Re: case sensitive solutions requested for odbc with oracle and/or sql-server.

Stef,

If you move to Oracle 8i you can define an index on a function (eg: UPPER(my_mixedcase_column)). Then when you use a comparison of UPPER(my_parameter) =
UPPER(my_mixedcase_column) the DBMS will recognise that it can use the function index to make the comparison... unfortunately I don't use 8i yet, but this is a very handy feature indeed. We have simply created additional columns to hold the mixed case data and made the original column UPPER case.

Regards, David.

Stef ten Bras wrote:

>
> Hi community,
>
> I am looking into solutions with case sensitive for
> storing data, (preferable mixed)
> sorting data, (lower case should be equal to upper case)
> indexing, (should be done case insensitive)
> finding data, (how to do that).
>
> Our application works through an ODBC interface, to an
> Oracle or
> SQL-server database.
> The SQL-server database can be defined case-insensitive.
> This is
> allright by us.
> What steps do we take to use the same solution (client
> application) to
> use Oracle in
> a case-insensitive matter ?
>
> I have tried to track old messages with Deja but didn't
> come up with a
> definitive answer and
> there were a lot of messages I couldn't reach or read
> (language
> barrier).
>
> Suggestions seen so far are using UPPER (or LOWER) but
> UPPER seems to be
> the preferred way,
> or converting the complete database to upper or lower. Or
> using extra
> colums which contain the
> converted data. To my understanding a mixed case index
> will not work if
> you don't use the exactly
> same mixed case while searching.
> What solutions do other database programmers use for this
> problem ?
>
> Thanks for your attention, have a nice day,
> stef,
>
> --
> E-mail:
> +------------------------------------------------------------------+
> | stef ten bras, HISCOM afdeling Systeem Ontwikkeling __ o ~ |
> | Schipholweg 97, Postbus 901, 2300 AX Leiden, Nederland _ _/[._ |
> | Tel. +31 71 5256758, Email stbras_at_hiscom.nl -- (_)>(_) |
> +------------------------------------------------------------------+
>
>
Received on Tue Jun 01 1999 - 18:34:24 CDT