fhirstore

Python library to manipulate fhir resources leveraging mongoDB as storage layer.

Installation

pip install fhirstore

Usage

from pymongo import MongoClient
from fhirstore import FHIRStore

client = MongoClient()
store = FHIRStore(client, "<my_database>")

# Dropping collections
store.reset()

# Parse json schema and create collections
store.bootstrap(depth=5)
# OR
# Get existing collections from the database
store.resume(depth=5)

# Create resources
store.create({
    "resourceType": "Patient",
    "id": "pat1",
    "gender": "male"
})

# Read resources
patient = store.read("Patient", "pat1")

# Update resources
updated_patient = store.update("Patient", "pat1", {"gender": "other"})

# Delete resources
deleted_patients_count = store.delete("Patient", "pat1")
deleted_patients_count = store.delete("Patient", resource_id="pyrogResouceId")

Bootstrap the database

1. Start the database
2. Drop and re-create all collections based on the provided schema

docker-compose up -d
python main.py

Development setup

1. Create a virtual environment and enter it
2. Install python dependncies

virtualenv .
. ./bin/activate
pip install -r requirements.txt
pip install -r requirements-dev.txt

Test

Test can be run using :

python -m pytest

Make sure dev dependencies are installed.

Benchmark

A benchmark based on example FHIR resources publicly available at can be run using :

python benchmark/benchmark.py

Note that you will need to have a MongoDB up in order for the benchmark to run. You can use the docker-compose file of this repository by running docker-compose up before launching the benchmark.

On a machine with 16GB RAM and an i7 (2.5GHz) processor, the results of the benchamrk were:

--- WRITES ---
insertions per second (on average): 267.17
average: 3.74 milliseconds
median: 1.66 milliseconds
min: 1.07 milliseconds
max: 724.65 milliseconds
spread: 0.00028004697751347234

--- READS ---
reads per second (on average): 378.93
average: 2.63 milliseconds
median: 1.50 milliseconds
min: 0.88 milliseconds
max: 481.18 milliseconds
spread: 0.0002154728657872756

On the same machine, here are the results for the same benchmark using a different library, fhirbase. The latter relies on postgreSQL rather than MongoDB as its storage layer. Also, it only supports FHIR resources 3.3.0 and below.

--- WRITES ---
insertions per second (on average): 152.38
average: 6.56 milliseconds
median: 4.43 milliseconds
min: 3.08 milliseconds
max: 158.88 milliseconds
spread: 7.568139030481407e-05

--- READS ---
reads per second (on average): 261.07
average: 3.83 milliseconds
median: 3.03 milliseconds
min: 2.17 milliseconds
max: 102.06 milliseconds
spread: 1.4565324857334904e-05

Publish

First, you need to have twine installedd

pip install --user --upgrade twine

Make sure you have bumped the version number in setup.py, then run the following:

python setup.py sdist bdist_wheel
python -m twine upload dist/*