[
https://issues.apache.org/jira/browse/SOLR-4130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18132061#comment-18132061
]
Nick Shanin commented on SOLR-4130:
-----------------------------------
>From code reading against apache/main b73f03976fa, addPhraseFieldQueries skips
>clauses that have a field or are phrases
>(solr/core/src/java/org/apache/solr/search/ExtendedDismaxQParser.java:308).
>For +sw(a b) +ti:(c d) with pf set, the phrase boost is built across terms in
>different fields ("b d"). Is this still an issue, and is that the parse you
>expect?
🤖 AI-generated 🤖
> eDismax: Terms are skipped for phrase boost when using parentheses
> ------------------------------------------------------------------
>
> Key: SOLR-4130
> URL: https://issues.apache.org/jira/browse/SOLR-4130
> Project: Solr
> Issue Type: Bug
> Components: query parsers
> Affects Versions: 4.0
> Reporter: Leonhard Maylein
> Priority: Major
>
> I've tried the following combination with the eDismax handler
> in SOLR 4.0.0:
> q: +sw(a b) +ti:(c d)
> qf: freitext exttext^0.5
> pf: freitext^6 exttext^3
> The result is:
> <str name="rawquerystring">+sw:(a b) +ti:(c d)</str>
> <str name="querystring">+sw:(a b) +ti:(c d)</str>
> <str name="parsedquery">(((sw:a sw:b) +(ti:c ti:d))
> DisjunctionMaxQuery((freitext:"b d"^6.0)) DisjunctionMaxQuery((exttext:"b
> d"^3.0)))/no_coord</str>
> All terms are (equally) qualified by a field (field sw for the terms a and b,
> field ti for the terms c and d).
> Why do the eDismax handler only use the terms b and d to build the phrase
> boost query?
> It appears that some terms have been skipped for phrase boost.
> Moreover, in my opinion, fielded terms should not be used in phrase boost
> except for the specified field.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
