Domino to SAP translation produces garbage text

You use an LSXLC agent for a connection from a Lotus® Domino® server to SAP using the SAP Connector, version 1.7.2. The data is in rich text format leaving Domino and enters SAP without any problems. When SAP returns the result of the query from Domino, however, the text is garbled and unreadable.

Example of garbled text:
IN_DOCUMENT_ITEMS_TABMATERIAL = ‘“”˜•€€€€€€€€
IN_DOCUMENT_ITEMS_TABEXT_ITM_NO = 
IN_DOCUMENT_ITEMS_TABCUST_MATRL = €€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€
IN_DOCUMENT_ITEMS_TABPO_NUMBER = €€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€
IN_DOCUMENT_ITEMS_TABPLANT = €€€€
IN_DOCUMENT_ITEMS_TABORDER_QTY = 5000
IN_DOCUMENT_ITEMS_TABSALES_UOM = €€€
IN_DOCUMENT_ITEMS_TABCOND_TYPE = €€€€
IN_DOCUMENT_ITEMS_TABRATE = 0
IN_DOCUMENT_ITEMS_TABPRICE_UNIT = 0
IN_DOCUMENT_ITEMS_TABCOND_UNIT = €€€
IN_DOCUMENT_ITEMS_TABREQ_DEL_DT = 06/12/2006

This problem occurs due to a character translation issue.

To correct the problem you must set the SAPReadCodePage to LCSTREAMFMT_NATIVE as documented in the IBM Lotus Connector for SAP Solutions Version 1.7.2 Release Notes. The relevant excerpt from that Release Notes (readme.txt) is as follows:

"The internal format for text in the SAP connector for scripts (LSXLC) is now by default UTF8. For non-UNICODE systems, this may cause character corruption. To fix this, set your LCConnection object for SAP property SAPReadCodePage to LCSTREAMFMT_NATIVE. This is the default format used in SAP connection documents for LEI and DECS, for all platforms except i5/OS, where the default is LCSTREAMFMT_IBMCP819, which corresponds to US ASCII."