A simple python script to scrape the FHIR schemas from http://hl7.org/fhir and transform them into Postgres tables. This approach creates a column for each field instead of stuffing the JSON representation into a column. By doing this you can better leverage the community of tools built on Postgres like Postgrest, Supabase or Hasura. References are handled through foreign keys and in cases where a reference references more than one resource we create a polymorphic table.
Setup your virtual env of choice.
pip install -r requirements.txt
python --sql-file output.sql --resources "CarePlan" "Patient" "Communication" "Practitioner" "Identifier" "CodeableConcept" "Observation" "Consent" "Goal" "HumanName" "ContactPoint"
If you'd like to generate tables for every FHIR resource exclude the --resources flag. Otherwise specify which subset you'd like to generate tables for.