Обсуждение: Split query

Hi,

On Tue, 2004-11-02 at 07:47, Katsaros Kwn/nos wrote:
> Hi!
>
> I want to parse a query (SFW) and create new queries: one for each
> table referenced, if that possible.
> I've written a function for this in the Query struct level after
> parse&rewrite (not very general ofcourse :-)) but I would like to
> know  whether there is any code written for this purpose at any level.
> Are there any algorithms for this?
> Following, is there any code or at least some algorithm about merging
> the individual results returned from these queries?

The latter is easy. Consider:

SELECT ... FROM
(SELECT ... first query here) AS table_a
JOIN
(SELECT ... second query here) AS table_b
USING (same_column);

(or ON table_a.column = table_b.column)

I dont think you should think about joining results
outside the database. You have Postgresql here, remember :-)

Regards
Tino

Well, the second question is trivial indeed :-) ,even though I had
pipelining in mind when writting it...
(which is not so query oriented but rather plan/executor oriented)
However, I'm more interested in the first question.

Thank you very much,
Ntinos


Tino Wildenhain wrote:
> Hi,
>
> On Tue, 2004-11-02 at 07:47, Katsaros Kwn/nos wrote:
> > Hi!
> >
> > I want to parse a query (SFW) and create new queries: one for each
> > table referenced, if that possible.
> > I've written a function for this in the Query struct level after
> > parse&rewrite (not very general ofcourse :-)) but I would like to
> > know  whether there is any code written for this purpose at any level.
> > Are there any algorithms for this?
> > Following, is there any code or at least some algorithm about merging
> > the individual results returned from these queries?
>
> The latter is easy. Consider:
>
> SELECT ... FROM
> (SELECT ... first query here) AS table_a
> JOIN
> (SELECT ... second query here) AS table_b
> USING (same_column);
>
> (or ON table_a.column = table_b.column)
>
> I dont think you should think about joining results
> outside the database. You have Postgresql here, remember :-)
>
> Regards
> Tino
>