Feature request: provide a Python function to load lambda env vars in local IDE for tests

[ran-isenberg]

I'm looking for feedback for a feature where:

1. In your integration tests you setup env vars since the decorator 'init_environment_variables' will require them to exist when you call the handler function.
2. instead of manually setting values, a better way would be to load the "real" env vars deployed to your function. That way, you know for sure they are correct even before running your e2e tests and you can remove the manual code setup from conftest.py

import os

import pytest

from cdk.service.constants import (
    CONFIGURATION_NAME,
    ENVIRONMENT,
    IDEMPOTENCY_TABLE_NAME_OUTPUT,
    POWER_TOOLS_LOG_LEVEL,
    POWERTOOLS_SERVICE_NAME,
    SERVICE_NAME,
    TABLE_NAME_OUTPUT,
)
from tests.utils import get_stack_output


@pytest.fixture(scope='module', autouse=True)
def init():
    os.environ[POWERTOOLS_SERVICE_NAME] = SERVICE_NAME
    os.environ[POWER_TOOLS_LOG_LEVEL] = 'DEBUG'
    os.environ['REST_API'] = 'https://www.ranthebuilder.cloud/api'
    os.environ['ROLE_ARN'] = 'arn:partition:service:region:account-id:resource-type:resource-id'
    os.environ['CONFIGURATION_APP'] = SERVICE_NAME
    os.environ['CONFIGURATION_ENV'] = ENVIRONMENT
    os.environ['CONFIGURATION_NAME'] = CONFIGURATION_NAME
    os.environ['CONFIGURATION_MAX_AGE_MINUTES'] = '5'
    os.environ['AWS_DEFAULT_REGION'] = 'us-east-1'  # used for appconfig mocked boto calls
    os.environ['TABLE_NAME'] = get_stack_output(TABLE_NAME_OUTPUT)
    os.environ['IDEMPOTENCY_TABLE_NAME'] = get_stack_output(IDEMPOTENCY_TABLE_NAME_OUTPUT)


@pytest.fixture(scope='module', autouse=True)
def table_name():
    return os.environ['TABLE_NAME']

I can provide a function that loads the env vars for you. It can accept one or more of the following parameters:

1. stack name
2. function name
3. function ARN

Usage:
before running an integration test for a specific lambda with pytest as show here:
you will create a pytest fixture that will load the env vars for that function.
An easy way to do in to use stack outputs for the names or provide stack name and lambda regex/name to look for.

What do you think of the feature? how would you like to use it? provide stack name and lambda name? or just ARN?

import pytest

@pytest.fixture
def init():
    load_env_vars_from(stack_name=..., lambda_arn=...)