Improve Error Messaging for Unconstructed Structs and Enum Variants in Generic Contexts

[George-lewis]

Improves error messaging for empty-tuple structs and enum variants in certain generic contexts. See new ui tests for examples.

Closes#87181

[rust-highfive]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @davidtwco (or someone else) soon.

Please see the contribution instructions for more information.

[George-lewis]

r? @estebank

[George-lewis]

Can we assume this won't fail here? Is it ok for it to fail?

[camelid]

Please also add a UI test (src/test/ui/...).

[camelid]

I think you need to run ./x.py test --bless src/test/ui/typeck/issue-87181-empty-tuple-fields.rs.

[George-lewis]

I think I might be missing something? I ran your command and it generated the stderr file, but the test still fails, and it's showing me some JSON.

Edit: I figured it out, I needed a //~^ ERROR in there

[George-lewis]

@camelid Did I pick an acceptable filename for the ui test? I couldn't find much guidance in the contribution guide

[camelid]

The name seems fine to me.

[George-lewis]

Proposal: Go beyond the scope of the ticket to improve error handling to support tuple-structs with elements, and also add the check to member access rather than just to method calls

For instance we can cover this case as so:

structBar<T>{bar:T}structFoo(u8);implFoo{fnfoo(){}}fnmain(){let thing = Bar{bar:Foo};
thing.bar.0;}

error[E0609]: no field `0` on type `fn(u8) -> Foo{Foo}`
--> test.rs:12:15
|
12 | thing.bar.0;
| --------- ^
| |
| help: call the constructor: `(thing.bar)(_)`
|
= help: placeholder
error: aborting due to previous error
For more information about this error, try `rustc --explain E0609`.

Also, if the original code has unnecessary brackets, like (thing.0).0, the current code will suggest a superfluous pair of brackets. Should we worry about that or just let the unnecessary-bracket lint pick it up for us?

[George-lewis]

I have expanded the scope a little in line with my proposal. Now, the same checks are applied to field accesses, and non-empty tuple constructors are handled. There are 2 new tests to cover these cases.

[bors]

💔 Test failed - checks-actions

[George-lewis]

curl: (6) Could not resolve host: ci-mirrors.rust-lang.org

It seems like there might have been a network error?

[estebank]

@bors retry

[bors]

🔒 Merge conflict

This pull request and the master branch diverged in a way that cannot be automatically merged. Please rebase on top of the latest master branch, and let the reviewer approve again.

How do I rebase?

Assuming self is your fork and upstream is this repository, you can resolve the conflict following these steps:

1. git checkout 87181(switch to your branch)
2. git fetch upstream master(retrieve the latest master)
3. git rebase upstream/master -p(rebase on top of it)
4. Follow the on-screen instruction to resolve conflicts (check git status if you got lost).
5. git push self 87181 --force-with-lease(update this PR)

You may also read Git Rebasing to Resolve Conflicts by Drew Blessing for a short tutorial.

Please avoid the "Resolve conflicts" button on GitHub. It uses git merge instead of git rebase which makes the PR commit history more difficult to read.

Sometimes step 4 will complete without asking for resolution. This is usually due to difference between how Cargo.lock conflict is handled during merge and rebase. This is normal, and you should still perform step 5 to update this PR.

Error message

Auto-merging compiler/rustc_typeck/src/check/method/suggest.rs
CONFLICT (content): Merge conflict in compiler/rustc_typeck/src/check/method/suggest.rs
Auto-merging compiler/rustc_typeck/src/check/expr.rs
Automatic merge failed; fix conflicts and then commit the result.

[estebank]

It seems like you'll have to rebase. You might want to squash all your commits into one before doing that to make it easier to resolve the merge conflicts.

[George-lewis]

It does seem that way. I'm looking at it now and it looks like some new diagnostics were implemented, so I'll have to figure out what it does so I can integrate it with my diagnostics 🤔

[bors]

☔ The latest upstream changes (presumably #96428) made this pull request unmergeable. Please resolve the merge conflicts.

[George-lewis]

Oh God

[George-lewis]

I think I messed up the rebase. At this point I just want to make a new branch

Edit: I'm in the process of making a new branch now. I guess I'll open a new PR when it's ready

[George-lewis]

Unfortunately I had to force push, but I think it should be ok now

[estebank]

@bors r+

[bors]

📌 Commit a6b570b has been approved by estebank

[bors]

⌛ Testing commit a6b570b with merge 99b70ee...