lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ard Schrijvers" <a.schrijv...@hippo.nl>
Subject RE: reg exp
Date Tue, 17 Jul 2007 07:48:58 GMT

> 
> In general, the type of feature you are interested in is 
> typcially caleld
> a "PrefixQuery" you can acomplish it with the SolrQueryParser 
> using "a*"
> 
> 	http://wiki.apache.org/solr/SolrQuerySyntax
> 
> : example show all docs starting with an "a" or "A", I would 
> just add an
> : extra field only containing the first letter. This is 
> *much* faster in
> 
> the EdgeNGramTokenizerFactory should actaully make this
> trivial for you ... you can copyField the value into an 
> "initial" field
> and let the Tokenizer do all the work of indexing only teh first
> character.

That is indeed much easier....never heard of the EdgeNGramTokenizerFactory but seems to be
doing exactly the thing you describe,

Thx, Ard

> 
> : lucene, and for many documents you don't risk the too many boolean
> : clauses (though you can set it very high, at the cost of memory)
> 
> this isn't actually a problem in Solr .. the SolrQueryParser 
> doesn't use
> regualr PrefixQueries, it rewritesthem into PrefixFilter's 
> which don't hit
> the boolean clauses limit.
> 
> 
> 
> -Hoss
> 
> 

Mime
View raw message