Re: [Q] Daves's top 10 SQL question

davidp3_at_soho.ios.com (David Pais) wrote:

>But if PL*SQL used, let say:
>cursor c1 is select my_any_column
>from my_table
>order by sort_rule;
>and then fetch first N record out of active set-well, it will do the job quite
>nicely

A general solution to finding the row(s) which have N copies in the given table across k columns is as follows:

select P, a1, a2,...,ak from
(

   select count(*) P ,a1,a2,....,ak from tablename

       group by a1,a2,....ak )
where P = N;

Now that we have away to get P and its row value(s), we can then derive any solution based on the value of P.

-Austin Received on Wed Apr 10 1996 - 00:00:00 CEST