Recursive Decorator

Decorator to apply a given decorator recursively on all function, inside a function/method, recursively.

What is recursive_decorator?

recursive_decorator is a decorator that allows us to decorate/trasform all functions along the stack call at runtime, motivated by the need to add/transform logics, to knownunknown functions, along the stack calls.

Notes:

• Functions/Methods will not be replaced, new instances will be returned.
• Function/Methods cannot be wrapped more then once with same transformer/decorator.

Installing

$ pip install recursive_decorator

Usage

import recursive_decorator

from recursive_decorator import recursive_decorator

define your decorator to apply recursively on all functions.

>>> def decorator(f):
...:    def wrapper(*args, **kwargs):
...:        print(f.__name__)
...:        return f(*args, **kwargs)
...:    return wrapper

Now using your decorator on function without using recursive_decorator will leads to the following output

>>> @decorator
...:def main_function():
...:   sub_function()

>>> main_function()
main_function

Using recursive_decorator leads to

>>> @recursive_decorator(decorator)
...:def main_function():
...:   sub_function()

>>> main_function()
main_function
sub_function

Furthermore, if sub_function has function calls, they will decorated to

>>> def sub_function():
...:    another_function()

>>> @recursive_decorator(decorator)
...:def main_function():
...:   sub_function()

>>> main_function()
main_function
sub_function
another_function

and so on...

Examples

Stop on Execption

We can wrap all functions with try except...

>>> import sys
>>> import ipdb
>>> from recursive_decorator import recursive_decorator

>>> def wrap_function_with_try_except(f):
...:    def transformed_func(*args, **kwargs):
...:        try:
...:            return f(*args, **kwargs)
...:        except:
...:            ipdb.set_trace(sys._getframe().f_back)
...:    return transformed_func

>>> def throws_exception():
...:    raise Exception


>>> @recursive_decorator(wrap_function_with_try_except)
...:def function():
...:    throws_exception()
...:    pass

>>> function()
   21     throws_exception()

---> 22 pass

23

If function will throw an error... ipdb session will start.

Profiler

We can set time profiler for all running functions.

>>> import time

>>> from recursive_decorator import recursive_decorator


>>> def duration_transformer(f):
...:    def transformed_func(*args, **kwargs):
...:        start_time = time.time()
...:        value = f(*args, **kwargs)
...:        end_time = time.time()
...:        print("function {} duration is {} minutes"
...:              .format(f.__name__, end_time - start_time))
...:        return value
...:    return transformed_func


>>> def waiting_function():
...:    time.sleep(5)


>>> @recursive_decorator(duration_transformer)
...:def function():
...:    waiting_function()

>>> function()
function waiting_function duration is 5.00511908531189 minutes
function function duration is 5.006134510040283 minutes