How to Handle List of Searches? about fhir-py

[mkizesov]

AsyncFHIRClient make asynchronous I/O calls, that means those calls are non-blocking. And of course it's ready to use it with any python asyncio features.
If I understand correctly, you're trying to run these tasks concurrently.
So the answer to your question is to use asyncio.gather.
I'll provide you with the basic example based on your code. It works and it's a good starting point, but it's not the best way to do it. For example, if any task throws an exception, you'll not be able to get any results in the patients list.

import asyncio
from fhirpy import AsyncFHIRClient
from aiohttp import BasicAuth


async def main():
    # Create an instance
    client = AsyncFHIRClient(
        'http://localhost:8080/',
        authorization=BasicAuth('user', 'password').encode()
    )

    pat_ids = [
        "patient1",
        "patient2",
        "patient3",
        "patient4",
        "patient5",
    ]

    resources = client.resources('Patient')

    # Create a list of coroutines (not awaited)
    tasks = [
        resources.search(_id=pat_id).first()
        for pat_id in pat_ids
    ]

    patients = await asyncio.gather(*tasks)
    print(patients)


if __name__ == '__main__':
    loop = asyncio.get_event_loop()
    loop.run_until_complete(main())

Also, there's no need to use client.schema = None and fhir_version='4.0.0', because we've removed schema validation since version 1.0.0.

[kylejbrk]

@mkizesov
That makes sense. It looks like this only work with .first()? I wouldn't be able to retrieve all the results from a search? Never mind. Just had to replace with .fetch()

One of the things I'd like to do is get all the encounters for many patient id's.

Also, from the code you posted it looks like I can authenticate with a username and password as well a Bearer Token? That's pretty neat if so.

[mkizesov]

The value of "authorization" argument is used in the Authorization header of each lib request. If your FHIR-server supports basic auth, BasicAuth('user', 'password').encode() will do its job. If your server use OAuth 2 (for example), you can use any 3rd party library to authenticate on the server and get Bearer token. Then provide this token to AsyncFHIRClient's "authorization" argument.

Yep, any fhirpy coroutine function (count, fetch, get, first, fetch_all, etc.) could be used there.

Another example suitable to speed up some independent tasks like retrieving and updating patient.

import asyncio
from fhirpy import AsyncFHIRClient
from aiohttp import BasicAuth


client = AsyncFHIRClient(
    'http://localhost:8080/',
    authorization=BasicAuth('root', 'secret').encode()
)
patient_resources = client.resources('Patient')


async def update_patient(patient_id):
    patient = await patient_resources \
        .search(_id=patient_id).first()
    if patient:
        # change patient here

        await patient.save()
        print(f'Patient "{patient_id}" was updated')
    else:
        print(f'Patient "{patient_id}" wasn\'t found')


async def main():
    pat_ids = [
        "patient1",
        "patient2",
        "patient3",
        "patient4",
        "patient5",
    ]

    resources = client.resources('Patient')

    # Create a list of coroutines (not awaited)
    tasks = [
        update_patient(pat_id) for pat_id in pat_ids
    ]

    await asyncio.gather(*tasks)


if __name__ == '__main__':
    loop = asyncio.get_event_loop()
    loop.run_until_complete(main())

[kylejbrk]

Oh thanks again. I could use something like the above to write all the results of the individual tasks to disk.

Our implementation uses OAuth2 so I've been using the bearer token method. Wasn't sure if the other would work as well.

Very helpful thank you.