omergertel/mongomock

Small library for mocking pymongo collection objects for testing purposes

What is this?

Mongomock is a small library to help testing Python code that interacts with MongoDB via Pymongo.

To understand what it's useful for, we can take the following code:

def increase_votes(collection):
    for document in collection.find():
        collection.update(document, {'$set' : {'votes' : document['votes'] + 1}})

The above code can be tested in several ways:

1. It can be tested against a real mongodb instance with pymongo.
2. It can receive a record-replay style mock as an argument. In this manner we record the expected calls (find, and then a series of updates), and replay them later.
3. It can receive a carefully hand-crafted mock responding to find() and update() appropriately.

Option number 1 is obviously the best approach here, since we are testing against a real mongodb instance. However, a mongodb instance needs to be set up for this, and cleaned before/after the test. You might want to run your tests in continuous integration servers, on your laptop, or other bizarre platforms - which makes the mongodb requirement a liability.

We are left with #2 and #3. Unfortunately they are very high maintenance in real scenarios, since they replicate the series of calls made in the code, violating the DRY rule. Let's see #2 in action - we might right our test like so:

def test_increase_votes():
    objects = [dict(...), dict(...), ...]
    collection_mock = my_favorite_mock_library.create_mock(Collection)
    record()
    collection_mock.find().AndReturn(objects)
    for obj in objects:
        collection_mock.update(document, {'$set' : {'votes' : document['votes']}})
    replay()
    increase_votes(collection_mock)
    verify()

Let's assume the code changes one day, because the author just learned about the '$inc' instruction:

def increase_votes(collection):
    collection.update({}, {'$inc' : {'votes' : 1}})

This breaks the test, although the end result being tested is just the same. The test also repeats large portions of the code we already wrote.

We are left, therefore, with option #3 -- you want something to behave like a mongodb database collection, without being one. This is exactly what this library aims to provide. With mongomock, the test simply becomes:

def test_increase_votes():
    collection = mongomock.Connection().db.collection
    objects = [dict(votes=1), dict(votes=2), ...]
    for obj in objects:
        obj['_id'] = collection.insert(obj)
    increase_votes(collection)
    for obj in objects:
        stored_obj = collection.find_one({'_id' : obj['_id']})
        stored_obj['votes'] -= 1
        assert stored_obj == obj # by comparing all fields we make sure only votes changed

This code checks increase_votes with respect to its functionality, not syntax or algorithm, and therefore is much more robust as a test.

To download, setup and perfom tests, run the following commands on Mac / Linux:

git clone <repo>
cd <reponame>
virtualenv venv --distribute
source venv/bin/activate
pip install nose
pip install pymongo
pip install pyexecjs
python setup.py install
nosetests

Important Note About Project Status & Development

MongoDB is complex. This library aims at a reasonably complete mock of MongoDB for testing purposes, not a perfect replica. This means some features are not likely to make it in any time soon.

Also, since many corner cases are encountered along the way, our goal is to try and TDD our way into completeness. This means that every time we encounter a missing or broken (incompatible) feature, we write a test for it and fix it. There are probably lots of such issues hiding around lurking, so feel free to open issues and/or pull requests and help the project out!

NOTE: We don't include pymongo functionality as "stubs" or "placeholders". Since this library is used to validate production code, it is unacceptable to behave differently than the real pymongo implementation. In such cases it is better to throw NotImplementedError than implement a modified version of the original behavior.

Acknowledgements

Many thanks go to the following people for helping out:

• Alec Perkins
• Austin W Ellis
• Andrey Ovchinnikov
• Arthur Hirata
• Corey Downing
• Craig Hobbs
• David Fischer
• Edward D'Souza
• Eugene Chernyshov
• Israel Teixeira
• Jacob Perkins
• Jason Sommer
• Jeff McGee
• Joël Franusic
• Mike Ho
• Nigel Choi
• Omer Gertel
• Scott Sexton
• Todd Tomkinson
• catty (ca77y _at live.com)
• emosenkis
• waskew (waskew _at narrativescience.com)
• zcarter