Redundant () on non-mutating calls.

Since Scala 3 enforces the distinction between calls without brackets, and with empty parameter list, some calls become somewhat ugly.

For example:

hasChildNodes()

scala-js-dom/dom/src/main/scala/org/scalajs/dom/Node.scala

Line 155 in 1acf153

def hasChildNodes(): Boolean = js.native

hasAttributes()

scala-js-dom/dom/src/main/scala/org/scalajs/dom/Node.scala

Line 144 in 1acf153

def hasAttributes(): Boolean = js.native

It would probably make sense to revisit all methods that use empty parameter lists and unify. Generally, if the method is non-mutating, it should drop the brackets.

Thanks for reporting.

Generally, if the method is non-mutating, it should drop the brackets.

Yes, this is generally good practice for Scala code. However, for Scala.js facades the brackets are an important distinction:

var, val and def definitions without parentheses all map to field access in JavaScript, whereas def definitions with parentheses (even empty) map to method calls in JavaScript.

Via https://www.scala-js.org/doc/interoperability/facade-types.html

Because these are method calls in JavaScript (and not field accesses), we have no choice but to put the brackets.

Indeed, declaring a method with or without () makes a semantic difference. Declaring them without () would break the meaning of the facade wrt. the JavaScript code behind it. So this very much as designed and must not be changed.