[
https://issues.apache.org/jira/browse/SOLR-10891?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Steve Rowe updated SOLR-10891:
------------------------------
Attachment: SOLR-10891.patch
Patch, the test passes for me now.
Thanks [~dsmiley] for the pointers.
Fortunately I didn't have to make any modifications to BBoxStrategy. The issue
was that the fix for the frozen Point fieldtype (see my first comment above)
stripped the identity of the Trie fieldtype, which caused other code to think
the fieldtype was neither fish nor fowl; I fixed it by only copying the
fieldtype if it's not a Trie fieldtype.
The unsupported operation exceptions were due to special case for "bbox"
fieldName only skipping of non-rectangles in checkHits(); I just added a case
for the new "pbbox" fieldName.
I'll commit this once precommit and all tests pass.
> BBoxField does not support point-based number sub-fields
> --------------------------------------------------------
>
> Key: SOLR-10891
> URL: https://issues.apache.org/jira/browse/SOLR-10891
> Project: Solr
> Issue Type: Sub-task
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Steve Rowe
> Assignee: Steve Rowe
> Attachments: SOLR-10891.patch, SOLR-10891.patch, tests-failures.txt
>
>
> I noticed while removing Trie fields from example schemas on SOLR-10760 that
> BBoxField uses Trie fields in at least one example schema.
> I went looking and there is theoretical machinery to support points, but when
> I added a point-based bbox variant to TestSolr4Spatial, I get test failures.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
