varchar2 vs. nvarchar2 and confusing unicode

I've got an question: What is the best practice to create a unicode supporting column?

I have to set up a Table with unicode support.

Sometime my Table will be installed into an existing Oracle Instance (witch may ot be with the startet with the correct NLS settings).

The Documentation is a litlee bit confusing:

If a create a table with following DDL:

create table MyUtf8Test ( vrchr2000Char varchar2(2000 CHAR) );

and insert a (Java String) with 2000 ¤-Signs, i get following error: Error: java.sql.SQLException: ORA-01704: Zeichenfolge zu lang, SQL State: 42000, Error Code: 1704

Did I cacluate right, that because of the 4000bytes limit, Strings longer than 1333 charackters may produce this error?

What is the best practice to create a unicode supporting column?

I have to store 2000 unicode-signs.
I have an own schema.

I've looked at the oracle website, but it didn't helped.

And where is the difference between varchar2(2000 char) an nvarchar2(2000) ?

Thanks,

Christian Kütbach Received on Tue Jan 19 2010 - 09:38:49 CST