Python Koans is a port of Edgecase's "Ruby Koans".
Python Koans is an interactive tutorial for learning Python by making tests pass.
Most tests are 'fixed' by filling the missing parts of assert functions. Eg:
which can be fixed by replacing the __ part with the appropriate code:
Occasionally you will encounter some failing tests that are already filled out. In these cases you will need to finish implementing some code to progress. For example, there is an exercise for writing some code that will tell you if a triangle is equilateral, isosceles or scalene.
As well as being a great way to learn some Python, it is also a good way to get a taste of Test Driven Development (TDD).
Downloading Python Koans
Python Koans is available through git on Github:
It is also mirrored on bitbucet for Mercurial users :
Either site will allow you to download the source as a zip/gz/bz2.
Installing Python Koans
Aside from downloading or checking out the latest version of Python Koans, all you need to install is Python.
At this time of writing there are two versions of the koans, one for use with Python 2.6 and one for Python 3.1. You should be able to work with newer Python versions, but older ones will likely give you problems.
You can download Python from here:
On installing Python make sure the folder containing the python executable is in the system path. In other words, you need to be able to be able to run Python from a command console. With Python 2 it will be called 'python' or 'python.exe' depending on the operating system. For Python 3 it will either be 'python3' or for windows it will be 'python.exe'.
If you have problems, this may help:
Jake Hebbert has created a couple of screencasts available here:
Or if you prefer to read:
From a *nix terminal or windows command prompt go to the python koanspython_VERSION folder and run:
In my case I'm using Python 3 with windows, so I fire up my command shell (cmd.exe) and run this:
Apparently a test failed:
AssertionError: False is not True
It also tells me exactly where the problem in, its an assert on line 12 of .koansabout_asserts.py. This one is easy, just change False to True to make the test pass.
Sooner or later you will likely encounter tests where you are not sure what the expected value should be. For example:
class Dog: pass def test_objects_are_objects(self): fido = self.Dog() self.assertEqual(__, isinstance(fido, object))
This is where the Python Command Line can come in handy. in this case I can fire up the command line, recreate the scenario and run queries:
Getting the Most From the Koans
Quoting the Ruby Koans instructions:
"In test-driven development the mantra has always been, red, green, refactor. Write a failing test and run it (red), make the test pass (green), then refactor it (that is look at the code and see if you can make it any better. In this case you will need to run the koan and see it fail (red), make the test pass (green), then take a moment and reflect upon the test to see what it is teaching you and improve the code to better communicate its intent (refactor)."
Python is a made up of about 2/3 Ruby Koans ported material and 1/3 Python specific tests. The content ported from Ruby Koans includes all the assignment projects.
Content for Python 3 is a little different to the Python 2 flavor due to big changes between the 2 different languages. For example in the Python 2 variant the differences between old and new style classes are covered. This loses relevance in in the Python 3 version, but there are some extra tests covering new functionality.
Finding More Koan Projects
Right now there are a lot of spinoff Koan projects out there for a great number of languages and frameworks. Many of them do not have that much content, but contributing to them is a great way to learn. At the moment most of them can be found by searching for 'koans' on github.
A couple of promising projects include DotNetKoans and TestMongoKoans.
Thanks go to Jim Weirich and Joe O'Brien for the original Ruby Koans that Python Koans is based on! Also the Ruby Koans in turn borrows from Metakoans so thanks also go to Ara Howard for that!
Also thanks to everyone who helped with the Python Koans conversion! In particular I got a great headstart on the project by forking from this Python Koans startup project: