I have a problem with transferring UTF-8 data from mainframe to Unix.
The mainframe receives the UTF-8(CCSID 1208) data (which contains Chinese data) via MQ and stores it into a DB2 table (tablespace defined with CCSID 1208).
A job runs and extracts this data from DB2 table and puts it into a dataset and then attempts to connect direct this data to a Unix server. (SYSOPTS on the destination file as DATATYPE=BINARY and XLATE=NO)
But, the data when reaches the destinations becomes unreadable.
However, when I tried to connect direct the same data to a local Windows server using the above options, the data was readable.
After some search on internet, I suspect there is some extra SYSOPTS paramter (like CODEPAGE) needs to be specified either on the source file or the destination file specially for transferring it to UNIX. Knowing the source codepage is UTF-8, but unsure of the destination codepage to be specified, I am unable to progress much further.
Any suggestions would be helpful?
I hope I have conveyed my query well. Please let me know if any more information is needed.
I have discussed with the Unix sysadmin team of the vendor who has provided this service.
So I now know the native codepage of the Unix server at destination (it is GB18030)
Though the source native codepage for mainframe is 500. But the source data is translated into UTF-8 (CP 1208)
The questions arise that:
In the COPY step of connect direct, Should I specify the codepage parameter in source file or destination file as CODEPAGE=(UTF-8,GB18030). I assume I also need to specify XLATE parameter as YES ?
Is the direct translation from UTF-8 to GB18030 possible OR there's a need to convert into an intermediate codepage (like its done using NATIONAL-OF and DISPLAY-OF in COBOL) ?
Again, I am unsure if I have conveyed my query well or if its to the right audience. But any inputs should be helpful.