firestore.md

Tutorial: Firebase Mock Basics

When writing unit tests with Firebase Mock, you'll typically want to focus on covering one of two scenarios:

1. Your client receives data from Firestore using a method like get
2. Your client writes data to Firestore using a method like set or update

While your application almost certainly does both reading and writing to Firestore, each test should try to cover as small a unit of functionality as possible.

Testing Reads

In this example, our source code will listen for new people on a reference we provide and call a function each time a new one is added.

Source

var collection;
var people = {
  collection: function () {
    if (!collection) collection = firebase.firestore().collection('people');
    return collection;
  },
  greet: function (person) {
    console.log('hi ' + person.first);
  },
  process: function () {
    people.collection().get().then(function(snaps) {
      snaps.forEach(function(doc) {
        people.greet(doc.data());
      });
    });
  }
};

In our tests, we'll override the greet method to verify that it's being called properly.

Test

MockFirebase.override();
var greeted = [];
people.greet = function (person) {
  greeted.push(person);
};
people.collection().add({
  first: 'Michael'
});
people.collection().add({
  first: 'Ben'
});
people.process();
people.collection().flush();
console.assert(greeted.length === 2, '2 people greeted');
console.assert(greeted[0].first === 'Michael', 'Michael greeted');
console.assert(greeted[1].first === 'Ben', 'Ben greeted');

We're calling MockFirebase.override to replace the real Firebase instance with Firebase Mock. If you're loading Firebase using Node or Browserify, you need to use proxyquire instead.

Notice that we queued up multiple changes before actually calling ref.flush. Firebase Mock stores these changes in the order they were created and then performs local updates accordingly. You'll only need to flush your changes when you need listeners, callbacks, and other asynchronous responses to be triggered.

Testing Writes

Testing writes is especially easy with Firebase Mock because it allows you to inspect the state of your data at any time. In this example, we'll add a new method to people that creates a new person with the given name:

Source

people.create = function (first) {
  return people.collection().add({
    first: first
  });
};

Test

var newPersonRef = people.create('James');
newPersonRef.then(function(doc) {
  console.assert(doc.get('first') === 'James', 'James was created');
});
people.collection().flush();