Take advantage of dictionary pages when performing rowgroup filters in the filter2 API

We currently only filter row groups via the min / max value in the row group.
We should additionally inspect the dictionary of unique values in a row group (if it has one) – this could dramatically increase our ability to drop entire rowgroups.

**Reporter**: [Alex Levenson](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=alexlevenson) / @isnotinvain

_{**Note**: *This issue was originally created as [PARQUET-40](https://issues.apache.org/jira/browse/PARQUET-40). Please see the [migration documentation](https://issues.apache.org/jira/browse/PARQUET-2502) for further details.*}

Start by tracing the filter2 API's existing row-group filtering path, which currently uses min/max values. Then inspect how row-group dictionaries and unique values are represented, and determine how they can inform row-group elimination. Done means filters can use available dictionary data in addition to min/max metadata, with coverage for the relevant cases.

Written by the indexing model from the issue text.