Allow : after pat_param

[coolreader18]

As on the tin. Perhaps this requires a more formal decision process, but in my view this is something of a bug. Originally, pat_param was just the bikeshodden name of pat2015, c.f. #83386 (comment):

• pat_no_or would be used for closure arguments; can we have something that talks about that (e.g., pat_arg, pat_param -- pattern for a closure argument)
• pat_param is preferred because it is a parameter, not the argument passed to a call

But if it truly is a matcher for "a pattern you can use for a parameter", there shouldn't be a problem allowing it to be followed by a colon, given that that's kind of the definitional constraint for that grammar. And if, in the distant future, the language decides to separate the name and type of a parameter with something other than a colon, well, that'll need an edition change anyway.

In my opinion, this would resolve #87724: I'd imagine that most-if-not-all use cases for a pat followed by a colon are for matching function signatures or let statements, which don't let you use | patterns anyway. But, we get to have our cake and eat it too by still reserving : for future extensions to match expression (pat) patterns. If it ever is desired to make pat match the whole $pat : $ty syntax, as mentioned in #87724 (comment), that's something that can be done over an edition boundary and/or just for pat matchers.

[rustbot]

r? @nnethercote

rustbot has assigned @nnethercote.
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 20 candidates

[nnethercote]

This definitely would need approval from the lang team.

cc @rust-lang/lang @rust-lang/lang-advisors
cc @petrochenkov

[traviscross]

Makes sense to me. Thanks @coolreader18, @nnethercote.

@rfcbot fcp merge lang

[rust-rfcbot]

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

• @joshtriplett
• @nikomatsakis
• @scottmcm
• @tmandry
• @traviscross

Concerns:

• extend-migration-lint (Allow : after pat_param #162264 (comment))

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!

cc @rust-lang/lang-advisors: FCP proposed for lang, please feel free to register concerns.
See this document for info about what commands tagged team members can give me.

[tmandry]

It's surprising we don't already allow this. Thanks for the PR.

@rfcbot reviewed

[rust-bors]

☔ The latest upstream changes (presumably #163095) made this pull request unmergeable. Please resolve the merge conflicts by rebasing.

[scottmcm]

This absolutely makes sense to me. And if for some reason there's an edge case, we have the option later to introduce a new matcher that doesn't allow this.

@rfcbot reviewed

---

There's potentially some surprises that this could cause if people are writing new macros in e2015 (where param means pat_param), but surprises for people doing that doesn't really bother me.

If future if we do want colons for things like Some(x: u32) that wouldn't be impacted by this either because that inner pattern is a full pat, not a pat param.

[rust-rfcbot]

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

[traviscross]

@rfcbot concern extend-migration-lint

Do we need to extend rust_2021_incompatible_or_patterns to handle pat followed by :? Let's please add a test for the migration here.

[traviscross]

@rustbot author

[rustbot]

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

[coolreader18]

There's potentially some surprises that this could cause if people are writing new macros in e2015 (where param means pat_param), but surprises for people doing that doesn't really bother me.

Yeah I considered making it such that only explicit pat_param allows this, but figured it didn't matter very much.

[coolreader18]

Actually maybe I will make it such that only explicit pat_param allows this? That way we don't need to edit a years-old migration lint to be now about "old versions of or-patterns OR param-followed-by-colon"