How to parse enum where the discriminant is not the first field?

[pastelmind]

How do I parse a data structure whose 'discriminant' comes after the data? e.g.

// bytes:      0        1
//        |--------|--------|
//           value     kind
//
// where:
//
// - if kind == 0, value can be one of: 0, 1, 2, 3, 255
// - if kind == 1, value is a regular u8 (between 0..256)
//
// I want to parse it to something like:
#[repr(C)]
enum Data {
    Status(Status),
    ReturnCode(u8),
}

enum Status {
    None = 0,
    Ready = 1,
    Error = 2,
    Pending = 3,
    Invalid = 255,
}

I guess this requires deku to defer parsing the first byte while it examines the second byte, then go "backwards" to parse the first byte.

Is this doable with derive macros and attributes? Or do I have to implement DekuReader and DekuWriter manually?

P. S. deku is an excellent library! Thanks to everyone who worked on it.