Prefix methods?

[Lumintorious]

We already have infix methods and we have limited unary operators, I think an option for prefix methods would allow for more expressivity, allowing some methods to behave like keywords and allowing more unary operators.

Use cases:

raise Exception(???)
await Future(???)
reflect IO(???)
assert ???
require ???
not true
print ???
pure ???

I don't know if those should be defined as instance methods or as parametrized methods.
One problem I see with this is differentiating between

f g x = f(g(x))
and
f g x = f.g(x)

If the implementation would be instance methods, a check for whether the first term name f is in scope, if it's in scope then find member f of g x if it's not try looking at member g of f

[LPTK]

I think this would be too ambiguous. The parser is supposed to resolve these things, but now how things eventually parse would have to be based on type checking info? Seems hardly doable in a clean way. Also, trying to parse f g x as f(g(x)) by default would probably break most existing code bases.

[SethTisue]

I think Scala is moving away from this level of syntactic flexibility, rather than moving towards it. (Consider the elimination of postfix syntax and the restriction of infix syntax to methods that specifically declared infix.)

[som-snytt]

I miss postfix, and I think with Scala 3 it could be robust again. Maybe also prefix!

Let's not be cowards!

That is, forwards or backwards, but wards the cow? I mean I like the cow that showed up in La Brea.