Clone the repo.

git clone https://github.com/tylertreat/Kanna.git

Ensure that pip is installed.

• http://www.pip-installer.org/en/latest/

We highly recommend using a virtual environment (virtualenv) to ease development.

• http://www.virtualenv.org/en/latest/

To help make working with virtualenv even easier we recommend checking out virtualenvwrapper.

• http://virtualenvwrapper.readthedocs.org/en/latest/
• Then you can cd to your directory.

  • Create your virtualenv and point it to your current directory

    mkvirtualenv kanna -a $PWD
    

  • From then on when you activate your virtualenv it will take you to your directory.

    workon kanna
    

To install all dependencies and correctly wire up lib for GAE deployment:

sudo make deps

• This will pip install all the libraries referenced in requirements.txt and requirements_dev.txt. All of the libraries within requirements.txt will also get symlinked into a lib directory. This will allow those libraries to be deployed to GAE.

Install a path configuration file in the virtualenv so you can use the appengine's runtime environment.

• Create a gae.pth file in your virtual env site-packages directory. /path/to/google_appengine should be the install location of the Google AppEngine SDK. For example, /usr/local/google_appengine.

  If you have created the symlinks for the App Engine SDK:

workon kanna
dirname $(readlink `which dev_appserver.py`) >> $VIRTUAL_ENV/lib/python2.7/site-packages/gae.pth
echo 'import dev_appserver; dev_appserver.fix_sys_path()' >> $VIRTUAL_ENV/lib/python2.7/site-packages/gae.pth

You should now be able to launch a Python interactive shell (with your virtualenv activated) and import google.appengine.

$ python
Python 2.7.3 (v2.7.3:70274d53c1dd, Apr  9 2012, 20:52:43)
[GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import google.appengine
>>>

Unit tests

make unit

• The unit tests just run the basic nosetests using all the configuration from the setup.cfg. By default we skip the tests marked with slow. To run the slow tests as well you need to run the integration tests shown below.

You could also just as easily run:

nosetests

Integration tests

make test

• This will run all the tests including the slow ones. By default we are ignoring all logs below ERROR level. You can switch this to whatever you prefer when needed. NDB outputs a ton of debugging level logs which makes it quite difficult to debug when you do hit errors in your tests. This will also clean out all your pyc files before running the tests to ensure running from a good state.

or

make integrations

• This will run the same as above just without cleaning the pyc files.

or

nosetests --logging-level=ERROR -a slow

• This is the command that integrations is wrapping.

To run the development environment you can run:

make run

• This runs the dev app server on the default port. This command also runs all dependency checks and installs any missing libraries.

You can also just run:

dev_appserver.py .

• This will no do any of the dependency checks.

To install all of the projects dependencies run:

make deps

To clean the dist logs:

make distclean

To update just the deployment dependencies:

make py_deploy_deps

To update just the development dependencies:

make py_dev_deps

Requirements to commit here:

• Branch off master, PR back to master.
• Your code should pass Flake8.
• Unit test coverage is required.
• Good docstrs are required.
• Good commit messages are required.