[
https://issues.apache.org/jira/browse/SOLR-1553?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12793813#action_12793813
]
Yonik Seeley commented on SOLR-1553:
------------------------------------
bq. why is "TO" listed as an operator when building up the phrase boost fields?
(line 296) ... if range queries are supported, then shouldn't the upper/lower
bounds also be striped out of the clauses list?
It seemed incrementally better than leaving it out.
Range queries aren't actually supported in the fallback mode. The phrase
boosting is also relatively rudimentary. We could try to recognize range
queries and removing the whole thing.
bq. i see that the boost queries built from the pf and pf3 fields are put in
BooleanQueries instead of DisjunctionMaxQueries ... but why?
I honestly don't recall (this was written a while ago).
bq. ExtendedAnalyzer feels like a really big hack ... i'm not certain, but i
don't think it works correctly if a CharFilter is declared.
Probably not - CharFilter didn't exist when this was written.
Should we use an alternative method for signaling that stopwords are optional?
The nice thing about putting the stopword filter in the query analyzer (and
having edismax detect it) is that something like the lucene query parser works
as it did in the past... stopwords are removed. But I could see the other side
of the argument too.
Anyway, I think I agree with pretty much everything you say - just haven't had
time to do anything about it.
> extended dismax query parser
> ----------------------------
>
> Key: SOLR-1553
> URL: https://issues.apache.org/jira/browse/SOLR-1553
> Project: Solr
> Issue Type: New Feature
> Reporter: Yonik Seeley
> Fix For: 1.5
>
> Attachments: SOLR-1553.patch
>
>
> An improved user-facing query parser based on dismax
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
