[ https://issues.apache.org/jira/browse/SOLR-9775?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17622496#comment-17622496 ]
Eric Pugh commented on SOLR-9775: --------------------------------- What if we change the logic in SolrIndexSearcher.java key = {color:#cc7832}new {color}QueryResultKey(q{color:#cc7832}, {color}cmd.getFilterList(){color:#cc7832}, {color}cmd.getSort(){color:#cc7832}, {color}flags{color:#cc7832}, {color}cmd.getMinExactCount()) So that Sort is never null and getFilterList is never null, and indeed, then also purge any nulls in the sort fields and filterlist? That might simplify a lot of the logic in QueryResultKey hashcode and equals methods..... > NPE in QueryResultKey constructor (when executing a clustering search query?) > ----------------------------------------------------------------------------- > > Key: SOLR-9775 > URL: https://issues.apache.org/jira/browse/SOLR-9775 > Project: Solr > Issue Type: Bug > Reporter: Christine Poerschke > Priority: Minor > Time Spent: 50m > Remaining Estimate: 0h > > https://github.com/apache/lucene-solr/pull/116 from Roman Kagan yesterday > (November 2016) has a fix. > On the solr-user mailing list (in March) Tim Hearn > [reported|http://mail-archives.apache.org/mod_mbox/lucene-solr-user/201603.mbox/%3CCAFoZJmC87Lbuj%2BaYwcVh%2B5ay_%3Dwi5n8TGs7gBPcF%3Djjo%2BW7vGg%40mail.gmail.com%3E] > encountering what sounds like the same NPE when executing a clustering > search query. > This ticket to tentatively link the two together. Open question: do we want > to include a reproducing test case along with the fix or just commit the fix > alone? -- This message was sent by Atlassian Jira (v8.20.10#820010) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org For additional commands, e-mail: issues-h...@solr.apache.org