pocha was created after spending sometime working in nodejs and finding
that mocha made testing very easy in
nodejs and there wasn't a similar option for python. pocha
isn't an
exact clone but has the various features I felt made writing tests in
python easier and cleaner.
pip install pocha
pip install -e http://github.com/rlgomes/pocha
Writing a test case is a simple as decorator an existing method with it
decorator. Lets write our first test which verifies that the builtin abs
function works as desired:
from pocha import it
@it('can verify that abs works for negative numbers')
def negative_numbers():
assert abs(-1) == 1
The above is all you'd have to do to make a python method become a pocha
test. With the above in a file called test.py
you can then run your test
by using the pocha
cli like so:
> pocha test.py
✓ can verify that abs works for negative numbers
1 passing (0ms)
That was your first test case and it couldn't have been simpler. The correct
way to use pocha though is to split up your tests into test suites and use the
@describe
decorator like so:
from pocha import describe, it
@describe('abs tests')
def _():
@it('can verify that abs works for negative numbers')
def negative_numbers():
assert abs(-1) == 1
Which produces the following output:
> pocha test.py
abs tests
✓ can verify that abs works for negative numbers
1 passing (0ms)
There are decorators for running a method before/after all the tests have
executed in a suite or also to run before/after each test in a suite. The
@before
and @after
decorators are used to provide hooks to run before and
after a whole suite and generally used for setting up or tearing down some
fixture used by all the tests in suite.
As an example the following:
from pocha import after, before, describe, it
@before
def _():
print('before hook')
@after
def _():
print('after hook')
@describe('suite')
def _():
@it('test1')
def test1():
print('from test1')
Which as expected would produce the following when executed:
> pocha foo.py
before hook
suite
from test1
✓ test1
after hook
1 passing (0ms)
Then there are the @before_each
and @after_each
decorators which create
hooks that will run before/after each test in a suite.
You can currently choose from 3 different builtin reporters. Here are the currently available ones and their output for the previous example test:
The default reporter which mimics the same output that mochajs
does with a few
tweaks that should make all white space around elements the same.
> pocha test.py --reporter spec
abs tests
✓ can verify that abs works for negative numbers
1 passing (0ms)
The dot
reporter is a minimalist reporter that can be quite useful if you have
a lot of test cases and want to minimize the amount of output from pocha
.
> pocha test.py --reporter dot
.
1 passing (0ms)
The xunit
reporter is of course for any continuous integration system to
consume in order to report test results at a higher level.
> pocha test.py --reporter xunit
<?xml version="1.0" ?>
<test suite errors="0" failures="0" name="Pocha Tests" skip="0" tests="1">
<test case classname="abs tests" name="can verify that abs works for negative numbers" time="0.000"/>
</test suite>
pocha
has the ability to tag tests and then allow the end user to run only
tests that satisfy a given filter expression. As an example lets first extend
the previous abs
test example to something a bit more elaborate:
from pocha import describe, it
@describe('abs tests')
def _():
@it('can verify that abs works for negative numbers', tags=['negative'])
def _():
assert abs(-1) == 1
@it('can verify that abs works for zero', tags=['zero'])
def _():
assert abs(0) == 0
@it('can verify that abs works for positive numbers', tags=['positive'])
def _():
assert abs(1) == 1
As you can see tagging is straightforward and running a specific test with a tag is also very straightforward:
> pocha test.py --filter negative
abs tests
✓ can verify that abs works for negative numbers
1 passing (0ms)
pocha
has the ability to skip tests easily by setting the skip=True
flag
on the describe
or it
decorator and then that test suite or test case will
simply not execute or appear in the output. The tagging itself would look like
so:
from pocha import describe, it
@describe('abs tests')
def _():
@it('can verify that abs works for negative numbers', tags=['negative'])
def _():
assert abs(-1) == 1
@it('can verify that abs works for zero', tags=['zero'], skip=True)
def _():
assert abs(0) == 0
@it('can verify that abs works for positive numbers', skip=True)
def _():
assert abs(1) == 1
The output of running the above test would look like so:
abs tests
✓ can verify that abs works for negative numbers
1 passing (0ms)
When debugging tests sometimes you simply want to run a single test suite or
test case and instead of tagging the test and then proceeding to specify the
exact tag on the command line you want to simply mark that test unit as the
"only" thing to execute. So there's an only
keyword you can set to True and
pocha
will run just that test suite or test case. Here's what the usage looks
like:
from pocha import describe, it
@describe('abs tests')
def _():
@it('can verify that abs works for negative numbers', only=True)
def _():
assert abs(-1) == 1
@it('can verify that abs works for zero')
def _():
assert abs(0) == 0
@it('can verify that abs works for positive numbers')
def _():
assert abs(1) == 1
Executing the above even when specifying a filter expression on the command line will always result in the following output:
abs tests
✓ can verify that abs works for negative numbers
1 passing (0ms)
You can easily run the existing unittests by installing the test dependencies:
pip install -r test-requirements.txt
Then using pocha
to run the unittests like so:
pocha
Also be aware that python setup.py test
works without having to install the
test dependencies in your python virtualenv.
If you have any issues to report please feel free to open a new issue
here. Also feel free to open PR's
and submit patches to make pocha
better for others.