You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by rfr <er...@gmail.com> on 2012/05/11 11:08:48 UTC

Index referenced node

Hi!

In our projet, we have some custom types with a few reference 
attributes. We tried to minimize the use of references but we could not 
avoid them totally.

Our client wants a powerfull search function and the use of lucene in 
jackrabbit helps us a lot. But we did not see a mean to configure 
aggregates on referenced nodes.

What we would like is to be able to retrieve a node through a full text 
search if another node, referenced by an attribute of the first node, 
matches the full text search.

Example:

[dummy:a] < nt:unstructured

[dummy:b] < nt:base
- dummy:val (REFERENCE) < dummy:a

if a node "node1" of type dummy:b reference a node2 of type dummy:a 
which dummy:val attributes is set to "hello", we would like to configure 
the indexing to allow a fulltext search with the value "hello" to return 
node1.

Is it possible directly? If not, is there a way to hook the indexing 
feature to allow our code to extract text from the referenced node for 
the indexer?

Thanks for your help!

Best regards,

Fred