persn/vscode-karma-test-adapter

Implementing a Test Adapter for Visual Studio Code

This repository contains an example for implementing a TestAdapter extension that works with the Test Explorer extension.

Setup

• install the Test Explorer extension
• fork and clone this repository and open it in VS Code
• run npm install
• run npm run watch or start the watch Task in VS Code
• start the debugger

You should now see a second VS Code window, the Extension Development Host. Open a folder in this window and click the "Test" icon in the Activity bar. Now you should see the fake example test suite in the side panel:

Basic implementation

• add any configuration properties that your Test Adapter needs to the contributes.configuration.properties section of package.json
• replace the loadFakeTests() call in src/adapter.ts with your code for loading the test definitions for the real test framework
• replace the runFakeTests() call in src/adapter.ts with your code for running the tests in a child process using the real test framework

Getting ready to publish

• search for all occurrences of the word "example" in this project and replace them with the name of the testing framework that your Test Adapter supports
• update package.json with your preferred values (at a minimum you should change author, publisher, homepage, repository and bugs)
• create an icon for your Test Adapter (there's an SVG version of the Test Explorer icon at img/test-explorer.svg) and reference it in package.json
• replace this README with your documentation

Now you're ready to publish the first version of your Test Adapter.

Completing the implementation

• implement the debug() method
• implement the cancel() method (it should kill the child process that was started by run() or debug())
• watch the configuration for any changes that may affect the loading of test definitions and reload the test definitions if necessary
• watch the workspace for any changes to the test files and reload the test definitions if necessary
• watch the configuration for any changes that may affect the results of running the tests and emit an autorun event if necessary
• watch the workspace for any changes to the source files and emit an autorun event if necessary
• ensure that only one test run is active at a time