Mimic is a mock library for python that is based on Google’s Pymox, a fanstastic testing library, which is in turn based on EasyMock – a Java mock object framework.
Mimic allows you to write true unit tests even in situations in which your code is dependent on external systems, in situations in which dependency injection won’t work, or would otherwise be too complicated.
Mimic is a bit more complex than many other mocking libraries. This is a strength and a weakness. The way Mimic tests are meant to be run is in the following order:
So the first part of your test ends up being about setting up the scenario for mimic, and then the second part – after you enter replay mode – is about calling the code you hope to test from you test function.
This two-step process is a little extra work from the onset, but it’s a hidden strength when you realize that Mimic holds you to the expectations you set: if you don’t call a method you mock out, you get an error; if you call a method you weren’t expecting, you get an error. It has a kind of symmetry that many developers find easy to trust because of its explicitness.
There are a couple of features that have been needed for a while, including:
Move the codebase over to github (and thereby git) in order to allow for more community participation
PEP8 compliant method names
Experimental Python 3 support
Complete, comprehensive documentation
Continuous Integration
Most importantly, though, a library this good needs active maintenance. It’s been a few years now since the latest release. While this is a relatively mature code-base, there are a number of outstanding issues, which don’t seem to be getting any traction.