Compiled by Frank Meineke (2022-02-15) as part of the MII Taskforce DDT.
Features
- load and start publicly available FHIR servers as docker containers.
- always load the "latest" tagged version
- load publicly available FHIR data
- use the databases that are currently delivered with the containers (derby etc)
- integrate server into the HAPI tester frontend (except IBM)
- Note: performance tests are only possible to a very limited extent / not at all
| Server | Endpoint | Remark |
|---|---|---|
| Hapi | http://localhost:8081/fhir | |
| Blaze | http://localhost:8082/fhir | |
| Vonk/Firely | http://localhost:8083/fhir | License needed |
| IBM | http://localhost:8084/fhir | Authorization needed |
- Load data (here: POLAR testdata). The zip formats are loaded and unpacked to json files.
make data - for vonk: get your personal trial license file from https://simplifier.net/firely-server
Goto specific server/XXX and type make start individually
or (for the brave) type make -C server start-all
make upload-all
make count-all
- Windows 10 (Ryzen 5, 32GB)
- wsl2 (using 16GB, necessary if all servers are running)
- DockerDesktop
- ubuntu (packages: install "jq" as json pretty printer, curl, wget, make, unzip)
To configure CQL for HAPI activate CQL in the HAPI application.yml or via .env var depending on your server setup.
In this repository change the my-application.yml in the HAPI folder and set cql_enabled: true.
Further HAPI will need extra helper CQL libraries to work. navigate to the cql-init-hapi folder of this repository and execute hapi-init-cql.sh to load the required libraries to your HAPI server.
Once HAPI is started up you can then use the python script example-cql-exec.py to evaluate the number of female patients in your server.