FloatRange: add options that exclude each endpoint

Thanks for developing this beautiful program.

At this time, FloatRange assumes that the passing argument value is always in closed interval [a, b], not open one. Sometimes, this may be harsh when the user want to accept a floating number exclude two endpoints, a and b. A user who uses IntRange can solve this problem fairly easily, but FloatRange cannot.

So, I think FloatRange should have the options which can exclude each endpoint. May I PR for this?