wrapper eliminating the unicode char from the input / output structure, again for this you will need to know the structure

with respect,

amit