github.com/sirupsen/logrus usage refactoring

[na--]

For some reason, my very sleep-deprived brain got stuck on these logging minutia today 🤦‍♂️ Still, the zombie state somewhat good at slogging through repetitive code modifications, so at least something semi-useful might come out of it in the long run... 😅 🤞

This is one small part of #2958. Because of grafana/xk6-browser#818, this PR requires #2994 to be merged first.

The important parts here are replacement of the *logrus.Logger concrete type usage with logrus.FieldLogger interface in a bunch of places, most importantly in the the VU lib.State and the lib.TestPreInitState 🎉 The testutils refactoring from manually constructed objects to constructor functions + removing any usage of github.com/sirupsen/logrus/hooks/test is both making it easier to write tests now, and hopefully also making it easier to remove logrus in the future.

[inancgumus]

Hey, @na--, I'll start working on grafana/xk6-browser#818.

In the first step, I switched to this branch, ran the tests at the root with the following command, and some failed. Are these known to be flaky?

$ go test ./... -v
...
=== RUN   TestTracer/Test_#2
    tracer_test.go:169: 
                Error Trace:    /Users/inanc/grafana/k6/lib/netext/httpext/tracer_test.go:169
                Error:          Should be true
                Test:           TestTracer/Test_#2
                Messages:       http_req_blocked is <= 0
--- FAIL: TestTracer (0.00s)
    --- PASS: TestTracer/Test_#0 (0.00s)
    --- PASS: TestTracer/Test_#1 (0.00s)
    --- FAIL: TestTracer/Test_#2 (0.00s)
--- FAIL: TestVUIntegrationClientCerts (0.00s)
    --- FAIL: TestVUIntegrationClientCerts/VerifyServerCert (0.00s)
        --- FAIL: TestVUIntegrationClientCerts/VerifyServerCert/Archive (0.00s)
        --- FAIL: TestVUIntegrationClientCerts/VerifyServerCert/Source (0.01s)
    --- PASS: TestVUIntegrationClientCerts/WithCert (0.00s)
        --- PASS: TestVUIntegrationClientCerts/WithCert/Archive (0.00s)
        --- PASS: TestVUIntegrationClientCerts/WithCert/Source (0.00s)
    --- PASS: TestVUIntegrationClientCerts/WithoutDomains (0.00s)
        --- PASS: TestVUIntegrationClientCerts/WithoutDomains/Source (0.00s)
        --- PASS: TestVUIntegrationClientCerts/WithoutDomains/Archive (0.00s)
    --- PASS: TestVUIntegrationClientCerts/WithoutCert (0.00s)
        --- PASS: TestVUIntegrationClientCerts/WithoutCert/Source (0.00s)
        --- PASS: TestVUIntegrationClientCerts/WithoutCert/Archive (0.00s)

[na--]

It seems it was flaky and was disabled on Windows before (#902), but it's currently not disabled 😕 In any case, I added a note about it in #2144 (comment)

[codebien]

@na-- LGTM, I've mostly requested the change for the browser

[codecov-commenter]

Codecov Report

❗ No coverage uploaded for pull request base (update-xk6-browser@084f4ac). Click here to learn what that means.
The diff coverage is n/a.

❗ Current head be6322c differs from pull request most recent head 30b57e7. Consider uploading reports for the commit 30b57e7 to get more accurate results

@@                  Coverage Diff                  @@
##             update-xk6-browser    #2960   +/-   ##
=====================================================
  Coverage                      ?   76.91%           
=====================================================
  Files                         ?      228           
  Lines                         ?    17043           
  Branches                      ?        0           
=====================================================
  Hits                          ?    13108           
  Misses                        ?     3089           
  Partials                      ?      846           

Flag	Coverage Δ
ubuntu	76.91% <0.00%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[na--]

This should now be ready for another review. I rebased on top of #2994, with some extra improvements.

There is one part of the codebase that I haven't refactored and instead used a type assertion hack to work around. It would require a much larger refactoring, since it has a bunch of problems with it. I opened #2968 specifically for that, so we can fix it in another PR

[codebien]

LGTM

[imiric]

Nice 👍

We'll still need some tedious work to switch to slog, as the interfaces don't quite match up, but this is a step in the right direction.