Changeset 35172 for gs3-extensions

Timestamp:

2021-05-20T14:26:35+12:00 (3 years ago)

Author:

kjdon

Message:

we noticed when doing a solr search at document level that we were getting a random section of the document appearing outside the section headings. This is because when doing a highlighting query at document level, it was returning the first TX element that matched, which was not the top level content element. That was getting added in as the top level content, which it shouldn't be. So only check one TX element for a match then stop.

File:

• gs3-extensions/solr/trunk/src/src/java/org/greenstone/gsdl3/util/SolrQueryWrapper.java (modified) (2 diffs)

gs3-extensions/solr/trunk/src/src/java/org/greenstone/gsdl3/util/SolrQueryWrapper.java

@@ -547 +547 @@
     public String runHighlightingQuery(String query,String hldocOID)
     {
-                    
         SolrQueryResult solr_query_result = new SolrQueryResult();
         solr_query_result.clear();
@@ -564 +563 @@
         //Set field for highlighting
         solrQuery.setParam("hl.fl", highlight_field);
+        
+        // this option only available for the OriginalHighlighter (hl.method=original, the default)
+        // if we are doing document level search, we only want to highlight the first section,
+        // (TX element) if applicable. Otherwise get a middle section displayed at the start of
+        // a document, outside the toc.
+        // if we are doing section level search, there will only be one TX element
+        solrQuery.setParam("hl.maxMultiValuedToExamine", "1");
         
         //Get whole highlighted field