[Lens] test suite - replace React specific props events with html events

[mbondyra]

Summary

This is a small improvement to our test suite - most of our tests checking the correctness of the invoked events were directly invoking passed props, eg. element.props().onClick(). This can be improved by instead simulating user events, eg. element.simulate('click'). This has couple of advantages:

• the test will work regardless of internal implementation. We check a bit more of a code (not just how it integrates with the middle component, but the path from the component we test to html. I know we don't have to check eui internals, but this way...
• ...we test the way user uses our app
• we don't have to assert types of the events, eg:

 wrapper.find('[data-test-subj="lnsIndexPatternFieldSearch"]').prop('onChange')!({
          target: { value: 'me' },
        } as ChangeEvent<HTMLInputElement>); 

I also simplified testing the Portal created by popover. I'll leave the comment close to the code.

[elasticmachine]

Pinging @elastic/kibana-vis-editors (Team:VisEditors)

[mbondyra]

Thanks to mocking the portal: ReactDOM.createPortal = jest.fn((element) => element) we can just assume that the content of the popover will be placed directly inside the popover. And thanks to that we can avoid the ugly lines above.

[kibanamachine]

💚 Build Succeeded

• continuous-integration/kibana-ci/pull-request
• Commit: 6f127b1
• Storybooks Preview

Metrics [docs]

✅ unchanged

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

[mbondyra]

I am a bit on the edge of replacing this with waitFor - I think it's more readable, but it comes from react/testing-library and we use in 99% enzyme. There are one more place where we use react-testing-library in lens (testing debouncing hook) and some other teams use it extensively in Kibana, so I am not sure. Thoughts?

[dej611]

The behaviour is different between the 2 things.
I think waitFor is more reliable than a manual arbitrary waiting. So 👍

[dej611]

Code review 👍

[dej611]

The behaviour is different between the 2 things.
I think waitFor is more reliable than a manual arbitrary waiting. So 👍

[kibanamachine]

💚 Backport successful

Status	Branch	Result
✅	7.x

This backport PR will be merged automatically after passing CI.