I want to separate DB configuration before test and test itself.
Let's say I have fixture_1 (user object) with method get_friends.
get_friends making calls to DB, well I should mock some calls beforehead.
I did the decorator below for this purpose, but I can't pass required fixture parameters (*a, **kw) to this function.
Pytest just not triggering on it, how I can do it?
The second decoration is just to avoid typing and wrapping decorator for every test_db_func.
P.S. please tell me if it's a bad idea.
I found this one question How do I make pytest fixtures work with decorated functions? but this will force me to wrap every test_db_func.
def decorator(test_db_func):
#functools_wraps(test_db_func)
def wrapper(unit_func, *a, **kw):
#functools_wraps(unit_func)
def unit_func_params(*unit_func_args, **unit_func_kwargs):
# Prepare DB for test
test_db_func(*a, **kw) # Should pass fixtures to this func
unit_func(*unit_func_args, **unit_func_kwargs)
test_db_func(*a, **kw) # Assertions after yield
return unit_func_params
return wrapper
#test_db_func
def test_x(fixture_1, fixture_2):
assert fixture_1.get_friends(fixture_2)
#decorator
def test_db_func(db_fixture):
db_fixture.return_value = 'foo'
yield
db_fixture.assert_called_once ...
Related
I use testinfra with ansible transport. It provides host fixture which has ansible, so I can do host.ansible.get_variables().
Now I need to create a parametrization of test based on value from this inventory.
Inventory:
foo:
hosts:
foo1:
somedata:
- data1
- data2
I want to write a test which tests each of 'data' from somedata for each host in inventory. 'Each host' part is handled by testnfra, but I'm struggling with parametrization of the test:
#pytest.fixture
def somedata(host):
return host.ansible.get_variables()["somedata"]
#pytest.fixture(params=somedata):
def data(request):
return request.param
def test_data(host, data):
assert 'data' in data
I've tried both ways:
#pytest.fixture(params=somedata) -> TypeError: 'function' object is not iterable
#pytest.fixture(params=somedata()) -> Fixture "somedata" called directly. Fixtures are not meant to be called directly...
How can I do this? I understand that I can't change the number of tests at test time, but I pretty sure I have the same inventory at collection time, so, theoretically, it can be doable...
After reading a lot of source code I have came to conclusion, that it's impossible to call fixtures at collection time. There are no fixtures at collection time, and any parametrization should happen before any tests are called. Moreover, it's impossible to change number of tests at test time (so no fixture could change that).
Answering my own question on using Ansible inventory to parametrize a test function: It's possible, but it requires manually reading inventory, hosts, etc. There is a special hook for that: pytest_generate_tests (it's a function, not a fixture).
My current code to get any test parametrized by host_interface fixture is:
def cartesian(hosts, ar):
for host in hosts:
for interface in ar.get_variables(host).get("interfaces",[]):
yield (host, interface)
def pytest_generate_tests(metafunc):
if 'host_interface' in metafunc.fixturenames:
inventory_file = metafunc.config.getoption('ansible_inventory')
ansible_config = testinfra.utils.ansible_runner.get_ansible_config()
inventory = testinfra.utils.ansible_runner.get_ansible_inventory(ansible_config, inventory_file)
ar = testinfra.utils.ansible_runner.AnsibleRunner(inventory_file)
hosts = ar.get_hosts(metafunc.config.option.hosts)
metafunc.parametrize("host_interface", cartesian(hosts, ar))
You should use helper function instead of fixture to parametrize another fixture. Fixtures can not be used as decorator parameters in pytest.
def somedata(host):
return host.ansible.get_variables()["somedata"]
#pytest.fixture(params=somedata()):
def data(request):
return request.param
def test_data(host, data):
assert 'data' in data
This assumes that the host is not a fixture.
If the host is a fixture, there is hacky way to get around the problem. You should write the parameters to a tmp file or in a environment variable and read it with a helper function.
import os
#pytest.fixture(autouse=True)
def somedata(host):
os.environ["host_param"] = host.ansible.get_variables()["somedata"]
def get_params():
return os.environ["host_param"] # do some clean up to return a list instead of a string
#pytest.fixture(params=get_params()):
def data(request):
return request.param
def test_data(host, data):
assert 'data' in data
In my test suite, I have certain data-generation fixtures which are used with many parameterized tests. Some of these tests would want these fixtures to run only once per session, while others need them to run every function. For example, I may have a fixture similar to:
#pytest.fixture
def get_random_person():
return random.choice(list_of_people)
and 2 parameterized tests, one which wants to use the same person for each test condition and one which wants a new person each time. Is there any way for this fixture to have scope="session" for one test and scope="function" for another?
James' answer is okay, but it doesn't help if you yield from your fixture code. This is a better way to do it:
# Built In
from contextlib import contextmanager
# 3rd Party
import pytest
#pytest.fixture(session='session')
def fixture_session_fruit():
"""Showing how fixtures can still be passed to the different scopes.
If it is `session` scoped then it can be used by all the different scopes;
otherwise, it must be the same scope or higher than the one it is used on.
If this was `module` scoped then this fixture could NOT be used on `fixture_session_scope`.
"""
return "apple"
#contextmanager
def _context_for_fixture(val_to_yield_after_setup):
# Rather long and complicated fixture implementation here
print('SETUP: Running before the test')
yield val_to_yield_after_setup # Let the test code run
print('TEARDOWN: Running after the test')
#pytest.fixture(session='function')
def fixture_function_scope(fixture_session_fruit):
with _context_for_fixture(fixture_session_fruit) as result:
yield result
#pytest.fixture(scope='class')
def fixture_class_scope(fixture_session_fruit):
with _context_for_fixture(fixture_session_fruit) as result:
yield result
#pytest.fixture(scope='module')
def fixture_module_scope(fixture_session_fruit):
with _context_for_fixture(fixture_session_fruit) as result:
yield result
#pytest.fixture(scope='session')
def fixture_session_scope(fixture_session_fruit):
with _context_for_fixture(fixture_session_fruit) as result:
# NOTE if the `_context_for_fixture` just did `yield` without any value,
# there should still be a `yield` here to keep the fixture
# inside the context till it is done. Just remove the ` result` part.
yield result
This way you can still handle contextual fixtures.
Github issue for reference: https://github.com/pytest-dev/pytest/issues/3425
One way to do this to separate out the implementation and then have 2 differently-scoped fixtures return it. So something like:
def _random_person():
return random.choice(list_of_people)
#pytest.fixture(scope='function')
def get_random_person_function_scope():
return _random_person()
#pytest.fixture(scope='session')
def get_random_person_session_scope():
return _random_person()
I've been doing this:
def _some_fixture(a_dependency_fixture):
def __some_fixture(x):
return x
yield __some_fixture
some_temp_fixture = pytest.fixture(_some_fixture, scope="function")
some_module_fixture = pytest.fixture(_some_fixture, scope="module")
some_session_fixture = pytest.fixture(_some_fixture, scope="session")
Less verbose than using a context manager.
Actually there is a workaround for this using the request object.
You could do something like:
#pytest.fixture(scope='class')
def get_random_person(request):
request.scope = getattr(request.cls, 'scope', request.scope)
return random.choice(list_of_people)
Then back at the test class:
#pytest.mark.usefixtures('get_random_person')
class TestSomething:
scope = 'function'
def a_random_test():
def another_test():
However, this only works properly for choosing between 'function' and 'class' scope and particularly if the fixture starts as class-scoped (and then changes to 'function' or is left as is).
If I try the other way around (from 'function' to 'class') funny stuff happen and I still can't figure out why.
I am fairly new to decorators but am experiencing unexpected behavior revolving around autoreload in an interactive workflow with decorated functions. Its best explained by example (note these are all cells in a jupyter notebook):
The decorator:
%%file testdec.py
def decorated(func):
print("decorating")
def wrapped(*args, **kwargs):
return func(*args, **kwargs)
return wrapped
Where the decorator is used:
%%file testmod.py
from testdec import decorated
#decorated
def thisfunc():
print("old output")
def _thisfunc1():
print("old output 1")
thisfunc1 = decorated(_thisfunc1)
I would use the following to call the decorated functions:
from testmod import *
thisfunc()
thisfunc1()
outputs:
decorating
decorating
old output
old output 1
Now updating testmod.py with:
%%file testmod.py
from testdec import decorated
#decorated
def thisfunc():
print("new output")
def _thisfunc1():
print("new output 1")
thisfunc1 = decorated(_thisfunc1)
and calling the functions again:
thisfunc()
thisfunc1()
gives the following, note the old output from the first method:
decorating
decorating
old output
new output 1
However, explicitly reimporting from this module:
from testmod import *
thisfunc()
thisfunc1()
results in:
new output
new output 1
Ideally the #decorated function (e.g. with the # and not the second method) would autoreload transparently as the second method does. Is there something I can do to achieve this? What am I missing for decorated functions. For now we're manually disabling decorators when editing interactively in order to have the benefits of autoreload.
Thanks.
I have a use case where I may use fixture multiple times inside a test in a "context manager" way. See example code below:
in conftest.py
class SomeYield(object):
def __enter__(self):
log.info("SomeYield.__enter__")
def __exit__(self, exc_type, exc_val, exc_tb):
log.info("SomeYield.__exit__")
def generate_name():
name = "{current_time}-{uuid}".format(
current_time=datetime.now().strftime("%Y-%m-%d-%H-%M-%S"),
uuid=str(uuid.uuid4())[:4]
)
return name
#pytest.yield_fixture
def some_yield():
name = generate_name()
log.info("Start: {}".format(name))
yield SomeYield()
log.info("End: {}".format(name))
in test_some_yield.py
def test_some_yield(some_yield):
with some_yield:
pass
with some_yield:
pass
Console output:
INFO:conftest:Start: 2017-12-06-01-50-32-5213
INFO:conftest:SomeYield.__enter__
INFO:conftest:SomeYield.__exit__
INFO:conftest:SomeYield.__enter__
INFO:conftest:SomeYield.__exit__
INFO:conftest:End: 2017-12-06-01-50-32-5213
Questions:
If I have some setup code in SomeYield.enter and cleanup code in
SomeYield.exit, is this the right way to do it using fixture for
multiple calls in my test?
Why didn't I see three occurrences of
enter and exit? Is this expected?
I would like to narrow the scope of the pytest xfail mark. As I currently use it, it marks the entire test function, and any failure in the function is cool.
I would like to narrow that down to a smaller scope, perhaps with a context manager similar to "with pytest.raises (module.Error)". For example:
#pytest.mark.xfail
def test_12345():
first_step()
second_step()
third_step()
This test will xfail if I assert in any of the three methods I call. I would like instead for the test to xfail only if it asserts in second_step(), and not elsewhere. Something like this:
def test_12345():
first_step()
with pytest.something.xfail:
second_step()
third_step()
Is this possible with py.test?
Thanks.
You can define a context manager yourself that does it, like this:
import pytest
class XFailContext:
def __enter__(self):
pass
def __exit__(self, type, val, traceback):
if type is not None:
pytest.xfail(str(val))
xfail = XFailContext()
def step1():
pass
def step2():
0/0
def step3():
pass
def test_hello():
step1()
with xfail:
step2()
step3()
Of course you can also modify the contextmanager to look for specific exceptions.
The only caveat is that you cannot cause an "xpass" outcome, i.e. a special result that the (part of the) test unexpectedly passed.