Replace underscores in facets of observational datasets with minuses

[schlunma]

Description

This PR replaces _ in facets of observational datasets with -.

TODO:

• Rename files on Levante (@axel-lauer)
• Rename files on Jasmin (@valeriupredoi?)

• Closes OBS datasets use underscores in the version number #3051

---

Before you get started

• ☝ Create an issue to discuss what you are going to do

Checklist

It is the responsibility of the author to make sure the pull request is ready to review. The icons indicate whether the item will be subject to the 🛠 Technical or 🧪 Scientific review.

• 🛠 This pull request has a descriptive title
• 🛠 Code is written according to the code quality guidelines
• 🛠 Documentation is available
• 🛠 Tests run successfully
• 🛠 The list of authors is up to date
• 🛠 Any changed dependencies have been added or removed correctly
• 🛠 All checks below this pull request were successful

---

To help with the number of pull requests:

• 🙏 We kindly ask you to review two other open pull requests in this repository

[axel-lauer]

Looks good to me. Thanks @schlunma !

[valeriupredoi]

cheers, Manu 🍺 I'll rename those on CEDA too 👍

[valeriupredoi]

so I was faffing about looking at what datasets need filename changes on CEDA-JASMIN, and then it dawned to me: what if someone wants to use an older version of ESMValTool (like we support on JASMIN), wouldn't be better to have these both supported? ie wait for the DKRZ rsync that will actually bring in the "new" datasets, instead of renaming existing ones

[schlunma]

what if someone wants to use an older version of ESMValTool (like we support on JASMIN), wouldn't be better to have these both supported? ie wait for the DKRZ rsync that will actually bring in the "new" datasets, instead of renaming existing ones

Yeah, the problem here is that we cannot have both versions if we want to use ESMValGroup/ESMValCore#1943 (which is the entire point of this little exercise here). If the old version is still there, facets will be assigned wrong.

I already thought about ways to make this fully backwards-compatible, but I couldn't think of one...The closest I got is creating an additional directory with links to the files (that use the old names), but this requires using a different configuration file.

[valeriupredoi]

...which I can do easily for central installs; also - is the current naming (ie this PR) gonna work if we merge it now? I see that Core#1943 is still a draft

[schlunma]

Yes, everything will work just fine. ESMValGroup/ESMValCore#1943 will enable reading facets off file names, which will for example allow finding the latest version of observational datasets.

...which I can do easily for central installs

Just out of curiosity - how do you do that? Changing the default configuration options?

[valeriupredoi]

Just out of curiosity - how do you do that? Changing the default configuration options?

Yup: different installations grab different user config files, that users can then change too (not the base values for data dirs though) 🍺

I'll wait for the rsync: then I can just move the underscorey ones in a separate dir 👍