[LUCENE-10466] IndexSortSortedNumericDocValuesRangeQuery unconditionally assumes the usage of the LONG-encoded SortField

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 9.0
• Fix Version/s: 9.2
• Component/s: core/query/scoring
• Lucene Fields: New
• Assignee: Unassigned
• Reporter: Andriy Redko
• Created: 14/Mar/22 19:04
• Updated: 28/Aug/22 16:37
• Resolved: 04/Apr/22 19:17

Description

We have run into this issue while migrating to OpenSearch and making changes to accommodate https://issues.apache.org/jira/browse/LUCENE-10087. It turned out that IndexSortSortedNumericDocValuesRangeQuery unconditionally assumes the usage of the LONG-encoded SortField, as could be seen inside static ValueComparator loadComparator method

    @SuppressWarnings("unchecked")
    FieldComparator<Long> fieldComparator = (FieldComparator<Long>) sortField.getComparator(1, 0);
    fieldComparator.setTopValue(topValue);
 

 

Using the numeric range query (in case of sorted index) with anything but LONG ends up with class cast exception:

   >     java.lang.ClassCastException: class java.lang.Long cannot be cast to class java.lang.Integer (java.lang.Long and java.lang.Integer are in module java.base of loader 'bootstrap')
   >         at org.apache.lucene.search.comparators.IntComparator.setTopValue(IntComparator.java:29)
   >         at org.apache.lucene.sandbox.search.IndexSortSortedNumericDocValuesRangeQuery.loadComparator(IndexSortSortedNumericDocValuesRangeQuery.java:251)
   >         at org.apache.lucene.sandbox.search.IndexSortSortedNumericDocValuesRangeQuery.getDocIdSetIterator(IndexSortSortedNumericDocValuesRangeQuery.java:206)
   >         at org.apache.lucene.sandbox.search.IndexSortSortedNumericDocValuesRangeQuery$1.scorer(IndexSortSortedNumericDocValuesRangeQuery.java:170)
 

Simple test case to reproduce (for TestIndexSortSortedNumericDocValuesRangeQuery):

  public void testIndexSortDocValuesWithIntRange() throws Exception {
    Directory dir = newDirectory();    
    IndexWriterConfig iwc = new IndexWriterConfig(new MockAnalyzer(random()));
    Sort indexSort = new Sort(new SortedNumericSortField("field", SortField.Type.INT, false));
    iwc.setIndexSort(indexSort);
    
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir, iwc);    
    writer.addDocument(createDocument("field", -80));    
    
    DirectoryReader reader = writer.getReader();
    IndexSearcher searcher = newSearcher(reader);    // Test ranges consisting of one value.
    assertEquals(1, searcher.count(createQuery("field", -80, -80)));    

    writer.close();
    reader.close();
    dir.close();
  } 

 

The expectation is that IndexSortSortedNumericDocValuesRangeQuery should not fail with class cast but correctly convert the numeric values.

Issue Links

• links to: GitHub Pull Request #749