[LUCENE-7769] UnifiedHighlighter doesn't highlight MTQs wrapped in BoostQuery

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 6.4.2, 6.5
• Fix Version/s: 6.5.1, 7.0
• Component/s: modules/highlighter
• Lucene Fields: New
• Assignee: David Smiley
• Reporter: Dmitry Malinin
• Created: 07/Apr/17 08:27
• Updated: 28/Aug/22 15:13
• Resolved: 18/Apr/17 18:43

Description

UnifiedHighlighter doesn't highlight MTQ wrapped in BoostQuery.
For example, suppose we have a doc with a field 'f' contains data 'lucene'.
UnifiedHighlighter highlights query (f:lucene*), but query (f:lucene*)^1 doesn't. Test code:

String field = "f";
String content = "lucene";
Term term = new Term(field, content);
UnifiedHighlighter highlighter = new UnifiedHighlighter(null, new StandardAnalyzer());
Query[] queries = {new PrefixQuery(term), new BoostQuery(new PrefixQuery(term), 1.0f)};
Object fragObj;		
for (Query query : queries)
{
	fragObj = highlighter.highlightWithoutSearcher(field, query, content, 1);
	System.out.printf("content=[%s]  Query=%s  frag=[%s]\n", content, query, fragObj);
}

My opinion it's because MultiTermHighlighting.extractAutomata() returns an empty automaton for BoostQuery. I think, should add some thing like:

if (query instanceof BoostQuery) 
{
      list.addAll(Arrays.asList(extractAutomata(((BoostQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc))) ; 
} 

to MultiTermHighlighting.extractAutomata()
Thanks.