Auger
Auger is a project to automatically generate unit tests for Python code.
See these slides or this blog entry for more information.
Installation
Install auger with:
pip install auger-python
Running Auger
To generate a unit test for any class or module, for Python 2 or 3, do this:
import auger
with auger.magic([ <any list of modules or classes> ]):
<any code that exercises your application>
A Simple Example
Here is a simple example that does not rely on Auger at all:
class Foo: # Declare a class with a method
def bar(self, x):
return 2 * x # Duplicate x and return it
def main():
foo = Foo() # Create an instance of Foo
print(foo.bar(32)) # Call the bar method and print the result
main()
Inside the main
function we call the bar
method and it will print 64.
Running Auger on our Simple Example
To generate a unit test for this class, we run the code again, but this time in the context of Auger:
import auger
with auger.magic([Foo]):
main()
This will print out the following:
64
Auger: generated test: tests/test_Foo.py
The test that is generated looks like this, with some imports and test for main removed:
import unittest
class FooTest(unittest.TestCase):
def test_bar(self):
foo_instance = Foo()
self.assertEquals(
foo_instance.bar(x=32),
64
)
if __name__ == "__main__":
unittest.main()
Running Auger in verbose mode
Rather than emit tests in the file system, Auger can also print out the test to the console,
by using the verbose
parameter:
import auger
with auger.magic([Foo], verbose=True):
main()
In that case, Auger will not generate any tests, but just print them out.
A larger example
Consider the following example, pet.py
, included in the sample
folder, that lets us create a Pet
with a name and a species:
from animal import Animal
class Pet(Animal):
def __init__(self, name, species):
Animal.__init__(self, species)
self.name = name
def getName(self):
return self.name
def __str__(self):
return "%s is a %s" % (self.getName(), self.getSpecies())
def createPet(name, species):
return Pet(name, species)
A Pet
is really a special kind of Animal
, with a name, which is defined in animal.py
:
class Animal(object):
def __init__(self, species):
self.species = species
def getSpecies(self):
return self.species
With those two definitions, we can create a Pet
instance and print out some details:
import animal
import pet
def main():
p = pet.createPet("Polly", "Parrot")
print(p, p.getName(), p.getSpecies())
main()
This produces:
Polly is a Parrot Polly Parrot
Calling Auger on our larger example
With Auger, we can record all calls to all functions and methods defined in pet.py
,
while also remembering the details for all calls going out from pet.py
to other modules,
so they can be mocked out.
Instead of saying:
if __name__ == "__main__":
main()
We would say:
import auger
if __name__ == "__main__":
with auger.magic([pet]): # this is the new line and invokes Auger
main()
This produces the following automatically generated unit test for pet.py
:
from mock import patch
from sample.animal import Animal
import sample.pet
from sample.pet import Pet
import unittest
class PetTest(unittest.TestCase):
@patch.object(Animal, 'get_species')
@patch.object(Animal, 'get_age')
def test___str__(self, mock_get_age, mock_get_species):
mock_get_age.return_value = 12
mock_get_species.return_value = 'Dog'
pet_instance = Pet('Clifford', 'Dog', 12)
self.assertEquals(pet_instance.__str__(), 'Clifford is a dog aged 12')
def test_create_pet(self):
self.assertIsInstance(sample.pet.create_pet(age=12,species='Dog',name='Clifford'), Pet)
def test_get_name(self):
pet_instance = Pet('Clifford', 'Dog', 12)
self.assertEquals(pet_instance.get_name(), 'Clifford')
def test_lower(self):
self.assertEquals(Pet.lower(s='Dog'), 'dog')
if __name__ == "__main__":
unittest.main()
Note that auger detects object creation, method invocation, and static methods. It automatically
generate mocks for Animal
. The mock for get_species
returns 'Dog' and get_age
returns 12.
Namely, those were the values Auger recorded when we ran our sample code the last time.
Benefits of Auger
By automatically generating unit tests, we dramatically cut down the cost of software development. The tests themselves are intended to help developers get going on their unit testing and lower the learning curve for how to write tests.
Known limitations of Auger
Auger does not do try to substitue parameters with synthetic values such as -1
, None
, or []
.
Auger also does not act well when code uses exceptions. Auger also does not like methods that have a decorator.
Auger only records a given execution run and saves the run as a test. Auger does not know if the code actually works as intended. If the code contains a bug, Auger will simply record the buggy behavior. There is no free lunch here. It is up to the developer to verify the code actually works.