Fix: #17783 useless_attribute false positive fixed

[GauravPawar101]

Fixes #17783

useless_attribute was short-circuiting whole items. As soon as it hit a permitted attribute or macro check, a bare return bailed on the rest of the item, leaving subsequent attributes completely unchecked.

Because of this early return, source order mattered:

#[allow(unused_imports)]          // Hits allowlist, returns early
#[allow(clippy::almost_swapped)]  // Never evaluated (false negative)
Swapping those two lines made the lint suddenly fire on almost_swapped.

Fix
Evaluated per-attribute instead of per-item: Extracted is_emitted_on_item and moved the external-macro check inline so processing continues past allowed attributes.

Expanded the allowlist: With early returns fixed, added clippy::allow_attributes, clippy::useless_attribute (self-suppression), and exported_private_dependencies.

Tests: Added tests/ui/useless_attribute_order.rs covering order permutations for use and extern crate.

changelog: [useless_attribute]: fix false positive/negative pair caused by early-return in the attribute-allowlist check

em dashes were just a force of habit and were for clear separation and emphasizing!?

[rustbot]

Thanks for the pull request, and welcome!

You should hear from one of our reviewers after this PR gets at least 2 reviews from the community.

Please see the contribution instructions for more information.

[rustbot]

⚠️ Warning ⚠️

• There are issue links (such as #123) in the commit messages of the following commits.
  Please move them to the PR description, to avoid spamming the issues with references to the commit, and so this bot can automatically canonicalize them to avoid issues with subtree.
  • 27aac8a

[rustbot]

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

[GauravPawar101]

@rustbot ready

[CommanderStorm]

Thanks for the response 😆

Community review: mostly LGTM with two questions

For the future (or now, if you have the time) please split this up into one refactoring and one bugfixing commit, since reviewing both code movement, code refactorigns and code fixes is kind of bug prown.
The me figuring out what moved/refactored/changed for every reviewer is kind of unnessary 😉

I also attached two questions about the tests and one refactoring request, both should be simple to answer

View changes since this review

[CommanderStorm]

why rustc_private?

[CommanderStorm]

I feel this can be written a bit clearer and more like you intened.
Can you push the ! down into the function? I feel it would increase the clarity a lot 😉

[CommanderStorm]

🤔
There must be a reason for this, but it is not entirely obvious from looking at the code.
Why is this here and not in am mod issue17783 in uselses_attributes.rs

[CommanderStorm]

please wrap this in an mod issue17562 like above code.

This makes future authors more sure which parts belong to what issue to read more

[CommanderStorm]

Curious: Documented where btw?

View changes since the review