[ https://issues.apache.org/jira/browse/LUCENE-7293?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15293507#comment-15293507 ]
Michael McCandless commented on LUCENE-7293: -------------------------------------------- The problem here is highlighter is "lying" to {{GeoPointInBBoxQuery}} by passing it textual terms that did not in fact come from a geopoint encoded field. I think it's fair game for the query to throw an exception in this case! In normal usage, e.g. if a user does this, it's a bad error case. I think to fix this we should change {{WeightedSpanTermExtractor.extract}} to skip all geopoint queries? > Highlighting with GeoPointInBBoxQuery can result in Exception > ------------------------------------------------------------- > > Key: LUCENE-7293 > URL: https://issues.apache.org/jira/browse/LUCENE-7293 > Project: Lucene - Core > Issue Type: Bug > Components: modules/highlighter > Affects Versions: 5.5, 6.0, master (7.0) > Reporter: britta weber > Attachments: LUCENE-7293.patch > > > Highlighter and GeoPointInBBoxQuery don't play well together. I wrote a test > here that throws an exception which I think it should not: > https://github.com/brwe/lucene-solr/commit/311f5527ffb6f3ef50e3f74b54456aa7b29d8898 > The problem seems to be that GeoPointInBBoxQuery calls rewrite with a reader > that contains a text field (which we want to highlight) and therefore has the > wrong encoding. > This is from an elasticsearch issue: > https://github.com/elastic/elasticsearch/issues/17537 -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org