[ 
https://issues.apache.org/jira/browse/LUCENE-10220?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17438662#comment-17438662
 ] 

Dawid Weiss commented on LUCENE-10220:
--------------------------------------

I think those static methods on Intervals are quite compact and well-defined. 
But maybe we should just consider moving intervals to the core. They're 
excellent.

> Unless you're thinking we should build an explicit intervals query parser? 
> Which would also be cool,

I have an extension of flexible QP in the works - this is ported code from a 
well-tried internal implementation that's been quite successful (1). Will post 
in a few hours.

(1) https://get.carrotsearch.com/lingo4g/latest/doc/#interval-functions

> Add an utility method to get IntervalSource from analyzed text (or token 
> stream)
> --------------------------------------------------------------------------------
>
>                 Key: LUCENE-10220
>                 URL: https://issues.apache.org/jira/browse/LUCENE-10220
>             Project: Lucene - Core
>          Issue Type: Improvement
>            Reporter: Dawid Weiss
>            Assignee: Dawid Weiss
>            Priority: Minor
>             Fix For: main (10.0)
>
>
> The Intervals has a number of utility methods that provide an IntervalSource 
> for tokens, phrases, etc. But it's missing an important bit: an interval 
> source matching tokens that are a result of some string applied to a full 
> analysis chain. This corresponds to actually resides in the index and is hard 
> to predict from the outside.
> This is an important omission in Intervals as a utility class.
> I borrowed the implementation from the then-ASL-licensed Elasticsearch code 
> at: 
> [https://github.com/elastic/elasticsearch/blob/7.10/server/src/main/java/org/elasticsearch/index/query/IntervalBuilder.java#L54-L106]
> I also modified it slightly to fit the static-method-based Lucene API. I also 
> added a small test that showcases how this method can be used in practice 
> (and why it's hard to accomplish the same result with existing methods).
> The only thing I'm not sure is how to attribute Elasticsearch properly - in 
> the notice file, perhaps?



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to