Stickybeak is an end to end test helper library that saves lots of testing endpoints and boilerplate code. Usually end to end testing is hard to debug when something goes wrong since modern microservice architecture can be quite convoluted. This library can flatten structure of even the most complex application and help developers write tests that are easy to debug and reflect real life scenarios.
This can save Hundreds of unit and integration tests.
Stickybeak uses code injection to execute arbitrary python code on remote servers in a local environment or dedicated centralized testing microservice. Code injection might sound scary but this solution is completely safe since code injection endpoints are only enabled in testing or staging environment. Results of executed code including all local variables or raised exceptions are pickled on remote server and sent back to the testing script where are unpickled and available for further testing and debugging - just like the code was executed locally.
Pretty amazing huh :) ?
At the moment stickybeak supports Django and Flask frameworks.
pip install stickybeak
To enable stickybeak in a django application add the following line to your urls.py
file:
path("stickybeak/", include("stickybeak.dj.urls")),
And add stickybeak
to INSTALLED_APPS
list.
To enable stickybeak in a flask application add setup(app)
as follows:
from flask import Flask
from stickybeak.flask_view import setup
app = Flask(__name__)
setup(app)
Function arguments and returns as well as default arguments are allowed.
from stickybeak.injector import DjangoInjector
injector = DjangoInjector(address='http://django-srv:8000',
django_settings_module='django_srv.settings')
@injector.function
def fun(a: int, b: int = 2) -> float:
return a / b
fun(2)
Exception are forwarded from remote to local server so the following piece of code raises ZeroDivisionError
:
from stickybeak.injector import DjangoInjector
injector = DjangoInjector(address='http://django-srv:8000',
django_settings_module='django_srv.settings')
@injector.function
def fun() -> float:
a = 1
b = 0
return a / b
fun() # raises ZeroDivisionError
Objects are pickled on the remote side and send back to the local script and are available for further inspection or use.
@self.injector.function
def fun():
from app.models import Currency
currency = Currency()
currency.name = "test_currency"
currency.endpoint = "test_endpoint"
currency.save()
return Currency.objects.all()[0] # noqa
ret: object = fun()
assert ret.name == "test_currency"
assert ret.endpoint == "test_endpoint"
The same concepts go to classes. Only classmethods are allowed at the moment.
@injector.klass
class Interface:
@classmethod
def fun(cls, x: int) -> int:
a = 1
b = 2
return a + b + x
@classmethod
def fun2(cls, x: int) -> int:
c = 3
d = 4
return c + d + x
@classmethod
def fun3(cls) -> int:
return cls.fun(5) + cls.fun2(x=5)
Interface.fun(1) # 4
Interface.fun2(2) # 9
Interface.fun3() # 20
Stickybeak uses pipenv. To install packages run:
pipenv install
honcho start