Re: Oracle Text: Indexing UTF8 or UTF16

From: Frank van Bortel <frank.van.bortel_at_gmail.com>
Date: Thu, 19 May 2005 16:35:33 +0200
Message-ID: <d6i7v8$m1r$1@news6.zwoll1.ov.home.nl>

Server Applications wrote:
> I have tried to install the database with UTF8 as the characterset, and
> UTF16 as national characterset. I have tried to load the data as UTF8, but I
> dont know wether this succeeded, since my dos-prompt, in which I run
> SQLPLUS, does not show unicode characters (like Arabic or Japanese) very
> well.
>
> I am using Oracle 10g (10.1.0.2.0)
>
>
>
> "Frank van Bortel" <frank.van.bortel_at_gmail.com> wrote in message
> news:d6hqnp$thv$1_at_news1.zwoll1.ov.home.nl...
>

>>What language did you install for Oracle Text?
>>The default is (US) English. You probably want to install
>>multiple languages.
>>
>>If I understand your post correctly, the data loaded is *not* UTF;
>>so actually, this is not about Text, but about character sets (UTF or
>>a fixed-8-byte character set)
>>
>>Please post versions.
>>
>>-- 
>>Regards,
>>Frank van Bortel

>
>
>

Please: no top posting!

OK - so the problem is you cannot get the data loaded correctly.

Points to remember:
- client side character set - what is it?

• use SET NLS_LANGUAGE=WE8AL32UTF8 (or WE8AL16UTF16)
  • what's your code page (chcp) - probably 437 (maybe 850): change to 1251, and see the difference :)
  • can you store the test text in a file, and use SQL*Loader?
• use SET NLS_LANGUAGE=WE8AL32UTF8 sqlldr scott/tiger control=TEST_BLOB.ctl log=TEST_BLOB.log

-- 
Regards,
Frank van Bortel

Received on Thu May 19 2005 - 09:35:33 CDT