Discussion for iter_nth_zero: would it improve readablility for most cases? should it be enabled by default?

[anatawa12]

.nth(0) and .next() has exactly same behavior, but clearly has different semantics in my opinion: retrieve 0-th element by advance iterator by one, and advance by one with getting current element.
The lint description says .next() is more readable but .nth(0) is more readable in some cases like example code below.

Therefore, I want to discuss if .next() is actually more readable than .nth(0) for all / most cases, and if it's not readable, I want to discuss for moving the lint to allow by default category.

I think if someone knows .nth(), it's likely to the person to know .next() and selected .nth for some reason easons in most cases.

Example code that I think intention is cleaner and more readable with nth(0).
link

pub fn target_arch(target: &str) -> &str {
    target.split('-').nth(0).unwrap()
}

pub fn target_vendor(target: &str) -> &str {
    target.split('-').nth(1).unwrap_or("unknown")
}

pub fn target_os(target: &str) -> &str {
    target.split('-').nth(2).unwrap_or("none")
}

pub fn target_abi(target: &str) -> &str {
    target.split('-').nth(3).unwrap_or("none")
}

[flip1995]

Seems like quite some projects are allowing this lint in specific situations: https://github.com/search?q=%2F%5C%5Ballow%28.*clippy%3A%3Aiter_nth_zero.*%29%5C%5D%2F&type=code&p=1

I don't have a strong opinion about this lint. I'd be fine with moving it to pedantic.