How to write a test for a function with optional arguments

Question

I want to test function calls with optional arguments.

Here is my code:

list_get()
list_get(key, "city", 0)
list_get(key, 'contact_no', 2, {}, policy)
list_get(key, "contact_no", 0)
list_get(key, "contact_no", 1, {}, policy, "")
list_get(key, "contact_no", 0, 888)

I am not able to parametrize it due to optional arguments, so I have written separate test functions for each api call in pytest.
I believe there should be better way of testing this one.

Can't you use some star-magic e.g. list_get(*args) to make it work with parameterised tests? — Tom Dalton, Mar 07 '16 at 13:27

score 15 · Accepted Answer · edited Nov 11 '17 at 14:56

You may be able to use the * operator:

@pytest.mark.parametrize('args,expected', [
    ([], expVal0),
    ([key, "city", 0], expVal1),
    ([key, 'contact_no', 2, {}, policy], expVal2)
    ([key, "contact_no", 0], expVal3)
    ([key, "contact_no", 1, {}, policy, ""], expVal4)
    ([key, "contact_no", 0, 888], expVal5)
])
def test_list_get(args, expected):
    assert list_get(*args) == expected

score 4 · Answer 2 · answered Mar 24 '16 at 14:51

In addition to the answers @forge and @ezequiel-muns I suggest using some sugar from pyhamcrest:

import pytest
from hamcrest import assert_that, calling, is_not, raises

@pytest.mark.parametrize('func, args, kwargs', [
    [list_get, (), {}],
    [list_get, (key, "city", 0), {}],
    [list_get, (key, "contact_no", 1, {}, policy, ""), {}],
    [list_get, (), {'key': key}],
])
def test_func_dont_raises(func, args, kwargs):
    assert_that(calling(func).with_args(*args, **kwargs), is_not(raises(Exception)))

Nice, going to check this library out! – Ezequiel Muns Mar 30 '16 at 06:28 — Ezequiel Muns, Mar 30 '16 at 06:28

score 1 · Answer 3 · answered May 27 '21 at 18:14

For future readers who come to this question trying to set up @parameterized tests to generate a Cartesian set of parameters AND sometimes do not want to pass a given parameter at all (if optional), then using a filter on None values will help

def function_under_test(foo="foo-default", bar="bar-default"):
    print([locals()[arg] for arg in inspect.getargspec(function_under_test).args])


@pytest.mark.parametrize("foo", [None, 1, 2])
@pytest.mark.parametrize("bar", [None, "a", "b"])
def test_optional_params(foo, bar):
    args = locals()
    filtered = {k: v for k, v in args.items() if v is not None}
    function_under_test(**filtered)  # <-- Notice the double star

Sample run:

PASSED   [ 11%]['foo-default', 'bar-default']
PASSED   [ 22%][1, 'bar-default']
PASSED   [ 33%][2, 'bar-default']
PASSED   [ 44%]['foo-default', 'a']
PASSED   [ 55%][1, 'a']
PASSED   [ 66%][2, 'a']
PASSED   [ 77%]['foo-default', 'b']
PASSED   [ 88%][1, 'b']
PASSED   [100%][2, 'b']

score 0 · Answer 4 · answered Mar 07 '16 at 13:46

I think what you are doing is good, but I would suggest to pass key-value arguments so you won't get the arguments order mixed up.

I assume your function header looks something like this:

def list_get(key=None, city=None, contact_no=None, policy=None):
    ...

In your test define the list of arguments combination you would like to test:

kwargs = {'key': '..', 'city': '..', 'contact_no': '..', 'policy': '..'} 
list_get(**kwargs) 

kwargs_all_default_values = {}
list_get(**kwargs_all_default_values)

How to write a test for a function with optional arguments

4 Answers4