Re: [HACKERS] Questions on using multi-byte character in a field of a table (BIG5)

>    Really thanks for your reply... I have been waiting for reply for a while.
>    I realy want to help out with this but I have some problems.
>    1. I am not familiar with ODBC Standards and internal.
>    2. I am not familiar with Language Coding and Convertion.
>
>    But I do used to programming in C, C++, Perl and both under UNIX and VC5
>    Maybe we can cooperate with some other East Asian Countries (Korean,
>Taiwan) to create customized ODBC driver for each language coding we have.
>    Besides, perl do work with Chinese, in fact, I only have problem with ODBC
>now. When I use bind variables in DBD:Pg, all things work. I think this is
>because when assigning variables in perl using single quote instead of double
>quote $var='sth'; would prevent perl from interpreting the value of the
>variable and thus everything works. Of course, I am using EUC_TW as my default
>encoding during initdb and createdb.
>    Can u tell me where can I find more info on language coding and writing
>ODBC dirver. I have read the source of the PsqlODBC and I think they are using
>Crygus GNU toolset.  Can u tell me more about what you guys have done.
>    Thanks.

I talked to the author of localized version of PostgreSQL ODBC Drive
(http://www.insightdist.com/download/) and found that he is
interesting in supporting Big5/EUC_TW. According to him, that
shouldn't be very difficult. I think major problems are:

(1) conversion algorism between EUC_TW and Big5
(2) test data (both Big5 and EUC_TW)
(3) testing (we do not understand Chinese)

For (1), we can refer to
ftp://ftp.ora.com/pub/examples/nutshell/ujip/doc/cjk.inf or whatever
in the Internet, and I believe this would not be a big concern.

So for us real problems are (2) and (3).
--
Tatsuo Ishii.