FacetCutter (Lucene 9.12.2 sandbox API)

All Known Implementing Classes:

DoubleRangeFacetCutter, LongRangeFacetCutter, LongValueFacetCutter, TaxonomyFacetsCutter
```
public interface FacetCutter
```
Creates LeafFacetCutter for each leaf.
TODO: do we need FacetCutterManager similar to CollectorManager, e.g. is createLeafCutter always thread safe?

WARNING: This API is experimental and might change in incompatible ways in the next release.

Method Summary

All Methods Instance Methods Abstract Methods Default Methods
Modifier and Type	Method	Description
`LeafFacetCutter`	`createLeafCutter(LeafReaderContext context)`	Get cutter for the leaf.
`default OrdinalIterator`	`getChildrenOrds(int ord)`	For facets that have hierarchy (levels), get all children ordinals for given ord.
`default OrdinalIterator`	`getOrdinalsToRollup()`	For facets that have hierarchy (levels), return all top level dimension ordinals that require rollup.

- Method Detail
  - createLeafCutter
```
LeafFacetCutter createLeafCutter(LeafReaderContext context)
                          throws IOException
```
    Get cutter for the leaf.
    
    Throws:
    
    IOException
  - getOrdinalsToRollup
```
default OrdinalIterator getOrdinalsToRollup()
                                     throws IOException
```
    For facets that have hierarchy (levels), return all top level dimension ordinals that require rollup.
    Rollup is an optimization for facets types that support hierarchy, if single document belongs to at most one node in the hierarchy, we can first record data for these nodes only, and then roll up values to parent ordinals.
    Default implementation returns null, which means that rollup is not needed.
    
    Throws:
    
    IOException
  - getChildrenOrds
```
default OrdinalIterator getChildrenOrds(int ord)
                                 throws IOException
```
    For facets that have hierarchy (levels), get all children ordinals for given ord.
    
    Throws:
    
    IOException