Subject | Re: [firebird-support] Re: character sets problems and i18n |
---|---|
Author | Daniel Albuschat |
Post date | 2004-07-05T16:41:36Z |
On Friday 25 June 2004 14:39, peter_jacobi.rm wrote:
I have a question to the ''NONE" character set: Is there any difference
in specifying no character set and specifying 'NONE' explicitly?
It seems to work if I declare the temporary text field as `character set
NONE', but doesn't without. Perhaps the databases default character set or
the connection's character set is used then?
Isn't there some in-depth documentation about firebird character sets?
It seems a rather complicated topic, though...
The trick with the character set cast works perfectly, thanks again.
cu,
Daniel Albuschat
--
eat(this); // delicious suicide
> Daniel,Thanks for your hints, Peter.
>
> Just one more small suggestion until I must switch
> back to day's work:
>
> You can do 'forbidden' charset conversions by intermediary
> casting to character set octets.
>
> An intra-table example:
>
> DDL:
> SQL> create table t2 (c1 varchar(8) , c2 varchar(8) character set
> ISO8859_1);
>
> ...inserting some non-ASCII chars into c1...
>
> This will fail:
> SQL> update t2 set c2 = c1;
>
> This will work:
> SQL> update t2 set c2 = cast (c1 as varchar(8) character set octets);
I have a question to the ''NONE" character set: Is there any difference
in specifying no character set and specifying 'NONE' explicitly?
It seems to work if I declare the temporary text field as `character set
NONE', but doesn't without. Perhaps the databases default character set or
the connection's character set is used then?
Isn't there some in-depth documentation about firebird character sets?
It seems a rather complicated topic, though...
The trick with the character set cast works perfectly, thanks again.
cu,
Daniel Albuschat
--
eat(this); // delicious suicide