google-fhir-r4 0.11.0

Protocol buffer conversion, validation and utilities for FHIR R4 resources.

Users in interested in analyzing FHIR data should reference the google.fhir.views package. This library contains underlying capabilities to work with FHIR data represented in protocol buffers in PYthon code.

FHIR JSON to and from Protocol Buffers

The json_format package supports converting FHIR protocol buffers to and from the FHIR JSON format. Here are some simple examples:

from google.fhir.r4 import json_format
from google.fhir.r4.proto.core.resources import patient_pb2

patient_json = """
{
  "resourceType" : "Patient",
  "id" : "example",
  "name" : [{
    "use" : "official",
    "family" : "Roosevelt",
    "given" : ["Franklin", "Delano"]
  }]
}
"""

# Read the JSON as a proto:
patient = json_format.json_fhir_string_to_proto(patient_json, patient_pb2.Patient)

# Get the FHIR JSON representation of the proto:
json_format.print_fhir_to_json_string(patient)

FHIRPath support

FHIRPath is the basis of the google-fhir-views logic, but can also be used directly against FHIR protos themselves. Here is an example:

from google.fhir.r4 import fhir_path
from google.fhir.r4 import r4_package
from google.fhir.core.fhir_path import context

# Create the FHIRPath context for use.
fhir_path_context = context.LocalFhirPathContext(r4_package.load_base_r4())

# Compile the FHIRPath expressions for evaluation. This validates the FHIRPath and returns
# an optimized structure that can be efficiently evaluated over protocol buffers.
# The compiled expression is typically created once and reused for many following invocations.
expr = fhir_path.compile_expression('Patient', fhir_path_context, "name.where(use = 'official').family")

# Now we evaluate the expression on the given resource.
result = expr.evaluate(patient)

# Check to see if the result matched anything.
if result.has_value():
  # Gets the result as protocol buffer messages.
  result_as_messages = result.messages
  # Converts the result to a simple string, if applicable.
  result_as_simple_string = result.as_string()

The library also supports a fluent Python builder for creating FHIRPath expressions, which is used extensively in the google.fhir.views package. Here is the above string being created using a Python builder pattern:

# FHIRPath builder for patients.
pat = fhir_path.builder('Patient', fhir_path_context)

# Build the expression and get it in string form.
fhir_path_string = pat.name.where(pat.name.use == 'official').family.fhir_path