Orbeon be configured to query lucene?

> On 4/10/06, Richard Braman <[hidden email]> wrote:
> > What would the architecture look like?
> > My guess is that a servlet would query the lucene index and display the
> > results as xml, for OPS to present.
>
> Hi Richard,
>
> We haven't done much with Lucene here, but I think Eric van der Vlist
> did, and maybe even wrote a processor for Lucene. Maybe Eric will
> comment directly on this.

> Hi,
>
> Le jeudi 13 avril 2006 à 15:45 -0700, Alessandro Vernet a écrit :
> > On 4/10/06, Richard Braman <[hidden email]> wrote:
> > > What would the architecture look like?
> > > My guess is that a servlet would query the lucene index and display the
> > > results as xml, for OPS to present.
> >
> > Hi Richard,
> >
> > We haven't done much with Lucene here, but I think Eric van der Vlist
> > did, and maybe even wrote a processor for Lucene. Maybe Eric will
> > comment directly on this.
>
> I have done two different implementations integrating Lucene with OPS.
>
> The first one is for XMLfr and you can play with it at
> http://beta.xmlfr.org/orbeon/lucene/cherche .
>
> This one uses an OPS processor to query Lucene indexes. The input of
> this query processor contains the parameters of the query and its output
> is a RSS 1.0 document (see
> http://beta.xmlfr.org/orbeon/lucene/rss?query=orbeon+presentationserver
> for an example of such an output).
>
> The indexing is done outside of OPS through a crontab process running a
> Java program.
>
> This implementation could be used as a basis for other applications, but
> some features are currently rather specific to XMLfr. For instance the
> you can select between two set of indexes, one of them calculating the
> level of pertinence taking the date and the document type into account
> (an article will come before a wire item which will come before a
> mailing list message and newer stuff come before older ones).
>
> The second implementation is still work in progress even if I have a
> proof of concept which is working pretty fine.
>
> The idea is to manage everything within OPS.
>
> The query processor is derived from the one developed for XMLfr and
> adapted to be less advanced but more generic.
>
> The indexing is done as a background task within OPS through the
> scheduler.
>
> The indexing uses a modified mime type XML database that associates OPS
> pipelines to media types and the algorithms to index different file
> types are defined as OPS pipelines.
>
> I have started developing processors (called from these pipelines) to
> index Word, Excel, OpenOffice, PDF, XML and HTML documents.
>
> The configuration (directories to index, priority of the indexing
> process, ...) is done through a XML configuration file that can be
> updated while the server is running through a configuration processor.
>
> This system could be used to develop a Java OPS based alternative to
> Beagle (http://beaglewiki.org/Main_Page) but I am sure many other
> applications could take advantage of these processors.
>
> I would be happy to publish all that under open source licences,
> unfortunately, all these developments are still very experimental and
> not very well documented and I don't expect to have enough time to work
> on them anytime soon (except of course if some funding could be raised
> that would allow me to postpone other payed activities).
>
> To summarize, I am sure Lucene + OPS make a very interesting couple but
> I don't have as much to share as I'd like to right now!
>
> Eric
> --
> GPG-PGP: 2A528005
> Did you know it? Python has now a Relax NG (partial) implementation.
>                                           http://advogato.org/proj/xvif/
> ------------------------------------------------------------------------
> Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
> (ISO) RELAX NG   ISBN:0-596-00421-4 http://oreilly.com/catalog/relax
> (W3C) XML Schema ISBN:0-596-00252-1 http://oreilly.com/catalog/xmlschema
> ------------------------------------------------------------------------
>
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.1 (GNU/Linux)
>
> iD8DBQBEP1XZDvn+ZCpSgAURAmLjAJwPhYt1XZnuJfHb9PA0O+zcioyKsQCdEBQ1
> 5c5qwo2TTMJi/anMIqJBcIY=
> =geVg
> -----END PGP SIGNATURE-----
>
>
>
> --
> You receive this message as a subscriber of the [hidden email] mailing list.
> To unsubscribe: mailto:[hidden email]
> For general help: mailto:[hidden email]?subject=help
> ObjectWeb mailing lists service home page: http://www.objectweb.org/wws
>
>
>