Re: Unexpected conversion from AL32UTF8 to WE8ISO8859P1
From: joel garry <joel-garry_at_home.com>
Date: Thu, 8 Jan 2009 13:05:56 -0800 (PST)
Message-ID: <3ca2a645-9f43-4352-a679-ed9c3caaa814_at_x16g2000prn.googlegroups.com>
On Jan 8, 2:16 am, "Jaap W. van Dijk" <j.w.vand..._at_hetnet.nl> wrote:
> Hi,
>
> I have two databases:
>
> the source, version 10.2.0.3.0 with AL32UTF8, and
> the target, version 9.2.0.5.0, with WE8ISO8859P1.
>
> I copy a table, containing single and double byte characters, from the
> source database to the target database with "insert into ... select
> from ..." over a database link. The copy is done in a database
> procedure in the target database that is started in a Windows sqlplus
> session.
>
> What happens is that in the target database all the originally double
> byte characters are converted to an inverted question mark, including
> characters that are defined within WE8ISO8859P1.
>
> For instance: a SMALL E ACUTE (code point C265) in the source database
> is converted to an INVERTED QUESTION MARK (code point BF) in the
> target database.
>
> Why is code point C265 not converted to code point E9, which is SMALL
> E ACUTE in WE8ISO8859P1?
>
> Regards,
> Jaap.
Date: Thu, 8 Jan 2009 13:05:56 -0800 (PST)
Message-ID: <3ca2a645-9f43-4352-a679-ed9c3caaa814_at_x16g2000prn.googlegroups.com>
On Jan 8, 2:16 am, "Jaap W. van Dijk" <j.w.vand..._at_hetnet.nl> wrote:
> Hi,
>
> I have two databases:
>
> the source, version 10.2.0.3.0 with AL32UTF8, and
> the target, version 9.2.0.5.0, with WE8ISO8859P1.
>
> I copy a table, containing single and double byte characters, from the
> source database to the target database with "insert into ... select
> from ..." over a database link. The copy is done in a database
> procedure in the target database that is started in a Windows sqlplus
> session.
>
> What happens is that in the target database all the originally double
> byte characters are converted to an inverted question mark, including
> characters that are defined within WE8ISO8859P1.
>
> For instance: a SMALL E ACUTE (code point C265) in the source database
> is converted to an INVERTED QUESTION MARK (code point BF) in the
> target database.
>
> Why is code point C265 not converted to code point E9, which is SMALL
> E ACUTE in WE8ISO8859P1?
>
> Regards,
> Jaap.
I note there are other problems over dblinks not fixed until 11 (like bug 3419260)... you may have hit a known or unknown bug. Worth trying in 9208 and creating a simple demo for support. Also see if exp/imp works right. Also print out your NLS from the procedure.
jg
-- _at_home.com is bogus. Webex advertising on radio?Received on Thu Jan 08 2009 - 15:05:56 CST