DISCOVERY-481: Gather browser-issued errors during UI testing

[mirekdlugosz]

During UI testing, gather errors that browser saw and fail a test if there were any.

Issues gathered:

• failed HTTP requests (time outs, broken pipes, network issues in the middle of reading response and other problems that prevented browser from receiving the whole response)
• HTTP responses with 4xx and 5xx status codes
• JavaScript errors (syntax errors, uncaught exceptions, but NOT console.error() messages)

This allows us to remove one of integration tests from quipucords.

This is implemented by gathering interesting events during test execution and calling pytest.fail() in function-scoped fixture teardown. pytest in command line shows that funny - for each test, it will be reported as pass and then as error (because technically test did pass - fixture teardown is executed after the test). Ibutsu and Jenkins show test as errored instead of failed, but this difference is irrelevant.

See https://url.corp.redhat.com/ibutsu-runs-43032926-7ccf-471a-942a-2e79ee36b4ee for sample test session with some tests failing due to injected errors.

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (a715664) 77.07% compared to head (fd0fdd6) 77.07%. Report is 1 commits behind head on main.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #466 +/- ## ======================================= Coverage 77.07% 77.07% ======================================= Files 4 4 Lines 205 205 ======================================= Hits 158 158 Misses 47 47 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.