api-datasette

Datasettes containing data related to api.existenz.ch.

LIVE: https://api-datasette.konzept.space

Structure

Run poetry run python generate.py

• Raw metadata gets downloaded from the Existenz-API.
• Gets put into existenz-api.db (SQLite file).
• datasette displays the database as webpages.

Start Datasette

poetry run datasette existenz-api.db -m metadata.json

Update

poetry add datasette:latest

poetry update

If Datasette has changed: Re-run plugin installations, i.e. poetry run datasette install datasette-cluster-map.

If any dependencies have changed: Run poetry export --without-hashes -f requirements.txt --output requirements.txt and commit the file.

If new plugins are required, add their installation to post_compile.

Initial setup

brew install pipx

pipx install poetry

pipx inject poetry poetry-plugin-export

Upgrade Python

See officially supported runtimes here: https://devcenter.heroku.com/articles/python-support#supported-runtimes

To switch to a new Python version on Dokku, update the string in runtime.txt and push to Dokku.

Deployment

Run git push dokku main

If without changes and you just want to rebuild: git commit --allow-empty -m "Rebuild."

Deployment setup

• Create a dokku app on konzept.space with dokku apps:create api-datasette; dokku letsencrypt:enable api-datasette
• Locally: git remote add dokku dokku@konzept.space:api-datasette

Dokku will identify this as a Python project, load the appropriate buildpack, run the post_compile script and then start serving the service defined in the Procfile. Pretty nifty.