Ivan Shcheklein pisze:
> Hi Antoni,
> 
> 
> As far as I understand you use Sedna to transform document from one
> representaion into another (XSLT like transformation). It's not the best
> task for XQuery, and unfortunately for Sedna also. It needs a lot of
> space to construct thousands of nodes. However, I should note that it
> works far from optimal in this and anyway we should work to decrease
> temporary memory consuming in such cases.

Indeed, it is an XSLT-like transformation, from XML to RDF, similar in
spirit to the WEESA framework [1], though without the limitation that
all XML data must be in memory. That's why I wanted to use XQuery, to
move all the memory-intensive processing to the database and get a
stream of SAX events that would allow me to produce RDF triples without
any buffering.

Now it seems that I simply moved the problem from RAM to disk, which
renders the "novelty" of my approach somewhat dubious :|.

I posted the original query in the mail that started this thread. Do you
think it would be more efficient to drop the one great query

for $x at $pos in /dblp/*

and instead have a loop and post many smaller queries

for $x at $pos in
/dblp/*[fn:position() >= offset && fn:position() < offset + limit]

or

for $x at $pos in /dblp/*[fn:position = (offset to (limit+offset))]

until some query returns an empty sequence? Will it be more efficient?
Can sedna deal with the fn:position() functions effectively, or will it
internally iterate over all results and return only the matching ones.

Antoni Mylka
[email protected]


[1]
http://www.ifi.uzh.ch/pax/uploads/pdf/publication/7/reif-www05.pdf

------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
_______________________________________________
Sedna-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sedna-discussion

Reply via email to