Add regression test for unstable path suggestions

[KevinA-cpu]

closes #78130

[rustbot]

r? @dingxiangfei2009

rustbot has assigned @dingxiangfei2009.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 75 candidates
• Random selection from 21 candidates

[JonathanBrouwer]

Hmmm wel it seems like rust is suggesting an unstable path, so it is not fixed

View changes since the review

[KevinA-cpu]

Hi @JonathanBrouwer , would it be acceptable to add a note for unstable features in nightly builds?

let note = if !is_stable {
    this.tcx.lookup_stability(did).and_then(|stab| match stab.level {
        StabilityLevel::Unstable { .. } => Some(format!(
            "'{}' is unstable in nightly Rust and is only available with the `#![feature({})]` attribute",
            path_names_to_string(&path),
            stab.feature
        )),
        _ => None,
    })
}

or should this be addressed in a different item for nightly builds?

[JonathanBrouwer]

From taking a look at the current code:

rust/compiler/rustc_resolve/src/diagnostics/impls.rs

Line 1932 in 0591d9a

suggestions.retain(|suggestion| suggestion.is_stable || self.tcx.sess.is_nightly_build());

It seems that showing unstable suggestions on nightly is intentional, and I like this.

I think there are two things I'd like to change here:

• IMO We should never be suggesting intrinsics to be imported, unless the core_intrinsics feature is already enabled. The current implementation still suggests them if there is nothing else to suggest, I don't see the point of this as nobody outside of the stdlib should be using these. This is implemented here

  rust/compiler/rustc_resolve/src/late/diagnostics.rs

  Line 957 in 07d9787

  let intrinsic_candidates: Vec<_> = candidates

  
  This was discussed previously here E0425: Don't recommend the use of intrinsics first #97618
  cc @compiler-errors @oli-obk if you have opinions as people who were previously involved in this issue
• As you mentioned, adding an extra note when a suggestion is unstable is a good idea to prevent people from adding the import and then immediately removing it again if they don't want to use features. Make sure to only add this note if the feature is not already enabled.

@KevinA-cpu Do you feel like implementing these two points? Preferably in separate PRs as they are mostly unrelated. Feel free to assign me as a reviewer

@rustbot author

[KevinA-cpu]

Hi @JonathanBrouwer , thank you for the suggestion. I'd love to try my hand at implementing those two points. As I am still new, let's hope it does not take 10 years 🤣

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.