Fix failing Meteonorm tests

[AdamRJensen]

• I am familiar with the contributing guidelines
• Tests added
• Adds description and name entries in the appropriate "what's new" file in docs/sphinx/source/whatsnew for all changes. Includes link to the GitHub Issue with :issue:`num` or this Pull Request with :pull:`num`. Includes contributor name and/or GitHub username (link with :ghuser:`user`).
• Pull request is nearly complete and ready for detailed review.
• Maintainer: Appropriate GitHub Labels (including remote-data) and Milestone are assigned to the Pull Request and linked Issue.

The Meteonorm iotools tests recently started failing due to changes on their end.

[RDaxini]

What's wrong with the original? For a single URL definition as the whole function, isn't it okay just to leave it as:

Suggested change

	demo_url = 'https://demo.meteonorm.com/v1/'
	return demo_url
	return 'https://demo.meteonorm.com/v1/'

Same for the demo_api_key
Maybe there's something about python (style?) here that I don't know

[AdamRJensen]

It's just a matter of convenience when I was hacking around locally. With the line split into two, you can run the single-line manually that defines the variable and then run the content of a test function.

Tests are a bit messy to debug sometimes because of all the magic they employ.

[kandersolar]

@AdamRJensen are we confident that we should switch to the new values in this PR, or do we need to wait to hear back from Meteonorm about the change?

[AdamRJensen]

@AdamRJensen are we confident that we should switch to the new values in this PR, or do we need to wait to hear back from Meteonorm about the change?

That is hard to say. I suggest that we merge it before making a new release if we do not hear back from them.

[kandersolar]

We heard back from Meteonorm that the data changes are intentional. Let's merge this and maybe consider different testing strategies later.

[kandersolar]

Test failures are unrelated (the python 3.9 issue which was fixed elsewhere).

Thanks @AdamRJensen!