Re: [bdbxml] collections

Jim,

DB XML doesn't natively support the notion of collections containing 
collections.  However, I am on the NXQD, or native xml query daemon, 
team.  It uses DB XML as the engine, wrapped up in a light weight 
multi-threaded server(daemon) written in c++.  It uses SOAP(gsoap 
toolkit) for the client/server protocol, and there is a java Db:Xml 
client layer as well.  From a performance issue, DB XML scales far 
better than eXist, Indice, etc.  It will meet you needs easily from the 
sounds of it.

We are releasing version 1 of nxqd Aug 31.  You can find out more at 
http://nxqd.sourceforge.net/, and if you just have to try something out 
sooner, i'm happy to help get you going. 

To clarify just a bit, we will support notions such as:

//*    => search all collections (depth wise first)
//*/y =>  search all leaf collections named y
/x/y =>  your example...

note: in DB XML they call collections containers just to make life 
interesting :)

-tom

ps - down the road, i will likely export the additional functionality 
(/x/y) for DB XML as a library, so it can be embedded as well avoiding 
the client/server overhead,e tc...i also hope to expose it via ICE, 
which is arguably much better than soap.

Jim Endicott wrote:

>I have beem using eXist to investigate using an XML DB and I am now looking for
>a more robust implmentation. I have tried to use the latest release of Oracle
>that supports XQuery but it crashes when I try to run my query. This either
>means that the Oracle XQuery support is not very good or the query that I am
>doing is somehow problematic.
>
>I have a large number of very small (500 bytes) documents organized in a
>hierarchy of collections (at least that is what they are called in eXist) so the
>actual xml files are /x/y/foo.xml /x/y/foo2.xml /x/z/foo3.xml.
>
>In my query I need to be able to have something like "for $i in collection(/x)"
>and have the query go across all 3 of the files or "for $i in collection(/x/y)"
>and have it look at the 2 files. Either of these work fine in eXist (if a little
>slowly) but the first doesn't work at all in Oracle and the second results in an
>internal error maximum open cursors exceeded (there are several hundred xml
>files in the collection when this happens). 
>
>My question is can I do this easily with Berkeley DB XML? Is it possible to make
>it perform?
>
>thanks
>Jim
>
>
>
>
>------------------------------------------
>To remove yourself from this list, send an
>email to [email protected]
>  
>



------------------------------------------
To remove yourself from this list, send an
email to [email protected]