Misleading compiler suggestion when trying to `match` NonZero type

[baumanj]

I tried this code (playground link):

    let n = std::num::NonZeroU32::new(1).unwrap();
    match n {
        std::num::NonZeroU32(i) => {},
    }

I expected this to work.

Instead, this happened:

error[E0532]: expected tuple struct or tuple variant, found struct `std::num::NonZeroU32`
   --> src/lib.rs:4:9
    |
4   |           std::num::NonZeroU32(i) => {},
    |           ^^^^^^^^^^^^^^^^^^^^^^^ help: use struct pattern syntax instead: `std::num::NonZeroU32 { 0 }

And trying that (playground link):

error[E0769]: tuple variant `std::num::NonZeroU32` written as struct variant
 --> src/lib.rs:4:9
  |
4 |         std::num::NonZeroU32 { 0 } => {},
  |         ^^^^^^^^^^^^^^^^^^^^^^^^^^
  |
help: use the tuple variant pattern syntax instead
  |
4 |         std::num::NonZeroU32(0) => {},
  |                             ^^^

Results are the same in beta and nightly.

I recognize the real issue here is that the field of the NonZero newtype structs is private, but the error message could be better.

As an aside, it would be nice to be able to do this matching. In the end I resorted to

        match std::num::NonZeroU32::new(1).unwrap().get() {
            0 => unreachable!(),
            1 => Ok(n),
            d => { … }
        }

which can't be compile-time verified

[leonardo-m]

Rust language needs a well implemented value range analysis.

[luqmana]

Dupe of #78708.