Add support for yoda-H5

[mhabedan]

As a follow-up to HEPData/hepdata-converter#64, this PR adds yoda-H5 support for hepdata itself.
Namely the changes are

• Include .yoda.h5 in supported file formats for download requests/urls.
• Update documentation accordingly.
• Add test cases for .yoda.h5.
• As discussed in Add yoda-hdf5 support hepdata-converter#64 (comment), this does not add .yoda.h5 to the dropdown menus.

Tested locally on commit f923a755 from hepdata-converter.
Requires HEPData/hepdata-converter-ws#17.

[mhabedan]

Hi @GraemeWatt!
Now that HEPData/hepdata-converter-ws#17 is merged, this PR should be good to go, too! (It's the second to last one for yoda-h5, the last one being HEPData/hepdata-cli#7)

[GraemeWatt]

Thanks, looks good. I agree not to add the new yoda.h5 format to the dropdown menus initially, but it can easily be added later if there is demand for it. I checked out your branch and I confirm it works locally with the latest hepdata-converter-ws Docker image (see docs). I updated your branch to include the latest changes from the main branch. Few comments:

1. The JSON format of a record contains links to various output formats of data tables. Probably the new yoda.h5 format should be added here.
2. The yoda.h5 filename of a whole record is e.g. HEPData-ins1847779-v1-yoda.h5.tar.gz containing HEPData-ins1847779-v1-yoda.h5.yoda.h5. Is this OK or would it be better to change the . to a - if not in the file extension, e.g. HEPData-ins1847779-v1-yoda-h5.tar.gz containing HEPData-ins1847779-v1-yoda-h5.yoda.h5? This could be done by replacing file_format with file_format.replace('.', '-') here and here. Another option is e.g. HEPData-ins1847779-v1-yodah5.tar.gz containing HEPData-ins1847779-v1-yodah5.yoda.h5. Which do you prefer?
3. I triggered the CI, but I think it's failing because your fork doesn't have access to the repository secrets, particularly OPENSEARCH_INITIAL_ADMIN_PASSWORD. The end-to-end tests also require SAUCE_ACCESS_KEY and SAUCE_USERNAME. We could change the GitHub Actions workflow to use a pull_request_target event, but I'd need to investigate how this works and if it is secure. I've invited you as a Collaborator with Write access to the main HEPData/hepdata repository, so in future you can create a new branch under this repository. I've sent you the necessary secrets privately that you can add to your fork to avoid creating a new branch and opening a new PR.

[mhabedan]

Hi @GraemeWatt!

Thanks for your review and apologies for being so slow for acting on it!

1. Done.
2. That's a good point, the previous notation was confusing. I opted for HEPData-ins1847779-v1-yoda-h5.tar.gz containing HEPData-ins1847779-v1-yoda-h5.yoda.h5.
3. I've added the secrets but the CI still fails because the connection was refused. Will follow-up via email.

[GraemeWatt]

I'm not sure why this PR is shown as merged after I merged the replacement PR #899, but the commits to the main branch look correct. I edited the CONTRIBUTING.rst file in commit e42a7e7 to request that future PRs should not be made from forks.