New logger noise during test runs

[mccalluc]

Tests runs used to just be ..... etc., but with the latest changes, I see this a dozen times:

INFO -- : EZID GetIdentifierMetadata -- success: ark:/88435/dsp01h415pd457

Not the end of the world, but with a lot of log noise during test runs it can be a bit harder to spot the real problems.

• Is this message set at the appropriate logging level?
• Or maybe the filter on log messages during tests can change?
• or ...

[hectorcorrea]

I hope we are not really minting new ARKs every time we run the tests.

[mccalluc]

• All the messages are the same: Pretty sure IDs are not being minted.
• Checking out earlier versions, I believe this was introduced by #907.
• Commenting out state { "draft" } here seems to eliminate the excess logging, though it introduces two test failures:

  rspec ./spec/system/work_edit_spec.rb:157 # Creating and updating works change log displays changes
  rspec ./spec/system/work_edit_spec.rb:131 # Creating and updating works when editing funding information allows the user to edit funding information

• At least one test is failing because after the form POST, it is being 302 redirected to the home page: It looks like an error should be rendered, but I don't see anything on the screen.

  elsif !@work.editable_in_current_state?(current_user)
    redirect_to root_path, notice: I18n.t("works.approved.uneditable")

• I tried adding work.state = "draft" to just this test. No change.
• Are flash messages actually being displayed on the homepage? In this case that doesn't seem to be the case.

[bess]

Not really an issue. We aren't minting real ARKs every time, we're faking it with webmock. Having these messages in the test logs isn't an issue.