Support BigFloat/BigInt?

[jariji]

After getting annoyed by all the fiddly details when testing with normal floats, I wanted to to try testing a claim using BigFloat. But Data.Floats{BigFloat}() doesn't work.

julia> Data.Floats{BigFloat}()
ERROR: TypeError: in Floats, in T, expected T<:Union{Float16, Float32, Float64}, got Type{BigFloat}

[Seelengrab]

Well, those fiddly details are just what make floating point floating point :)

BigFloat (like BigInt) is currently unsupported since managing its state is quite a bit more challenging. There are not a lot of ways to construct a BigFloat/BigInt, and generating one from scratch is itself challenging. Hypothesis has this limitation too, it only supports up to 128-bit wide integers as far as I can tell and 16/32/64-bit floating point numbers.

I'm not opposed in principle to add this at some point, but it's going to require at the very least #3. There isn't really a good way to construct a BigInt/BigFloat from the raw choicesequence, without copying those (potentially huge) numbers around a lot.