Fix #[derive(Default)] on a generic #[default] enum adding unnecessary Default bounds

[danielhenrymantilla]

That is, given something like:

// #[default] on a generic enum does not add `Default` bounds to the type params.
#[derive(Default)]
enum MyOption<T> {
    #[default]
    None,
    Some(T),
}

then MyOption<T> : Defaultas currently implemented only holds when T : Default, as reported by @5225225 over Zulip.

This is contrary to what the accepted RFC proposes (i.e., that T be allowed not to be itself Default), and indeed seems to be a rather unnecessary limitation.

[joshtriplett]

Nominating for discussion. Also kicking off an FCP: shall we remove the apparently unintended bound here?

@rfcbot merge

[rfcbot]

Team member @joshtriplett has proposed to merge this. The next step is review by the rest of the tagged team members:

• @cramertj
• @joshtriplett
• @nikomatsakis
• @pnkfelix
• @scottmcm

No concerns currently listed.

Once a majority of reviewers approve (and at most 2 approvals are outstanding), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

[joshtriplett]

We discussed this in today's @rust-lang/lang meeting. We'd like to hand this over to @rust-lang/libs-api, on the basis that there's no special language functionality needed to implement this.

@rfcbot cancel

[rfcbot]

@joshtriplett proposal cancelled.

[joshtriplett]

@rfcbot merge

[rfcbot]

Team member @joshtriplett has proposed to merge this. The next step is review by the rest of the tagged team members:

• @Amanieu
• @BurntSushi
• @dtolnay
• @joshtriplett
• @m-ou-se
• @yaahc

No concerns currently listed.

Once a majority of reviewers approve (and at most 2 approvals are outstanding), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

[Mark-Simulacrum]

I think it's a good idea to update src/test/ui/macros/macros-nonfatal-errors.rs to make sure that #[non_exhaustive] #[default] (currently an error, it seems) when supported does generate the right bound. I think given the non-fatal there it's possible that we're emitting code, just later erroring? (In any case seems good to write a test to avoid regressing if/when we add non_exhaustive support).

[danielhenrymantilla]

So, to clarify, you'd like me to add something akin to:

rust/src/test/ui/macros/macros-nonfatal-errors.rs

Lines 89 to 95 in 5b4bd15

	#[derive(Default)]
	enum NonExhaustiveDefault {
	#[default]
	#[non_exhaustive]
	Foo, //~ ERROR default variant must be exhaustive
	Bar,
	}

but with a, say, <T>-generic type, and then assert that NonExhaustiveDefaultGeneric<NotDefault> does not implement Default?

EDIT: I hope 3d4980b is what you wanted

[joshtriplett]

@Mark-Simulacrum In theory it should be OK to explicitly mark a non_exhaustive variant as Default, as long as you're committing in your API to making it stay that way in the future. But I do think that should be a separate decision, and not part of this PR. So, this seems fine.

[bors]

☔ The latest upstream changes (presumably #100869) made this pull request unmergeable. Please resolve the merge conflicts.

[yaahc]

@joshtriplett is this related to the "perfect derive" stuff I remember the lang team talking about a while back?

[scottmcm]

@yaahc Sortof, but not really, because the feature is intentionally limited right now to only work when the #[default] is on a fieldless variant. Rather the RFC discussion for this explicitly decided to make this derive work differently from the others.

[m-ou-se]

Yeah, that derived Default impl was never supposed to have that bound. That seems to be an accident we should fix. :)

(Haven't checked the code, that's for the compiler team. But the resulting change looks good to me.)

[rfcbot]

🔔 This is now entering its final comment period, as per the review above. 🔔

[rfcbot]

The final comment period, with a disposition to merge, as per the review above, is now complete.

As the automated representative of the governance process, I would like to thank the author for their work and everyone else who contributed.

This will be merged soon.

[joshtriplett]

@bors r+ rollup

[bors]

📌 Commit 3d4980b has been approved by joshtriplett

It is now in the queue for this repository.