public class DocValuesFacets extends Object
This is basically a specialized case of the code in SimpleFacets. Instead of working on a top-level reader view (binary-search per docid), it collects per-segment, but maps ordinals to global ordinal space using MultiDocValues' OrdinalMap.
This means the ordinal map is created per-reopen: O(nterms), but this may perform better than PerSegmentSingleValuedFaceting which has to merge O(nterms) per query. Additionally it works for multi-valued fields.
|Modifier and Type||Method and Description|