GSoC’22: Port fortran-lang.org to Sphinx

[henilp105]

This pr is to port the Existing site of fortran-lang.org to sphinx.
This GSoC’22 project is done by Henil Shalin Panchal under the Guidance of GSoC mentors Sebastian Ehlert @awvwgk and Rohit Goswami @HaoZeke .
https://summerofcode.withgoogle.com/programs/2022/projects/L28fukIz

Thanks and Regards,
Henil Shalin Panchal

[henilp105]

Thanks sir, I will surely patch this.

Thanks and Regards,
Henil Shalin Panchal

[awvwgk]

The data files like contributors.json shouldn't be checked into git, if we are able to generate them in conf.py. Instead we should explicitly ignore them in .gitignore.

[henilp105]

Sir, but they aren't made at the build time, they are generated by workflow (fortran_packages) by using cron .

[awvwgk]

Can we generate them in conf.py instead? This wouldn't require to continuously push new commits to the repository.

The step can be guarded to be done only if requested, to avoid refetching the data to frequently in local builds, while in CI the environment variable will be set accordingly:

if os.environ.get('SPHINX_GET_DATA', None) == 'True':
    ...

We than can put the build of the page itself on a cron job.

[henilp105]

We can surely do that sir, but that would lead to increase in the build time by 2.5 mins as this data fetching script takes about 2.5 mins to fetch data from github api. and in local builds the data is fetched directly from json files from repo by a get request .

[awvwgk]

We already have data files checked into the repository, so this should be okay to keep for the time being. Still it would be preferable to not have to check-in those data files in the first place, but we have to store them somewhere, of course.

[henilp105]

Sure sir, I will be finding a balanced way out.

Thanks and Regards,
Henil Shalin Panchal

[awvwgk]

This has been resolved with the migration to the new repository (see #415)