API: Removed the CallableData trait

[xFrednet]

The CallableData trait, was designed with the idea to have a general trait for function like things, including FnItem, ClosureExpr, SenFnPrtTy. However, after working with Marker for a bit, I learned that these AST nodes are just too different. At least, too different, to have the CallableData trait as the main API.

I'm also not sure what the use case would be, where a syntactic type should be handled the same way as a FnItem.

This PR simply removes the trait and moves the remaining functions to the SenFnPrtTy type. #194 already removed the trait from FnItem and ClosureExpr.

This PR also deletes the SynClosureTy since it's not actually a syntactic type that can be written ^^

[Veetaha]

That's weird how there could be a syntactic representation of a closure type because there is no syntax to denote a type of the closure except for impl Fn*. However, impl Fn* should not be coupled with closures, that's a broader concept.

So I wonder what this struct represented then

[xFrednet]

I don't actually know anymore why I added this struct. My guess is that this is a left over from an old architecture. There was a time, where I planned to merge both syntactic and semantic types into a single representation. My idea was that this would make it easier, to work with types. Then you could compare syntactic and semantic types etc.

However, while working on semantic types, I quickly noticed that these are way more different than I thought. It also became clear that this combination will most likely be painful for everyone involved, with very few benefits.

During the move to split semantic and syntactic types, I just added the Syn prefix to most types. The struct might have been intended to be a semantic type, but got caught in the crossfire.