Add #[allow(unused_imports)] lint to unstable reexports

[bugadani]

The example shows what the PR does - we no longer have to mark reexports with #[allow(unused_imports)].

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 88.00000% with 3 lines in your changes missing coverage. Please review.

Project coverage is 89.93%. Comparing base (cf84fbe) to head (6669567).

Files with missing lines	Patch %	Lines
src/item_like.rs	83.33%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #21      +/-   ##
==========================================
+ Coverage   89.72%   89.93%   +0.21%     
==========================================
  Files           4        4              
  Lines         613      626      +13     
==========================================
+ Hits          550      563      +13     
  Misses         63       63              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[joshka]

Hey there, this makes sense. Thanks

I tweaked this a bit in 0063cc1

• renamed impl_has_visibility to impl_item_like
• reordered matches so they flow better top to bottom
• added back whitespace (makes things a bit easier to read)
• added comments to document what's happening
• renamed allows macro param to lint
• made the syntax for specifying allowed lints like a normal attribute (#[allow(...))
• removed the changelog entry (will be autopopulated from the commit message - I need to remove that unreleased section to make that more obv.)

Does this seem ok to you? Anything else you'd modify?

I did these changes myself rather than throwing them back your way as I wasn't sure if they'd work / be simpler and the time spent to confirm that was the same as implementing it. The comments were there from trying to understand the change.

For the squash commit mesage, something like:

Add #[allow(unused_imports)] lint to unstable reexports

Reexports marked as unstable no longer need to explicitly allow
the unused_imports lint.

[joshka]

I'll merge and release this as-is if I don't hear from you in a few days, or sooner if you're fine with my changes.

[bugadani]

Thanks, I don't think I mind the changes, except that the macro's usage looks very weird to me.

impl_item_like!(
    syn::ItemType,
    syn::ItemEnum,
    syn::ItemFn,
    syn::ItemMod,
    syn::ItemTrait,
    syn::ItemConst,
    syn::ItemStatic,
    #[allow(unused_imports)]
    syn::ItemUse,
);

This looks like the allow is on the macro's invocation, not something that is part of the syntax itself. I'm slightly worried that extending this syntax to cover other attributes may have surprising effects. If you're okay with this, I don't have objections. Thanks.

[joshka]

This looks like the allow is on the macro's invocation, not something that is part of the syntax itself. I'm slightly worried that extending this syntax to cover other attributes may have surprising effects. If you're okay with this, I don't have objections. Thanks.

I think that worry is a reasonable and I agree it does look a bit odd. I also thought that the change with just the square brackets looked similarly odd. I think most of this stems from the way the macro is constructed, so resolving it would be a more major rewrite. But I think for now (with the comments) the implementation is pragmatic enough. Because this isn't a user facing macro / part of the code, it can be changed in future if needed.