[RFC] feat: Middlewares addition

[petarvujovic98]

Add middlewares decorator
Add middlewares test

Keep in mind that right now the middleware decorator needs to be applied after the other function decorators to preserve the capturing of the respective methods. This, however, could be changed if we want to change the way we capture the decorated methods, e.g. by adding a brand property to the decorated functions rather than looking for the name of the function decorator.

Addresses #256

[ailisp]

This is great implementation! It's using a different approach as in #256, but achieves same semantics!

[petarvujovic98]

One thing that I am not satisfied with is type inference in the decorator. It would be ideal if I could get the editor/language server to tell the developer what type the middleware arguments will be when the middleware decorator is attached to a method. I have not yet seen something similar, but I will try to come up with a solution.

[ailisp]

One thing that I am not satisfied with is type inference in the decorator. It would be ideal if I could get the editor/language server to tell the developer what type the middleware arguments will be when the middleware decorator is attached to a method. I have not yet seen something similar, but I will try to come up with a solution.

This sounds challenging. I think we can address in future PRs. Your current middleware implementation is general, easy-to-use and well tested. IMO, it's ready for merge.

@volovyks Do you observe any issue regarding Petar's implementation?

[volovyks]

No objections @ailisp , it is a great implementation.