Open NOP0 opened 4 years ago
Redid this, tried in principle to do it like the different variants of Integer. What do you think?
Edit: I'm used to the convention asserteq(expected, actual)
, but in ast.rs
it's used asserteq(got, should_be)
. Is it there any reason for this?
I think I'm on thin ice and might need some mentoring here. Trying to implement "engineering floats" that is
3.14E-1 // 0.314
There's a lot of unwrapping going on, is this ok? The exponent is max u32, but I think that is large enough by the IEEE standard?
Another thing I see now, is that this rule might need to be over the float rule? (Most specific first?)