refactor(webconnectivity*): don't call httpapi directly #1582
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This diff includes a first attempt at consolidating the patterns with which we invoke OONI and third-party API calls. I have refactored the code of httpx and httpapi into a new package called httpclientx, added some tests, started converting some parts of the tree, and explained myself in a design document. Part of ooni/probe#2700
Spotted thanks to a very paranoid check inside ./internal/oonirun. Was not a problem before for `httpx` because of its usage pattern and may or may not be a problem for the `httpapi` package (did not check since this work is focused on replacing both `httpx` and `httpapi`).
Previously, we were gracefully handling this case, but honestly it is not the best approach to pretend there's an empty structure if the server breaks the API and returns `"null"` rather than an object. That said, it was still awesome to have this test in place because it helped us to figure out this extra condition of which httpclientx should be aware and that this problem needs to be dealt with systematically inside the httpclientx package.
Conflicts: internal/enginelocate/cloudflare.go internal/enginelocate/ubuntu.go
Conflicts: internal/oonirun/v2_test.go
As before, here I am going to ensure there's redundancy.
Conflicts: internal/httpclientx/DESIGN.md internal/httpclientx/getraw.go internal/httpclientx/overlapped.go
Conflicts: internal/probeservices/login_test.go
Conflicts: internal/testingx/oonicollector.go
Conflicts: internal/probeservices/collector_test.go
Conflicts: internal/probeservices/bouncer.go internal/probeservices/collector.go internal/probeservices/login.go internal/probeservices/measurementmeta.go internal/probeservices/psiphon.go internal/probeservices/register.go internal/probeservices/tor.go
While there, add additional tests ensuring that we're actually writing into the key-value store after check-in. Also, ensure that failure to write into the key-value store does not break check-in for good.
Conflicts: internal/legacy/mockable/mockable.go internal/mocks/session.go
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The #1581 pull request missed that adding a
CallWebConnectivityTestHelpermethod to*engine.Sessionmeans we need to mock such a method inwebconnectivityqa.I could have mocked the method, but I was not super happy about doing this, because I would rather have wanted to add more QA tests inside of
webconnectivityqamaking sure we're indeed falling back when using test helpers. (We have unit tests testing that, but it would be nice to have integration tests, so we could observe the overall code behavior.)My initial approach for allowing this was to create a
CallWebConnectivityTestHelperfunction inside the engine, taking amodel.ExperimentSessionas its fourth argument and then modifying theCallWebConnectivityTestHelpermethod to just invoke the function passing to it the session. However, this approach created import cycles in tests.To avoid import cycles, I moved the
CallWebConnectivityTestHelperfunction towebconnectivityalgoand modified the mocks insidewebconnectivityqato call such a function. This is where I paused and realized that the cleanest solution is actually to just always call theCallWebConnectivityTestHelperfunction inwebconnectivityalgoand undo part of the changes in #1581 to avoid having aCallWebConnectivityTestHelpermethod inside ofmodel.ExperimentSession. And so I also implemented this change.All of this leads us to the current diff, which is part of ooni/probe#2725.