Enabled

UnnecessaryParentheses

rule and it reported this code

offset.y.toInt() in item.offset..(item.offset + item.size)

Sure, technically those can be removed, but wouldn't it look less readable? Do you think this might be an exception or config option? If yes, I can create issue and occasionally a PR.

I think this is probably worth an exception in the rule config or a local suppress. 🤔 It’s essentially equivalent to:

i in 0..(1+2)
// vs
i in 0..1+2

yeah, maybe it should be a local suppress. by the way, another one I found "controversial" is this code

(someClass.value.toInt() + otherClass.value) to (size + count)

vs

someClass.value.toInt() + otherClass.value to size + count

I find the latter less readable.

if infix precedence confuses you, write it without

offset.y.toInt() to item.offset.rangeTo(item.offset + item.size)

(someClass.value.toInt() + otherClass.value).to(size + const)

oh, nice idea!

that being said, while Kotlin's rules are straightforward (infix has precedence between arithmetic and logic, and always associates to the left), it does work differently enough from other languages that I can understand wanting a configurable warning about potential confusion, similar to Clippy's https://rust-lang.github.io/rust-clippy/stable/index.html#precedence. certainly it took me a while to get used to bitwise

and

and

or

having the same precedence, unlike

&

and

|

in every other language…

IIRC in haskell you can even override operator fixity (precedence) locally. And when creating new operators (which are functions of course), you can specify precedence.

yes, and left/right/neither associativity can be specified as well. but I don't expect Kotlin to ever do that

The problem with this rule is that "readability" is a broad term so it's difficult to write a line. I agree with you, those parentheses help. I know that IntelliJ now have a rule that even tells you "It's true that you don't need parentheses here but we recommend to add them to improve readability". A port from that code could be a good compromise.