Second invocation of ClusteredQueryImpl.lazyIterator() yields results in reverse order

When you invoke lazyIterator() for the 2nd time on the same ClusteredQueryImpl instance, the iterator returns results in reverse order. This only occurs when the query has a Sort specified.

Caused by DistributedIterator.setTopDocs(), which inverts the reverse flag on SortField.

The same is probably also true for ClusteredQueryImpl.iterator()