Stabilize tests

[carterworks]

Description

During the creation of the create-browser-package branch, I discovered a couple of patterns that were causing a number of instability in our integration and unit tests.

This includes

• Mock's on important globals like fetch or querySelector That weren't getting cleaned up.
• Cookies that were being written and not deleted.
• URL parameters and URL changes that were happening and not getting returned to the original URL.
• Dom parameters and nonces that were getting set and not cleaned up.

Because we run our tests in a non isolated and nonparallel manner on our continuous integration actions runners, this was causing the effects of one test to leak into others, and they would fail sometimes because the order was non deterministic.

Related Issue

Motivation and Context

Screenshots (if appropriate):

Checklist:

• I have signed the Adobe Open Source CLA or I'm an Adobe employee.
• I have made any necessary test changes and all tests pass.
• I have run the Sandbox successfully.
• I have added a Changeset (pnpm changeset) or it is not necessary because this PR is not consumer-facing.

[changeset-bot]

🦋 Changeset detected

Latest commit: d831b02

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 0 packages

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[dompuiu]

In serviceWorker.js you already have a makeSendServiceWorkerTrackingData. Can you pass that to this module? When you need to create a new instance?

[dompuiu]

The tests are still failing.

[dompuiu]

I was thinking at this, and the refactoring seems to circumvent the errors but we still not know why is causing the tests to fail. We should probably still try to find out why the tests are flaky.

[carterworks]

I was thinking at this, and the refactoring seems to circumvent the errors but we still not know why is causing the tests to fail. We should probably still try to find out why the tests are flaky.

I think they are failing because two or more tests import the same file, but one test mocks the file and the other does not. Our tests are run all in the same environment to use fewer resources on the CI environment. So the tests are order-dependent, I think. This PR removes the vi.mock(modulePath) usage and the tests pass on CI now.

[jonsnyder]

Should we get the Ad-Tech contributors to look at your changes to the advertising component?

[carterworks]

Should we get the Ad-Tech contributors to look at your changes to the advertising component?

I don't think that's necessary, since the changes are organizational, not functional

[Spencer-Smith]

What's up with the "fixme" notes? Is that things that were fixed or need to be fixed?

[carterworks]

What's up with the "fixme" notes? Is that things that were fixed or need to be fixed?

Forgot about those. Fixed