An example of a generic KeyIndex which accounts for sharding that could occur
in a data store like Apache Accumulo. Given a shard count (pre-configured),
this index adds a "shard prefix" to the true index as given by the inner
argument. Accumulo can shard based off of the first digit of a row ID,
and since we use a round-robin approach to generate prefixes, this distributes
spatially close indices across different shards, and thus helps avoid hot spots.
Some(keybounds) if the indexed space is bounded; None if it is unbounded