Skip to content

Latest commit

 

History

History
164 lines (134 loc) · 7.06 KB

README.md

File metadata and controls

164 lines (134 loc) · 7.06 KB
Raw

Exercise 2: Testing the Translation Workflow

During this exercise, you will

  • Run a unit test provided for the translate_term Activity
  • Develop and run your own unit test for the translate_term Activity
  • Write assertions for a Workflow test
  • Uncover, diagnose, and fix a bug in the Workflow Definition
  • Observe the time-skipping feature in the Workflow test environment

Make your changes to the code in the practice subdirectory (look for TODO comments that will guide you to where you should make changes to the code). If you need a hint or want to verify your changes, look at the complete version in the solution subdirectory.

Prerequisite: Ensure the Microservice is running and your Virtual Environment is activated

If you haven't already started the Translation Microservice used by this exercise, do so in a separate terminal.

Note: If you're using the Gitpod environment to run this exercise you can skip this step. An instance of the Microservice is already running in your environment

Activate the Virtual Environment

  1. Ensure that the virtual environment you setup at the beginning of the course is activated as detailed in the course README

Start the Microservice

  1. Navigate to the utilities directory at the root level of the course
  2. Change directories into the microservice directory
    1. cd utilities/microservice
  3. Run the microservice
    1. python microservice.py

Part A: Running a Test

We have provided a unit test for the translate_term Activity to get you started. This test verifies that the Activity correctly translates the term "Hello" to German. Take a moment to study the test, which you'll find in the test_activities.py file in the tests directory. Since the test runs the Activity, which in turn calls the microservice to do the translation, you'll begin by starting that.

  1. Run the python -m pytest command to execute the provided test

Part B: Add another test case for the Activity

In the previous step you verified that running translate_term for Hello was invoking properly. Now it is time to test if Goodbye is also working.

  1. Edit the test/test_activities.py file
  2. Copy the tuple in the pytest.mark.parametrize list and paste it into the list, effectively creating a second entry to test the method on.
  3. Change the term for the TranslationActivityInput object in this second tuple from Hello to Goodbye
  4. Change the language code for the TranslationActivityInput object in this second tuple from de (German) to lv (Latvian)
  5. Change the translation for the TranslationActivityOutput object in this second tuple from Hallo to Ardievu
  6. Run the python -m pytest command to verify that the method was tested twice and that the test completed successfully.

Part C: Test the Activity with Invalid Input

In addition to verifying that your code behaves correctly when used as you intended, it is sometimes also helpful to verify its behavior with unexpected input. The example below does this, testing that the Activity returns the appropriate error when called with an invalid language code.

@pytest.mark.asyncio
async def test_failed_translate_activity_bad_language_code():
    with pytest.raises(Exception) as e:
        input = TranslationActivityInput("goodbye", "xq")
        async with aiohttp.ClientSession() as session:
            activity_environment = ActivityEnvironment()
            activities = TranslationActivities(session)
            await activity_environment.run(activities.translate_term, input)
    assert "Invalid language code" in str(e)

Take a moment to study this code, and then continue with the following steps:

  1. Edit the test_activities.py file
  2. Copy the entire test_failed_translate_activity_bad_language_code function provided above and paste it at the bottom of the test_activities.py file
  3. Save the changes
  4. Run python -m pytest again to run this new test, in addition to the others

Part D: Test a Workflow Definition

  1. Edit the tests/test_workflow.py file
  2. Add assertions for the following conditions
    • The hello_message field in the result is Bonjour, Pierre
    • The goodbye_message field in the result is Au revoir, Pierre
  3. Save your changes
  4. Run python -m pytest. This will fail, due to a bug in the Workflow Definition.
  5. Find and fix the bug in the Workflow Definition
  6. Run the python -m pytest command again to verify that you fixed the bug

There are two things to note about this test.

First, the test completes in under a second, even though the Workflow Definition contains a await asyncio.sleep(15) call that adds a 15-second delay to the Workflow Execution. This is because of the time-skipping feature provided by the test environment.

Second, calls to RegisterActivity near the top of the test indicate that the Activity Definitions are executed as part of this Workflow test. As you learned, you can test your Workflow Definition in isolation from the Activity implementations by using mocks. The optional exercise that follows provides an opportunity to try this for yourself.

This is the end of the exercise.

(Optional) Using Mock Activities in a Workflow Test

If you have time and would like an additional challenge, continue with the following steps.

  1. Make a copy of the existing Workflow Test by running cp tests/test_workflow.py tests/test_workflow_with_mocks.py
  2. Edit the test_workflow_with_mocks.py file
  3. Add the following imports 
    from temporalio import activity
from shared import (
    TranslationWorkflowInput,
    TranslationActivityInput,
    TranslationActivityOutput,
)
  4. Create a new async function to mock your Activity.
    1. Name it translate_term_mocked_french and decorate it with @activity.defn(name="translate_term"). The function should take TranslationActivityInput
    2. In the body of your new mocked Activity, write an if statement that returns a new TranslationActivityOutput object containing Bonjour if the term that was passed in via the TranslationActivityInput was hello. Otherwise return a new TranslationActivityObject containing Au revoir.
    3. The full mocked Activity is shown below:
    @activity.defn(name="translate_term")
async def translate_term_mocked_french(input: TranslationActivityInput):
  if input.term == "hello":
      return TranslationActivityOutput("Bonjour")
  else:
      return TranslationActivityOutput("Au revoir")
  5. Delete the context manager and the creation of a TranslationActivites object, these are now unnecessary due to the use of a mocked Activity. 
    async with aiohttp.ClientSession() as session:
    activities = TranslationActivities(session)
    Once deleted realign your Python code properly now that a scope has been removed.
  6. In the Worker creation, replace the statement activities=[activities.translate_term], with activities=[translate_term_mocked_french],
  7. Save your changes
  8. Run python -m pytest to run the tests