Create NewTrafficPolicyFromJson deserializer checking for unknown keys and implement it in translating ingresses

[sabrina-ngrok]

#K8SOP-147

What

it looks like we don't handle traffic policy with an inbound phase appropriately. Instead we just discard all of it, leading to no user supplied policy being merged with our generated traffic policy.

This could just be an edge case that you only hit with the inbound key

How

Amp suggested that we create a new JSON deserializer that checks the valid traffic policy keys (on_http_request, on_http_response, and on_tcp_connect) and have a bunch of error handling for edge cases such as this.

Breaking Changes

¯\(ツ)/¯

[codecov]

Codecov Report

❌ Patch coverage is 70.83333% with 7 lines in your changes missing coverage. Please review.
✅ Project coverage is 49.14%. Comparing base (029fd83) to head (e7c146e).

Files with missing lines	Patch %	Lines
pkg/managerdriver/translate_ingresses.go	37.50%	3 Missing and 2 partials ⚠️
internal/trafficpolicy/policy.go	87.50%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #718      +/-   ##
==========================================
+ Coverage   49.13%   49.14%   +0.01%     
==========================================
  Files          96       96              
  Lines       10604    10616      +12     
==========================================
+ Hits         5210     5217       +7     
- Misses       5037     5038       +1     
- Partials      357      361       +4     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[jonstacks]

Overall this looks good. Just a few changes around error messaging.

The place where this could break someone is if they had a traffic policy like:

bad_key: "I'm ignored today"
on_http_request:
   (fully valid policy here)

Today we ignore the bad_key completely. This would now throw an error for the bad key. I think this is expected behavior, but something we'll want to call out in the release notes.

Did the changes requested in the last commit of this PR

[sabrina-ngrok]

I tested the changes in my local, but I was having issues with SSH to push my commit and it was a small change so I just did a commit via suggestions on here.

[jonstacks]

I think we would expect the original traffic policy {invalid} to be printed so the user knows what wasn't able to be parsed. I think what you had before was close, it just needed to be cast as a string rather than a []byte