My first approach would be to have the application strip off the extra
whitespace, and then have the application format the string to display
the appropriate whitespace, without ever passing the extra whitespace to
the database.

On Mon, 2005-03-07 at 02:55, Daniel Albuschat wrote:
>
>
>
> Hello,
>
> I have the following problem:
>
> We have a rather huge address database. When an incoming
> call arrives at a workstation, it looks up if the phone number is
> known in this address database, and if so, displays the according
> address on the screen.
> But there's a problem with the lookup:
> There may be numbers that end in direct access numbers,
> like this one:
> +49 55544 55
>
> The last whitespace block is the direct call number.
> When this certain telephone number arrives from the
> workstation's phone, it looks like this:
> +49 5554455
>
> Well, I guess you see the problem. I don't know how
> I could find the number "+49 55544 55" when I only
> have "+49 5554455". Of course, I could try to match
> 5554455, 55544 55 and 5554 455 for 2 or 3 digit direct call numbers.
> But this would only slow things down by the factor of 3...
>
> Now my actual question is, if there's any UDF or built-in function
> to strip off certain characters, like everything that's not a number
> or all whitespaces or so. Or can I build this as a stored procedure
> somehow?
> I would like to either have a redundant field filled in a trigger that holds the
> number without the stripped characters, or directly call it in the
> select statement, like
> select * from addresses where strip(telephone)="+49 5554455"
> or similar.
>
> Any ideas how to solve my problem?
>
> Thanks,
> Daniel
>
> --
> eat(this); // delicious suicide
>
>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
>