模拟、存根与测试数据
学习隔离代码依赖、使用模拟和存根，以及高效管理测试数据的技术。
模拟、存根与测试数据 是 CoddyKit 上的免费 Elixir & Phoenix: Scalable Backend Development 课时。 这是第 3 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 Elixir & Phoenix: Scalable Backend Development 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Elixir & Phoenix: Scalable Backend Development 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
Why Isolate Dependencies?
In real-world applications, your code often relies on other components or external services. These are called dependencies.
Imagine your application needs to:
- Save data to a database.
- Send emails via an email service.
- Fetch data from a third-party API.
Testing these directly can be slow, unreliable (what if the API is down?), or costly. We need a way to test our code in isolation.
Understanding Test Doubles
When testing code that interacts with external services or complex components, we often use test doubles. These are generic objects that stand in for real dependencies during a test.
Common types include:
- Stubs: Provide predefined answers to method calls.
- Mocks: Verify that certain actions (method calls) occurred.
- Fakes: Lightweight working implementations (e.g., an in-memory database).
In this lesson, we'll focus on Stubs and Mocks, primarily using the Mox library.
What's a Stub?
A stub is a stand-in for a real dependency that provides pre-programmed responses to method calls. It's like a script for a supporting actor: when your code 'asks' the stub something, the stub simply gives a canned answer.
Stubs help you control the environment your code is tested in, ensuring consistent results without external side effects.
Try this example to see the concept:
defmodule Notifier do
@callback send_notification(message :: String.t) :: :ok
end
defmodule RealNotifier do
@behaviour Notifier
def send_notification(message) do
IO.puts("Real Notifier sent: #{message}")
:ok
end
end
defmodule StubNotifier do
@behaviour Notifier
def send_notification(_message) do
IO.puts("Stub Notifier received message (but didn't send for real).")
:ok
end
end
defmodule Client do
def trigger_notification(message, notifier_module) do
notifier_module.send_notification(message)
end
end
IO.puts("--- Using Real Notifier ---")
Client.trigger_notification("Hello World!", RealNotifier)
IO.puts("\n--- Using Stub Notifier in a test-like scenario ---")
Client.trigger_notification("Test Message", StubNotifier)Introducing Mox for Stubbing
In Elixir, the Mox library is a popular choice for creating mocks and stubs. It integrates well with ExUnit and helps ensure your test doubles conform to defined behaviours.
First, you define a mock module using Mox.defmock/2, specifying the behaviour it should implement. This acts as a proxy for your real module.
# In test/support/mocks.ex
defmodule MyApp.NotifierMock do
use Mox
# This mock will implement the MyApp.Notifier behaviour
# (Assuming MyApp.Notifier defines @callback notify/1)
@behaviour MyApp.Notifier
end
# In test_helper.exs, you'd typically define:
# Mox.defmock(MyApp.NotifierMock, for: MyApp.Notifier)Stubbing Responses with Mox
Once your mock module is defined, you can use Mox.stub/3 or Mox.expect/3 (which also stubs) within your tests to define what functions should return. This allows you to simulate success or failure scenarios from your dependencies.
Mox.stub/3 takes the mock module, the function name, and a function that defines the return value based on arguments.
# Assume MyApp.NotifierMock is defined and MyApp.Client uses it.
# In an ExUnit test case:
# use Mox
# setup :verify_on_exit!
# Stub the :notify function to always return :ok
Mox.stub(MyApp.NotifierMock, :notify, fn _message ->
IO.puts("Mox stub received message, returning :ok.")
:ok
end)
# If MyApp.Client.send_message("Hello", MyApp.NotifierMock) is called,
# it will receive :ok as the result, without calling the real notifier.What's a Mock?
A mock is a test double that not only provides predefined responses (like a stub) but also verifies that specific interactions occurred. It answers the question: 'Was this function called, with these arguments, and how many times?'
Mocks are useful when you want to ensure your code correctly interacts with its dependencies, for example, making sure an email was indeed sent or a specific log entry was created.
Verifying Interactions with Mox
Mox.expect/3 is used for both stubbing responses and setting expectations for function calls. If an expected call isn't made, or is made with different arguments, the test will fail.
This helps you assert the 'side effects' of your code, ensuring it communicates correctly with its dependencies.
# Assume MyApp.NotifierMock is defined and MyApp.Client uses it.
# In an ExUnit test case:
# use Mox
# setup :verify_on_exit!
# Expect :notify to be called exactly once with "Urgent Message"
Mox.expect(MyApp.NotifierMock, :notify, fn "Urgent Message" ->
IO.puts("Mox mock received expected message: Urgent Message")
:ok
end)
# If MyApp.Client.send_message("Urgent Message", MyApp.NotifierMock) is called,
# the expectation is met. If not, the test fails at the end.Managing Test Data
Beyond isolating dependencies, managing test data is crucial. Hardcoding data directly in tests can lead to:
- Brittleness: Changing a schema or constraint breaks many tests.
- Repetition: Copy-pasting complex data structures.
- Inconsistency: Tests relying on different, potentially invalid, data.
We need a better way to generate realistic and valid data for our tests.
Test Data Factories
Test data factories are functions or modules designed to generate valid, customizable data for your tests. Instead of manually creating maps or structs, you call a factory function.
This approach makes your tests:
- Robust: Factories handle defaults and ensure validity.
- Flexible: You can override specific attributes as needed.
- Concise: Less boilerplate for data creation in tests.
Here's a simple factory example:
defmodule UserFactory do
def build_user(attrs \\ %{}) do
defaults = %{
id: "user-#{:rand.uniform(100000)}",
name: "User Name #{:rand.uniform(100)}",
email: "user#{:rand.uniform(100)}@example.com",
age: :rand.uniform(50) + 18
}
Map.merge(defaults, attrs)
end
end
user1 = UserFactory.build_user()
IO.puts("Generated User 1:")
IO.inspect(user1)
user2 = UserFactory.build_user(%{name: "Alice", email: "alice@example.com"})
IO.puts("\nGenerated User 2 (customized):")
IO.inspect(user2)Quick Check
You've learned about stubs, mocks, and how to manage test data. Let's test your understanding of the core concepts.
Recap & Next Steps
In this lesson, we explored crucial techniques for writing more effective and isolated tests:
- We learned about test doubles, specifically stubs (for predefined responses) and mocks (for verifying interactions).
- We saw how the
Moxlibrary helps implement these concepts in Elixir, especially when working with behaviours. - Finally, we discussed strategies for managing test data using factories to create robust and flexible test environments.
These tools are essential for building reliable Elixir and Phoenix applications. Keep practicing to master them!
常见问题解答
「模拟、存根与测试数据」课时是免费的吗？
是的 — 「模拟、存根与测试数据」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Elixir & Phoenix: Scalable Backend Development 课程的其余内容，请升级到 CoddyKit PRO。 Elixir & Phoenix: Scalable Backend Development 课程共包含 4 节课。
「模拟、存根与测试数据」这节课中我会学到什么？
学习隔离代码依赖、使用模拟和存根，以及高效管理测试数据的技术。 你通过在浏览器中直接运行的动手代码来练习 Elixir & Phoenix: Scalable Backend Development，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 Elixir & Phoenix: Scalable Backend Development 需要有经验吗？
无需任何先前经验。CoddyKit 上的 Elixir & Phoenix: Scalable Backend Development 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 3 节课，共 4 节。
「模拟、存根与测试数据」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 Elixir & Phoenix: Scalable Backend Development 课中编写并运行代码吗？
能。每节 Elixir & Phoenix: Scalable Backend Development 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。