Function to simulate int32 wrapping | Voters

Function to simulate int32 wrapping

planned

Tapple Gao

Please add a function that converts a float64 to a signed 32-bit integer, compatable with lsl. It would be equivilant to

in luau:

function bit32.s32(n: number)
    return (math.modf(n) + 0x8000_0000) % 0x1_0000_0000 - 0x8000_0000
end

in c:

double s32(double n) {
    return (double)(s32)n;
}

Code that dealt with integers

>= 0x8000_0000

behave differently in lsl and luau, due to s32 wrapping. I imagine there's already a function like this for the lsl compiler, but exposing it to slua scripters would be helpful for converting old scripts

6 days ago

Harold Linden

marked this post as

planned

Thanks! I'll get this in. The C++ implementation may be more subtle since you don't really want clamping there and float->int casts outside of int range are Undefined Behavior in C++ as I've recently come to learn. Something more like the Lua version should work well.