Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: pass (asf.osuosl.org: local policy)
Mime-Version: 1.0 (Apple Message framework v752.3)
In-Reply-To: <45332A6A.50602@teamware.com>
References: 
 <OFA797CB46.98E990ED-ON88257209.001682FD-88257209.0016FEF6@il.ibm.com>
 <45332A6A.50602@teamware.com>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <6DCBC109-7B4F-4836-88D7-28454D9E05BF@ehatchersolutions.com>
Content-Transfer-Encoding: 7bit
From: Erik Hatcher <erik@ehatchersolutions.com>
Subject: Re: Query not finding indexed data
Date: Mon, 16 Oct 2006 04:24:57 -0400
To: java-user@lucene.apache.org


On Oct 16, 2006, at 2:44 AM, Antony Bowesman wrote:

> Doron Cohen wrote:
>> Hi Antony, you cannot instruct the query parser to do that. Note  
>> that an
>
> Thanks, I suspected as much.  I've changed it to make the field  
> tokenized.
>
>> field name. This is an application logic to know that a certain  
>> query is
>> not to be tokenized. In this case you could create your query with:
>>   query = new TermQuery(fieldName, "IqTstAdminGuide2.pdf");
>
> The query is user driven, so I can't know without parsing whether  
> it should be tokenised or not.  I would have to extend the parser  
> to make use of TermQuery - it's easier just to tokenize the field  
> now I understand Lucene's behaviour.

You can also use PerFieldAnalyzerWrapper as the analyzer for  
QueryParser, and for all your untokenized fields, specify a  
KeywordAnalyzer.  That will keep untokenized fields from being split  
(as best it can given QueryParser meta-syntax).

	Erik


---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org