Python decorators are things like @decorator_name that you sometimes see before function definitions.
They can be really useful for things like:
- timing your function to see how long it takes to run
- caching the output from a function so it runs more quickly the second time
- retrying web/API calls if they go wrong or timeout
- logging stages of a pipeline
- modifying the input/output from the function
They are quite a thing to get your head around at first, but if you remember that a decorator is "a function that you pass a function to, that then returns another function" along with the following example, then you have most of the mental model already:
# This code...
@my_decorator
def my_function():
print("hello")
# is EXACTLY the same as this code...
def my_function():
print("hello")
my_function = my_decorator(my_function)
# (my_decorator replaces the my_function variable with a modified function)For some more examples, see: