Decorators Kata

This mini-kata explores Python decorators.

Introduction

This project is super-simple. All code should happen in main.py.

• Initial setup: poetry install
• Running: poetry run python main.py

All the steps below have corresponding branches—step1-implemented, step2-implemented, etc.—that contain one possible implementation of that particular step. Feel free to use those branches or to create your own implementations. (And if you do create your own, may I suggest branches like step1, step2?)

Run the basic program once without any changes. What is the output? What does it show for the function metadata?

Steps

1. Wrap the hello() function—without using the decorator syntax—in a trace() function so that you get the following output:

   🤖: function called
   👋: Hello, world!
   🤖: function returning
   

   What happened to the function metadata?
2. Use the decorator syntax to wrap hello() with @trace. Now what happened to the function metadata? Hint: your trace() needs to use a nested function.
3. Modify hello() to accept a name string. Use that name in the greeting, so that hello('John') produces:

   🤖: function called
   👋: Hello, John!
   🤖: function returning
   

   Hint: you can pass *args, **kwargs from the inner function through to the wrapped function.
4. Support passing an enabled keyword argument to the @trace decorator.
   • if True, @trace should print out the additional logging
   • if False, @trace should not print out the additional logging
   • enabled should be set to False by default Hint: you'll need to add another layer of nested functions.
5. Use the decorator @decorator to simplify the implementation. Note that you'll need to add this dependency first via poetry add decorator. When we use @decorator, what happens to our function metadata? Hint: all nested functions collapse, including their arguments, into a single, un-nested function.

One final note: we could also use the built-in @wraps decorator from the functools module to simplify our decorator. Like @decorator, @wraps preserves function metadata; however, it does not collapse the nested functions. It is for that reason that I still prefer @decorator for all your custom decorating needs.