distributed pivot faceting may not work well with some custom FieldTypes

A limitiation of the distributed pivot faceting code is that it makes some explicit assumptions about the datatypes of the pivot values for the purposes of "serializing" the values in order to make refinement requests to the individual shards for those values.

This logic works fine for String based fields, dates, and primitive numerics – but any custom FieldType that has a toObject() method which does not return one of those data types may have problems. While pivot faceting uses the typed objects returned by toObject() in it's responses, there is no general FieldType method for converting those objects back into Strings suitable for the refinement requests.

Untill we have some abstract, FieldType based, method for converting the value Objects into Strings thta can be included in the refinement requests for use in FieldType.getFieldQuery() there isn't really a good solution for this.