Add more target tests, add invalid keyID error code

[tanyav2]

No description provided.

[coveralls]

Pull Request Test Coverage Report for Build 552944478

Warning: This coverage report may be inaccurate.

This pull request's base commit is no longer the HEAD commit of its target branch. This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.

• For more information on this, see Tracking coverage changes with pull request builds.
• To avoid this issue with future PRs, see these Recommended CI Configurations.
• For a quick fix, rebase this PR at GitHub. Your next report should be accurate.

Details

• 8 of 13 (61.54%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+4.4%) to 51.844%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
target.go	8	13	61.54%

Totals
Change from base Build 549355364:	4.4%
Covered Lines:	239
Relevant Lines:	461

---

💛 - Coveralls

[chris-wood]

This type of dependency shouldn't be in a unit test. Why did we go with 1.1.1.1 here?

[tanyav2]

Likelihood of 1.1.1.1 being unavailable seems extremely low, and I wanted to exercise the resolve(): https://github.com/cloudflare/odoh-server-go/blob/master/resolver.go#L47 codepath. I figured that the benefits of having atleast one test for that codepath far outweighed the risk of creating a 1.1.1.1 dependency.

[chris-wood]

I get that, but I think we can exercise that path by mocking things (or wrapping resolve() appropriately). Adding a network dependency doesn't seem to be the right answer.

[tanyav2]

re: mocking it, personally I think exercising the networking codepath in resolve() is pretty valuable

[tanyav2]

But I can remove it if you think it's fine without.

[chris-wood]

Testing the implementation of resolve() isn't something this project should do. We should only be testing that the code that uses resolve is correct, which is why I suggested mocking it.

[chris-wood]

The key ID mismatch change looks good! I'm requesting changes because I don't think it's appropriate for unit tests to hit the network. (If we have a 1.1.1.1 dependency, let's mock it.)