jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ard Schrijvers" <a.schrijv...@hippo.nl>
Subject RE: IndexingConfiguration jr 1.4 release, analyzing, searching and synonymprovider
Date Mon, 20 Aug 2007 12:29:29 GMT
> Marcel Reutegger wrote:
> This would work quite well for jcr:contains functions that 
> operate on a 
> property. However I'm not sure what to do with this:
> //*[jcr:contains(., 'hägar')]
> the node scope does not indicate which analyzer to use for 
> the query statement. 
> Would we just run the statement through all analyzers and 
> combine them in an OR 
> query?

As said, I would give it an extra thought. Reminder: It is about my suggestion to support
indexing properties with configurable analyzers in the IndexingConfiguration, for example
to enable multilingual indexing for nodes having properties that contains language specific
texts. The idea is, that when searching for 


and you defined in the IndexingConfiguration a GermanAnalyzer for 'body_de' property, you
are searching the text with this analyzer.

Now, Marcel correctly pointed about the problem when referring to the node scope search:

//*[jcr:contains(., 'hägar')] 

As I have given it another thought, it does make sense to me, that when searching in the node
scope, you use the global/default analyzer, and when you are searching in a property, you
might use a specific analyzer you used for that property. In this way, we can actually do
stemming and stopwords, and implement synonym analyzer for any language you want to support
per property.  IMHO, configurable analyzers in IndexingConfiguration are a valuable add. Also,
people who don't want it, because they do not like the idea that you might find a hit for
some word in a property, while it is possible that there is no hit on the same word for the
node scope (due to analyzers), just do not configure it. WDOT? 

Regards Ard


View raw message