If you don't care about its contribution to scoring, one option is to
move the clause you want evaluated to an fq clause sitn {!cache=false
cost=101}. see: http://yonik.com/advanced-filter-caching-in-solr/

Best,
Erick

On Wed, Feb 7, 2018 at 12:05 PM, Emir Arnautović
<emir.arnauto...@sematext.com> wrote:
> Hi,
> Also note that score is different if only one term match and if both terms 
> are matched. Your case would make sense if you do not plan to order by score, 
> but as Walter explained, Solr does not go document by document and evaluate 
> query conditions, but it gets list of documents matching each part of boolean 
> query (you can think of it as bitsets) and do union/intersection to get the 
> final result.
>
> HTH,
> Emir
> --
> Monitoring - Log Management - Alerting - Anomaly Detection
> Solr & Elasticsearch Consulting Support Training - http://sematext.com/
>
>
>
>> On 7 Feb 2018, at 19:38, Walter Underwood <wun...@wunderwood.org> wrote:
>>
>> That doesn’t really make sense for Solr query evaluation. It fetches the 
>> posting lists for each term, then walks through them evaluating the query 
>> against all the documents.
>>
>> It can skip a document as soon as it fails the query, but it still has to 
>> fetch the posting lists.
>>
>> So, that feature doesn’t exist because it isn’t useful.
>>
>> wunder
>> Walter Underwood
>> wun...@wunderwood.org
>> http://observer.wunderwood.org/  (my blog)
>>
>>> On Feb 7, 2018, at 9:50 AM, bbarani <bbar...@gmail.com> wrote:
>>>
>>>
>>> I am trying to figure out a way to form boolean (||) query in SOLR.
>>> Ideally my expectation is that with boolean operator ||, if first term is
>>> true second term shouldn't be evaluated.
>>>
>>> &q=searchTerms:"testing" || matchStemming:"stemming"
>>> works same as
>>> &q=searchTerms:"testing" OR matchStemming:"stemming"
>>>
>>> Is there a way to form a boolean query such that it wont evaluate the right
>>> hand side if it isn't necessary?
>>>
>>>
>>>
>>> --
>>> Sent from: http://lucene.472066.n3.nabble.com/Solr-User-f472068.html
>>
>

Reply via email to