Hi Lennart,

Lennart Regebro wrote:
> I'm working on the plone port of CPSLuceneCatalog. It's pretty much
> working, but I have realized that Plone itself uses an
> ExtendedPathIndex which has a lot of fancy features, like a max depth
> to return documents, the possibility to also return all the parents
> (handy when making a breadcrumb) and such.
>
> I'm currently looking into how to support this with Lucene, and it seems
> tricky.
> 
> I think I either need to change the behaviour of the path index to
> support all these options, or maybe make a new index type. Do you have
> any pointers on how to implement a new index type? Is it a new Term or
> a new TermQuery, or something else?

I guess, for BBB sake, it would be better to implement a new index type
in NXLucene. For the moment, this part of NXLucene (top level app
fields) is not that great and would need a spring cleanup that we won't
have time to perform in here.

If you wanna implement this NXLucene side do the following :

 - implement a case in indexDocument() within server/core.py

 - Implement a case in searchDocuments() within the same server/core.py

 - A generic setup wrapper CPS side (CPSLuceneCatalog) and extend the
possible options forward from Zope to NXLucene in nuxeo.lucene.

Of course, this is a rough roadmap.

> Possibly I can fake it in plonelucenecatalog by actually using two
> indexes in Lucene,  but it doesn't seem any easier...

nope do it if you wanna. Just as usual create a branch for the
implementation and test it properly. (well, you know the house right ;))

Cheers,

        J.

P.S : Please, use cps-devel if you wanna discuss the implementation.

-- 
Julien Anguenot | Nuxeo R&D (Paris, France)
Open Source ECM - http://www.nuxeo.com
Nuxeo 5 : http://www.nuxeo.org
Mobile: +33 (0) 6 72 57 57 66

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
cps-devel mailing list
http://lists.nuxeo.com/mailman/listinfo/cps-devel

Reply via email to