Array (and map, set) access and update syntax

[osa1]

We use square brackets for type parameters and type applications:

-- Function definition with type parameter
fn id[T](a: T): T = a

-- Type application in expression context
id[I32](123)

We could still use square brackets for array access, but that reuses [...] a bit too much I think.

I think Scala uses function call syntax for array access, something like:

array(0)          // desugared to `array.apply(0)`, which for the array type gets the element at index
array(0) = "hi"   // desugared to `array.update(0, "hi")

which we could copy.

This is not urgent as we can always just call the methods: map.get(key), map.set(key, value) etc.