> I've found a very interesting writing about character encodings:
> http://w6.metronet.com/~wjm/tomcat/FromFeb11/msg01556.html
>
> I think that the FB driver should be similar to this. If lc_ctype is
> given it should use the appropriate Java encoding when Strings are
> returned.

FB has already support of the character sets and encodings inside. We
suspect (but have not tested it yet) it will return the characters in
the encoding you specified in the DPB. If FB is not able to handle
character conversion inside (which seems to me is all the time when
the encoding you ask is not an encoding of stored data or NONE) will
throw an exception "Cannot convert characters between character
sets". The only requirement is to implement the charset thing
correctly.

I have missed the issue of String to byte[] conversion (thanks to
Alejandro, he pointed this out), and therefore I will not commit code
that allows to pass the DPB params to the CVS until we have the
correct implementation. This might take some time since Easter is
comming.

If anybody wants to make some research, then please, give answers on
the following questions:

- mapping from the InterBase character sets into Java character
encoding, especially to national encodings;

- mapping from UNICODE_FSS to Java character encoding (I suspect this
to be UTF-16LE, but never checked though).

Thanks,
Roman