Review test warnings and messages about epinowcast

[Bisaloo]

I'm happy to do the temporary fixing as described (unless you can think of a better solution?).

No better solution, this sounds like the most sensible option.

Perhaps just opening an issue describing what needs to be done and what the blocker is?

Yes, this sounds good. Although the multiline description felt a bit strange anyways so maybe keeping the # nolint comments there is also an acceptable solution long-term 🤷.

[Bisaloo]

My apologies for breaking this 😞. I see you already self-assigned but let me know if I should deal with it myself since I'm the one who caused it.

For a potential long-term solution, I see there is already a report for this in testthat bug tracker: r-lib/testthat#1900.

#343 should help with the second part but has been stuck for a little while. Let's continue the discussion there on how to move forward.

[seabbs]

This shows up as a warning for a new snapshot for snapshots that have a testthat message with a linebreak. Solution for now is to remove those line breaks (and use nolint) and then check the status of all snapshots. I think as part of this we should also look at making warnings cause errors in our CI so that we can more easily detect this in the future.

[seabbs]

Oh no it's my bad. I knew this was a problem but completely lost track of it and forgot to document it. I'm happy to do the temporary fixing as described (unless you can think of a better solution?).

Do you have a suggestion for recording that this needs to be fixed once r-lib/testthat#1900 is resolved? Perhaps just opening an issue describing what needs to be done and what the blocker is?

[seabbs]

#343 should help with the second part but has been stuck for a little while. Let's continue the discussion there on how to move forward.

Also yes definitely and I am happy with the proposed solution.

FYI having this in place was really helpful for #407 as that needed to not change the environment variables outside of the tests.