ELIXIR’s Interoperability Platform 2024–26 activities

Repository for content produced and developed during ELIXIR’s Interoperability Platform 2024–26 activities.

Testing the site locally

With Docker Compose installed:

docker compose up

Then open http://localhost:4000. The site rebuilds automatically as you edit files.

Stop it with:

docker compose down -v

-v also removes the gems/site cache volumes, so the next docker compose up reinstalls everything from scratch.

If your checkout is a git worktree rather than the main clone, run ./docker-worktree-setup.sh once first.

docker compose up (docker-compose.yml) runs jekyll serve (dev server, no --baseurl, JEKYLL_ENV=development), which is not what the GitHub Actions workflow builds and deploys. To check a page the way it will actually render on GitHub Pages, build and serve it the same way CI does, using the separate production compose file:

docker compose -f docker-compose.prod.yml up

Then open http://localhost:4001/interoperability_2024-26/ (note the path — GitHub Pages serves this site under /interoperability_2024-26/, not /). This is a one-shot build, not a live-reloading dev server; re-run the command after making changes.

If your checkout is a git worktree, ./docker-worktree-setup.sh also generates a prod override file, which must be passed explicitly (unlike the dev override, it isn't merged automatically):

docker compose -f docker-compose.prod.yml -f docker-compose.prod.override.yml up

Regenerating _data/library-interop-stories.json

_data/library-interop-stories.json (interoperability stories, FAIR Metroline, FAIR Cookbook, RDMkit, FAIRification Template, DSM and Process data, plus their cross-mappings) is sourced from a Google Sheets document. It used to be exported by hand via an Apps Script bound to that document; scripts/generate_library_interop_stories.py does the same export locally instead, with no dependencies beyond the Python 3 standard library.

Run it from the repo root:

python3 scripts/generate_library_interop_stories.py SOURCE_URL

It reads each relevant tab straight from Google's public CSV export endpoint (so the document must stay shared as "Anyone with the link can view"), rebuilds the same JSON structure the old Apps Script produced, and overwrites _data/library-interop-stories.json. Review the resulting diff before committing.