BitArray syntax badarg runtime error

[schurhammer]

Just creating an issue to discuss this, there might not be much we can or should do about it.

Basically the bitarray syntax allows you to read more bits than is in the bit array, leading to a runtime error.

let a = <<1, 2, 3>>
let b = <<a:32-bits>>

erlang:error(Badarg)

If this was an integer instead of a bit array, it would be zero padded. Maybe we could do that.

[lpil]

We could statically reject the specific case of <<a:32-bits>> as we can tell the size is longer than the array?

What code would we generate if we were to pad it?