Re: [Fuseki] Free Text Indexing for SPARQL

[Paolo Castagna <ca...@googlemail.com>]

Andy Seaborne wrote:
> 
> 
> On 01/03/11 08:11, Camel Christophe wrote:
>> Hello again,
>>
>>
>>
>> It seems that by default Fuseki doesn't handle text matching
>> (pf:textMatch) in sparql queries (Error 500: Index not found).
>>
>> Is it possible to configure it to support this feature?
> 
> Not easily.
> 
> The ideal would be to have an assembler description for the index and 
> attach that to relevant dataset.  That's what assemblers are for - 
> setting up particular configurations.
> 
> But there is no assembler support for LARQ
> 
> Paolo - that should go in the new LARQ module, shouldn't it?

Absolutely.

There was something in ARQ but I have not tested/used extensively so
I am not sure how complete/good it is. It's here:
https://jena.svn.sourceforge.net/svnroot/jena/LARQ/trunk/src/main/java/org/apache/jena/larq/assembler/

Paolo

> So for now you left with trickery.
> 
> an assmbler can load and execute arbitrary code via ja:loadClass.
> 
> 
> For example:
> [] ja:loadClass "com.hp.hpl.jena.tdb.TDB" .
> 
> kicks the TDB initialization code.
> 
> You can write a class that sets up the LARQ index.  I think someone (or 
> more than one) have done that but I don't which project the code is in 
> (it's not in the Jena system).  Be nice to get hold of it as a temporary 
> solution.
> 
>     Andy
> 
>>
>>
>>
>> Thanks,
>>
>> Chris
>>
>>