[ 
https://issues.apache.org/jira/browse/SOLR-15833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17454722#comment-17454722
 ] 

ASF subversion and git services commented on SOLR-15833:
--------------------------------------------------------

Commit 187fc9c1b0054d92f9c842cfe4b8d8cb2fbd5a4b in lucene-solr's branch 
refs/heads/branch_8_11 from Mike Drob
[ https://gitbox.apache.org/repos/asf?p=lucene-solr.git;h=187fc9c ]

SOLR-15833 Spatial types should work with exists query (#2624)

Co-Authored-By: Houston Putman <hous...@apache.org>

> Exists query does not work for SRPT
> -----------------------------------
>
>                 Key: SOLR-15833
>                 URL: https://issues.apache.org/jira/browse/SOLR-15833
>             Project: Solr
>          Issue Type: Bug
>    Affects Versions: 8.6.3
>            Reporter: Mike Drob
>            Assignee: Mike Drob
>            Priority: Major
>             Fix For: 9.0, 8.11.1
>
>          Time Spent: 40m
>  Remaining Estimate: 0h
>
> {noformat}
> @@ -172,6 +173,8 @@ public class TestSolr4Spatial extends SolrTestCaseJ4 {
>      checkHits(fieldName, "0,0", 100, DistanceUtils.EARTH_MEAN_RADIUS_KM, 
> 0);//doesn't error
>    }
> +  @Test public void testExistsQuery() { assertQ(req("q", fieldName + ":*", 
> "fl", "id," + fieldName)); }
> +
>    private void checkHits(String fieldName, String pt, double distKM, double 
> sphereRadius, int count, int ... docIds) throws ParseException {
>      checkHits(fieldName, true, pt, distKM, sphereRadius, count, docIds);
>    }
> {noformat}
> This test passes against 8.5.0 release and fails when running 8.6.3.
> I initially suspected SOLR-11746, but that doesn't line up with the release 
> info, unless some of that revert/recommit stuff came in later.
> This error happens because the {{exists}} query for spatial types gets turned 
> into a range query with inclusive null bounds, which then causes the bounding 
> box parsing to fail. And generating the exception incidentally fails because 
> we try to pass a null argument to String.contains!
> Some options for solving this:
> * See null bounds when creating our shape and return a new shape that matches 
> everything
> * Go back to using a prefix query
> * Find some other extra "clever" data like norms that the field leaves around
> We might also need to check additional fields for this same behavior.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org
For additional commands, e-mail: issues-h...@solr.apache.org

Reply via email to