Japanese or Traditional Chinese EUC, and UCS-2 considerations for COBOL embedded SQL applications

Any graphic data that is sent from your application running under an eucJp or eucTW code set, or connected to a UCS-2 database, is tagged with the UCS-2 code page identifier. Your application must convert a graphic-character string to UCS-2 before sending it to the database server.

Likewise, graphic data retrieved from a UCS-2 database by any application, or from any database by an application running under an EUC eucJP or eucTW code page, is encoded using UCS-2. This requires your application to convert from UCS-2 to your application code page internally, unless the user is to be presented with UCS-2 data.

Your application is responsible for converting to and from UCS-2 because this conversion must be conducted before the data is copied to, and after it is copied from, the SQLDA. Db2 does not supply any conversion routines that are accessible to your application. Instead, you must use the system calls available from your operating system. In the case of a UCS-2 database, you might also consider using the VARCHAR and VARGRAPHIC scalar functions.