CODEWITHSUNDEEP
pythonunit-testingpytest
anarcat
anarcat

Reputation: 6197

import runs tests twice in pytest

why does py.test run the TestFoo.test_foo() test there? I understand it runs TestBar.test_foo().

Contents of test_foo.py:

import unittest

class TestFoo(unittest.TestCase):
    def test_foo(self):
        print "in test_foo"

Contents of test_bar.py:

from test_foo import TestFoo

class TestBar(TestFoo):
    def test_bar(self):
        print "in test_bar"

Output:

[999]anarcat@marcos:t$ pytest -v
no test dir found testing here: /tmp/t
===========================  test_bar.py  ============================
test_bar (test_bar.TestBar) ... in test_bar
ok
test_foo (test_bar.TestBar) ... in test_foo
ok
test_foo (test_foo.TestFoo) ... in test_foo
ok

===========================  test_foo.py  ============================
test_foo (test_foo.TestFoo) ... in test_foo
ok

*******************************************************************************
Ran 4 test cases in 0.00s (0.00s CPU)
All 2 modules OK

If TestBar is put in the same file as TestFoo, the TestFoo.test_foo() test get ran only once:

import unittest

class TestFoo(unittest.TestCase):
    def test_foo(self):
        print "in test_foo"

class TestBar(TestFoo):
    def test_bar(self):
        print "in test_bar"

Output:

[1001]anarcat@marcos:t$ pytest -v
no test dir found testing here: /tmp/t
===========================  test_foo.py  ============================
test_bar (test_foo.TestBar) ... in test_bar
ok
test_foo (test_foo.TestBar) ... in test_foo
ok
test_foo (test_foo.TestFoo) ... in test_foo
ok

*******************************************************************************
Ran 3 test cases in 0.00s (0.00s CPU)
All 1 modules OK

Shouldn't py.test ignore the tests that are found behind imports?

Upvotes: 5

Views: 3992

Answers (3)

Mahesh_Kokare
Mahesh_Kokare

Reputation: 89

from test_foo import TestFoo :- this sentence will make TestFoo class available to test_bar.py

and

class TestBar(TestFoo): this makes TestBar Class extends TestFoo....

Contents of test_bar.py

this is because u are importing class

class TestFoo(unittest.TestCase): def test_foo(self): print ("in test_foo")

class TestBar(TestFoo): #this is because u are extending TestFoo and hence its function are also available to TestBar def test_foo(self): print ("in test_foo")

def test_bar(self):
    print ("in test_bar")

now its crystal clear what would be the output if u run this particular module..

Upvotes: -1

Andrew Seaman
Andrew Seaman

Reputation: 390

An easy way to fix this is to import the module instead of importing the test class.

import test_foo

class TestBar(test_foo.TestFoo):
    def test_bar(self):
        print "in test_bar"

This will allow you to access the TestFoo class without having the tests run twice.

Upvotes: 8

D.Shawley
D.Shawley

Reputation: 59593

No. It doesn't have a good way to ignore imports. Test runners simply enumerate the names defined in a module and execute ones that look like a test. For example, if you import the first test_bar.py and dir the module, it defines both TestFoo and TestBar. The test runner sees both tests and executes them.

Similarly, TestBar has two methods - test_bar and test_foo. The test runner does not distinguish between names defined by the test class and those inherited from base classes.

Upvotes: 1

Related Questions