string_add is overly sensitive when a String is constructed purely to be passed as an argument

Summary

The string_add lint triggers in situations where it makes no sense to put a push_str on a separate line because the String's only purpose is to be fed as an argument.

At the very least, I think this deserves to be acknowledged in the docs.

Lint Name

string_add

Reproducer

I tried this code:

return Url::parse(&("https://".to_owned() + word)).ok();

(it's part of an HTML form spam pre-filter that normalizes loose domain names into Url objects with the help of the Public Suffix List so they get properly evaluated when applying spam rules.)

I saw this happen:

warning: you added something to a string. Consider using `String::push_str()` instead
   --> src/lib.rs:210:36
    |
210 |                 return Url::parse(&("https://".to_owned() + word)).ok();
    |                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |
    = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#string_add

I expected to see this happen:

Nothing.

In the code the reproducer is from, word is an immutable slice of a larger string, so String::insert_str isn't an option either.

Version

rustc 1.62.1 (e092d0b6b 2022-07-16)
binary: rustc
commit-hash: e092d0b6b43f2de967af0887873151bb1c0b18d3
commit-date: 2022-07-16
host: x86_64-unknown-linux-gnu
release: 1.62.1
LLVM version: 14.0.5

Additional Labels

No response

Start with the string_add lint and the reproducer at src/lib.rs:210, then review the lint guidance linked in the issue. Determine whether this argument-only String construction should be exempted or documented, and validate the chosen behavior against the reproducer.