Boring Sheep

The mind machine sheep.

(TODO: Copy Sheep materials over, and copy Shepherd docs)

(TODO: What are the most common ways to extend Sheep? essential methods/call traces?)

The Sheep is the class that represents a single user account on the third party service the mind machine is defining. In general, the pattern that mind machine bot flocks follow is: 1 Sheep = 1 bot account = 1 action.

The BoringSheep class is really boring, because we don't want to impose restrictions or make assumptions by implementing a bunch of functionality.

How to use the Boring Sheep class? (...don't)

The constructor is where you create the Sheep's API instance, and should be defined at the package level.

The Boring Sheep class has a virtual constructor, so it cannot be created directly. You should instead define a new Sheep class that inherits from Boring Sheep and defines a constructor, which should initialize an API instance.

How to extend the Boring Sheep class?

See the examples/ directory.

What does the Boring Sheep class define?

In addition to mentioning what you have to define, let's mention what you don't have to define.

The Boring Sheep class implements a dispatcher pattern, which is a way of passing the name of an action as a string, and turning that into a function call.

For example, if the user asks for the 'dummy' action via sheep.perform_action('dummy',**kwargs), this will call sheep.dummy(**kwargs).

See command pattern (wikipedia) for a detailed description of the dispatcher pattern.

    def perform_action(self,action,**kwargs):
        # Dispatcher pattern
        if hasattr(self, action):
            method = getattr(self, action)
            method(**kwargs)