Re: MS .NET Data Provider for Oracle: Chinese / Cyrillic Characters

Andreas Walthes wrote:
> Hi,
>
> we need to write to and read from an Oracle 8i database.
> The strings we deal with may contain Chinese or Russian
> Unicode characters from a .NET client application.
>
> We use the MS .NET Data Provider for Oracle. We do not use
> the DataSet / DataAdapter approach but use SQL commands,
> via the IDbCommand interface.
>
> Our problem is that the chinese characters are converted
> to reversed question marks (¿¿¿¿) when written to the
> database with an insert command, regardless of whether the
> characters are contained in the SQL string inline or
> passed to the command via the Parameters property.
>
> Does anybody have skills in using the MS .NET Provider for
> Oracle with non-ISO-latin unicode characters?
>
> Any help welcome,
>
> Thank u & merry Xmas
> Andreas

See a previous reply about charactersets in c.d.o.server. Define the correct client characterset (which would be the server, in this case - as you use the IDbCommand interface). UTF-8 jumps to minds - make sure your database can cope with that, and allow fields (columns) to be 50 to 100% wider than a simple characterset would use. I may even be on the narrow here; actually I wouldn't know how many bytes are needed for a Chinese character.

I know I have been in troubble with a name varchar2(30) column, where a 28 character name, containing diarethis (is that correct for é an ï?) did not fit...

Grtz, Frank Received on Sat Dec 21 2002 - 05:00:11 CST