quonfig-openfeature

OpenFeature provider for Quonfig -- Python server-side SDK.

This package wraps the quonfig native SDK and implements the OpenFeature Python server-side AbstractProvider interface.

Install

pip install quonfig-openfeature quonfig openfeature-sdk

Usage

from openfeature import api
from openfeature.evaluation_context import EvaluationContext
from quonfig_openfeature import QuonfigProvider

provider = QuonfigProvider(
    sdk_key="qf_sk_production_...",
    # targeting_key_mapping="user.id",  # default
)

api.set_provider(provider)
client = api.get_client()

# Boolean flag
enabled = client.get_boolean_value("my-feature", False)

# String config
welcome = client.get_string_value("welcome-message", "Hello!")

# Number config
timeout_ms = client.get_integer_value("request-timeout-ms", 5000)

# Object config (JSON or string_list)
allowed_plans = client.get_object_value("allowed-plans", [])

# With evaluation context (per-request)
is_pro = client.get_boolean_value(
    "pro-feature",
    False,
    EvaluationContext(
        targeting_key="user-123",          # maps to user.id by default
        attributes={
            "user.plan": "pro",
            "org.tier": "enterprise",
        },
    ),
)

Context mapping

OpenFeature context is flat; Quonfig context is nested by namespace. This provider maps between them using dot-notation:

OpenFeature context key	Quonfig namespace	Quonfig property
targeting_key	user	id (configurable via targeting_key_mapping)
"user.email"	user	email
"org.tier"	org	tier
"country" (no dot)	"" (default)	country
"user.ip.address"	user	ip.address (first dot only)

Customizing targeting_key mapping

provider = QuonfigProvider(
    sdk_key="qf_sk_...",
    targeting_key_mapping="account.id",  # maps targeting_key to {account: {id: ...}}
)

Accessing native SDK features

The get_client() escape hatch returns the underlying quonfig.Quonfig client for features not available in OpenFeature:

native = provider.get_client()

# Log level integration
should_log = native.should_log(
    config_key="log-level.auth",
    desired_level="DEBUG",
    contexts={"user": {"id": "user-123"}},
)

# List all config keys
keys = native.keys()

What you lose vs. the native SDK

OpenFeature is designed for feature flags, not general configuration. Some Quonfig features require the native quonfig SDK:

1. Log levels -- should_log() is native-only.
2. string_list configs -- accessed via get_object_value(), returned as a Python list[str].
3. duration configs -- return the raw float seconds via get_float_value() (or use get_duration() natively).
4. bytes configs -- not accessible via OpenFeature (no binary type in OF).
5. keys() and raw config access -- native-only via get_client().
6. Context keys use dot-notation -- "user.email", not nested objects.
7. targeting_key maps to user.id by default -- configure targeting_key_mapping if different.

Configuration changed events

The provider extends OpenFeature's AbstractProvider, so you can register event handlers via the standard api if you want to know when configs update. The native quonfig SDK pushes live updates over SSE; surfacing those as PROVIDER_CONFIGURATION_CHANGED events is on the roadmap.

Development

poetry install
poetry run ruff check .
poetry run pytest -v

The package depends on the local sibling sdk-python/ repo via a Poetry path dependency; CI checks that out as a sibling directory and installs the same way.