Writing and Running Tests for the In-house Speech Service

This guide outlines the process for writing, running, and debugging tests for the in-house Speech Service.

Adding new tests

When developing new features for Speech Service, it's crucial to add corresponding unit tests. These tests should be placed in the following directory: optional_components/speech_api/tests.

Executing unit Tests

You need to have installed Docker

To run the unit tests, simply use the following command:

make tests

This command handles all necessary steps, including:

1. Building the Docker container for testing
2. Running the tests inside the container
3. Cleaning up after the tests are complete

The entire process is automated, so you don't need to worry about setting up the environment or cleaning up afterwards.

Debugging unit tests

For debugging, you can modify the process to run an interactive container:

1. First, build the Docker image:

make setup-tests

2. Then, run an interactive container:

docker run -it --rm --name speech_test_container speech_test /bin/bash

3. Once inside the container, you can run tests manually:

pytest --color=yes -rPQ tests/*

Or run specific tests as needed.

1. Exit the container when done (the container will be automatically removed due to the --rm flag).

Configs for Visual Studio Code

Add this to your .vscode/launch.json file:

Add the following configuration to your .vscode/launch.json file to set environment variables for debugging:

{
    "version": "0.2.0",
    "configurations": [
        {  // configuration for debugging
            "name": "Python: Tests in current file",
            "purpose": ["debug-test"],
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "args": ["--color=yes"],
            "envFile": "${workspaceFolder}/optional_components/tests/test.env",
            "console": "integratedTerminal",
            "justMyCode": false
        }
    ]
}

Add this to your .vscode/settings.json file:

Add the following configuration to .vscode/settings.json to set the correct pytest working directory and environment variables:

{
    "python.testing.cwd": "${workspaceFolder}/optional_components/speech_api",
    "python.testing.pytestArgs": [
        "tests",
        "--rootdir=${workspaceFolder}/optional_components/speech_api"
    ],
    "python.testing.unittestEnabled": false,
    "python.testing.pytestEnabled": true,
    "python.envFile": "${workspaceFolder}/optional_components/speech_api/tests/test.env"
}

With these configurations in place, you'll be able to efficiently run and debug your Speech Service tests within Visual Studio Code.