rustc_parse: improve the error diagnostic for "missing let"

[Unique-Usman]

No description provided.

[rustbot]

r? @fee1-dead

rustbot has assigned @fee1-dead.
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

[Unique-Usman]

r @estebank

[fmease]

Wait, this renders if Some(_) = x {} syntactically invalid whereas on main it's merely semantically invalid, right? If so, this constitutes a breaking change and is generally not acceptable¹.

Footnotes

1. Unless T-lang is in favor of it (which I doubt) and crater is happy ↩

[Unique-Usman]

The parser is going to sucessfully parse this, but, the checking happens through the CondCheker(during the parsing process though) as it for all the cond after if and while https://github.com/rust-lang/rust/pull/151960/changes#diff-a278ef012eea7cc71e4c028d8c61c358e7a7cd7e35f846a41c75db09f893b998R4262. Does that constitute to changing from semantically invalidd to synthatically correct ?

If it is, I can limit the changes to

  if Some(_) = y &&
        let Some(_) = z

[fmease]

Check if your PR makes rustc reject the following code. If it does, then it's a breaking change.

fn main() {
    #[cfg(false)]
    if Some(_) = x {}
}

[Unique-Usman]

Yeah, it is a breaking change, I think I should remove it. What do you think ?

[estebank]

Yes, we should not introduce any breaking changes.

[rustbot]

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

[Unique-Usman]

This is a continuation of #151493 as commented in #151493 (comment)

[Unique-Usman]

r? @estebank

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.