[KinoSearch] question on querying by url

[Marvin Humphrey]

On Apr 18, 2006, at 4:46 AM, Filippo A. Salustri wrote:

> Doesn't this assume the PolyAnalyzer was called?  "analyzed" was on  
> originally, but before I did the tests I reported, I'd turned it  
> off for the 'url' field with "analyzed=>0" in spec_field.  My  
> understanding is that in that case, the PolyAnalyzer wouldn't be  
> called, even if QueryParser was on duty.

When you say "turned it off", do you mean that you regenerated the  
entire index from scratch?

The QueryParser doesn't know whether a field was analyzed or not at  
index-time.  Whatever analyzer you supply to QueryParser is always  
used, regardless of whether it is using the default field, or whether  
it is using a field which has it parsed out of a "fieldname:termtext"  
construct.

Behind the scenes in the Searcher, when you supply a query string  
instead of a full Query object to search(), it actually creates  
several QueryParsers -- one for each indexed field.  Each one of  
these QueryParsers uses the supplied Analyzer.  All of the Query  
objects that these QueryParsers generate are fused together into a  
big BooleanQuery, and search() returns the results of that.

I've now added a new unit test, <http://www.rectangular.com/svn/ 
kinosearch/trunk/t/214-spec_field.t>, which tests for whether the  
settings for "analyzed", "analyzer", and "indexed" are being obeyed.   
It hasn't turned up any surprises... so far. :)  Please let me know  
if you experience unexpected behavior along these lines again.

Thanks for the feedback,

Marvin Humphrey
Rectangular Research
http://www.rectangular.com/


_______________________________________________
KinoSearch mailing list
KinoSearch at rectangular.com
http://www.rectangular.com/mailman/listinfo/kinosearch