lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Erick Erickson" <erickerick...@gmail.com>
Subject Re: Field.Store.YES and Field.Store.Tokenized with CustomAnalyzer - Double Hit
Date Tue, 29 Aug 2006 17:49:48 GMT
The only reason you need to store a token is if you need to retrieve it from
the document, storing is completely unnecessary for answering the question
"is this term in the document?". So I guess I'm wondering why you don't just
use Field.Store.NO on *both* of the fields.......

On 8/29/06, Furash Gary <furashg@mcao.maricopa.gov> wrote:
>
> The behavior I want is that if I store a name (Gary Furash), a user who
> searches for "Gary Furash" gets a strong hit, wheras a user who seaches
> for "Gray Furish" gets a moderate hit.  I currently achieve this by
>
> 1. using a custom analyzer on insertion/search that tokenizes a
> "soundex" version of the name field in the document (so "Gray" and
> "Gary" become the same soundex code);
> 2. creating two separate fields in the document (one tokenized and one
> plain text) -using a multiquery
>
> So, when I search against the two fields, "Gary Furash" hits against two
> fields in the document, but "Gray Furish" hits only one.
>
> However, I noticed that when I say "Field.Store.YES" it stores the
> original, pre-tokenized version, so it seems like I'm doubling up here.
> Is there a better way to do this?
>
> Gary Furash, MBA, PMP
> Applications Manager, Maricopa County Attorney's Office
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message