Problems with WordDelimiterFilterFactory

2008-12-23 Thread GPS.

I am using a fieldType, with following configuration:



  







  


I have 

When I try searching with :
http://localhost:8001/solr/select/?q=Id:ARMZ
It gives me complete list, where Id is: ARMZ or ARMZ117 or ARMZ129

What I want is if I search for ARMZ, it should tightly match only ARMZ and
shouldn't return ARMZ117 OR ARMZ129
Similarly, If I try searching for ARMZ1, it shouldn't give me any of ARMZ117
OR ARMZ129

Is it possible to achieve this, by somehow strictly mapping the input text
with field Id?
Any help on this matter would be deeply appreciated.

Thanks
GPS.

-- 
View this message in context: 
http://www.nabble.com/Problems-with-WordDelimiterFilterFactory-tp21149384p21149384.html
Sent from the Solr - User mailing list archive at Nabble.com.



Re: Problems with WordDelimiterFilterFactory

2008-12-24 Thread GPS.

Tried this once again, with different combinations, none worked actually..


David Smiley @MITRE.org wrote:
> 
> It seems you want Id to only match on complete field values.  If that is
> the case then you should not do tokenization nor perhaps any text analysis
> altogether.  Consider removing the whole  block or using
> KeywordTokenizerFactory plus a modicum of other stuff (perhaps
> lowercasing).  For the particular examples you gave... it would be
> sufficient to simply remove “WordDelimeterFilterFactory” as an analysis
> step.
> 
> ~ David
> 
> 
> GPS. wrote:
>> 
>> I am using a fieldType, with following configuration:
>> 
>> 
>> > positionIncrementGap="100" >
>>   
>> 
>> > ignoreCase="true" expand="false"/>
>> > words="stopwords.txt"/>
>> > generateWordParts="0" generateNumberParts="0" catenateWords="1"
>> catenateNumbers="1" catenateAll="0"/>
>> 
>> > protected="protwords.txt"/>
>> 
>>   
>> 
>> 
>> I have > omitNorms="true"/>
>> 
>> When I try searching with :
>> http://localhost:8001/solr/select/?q=Id:ARMZ
>> It gives me complete list, where Id is: ARMZ or ARMZ117 or ARMZ129
>> 
>> What I want is if I search for ARMZ, it should tightly match only ARMZ
>> and shouldn't return ARMZ117 OR ARMZ129
>> Similarly, If I try searching for ARMZ1, it shouldn't give me any of
>> ARMZ117 OR ARMZ129
>> 
>> Is it possible to achieve this, by somehow strictly mapping the input
>> text with field Id?
>> Any help on this matter would be deeply appreciated.
>> 
>> Thanks
>> GPS.
>> 
>> 
> 
> 

-- 
View this message in context: 
http://www.nabble.com/Problems-with-WordDelimiterFilterFactory-tp21149384p21156935.html
Sent from the Solr - User mailing list archive at Nabble.com.