RE: Re: "Oracle's ROWNUM"

> -----Original Message-----
> From: Tom Lane [mailto:tgl(at)sss(dot)pgh(dot)pa(dot)us]
>
> Hiroshi Inoue <Inoue(at)tpf(dot)co(dot)jp> writes:
> > Oracle doc says.
>
> > If you embed the ORDER BY clause in a subquery and place the ROWNUM
> > condition in the top-level query, you can force the ROWNUM condition
> > to be applied after the ordering of the rows. For example, the
> > following query returns the 10 smallest employee numbers. This
> > is sometimes referred to as a "top-N query":
>
> > SELECT * FROM
> > (SELECT empno FROM emp ORDER BY empno)
> > WHERE ROWNUM < 11;
>
> This thing gets more poorly-defined every time I hear about it!?
>
> Based on what's been said so far, ROWNUM in a WHERE clause means
> something completely different from ROWNUM in the SELECT target list:
> it seems they mean input row count vs output row count, respectively.

They mean output row count AFAIK.

> If I do
> SELECT rownum, * FROM foo WHERE rownum > 10 and rownum < 20;
> will the output rows be numbered 1 to 9, or 11 to 19?

No rows are returned because rownum 2 doesn't exist without rownum 1
and so on.

> If I add a
> condition, say "AND field1 < 100", to the WHERE clause, does the rownum
> count include the rows rejected by the additional clause, or not?

Not.

> And how do you justify any of these behaviors in a coherent fashion?
>
> Dare I ask how it behaves in the presence of GROUP BY, HAVING,
> aggregates, DISTINCT, UNION, ... ?
>

I don't know the details about it unfortunately.

regards,
Hiroshi Inoue