Ever deal with a large test suite before, monkey patching functions to figure out whether it was called as expected? It's a dirty job. If you're not careful, you can make a mess of things. Leave behind evidence.
kgb's spies will take care of that little problem for you.
Spies intercept and record calls to functions. They can report on how many times a function was called and with what arguments. They can allow the function call to go through as normal, to block it, or to reroute it to another function.
Spies are awesome.
(If you've used Jasmine, you know this.)
Spies are like mocks, but better. You're not mocking the world. You're replacing very specific function logic, or listening to functions without altering them. (See the FAQ below.)
Anything Python-based:
You can even use it outside of unit tests as part of your application. If you really want to. (Probably don't do that.)
- liveswot-api -- REST API Backend for liveswot
- phabricator-emails -- Mozilla's utilities for converting Phabricator feeds to e-mails
- projector -- Takes the overhead out of managing repositories and development environments
- ynab-sdk-python -- Python implementation of the YNAB API
Plus our own products:
- Django Evolution -- An alternative approach to Django database migrations
- Djblets -- An assortment of utilities and add-ons for managing large Django projects
- Review Board -- Our open source, extensible code review product
- RBCommons -- Our hosted code review service
- RBTools -- Command line tools for Review Board
- Power Pack -- Document review, reports, and enterprise SCM integrations for Review Board
- Review Bot -- Automated code review add-on for Review Board
If you use kgb, let us know and we'll add you!
Before you can use kgb, you need to install it. You can do this by typing:
$ pip install kgb
kgb supports Python 2.7 and 3.6 through 3.11, both CPython and PyPy.
Spying is really easy. There are four main ways to initiate a spy.
A SpyAgency manages all your spies. You can create as many or as few as you
want. Generally, you'll create one per unit test run. Then you'll call
spy_on()
, passing in the function you want.
from kgb import SpyAgency
def test_mind_control_device():
mcd = MindControlDevice()
agency = SpyAgency()
agency.spy_on(mcd.assassinate, call_fake=give_hugs)
A SpyAgency
can be mixed into your unittest-based test suite, making
it super easy to spy all over the place, discretely, without resorting to a
separate agency. (We call this the "inside job.")
from kgb import SpyAgency
# Using Python's unittest:
class TopSecretTests(SpyAgency, unittest.TestCase):
def test_weather_control(self):
weather = WeatherControlDevice()
self.spy_on(weather.start_raining)
# Using pytest with the "spy_agency" fixture (kgb 7+):
def test_weather_control(spy_agency):
weather = WeatherControlDevice()
spy_agency.spy_on(weather.start_raining)
If you're creating a spy that calls a fake function, you can simplify some
things by using the spy_for
decorator:
from kgb import SpyAgency
# Using Python's unittest:
class TopSecretTests(SpyAgency, unittest.TestCase):
def test_doomsday_device(self):
dd = DoomsdayDevice()
@self.spy_for(dd.kaboom)
def _save_world(*args, **kwargs)
print('Sprinkles and ponies!')
# Give it your best shot, doomsday device.
dd.kaboom()
# Using pytest:
def test_doomsday_device(spy_agency):
dd = DoomsdayDevice()
@spy_agency.spy_for(dd.kaboom)
def _save_world(*args, **kwargs)
print('Sprinkles and ponies!')
# Give it your best shot, doomsday device.
dd.kaboom()
If you just want a spy for a quick job, without all that hassle of a full
agency, just use the spy_on
context manager, like so:
from kgb import spy_on
def test_the_bomb(self):
bomb = Bomb()
with spy_on(bomb.explode, call_original=False):
# This won't explode. Phew.
bomb.explode()
A spy can do many things. The first thing you need to do is figure out how you want to use the spy.
spy_agency.spy_on(obj.function)
When your spy is called, the original function will be called as well. It won't even know you were there.
spy_agency.spy_on(obj.function, call_original=False)
Useful if you want to know that a function was called, but don't want the original function to actually get the call.
def _my_fake_function(some_param, *args, **kwargs):
...
spy_agency.spy_on(obj.function, call_fake=my_fake_function)
# Or, in kgb 6+
@spy_agency.spy_for(obj.function)
def _my_fake_function(some_param, *args, **kwargs):
...
Fake the return values or operations without anybody knowing.
obj.function.unspy()
Do your job and get out.
for call in obj.function.calls:
print(calls.args, calls.kwargs)
See how many times your spy's intercepted a function call, and what was passed.
# Check the latest call...
print(obj.function.last_call.args)
print(obj.function.last_call.kwargs)
print(obj.function.last_call.return_value)
print(obj.function.last_call.exception)
# For an older call...
print(obj.function.calls[0].args)
print(obj.function.calls[0].kwargs)
print(obj.function.calls[0].return_value)
print(obj.function.calls[0].exception)
Also a good way of knowing whether it's even been called. last_call
will
be None
if nobody's called yet.
Mixing in SpyAgency
into a unittest-based test suite:
# Either one of these is fine.
self.assertSpyCalled(obj.function)
self.assertTrue(obj.function.called)
# Or the inverse:
self.assertSpyNotCalled(obj.function)
self.assertFalse(obj.function.called)
Or using the pytest spy_agency
fixture on kgb 7+:
spy_agency.assert_spy_called(obj.function)
spy_agency.assert_spy_not_called(obj.function)
Or using standalone assertion methods on kgb 7+:
from kgb.asserts import (assert_spy_called,
assert_spy_not_called)
assert_spy_called(obj.function)
assert_spy_not_called(obj.function)
If the function was ever called at all, this will let you know.
Mixing in SpyAgency
into a unittest-based test suite:
# Check if it was ever called with these arguments...
self.assertSpyCalledWith(obj.function, 'foo', bar='baz')
self.assertTrue(obj.function.called_with('foo', bar='baz'))
# Check a specific call...
self.assertSpyCalledWith(obj.function.calls[0], 'foo', bar='baz')
self.assertTrue(obj.function.calls[0].called_with('foo', bar='baz'))
# Check the last call...
self.assertSpyLastCalledWith(obj.function, 'foo', bar='baz')
self.assertTrue(obj.function.last_called_with('foo', bar='baz'))
# Or the inverse:
self.assertSpyNotCalledWith(obj.function, 'foo', bar='baz')
self.assertFalse(obj.function.called)
Or using the pytest spy_agency
fixture on kgb 7+:
spy_agency.assert_spy_called_with(obj.function, 'foo', bar='baz')
spy_agency.assert_spy_last_called_with(obj.function, 'foo', bar='baz')
spy_agency.assert_spy_not_called_with(obj.function, 'foo', bar='baz')
Or using standalone assertion methods on kgb 7+:
from kgb.asserts import (assert_spy_called_with,
assert_spy_last_called_with,
assert_spy_not_called_with)
assert_spy_called_with(obj.function, 'foo', bar='baz')
assert_spy_last_called_with(obj.function, 'foo', bar='baz')
assert_spy_not_called_with(obj.function, 'foo', bar='baz')
The whole callkhistory will be searched. You can provide the entirety of the arguments passed to the function, or you can provide a subset. You can pass positional arguments as-is, or pass them by name using keyword arguments.
Recorded calls always follow the function's original signature, so even if a keyword argument was passed a positional value, it will be recorded as a keyword argument.
Mixing in SpyAgency
into a unittest-based test suite:
# Check if the function ever returned a certain value...
self.assertSpyReturned(obj.function, 42)
self.assertTrue(obj.function.returned(42))
# Check a specific call...
self.assertSpyReturned(obj.function.calls[0], 42)
self.assertTrue(obj.function.calls[0].returned(42))
# Check the last call...
self.assertSpyLastReturned(obj.function, 42)
self.assertTrue(obj.function.last_returned(42))
Or using the pytest spy_agency
fixture on kgb 7+:
spy_agency.assert_spy_returned(obj.function, 42)
spy_agency.assert_spy_returned(obj.function.calls[0], 42)
spy_agency.assert_spy_last_returned(obj.function, 42)
Or using standalone assertion methods on kgb 7+:
from kgb.asserts import (assert_spy_last_returned,
assert_spy_returned)
assert_spy_returned(obj.function, 42)
assert_spy_returned(obj.function.calls[0], 42)
assert_spy_last_returned(obj.function, 42)
Handy for checking if some function ever returned what you expected it to, when you're not calling that function yourself.
Mixing in SpyAgency
into a unittest-based test suite:
# Check if the function ever raised a certain exception...
self.assertSpyRaised(obj.function, TypeError)
self.assertTrue(obj.function.raised(TypeError))
# Check a specific call...
self.assertSpyRaised(obj.function.calls[0], TypeError)
self.assertTrue(obj.function.calls[0].raised(TypeError))
# Check the last call...
self.assertSpyLastRaised(obj.function, TypeError)
self.assertTrue(obj.function.last_raised(TypeError))
Or using the pytest spy_agency
fixture on kgb 7+:
spy_agency.assert_spy_raised(obj.function, TypeError)
spy_agency.assert_spy_raised(obj.function.calls[0], TypeError)
spy_agency.assert_spy_last_raised(obj.function, TypeError)
Or using standalone assertion methods on kgb 7+:
from kgb.asserts import (assert_spy_last_raised,
assert_spy_raised)
assert_spy_raised(obj.function, TypeError)
assert_spy_raised(obj.function.calls[0], TypeError)
assert_spy_last_raised(obj.function, TypeError)
You can also go a step further by checking the exception's message.
# Check if the function ever raised an exception with a given message...
self.assertSpyRaisedWithMessage(
obj.function,
TypeError,
"'type' object is not iterable")
self.assertTrue(obj.function.raised_with_message(
TypeError,
"'type' object is not iterable"))
# Check a specific call...
self.assertSpyRaisedWithMessage(
obj.function.calls[0],
TypeError,
"'type' object is not iterable")
self.assertTrue(obj.function.calls[0].raised_with_message(
TypeError,
"'type' object is not iterable"))
# Check the last call...
self.assertSpyLastRaisedWithMessage(
obj.function,
TypeError,
"'type' object is not iterable")
self.assertTrue(obj.function.last_raised_with_message(
TypeError,
"'type' object is not iterable"))
obj.function.reset_calls()
Wipe away the call history. Nobody will know.
result = obj.function.call_original('foo', bar='baz')
Super, super useful if you want to use call_fake=
or
@spy_agency.spy_for
to wrap a function and track or influence some part of
it, but still want the original function to do its thing. For instance:
stored_results = []
@spy_agency.spy_for(obj.function)
def my_fake_function(*args, **kwargs):
kwargs['bar'] = 'baz'
result = obj.function.call_original(*args, **kwargs)
stored_results.append(result)
return result
Why start from scratch when setting up a spy? Let's plan an operation.
(Spy operations are only available in kgb 6 or higher.)
spy_on(pen.emit_poison, op=kgb.SpyOpRaise(PoisonEmptyError()))
Or go nuts, have a different exception for each call (in kgb 6.1+):
spy_on(pen.emit_poison, op=kgb.SpyOpRaiseInOrder([
PoisonEmptyError(),
Kaboom(),
MissingPenError(),
]))
spy_on(our_agent.get_identity, op=kgb.SpyOpReturn('nobody...'))
Maybe a different value for each call (in kgb 6.1+)?
spy_on(our_agent.get_identity, op=kgb.SpyOpReturnInOrder([
'nobody...',
'who?',
'not telling...',
]))
Now for something more complicated.
If you're dealing with many calls to the same function, you may want to return
different values or only call the original function depending on which
arguments were passed in the call. That can be done with a SpyOpMatchAny
operation.
spy_on(traps.trigger, op=kgb.SpyOpMatchAny([
{
'args': ('hallway_lasers',),
'call_fake': _send_wolves,
},
{
'args': ('trap_tile',),
'op': SpyOpMatchInOrder([
{
'call_fake': _spill_hot_oil,
},
{
'call_fake': _drop_torch,
},
]),
},
{
'args': ('infrared_camera',),
'kwargs': {
'sector': 'underground_passage',
},
'call_original': False,
},
]))
Any unexpected calls will automatically assert.
You can combine that with requiring the calls to be in the order you want
using SpyOpMatchInOrder
.
spy_on(lockbox.enter_code, op=kgb.SpyOpMatchInOrder([
{
'args': (1, 2, 3, 4, 5, 6),
'call_original': False,
},
{
'args': (9, 0, 2, 1, 0, 0),
'call_fake': _start_countdown,
},
{
'args': (42, 42, 42, 42, 42, 42),
'op': kgb.SpyOpRaise(Kaboom()),
'call_original': True,
},
{
'args': (4, 8, 15, 16, 23, 42),
'kwargs': {
'secret_button_pushed': True,
},
'call_original': True,
}
]))
kgb's spies and mock's patching are very different from each other. When
patching using mock, you're simply replacing a method on a class with
something that looks like a method, and that works great except you're limited
to methods on classes. You can't override a top-level function, like
urllib2.urlopen
.
kgb spies leave the function or method where it is. What it does do is replace the bytecode of the function, intercepting calls on a very low level, recording everything about it, and then passing on the call to the original function or your replacement function. It's pretty powerful, and allows you to listen to or override calls you normally would have no control over.
It is! It really is! We've been using it for years across a wide variety of codebases. It's pretty amazing.
Python actually allows this. We're not scanning your RAM and doing terrible
things with it, or something like that. Every function or method in Python has
a func_code
(Python 2) or __code__
(Python 3) attribute, which is
mutable. We can go in and replace the bytecode with something compatible with
the original function.
How we actually do that, well, that's complicated, and you may not want to know.
I'm going to level with you, I was going to say "hell no!", and then decided to give it a try.
Hell yes! (But only accidentally. YMMV... We'll try to officially support this later.)
Lots of things. Check out some of our other open source projects.