Add builds for windows and macOS

[Zeitsperre]

Pull Request Checklist:

• This PR addresses an already opened issue (for bug fixes / features)
  • This PR doesn't address Click commands fail on Windows #1500 finally
• Tests for the changes have been added (for bug fixes / features)
  • (If applicable) Documentation has been added / updated (for bug fixes / features)
• CHANGES.rst has been updated (with summary of main changes)
  • Link to issue (:issue:number) and pull request (:pull:number) has been added

What kind of change does this PR introduce?

• Adds both a Windows and macOS build to check the testing suite against different platforms.
• the .lock file generated in the cache is now removed after the testing data has been copied, and the file lock has been released.
• The prefetch argument for tox is now available for general testing environments.

Does this PR introduce a breaking change?

No.

Other information:

The approved tag is listed here to ensure that the relevant builds are being triggered. This PR still requires review approval.

In order to run Windows build tests, $ xclim prefetch_testing_data must be called first to gather the testing data.

Did you know that Windows doesn't support Unix-style File Locking (fcntl)? This was what was preventing asynchronous handling of testing data fetching from working.

[github-actions]

Note
It appears that this Pull Request modifies the main.yml workflow.

On inspection, the XCLIM_TESTDATA_BRANCH environment variable is set to the most recent tag (v2023.12.14).

No further action is required.

[aulemahal]

Nice! Thanks!

I see the test pass, does this mean #1500 can be closed ? (I should read the PR I review)

[Zeitsperre]

No, you're right, this unfortunately doesn't resolve #1500, but I now know why the windows call to pytest on conda may have been failing previously.

I'm going to try running $ xclim prefetch_testing_data as one of the setup commands prior to testing when we build on conda-forge from now, and that should fix any potential problems from that source.

I'm going to add a call to $ xclim --help here to get more information.

[coveralls]

coverage: 90.171% (+0.01%) from 90.16%
when pulling d27f9a9 on multi-os-builds
into d09bd4b on master.