Python Function Annotations: Adding Metadata to Functions

The function annotation feature of Python enables you to add additional explanatory metada about the arguments declared in a function definition, and also the return data type.

Although you can use the docstring feature of Python for documentation of a function, it may be obsolete if certain changes in the function's prototype are made. Hence, the annotation feature was introduced in Python as a result of PEP 3107.

The annotations are not considered by Python interpreter while executing the function. They are mainly for the Python IDEs for providing a detailed documentation to the programmer.

Annotations are any valid Python expressions added to the arguments or return data type. Simplest example of annotation is to prescribe the data type of the arguments. Annotation is mentioned as an expression after putting a colon in front of the argument.

Remember that Python is a dynamically typed language, and doesn't enforce any type checking at runtime. Hence annotating the arguments with data types doesn't have any effect while calling the function. Even if non-integer arguments are given, Python doesn't detect any error.

It will produce the following output −

30
Hello Python

Annotations are ignored at runtime, but are helpful for the IDEs and static type checker libraries such as mypy.

You can give annotation for the return data type as well. After the parentheses and before the colon symbol, put an arrow (->) followed by the annotation. For example −

As using the data type as annotation is ignored at runtime, you can put any expression which acts as the metadata for the arguments. Hence, function may have any arbitrary expression as annotation as in following example −

If you want to specify a default argument along with the annotation, you need to put it after the annotation expression. Default arguments must come after the required arguments in the argument list.

The function in Python is also an object, and one of its attributes is __annotations__. You can check with dir() function.

This will print the list of myfunction object containing __annotations__ as one of the attributes.

['__annotations__', '__builtins__', '__call__', '__class__', '__closure__', '__code__', '__defaults__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__get__', '__getattribute__', '__getstate__', '__globals__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__kwdefaults__', '__le__', '__lt__', '__module__', '__name__', '__ne__', '__new__', '__qualname__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__']

The __annotations__ attribute itself is a dictionary in which arguments are keys and anootations their values.

It will produce the following output −

{'a': 'physics', 'b': 'Maths', 'return': <class 'int'>}

You may have arbitrary positional and/or arbitrary keyword arguments for a function. Annotations can be given for them also.

It will produce the following output −

{'args': 'arbitrary args', 'kwargs': 'arbitrary keyword args', 'return': <class 'int'>}

In case you need to provide more than one annotation expressions to a function argument, give it in the form of a dictionary object in front of the argument itself.

It will produce the following output −

{'num': {'type': <class 'float'>, 'msg': 'numerator'}, 'den': {'type': <




The End! should you have any inquiries, we encourage you to reach out to the Vercaa Support Center without hesitation.