Coerce int range filter values for accuracy and rename some methods for clarity

[ryanrdoherty]

Main goal here is to make sure passed integer range filter values don't produce results (i.e. not filter) values outside the range, even if the values passed are floating point. Important to note that range filters are inclusive! Thus, for example, a passed range filter of [4.5, 6.5] should result in integer conversions of [5, 6] so we don't include 4s or 7s. To do this, we use floor on max values and ceil on min values. Unlike truncation (which we were doing before via longValue()), which does mirror-image conversion for negative numbers, floor and ceil are consistent so we don't need special logic for negative values.