fix: preserve Decimal precision during serialization

[rtmalikian]

Summary

Fixes Decimal precision loss during serialization. Previously, Decimal('19') was serialized as 19.0 in both model_dump() and model_dump_json() because the serializer unconditionally converted to float.

Per the FHIR spec, decimal precision is semantically significant: 19, 19.0, and 19.00 are clinically distinct values representing different measurement precisions.

Root Cause

The _serialize function in Decimal.__get_pydantic_core_schema__ used return float(value), which converts Decimal('19') to 19.0 — losing the precision information.

Fix

Check the Decimal's internal exponent (from as_tuple()) to determine the correct serialization format:

• exponent >= 0 (e.g., Decimal('19'), Decimal('0')): serialize as int → JSON renders 19, 0
• exponent < 0 (e.g., Decimal('19.0'), Decimal('19.5')): serialize as float → JSON renders 19.0, 19.5

This preserves the original precision through both model_dump() and model_dump_json().

Also removed the return_schema=core_schema.decimal_schema() constraint from the serializer, since it now returns int/float instead of Decimal.

Test Results

tests/test_types.py::test_decimal_type PASSED
tests/test_types.py::test_decimal_whole_number_inputs_serialize_as_float[1.0-1.0] PASSED
tests/test_types.py::test_decimal_whole_number_inputs_serialize_as_float[150.0-150.0] PASSED
tests/test_types.py::test_decimal_whole_number_inputs_serialize_as_float[100.0-100.0] PASSED
tests/test_types.py::test_decimal_whole_number_inputs_serialize_as_float[1234.5-1234.5] PASSED
tests/test_types.py::test_decimal_whole_number_inputs_serialize_as_float[1.01-1.01] PASSED
tests/test_fhir_resource_issue_203.py::test_fhir_core_decimal_serialization_issue_230 PASSED
======================== 81 passed, 1 skipped in 0.91s ========================

Verification

from decimal import Decimal
from fhir.resources.observation import Observation
import json

obs = Observation(**{
    "resourceType": "Observation",
    "status": "final",
    "code": {"coding": [{"system": "http://loinc.org", "code": "12345-6"}]},
    "component": [{
        "code": {"coding": [{"system": "http://loinc.org", "code": "12345-6"}]},
        "valueQuantity": {"value": Decimal("19"), "unit": "mg/dL",
                          "system": "http://unitsofmeasure.org", "code": "mg/dL"},
    }],
})

data = json.loads(obs.model_dump_json())
assert data["component"][0]["valueQuantity"]["value"] == 19        # ✅ was 19.0
assert isinstance(data["component"][0]["valueQuantity"]["value"], int)  # ✅ was float

Closes nazrulworld/fhir.resources#203

---

About the Author: Raphael Malikian — Clinical AI Solutions Architect. I specialise in building and fixing AI/ML systems for healthcare, including vector databases, RAG pipelines, and clinical NLP. If you need help with your project or think I can add value to your organisation, feel free to reach out — I'd love to connect.

📧 rtmalikian@gmail.com
🔗 GitHub: https://github.com/rtmalikian
🔗 LinkedIn: http://www.linkedin.com/in/raphael-t-malikian-mbbs-bsc-hons-71075436a

---

Disclosure: This code was developed with assistance from mimo-v2.5-pro (Xiaomi) via Hermes Agent (Nous Research). All changes were reviewed, tested against the actual codebase, and verified for correctness.

[nazrulworld]

Hi @rtmalikian thanks a lot for your PR.
We had similar PR #20 ,that is merged. Please check if that PR is solving same problem.