FreeTDS Development Group

["Kirill Gribunin" <gribunin AT zelnet.ru>]

Hello, everybody!

I know that UNICODE issues have been discussed a lot in this mailing list,
but I haven't managed to find solution to my particular problem, though I
have rummaged about in all threads.

I have C++ client working under FreeBSD and MS SQL. I call procedures from
C++ client, which require NVARCHAR parameters. My client program operates
with wide char strings (uses wchar_t and wstring). The problem is that I
don't know how to build and execute query which contains wide char
parameters.

I need something like that:

const wchar_t* query = L"EXEC SomeProc 'national_symbols_here'";
tds_submit_query(query);

But tds_submit_query expects only 1-byte strings, so I have no idea how to
implement it.

I know TDS uses UCS-2 encoding in its protocol. I would be glad to convert
all my strings to UCS-2 and pass it to FreeTDS functions but, again, I don't
know how to do that. 
My program is supposed to be multi-language so I can't afford to myself to
stick to one particular codepage. And the most convenient way to handle
multi-language string is using wchar_t based strings.

Have I missed any obvious solution to my problem? Or may be somebody knows
some kind of workaround? I have already started to think about encoding all
string parameters to Base64 and decoding them in database, and this way
definitely looks UGLY for me.

Thank you in advance,
Kirill