Quickstart

Conditional method selection

import os

from conditional_method import cfg

ENV = os.environ.get("ENV", "development")


class Worker:
    @cfg(condition=ENV == "production")
    def work(self, *args, **kwargs):
        return "production"

    @cfg(condition=ENV == "development")
    def work(self, *args, **kwargs):
        return "development"

    @cfg(condition=ENV == "staging")
    def work(self, *args, **kwargs):
        return "staging"


worker = Worker()
print(worker.work())  # development

# Only ONE `work` survives on the class — it is "well-formed":
print(Worker.__dict__.keys())
# dict_keys(['__module__', '__qualname__', 'work', ...])

Callable conditions

Conditions can be callables evaluated lazily per call:

import os

from conditional_method import cfg


class DatabaseConnector:
    @cfg(condition=lambda f: os.environ.get("ENV") == "production")
    def connect(self, config):
        return "production connection"

    @cfg(condition=lambda f: os.environ.get("ENV") != "production")
    def connect(self, config):
        return "default connection"

Conditional decorators with @cfg_attr

import os

from conditional_method import cfg_attr

os.environ["FEATURE_FLAG"] = "enabled"


def log_calls(fn):
    def wrapper(*args, **kwargs):
        print(f"calling {fn.__name__}")
        return fn(*args, **kwargs)

    return wrapper


@cfg_attr(
    condition=os.environ.get("FEATURE_FLAG") == "enabled",
    decorators=[log_calls],
)
def experimental(value):
    return value * 2

Global functions and classes

@cfg also works on module-level functions and on classes themselves:

from conditional_method import cfg


@cfg(condition=True)
class Person:
    def greet(self):
        return "hi"

See Usage and API Reference for the full surface.