Fixes model selection response in LLMOpenAIModelOnboardingStep picker

[vishnuravi]

Fixes model selection response in LLMOpenAIModelOnboardingStep picker

♻️ Current situation & Problem

The LLMOpenAIModelOnboardingStep contains a picker that currently tags model options with the formatted description of the model (e.g. "GPT 4 Turbo") not the raw value of the model itself, (e.g. "gpt-4-turbo"). This results in the formatted description being passed to the action, and subsequent failure to create a functioning LLMOpenAISchema from this value because it expects the raw value instead.

⚙️ Release Notes

Fixes the picker to pass the raw value of the model to the action so that it can be used to create an LLMOpenAISchema.

📚 Documentation

Documentation is up to date.

✅ Testing

• Adds a UI test to ensure that the LLMOpenAIModelOnboardingStep is passing to the action the correct raw value for the user's choice.
• Adds a --resetSecureStorage flag to fix an issue with the existing UI tests failing due to old values for the OpenAI API key remaining in secure storage between tests.

📝 Code of Conduct & Contributing Guidelines

By submitting creating this pull request, you agree to follow our Code of Conduct and Contributing Guidelines:

• I agree to follow the Code of Conduct and Contributing Guidelines.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 30.17%. Comparing base (ca37910) to head (656eaed).

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #50   +/-   ##
=======================================
  Coverage   30.17%   30.17%           
=======================================
  Files          66       66           
  Lines        2891     2891           
=======================================
  Hits          872      872           
  Misses       2019     2019           

Files	Coverage Δ
...enAI/Onboarding/LLMOpenAIModelOnboardingStep.swift	100.00% <100.00%> (ø)

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ca37910...656eaed. Read the comment docs.

[vishnuravi]

The UI test failure is unrelated to the current PR, it appears to be due to the token being cached in the keychain from previous tests and not being reset at the beginning of each test.

[philippzagar]

Thanks @vishnuravi for the fix, greatly appreciated! Only had some minor comments, feel free to address them and then merge the PR, no additional review needed 👍

[philippzagar]

The UI test failure is unrelated to the current PR, it appears to be due to the token being cached in the keychain from previous tests and not being reset at the beginning of each test.

Thanks for raising the issue!
As of now, this isn't a big problem as the runners reset the simulator instance for every UI test, therefore only affects local UI tests. Still, very nice to see that you included an easy workaround by resetting the keychain!

[vishnuravi]

The UI test failure is unrelated to the current PR, it appears to be due to the token being cached in the keychain from previous tests and not being reset at the beginning of each test.

Thanks for raising the issue! As of now, this isn't a big problem as the runners reset the simulator instance for every UI test, therefore only affects local UI tests. Still, very nice to see that you included an easy workaround by resetting the keychain!

The screenshot is from the artifact from the runner, so it looks like the runner didn't reset the simulator in that instance.

[philippzagar]

The UI test failure is unrelated to the current PR, it appears to be due to the token being cached in the keychain from previous tests and not being reset at the beginning of each test.

Thanks for raising the issue! As of now, this isn't a big problem as the runners reset the simulator instance for every UI test, therefore only affects local UI tests. Still, very nice to see that you included an easy workaround by resetting the keychain!

The screenshot is from the artifact from the runner, so it looks like the runner didn't reset the simulator in that instance.

That's interesting, never experienced that before on the runners.. So, the keychain state is kept between different UI tests?
Maybe related to (but should only affect macOS keychain): StanfordSpezi/SpeziStorage#19

[PSchmiedmayer]

Thank you for the fix and additions @vishnuravi 🚀
Thank you for the in-depth review @philippzagar 💪