Allow [] for macro invocation in item places

[bossmc]

The following code doesn't compile:

macro_rules! unit {
    ($name:ident) => { struct $name; }
}

unit!(Parens);
unit!{Braces}
unit![Brackets];

With the following compliant:

src/test/run-pass/macro-delimiters-in-item-position.rs:17:1: 17:5 error: expected item, found `unit`
src/test/run-pass/macro-delimiters-in-item-position.rs:17 unit![Brackets];
                                                          ^~~~
error: aborting due to previous error

It seems that, unlike in expr position, item macros couldn't be invoked with name![...]. This seems to be caused by a copy-paste issue, where the "could this be a macro" test is repeated in various places but with slightly different logic in each.

[rust-highfive]

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

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

Please see the contribution instructions for more information.

[pnkfelix]

cc @rust-lang/lang especially @nrc ; are we okay with this generalization to macro syntax?

[nrc]

We should not do this. Macro use syntax follows from the regular syntax allowed in that context - so macros in statement position need a ; and those in expression position do not, etc. For item position, we have items which are braced and those with (...);, we do not have any items that have square brackets, so we should not allow macros with square brackets either.

[bossmc]

Interesting, I wasn't aware of that symmetry between macros and their resolved forms.

I'm happy either way, but as a counter-point argument for allowing [], I often chose to use macro![...] to make it easier to see where a scope is a macro body as opposed to another (parenthetical) scope.

Feel free to close this PR if you want to stick with the restricted syntax.

[bors]

☔ The latest upstream changes (presumably #36764) made this pull request unmergeable. Please resolve the merge conflicts.

[nikomatsakis]

I agree that the intention was basically to mirror existing forms (e.g., tuple structs) and that extending to [] seems to defeat that intention.

[pnkfelix]

@rfcbot fcp close

[pnkfelix]

oh ha ha this is the rust repo, so rfcbot is irrelevant

---

(or maybe something will happen ... in .... time ....)

[pnkfelix]

This seems like a sufficiently significant change that it would require an RFC.

(you are welcome to compose such an RFC, but the lang team currently does not seem terribly disposed to the idea, so you should probably first see about building up a solid argument for it, perhaps via community discussion either on the IRC channels or in the internals.rust-lang.org forum.)

[rfcbot]

FCP proposed with disposition to close. Review requested from:

• @aturon
• @eddyb
• @nikomatsakis
• @nrc
• @pnkfelix

No concerns currently listed.
See this document for info about what commands tagged team members can give me.

[nikomatsakis]

@rfcbot fcp cancel

[rfcbot]

@nikomatsakis FCP proposal cancelled.