Fix false positive on semicolon_if_nothing_returned

[1c3t3a]

Currently the semicolon_if_nothing_returned lint fires in unwanted situations where a block only spans one line. An example of this was given in #7324. This code:

use std::mem::MaybeUninit;
use std::ptr;

fn main() {
    let mut s = MaybeUninit::<String>::uninit();
    let _d = || unsafe { ptr::drop_in_place(s.as_mut_ptr()) };
}

yields the following clippy error:

error: consider adding a `;` to the last statement for consistent formatting
 --> src/main.rs:6:26
  |
6 |     let _d = || unsafe { ptr::drop_in_place(s.as_mut_ptr()) };
  |                          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: add a `;` here: `ptr::drop_in_place(s.as_mut_ptr());`
  |
  = note: `-D clippy::semicolon-if-nothing-returned` implied by `-D clippy::pedantic`
  = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#semicolon_if_nothing_returned

I updated the lint to check if the statement is inside an unsafe block, a closure or a normal block and if the block only spans one line, in that case the lint is not emitted.

This closes #7324.

changelog: enhanced semicolon if nothing returned according to #7324.

[rust-highfive]

r? @flip1995

(rust-highfive has picked a reviewer for you, use r? to override)

[flip1995]

Please rebase on master. You have a commit in this PR that is already on master.

[1c3t3a]

Sorry, just rebased moments before the PR, guess it was bad timing...

[1c3t3a]

Looking at #7322, this PR would favor the unsafe { f(x) }; style. But as described in the issue, it might be better to have this in a separate, configurable lint.

[1c3t3a]

Alright, I added the changes discussed on Zulip, this PR is now ready for review @flip1995!

[flip1995]

Thanks!

[flip1995]

@bors r+

[bors]

📌 Commit 5ec80f3 has been approved by flip1995

[bors]

⌛ Testing commit 5ec80f3 with merge dd02468...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: flip1995
Pushing dd02468 to master...

[1c3t3a]

Looking at #7322, this PR would favor the unsafe { f(x) }; style. But as described in the issue, it might be better to have this in a separate, configurable lint.

What do you think about that @flip1995? Same lint or a new one?

[flip1995]

That should be a new pedantic lint.