Consuming AMQP/RabbitMQ messages in Python is simple if you use Sparkplug. I read “Building RabbitMQ apps using Python“, and there’s just too much code in the monitor method and the separation of concerns aren’t very clear.

With Sparkplug, you create a configuration file, then define Consumer class to handle each message. As messages are pushed into the queue, Sparkplug will call your class and you decide how you want handle the message in the __call__ method. The great thing about this is that all the connection logic is removed from your code letting you concentrate on how you want to handle each message.

Now you can run your app using sparkplug --daemon consumer.ini. Configuration is boring, so I’ve left the consumer.ini out of the blog post and put it up on github instead.

Jesse has written a great blog post on using test builders to improve the expressiveness of unit tests. He has convinced me that builders are useful and that they provide an easy way to construct objects that are in a valid state. Testing can be hard; I find defining a method for each of the object’s properties to be tedious and repetitive. This is especially problematic if you’re working with an object with dozens of properties.

To balance expressive with conciseness, I’ve refactored the builder to use Python dictionaries

I find this code to be just as expressive, easier to use and most importantly not repetitive.