I probably don’t need to tell you, how important testing your code is in software developement. One way of doing so is writing unit tests. Unit tests are an automated set of test cases. These test cases are there to make sure, that the code does exactly what it should, nothing more and nothing less.
Every once in a while you change something in the software, you simply just run the tests and see if the changes didn’t break something that was working fine beforehand. A set of good tests will increase your confidence while releasing your software and (believe me) those stupid unit tests of class interface can save literally weeks of debugging.
In agile, you’re supposed to write the tests before you implement the actual functionality. That’s how the tests are important :-). What are our testing possibilities in Python? Well, let’s have a look …
Python itself (all hail!) provides pretty good testing framework caled unittest. Let’s have a look how it works. In the center of it all is a class unittest.TestCase. While creating an unit test, you simply create a derived class from it. Here’s an example how could a test file look like:
You should create a file like this for every module in your code. Now,
let’s have a look on how to organize the tests in your project. The best
practice is to separate them all into a folder called
test . Directory tree
of your project might look like this:
This is nice, but now that you’re sources are in a separate directory it’s a
little tricky to be able to import your modules into your tests and then run
all the tests. For that I have written a simple script
run_tests.py that will
do it for you. Here it goes:
For more on this subject, see the unittest documentation.
is an example setup of unit tests for your project. It contains a convenient
script, that runs all the tests in the
One of key features of an unit testing framework is simplicity. You need to be able to create new tests fast and efficiently, because you’re gonna be doing it A LOT. The harder it is to write new tests, the more you’ll be skipping them and that’s a highway to coder’s hell.
Using the Python’s built-in
unittest framework isn’t that time-consuming,
but there is still some space for improvement. That’s where
nose comes in.
It’s a third-party testing framework for Python and it comes with a load of other cool featuers. I personaly have never used it, because I find the built-in framework quite sufficient for my needs, but if you look for something more robust, nose is definitely worth trying.