On Thu, Jul 28, 2005 at 04:14:16PM +0200, Alexander Foken wrote:
> Hello,
>
> A large Perl application (10^4 .. 10^5 lines of code) has to work with
> data on an MS SQL Server 2000 containing Unicode data in NVARCHAR
> columns. Perl 5.8.6 (ActivePerl) runs on the same or another Windows
> 2000 Server (usually english or german). The Unicode data may contain
> characters from all over the world, e.g. russian, chinese, arabic or
> hebrew, among the usual latin characters. A separate application also
> accesses the Database, performs some reporting and expects the data to
> be valid Unicode.
>
> DBD::ODBC returns all latin characters, german umlauts and french
> accents, but replaces all non-latin characters with question marks
> ("?", ASCII 63). Also, all data returned from DBD::ODBC has the UTF8
> flag off.
> I spent three days searching Google, DBI docs and FAQ, and DBI and
> DBD::ODBC sources for any hints, but without success.
>
> Any ideas, hints, tips?
I'd suggest you work with the maintainers of DBD::ODBC to fix it and
help save the world.
Tim.