analytickit-js v1.29.5

AnalyticKit Browser JS Library

Please see AnalyticKit Docs. Specifically, browser JS library details.

Testing

Unit tests: run yarn test. Cypress: run yarn serve to have a test server running and separately yarn cypress to launch Cypress test engine.

Running TestCafe E2E tests with BrowserStack

Testing on IE11 requires a bit more setup.

1. Run analytickit locally on port 8000 (DEBUG=1 TEST=1 ./bin/start).
2. Run python manage.py setup_dev --no-data on analytickit repo, which sets up a demo account.
3. Optional: rebuild array.js on changes: nodemon -w src/ --exec bash -c "yarn build-rollup".
4. Export browserstack credentials: export BROWSERSTACK_USERNAME=xxx BROWSERSTACK_ACCESS_KEY=xxx.
5. Run tests: npx testcafe "browserstack:ie" testcafe/e2e.spec.js.

Tiers of testing

1. Unit tests - this verifies the behavior of the library in bite-sized chunks. Keep this coverage close to 100%, test corner cases and internal behavior here
2. Cypress tests - integrates with a real chrome browser and is capable of testing timing, browser requests, etc. Useful for testing high-level library behavior, ordering and verifying requests. We shouldn't aim for 100% coverage here as it's impossible to test all possible combinations.
3. TestCafe E2E tests - integrates with a real analytickit instance sends data to it. Hardest to write and maintain - keep these very high level

Developing together with another repo

Using Yarn link

Use yarn link. Run yarn link in analytickit-js, and then yarn link analytickit-js in analytickit. Once you're done, remember to yarn unlink analytickit-js in analytickit, and yarn unlink in analytickit-js.

An alternative is to update dependency in package.json to e.g. "analytickit-js": "link:../analytickit-js", yarn and run yarn build && yarn build-module

Developing with main AnalyticKit repo

The analytickit-js snippet for a website loads static js from the main AnalyticKit/analytickit repo. Which means, when testing the snippet with a website, there's a bit of extra setup required:

1. Run AnalyticKit/analytickit locally
2. Link the analytickit-js dependency to your local version (see above)
3. Run yarn serve in analytickit-js. (This ensures dist/array.js is being generated)
4. In your locally running AnalyticKit/analytickit build, run yarn copy-scripts. (This copies the scripts generated in step 3 to the static assets folder for AnalyticKit/analytickit)

Further, it's a good idea to modify start-http script to add development mode: webpack serve --mode development, which doesn't minify the resulting js (which you can then read in your browser).

Using Yalc (Alternative to yarn link)

Run npm install -g yalc

• In the analytickit-js repo
  • Run yalc publish
• In the analytickit repo
  • Run yalc add analytickit-js
  • Run yarn
  • Run yarn copy-scripts

When making changes

• In the analytickit-js repo
  • Run yalc publish
• In the analytickit repo
  • Run yalc update
  • Run yarn
  • Run yarn copy-scripts

To remove the local package

• In the analytickit repo
  • run yalc remove analytickit-js
  • run yarn install

Releasing a new version

Just bump up version in package.json on the main branch and the new version will be published automatically, with a matching PR in the main AnalyticKit repo created.

It's advised to use bump patch/minor/major label on PRs - that way the above will be done automatically when the PR is merged.

Courtesy of GitHub Actions.

Manual steps

To release a new version, make sure you're logged into npm (npm login).

We tend to follow the following steps:

1. Merge your changes into master.
2. Release changes as a beta version:
   • npm version 1.x.x-beta.0
   • npm publish --tag beta
   • git push --tags
3. Create a PR linking to this version in the main AnalyticKit repo.
4. Once deployed and tested, write up CHANGELOG.md, and commit.
5. Release a new version:
   • npm version 1.x.x
   • npm publish
   • git push --tags
6. Create a PR linking to this version in the main AnalyticKit repo.

Questions?

Join our Slack community.