Skip to main content

Python DI/IoC container

Project description

Python DI / IoC container

Dependency injection means giving an object its instance variables.

-- James Shore, Dependency Injection Demystified

Initially it was created for AWS Lambda Functions to simplify development with loose coupling component and improved testing capabilities.

Inspiration comes from python:design_patterns:inversion_of_control summary article.

This modules is an optimized version of a container used in p3.4.advanced-di-ioc section of my lambda-modular-python MOB session @cloudreach.


Orchestrate container in file:

import pydioc
import boto3
import some_sdk
from . import services

def build_container(ssm_param_api_key: str, ddb_status_table: str, format_type: str) -> pydioc.Container:
    return pydioc.Container(
        ("_boto3_session", boto3.Session),
        ("lambda_context", pydioc.ContextProxy),

        ("_api_key_loader", services.api_key_loader, ["_boto3_session", lambda: ssm_param_api_key]),
        ("_status_updater", services.status_updater, ["_boto3_session", "lambda_context", lambda: ddb_status_table]),

        ("_sdk_data_formatter", some_sdk.FormatterFactory, [lambda: format_type]),
        ("_sdk_client_factory", services.sdk_client_factory, ["_api_key_loader"]),

        ("_sdk_data_transfer", services.sdk_data_transfer, ["_sdk_client_factory", "_sdk_data_formatter", "lambda_context"]),

        ("event_handler", services.event_handler, ["_sdk_data_transfer", "_status_updater"]),

Define services in file:

import boto3
import some_sdk

# ...

def api_key_loader(session: boto3.Session, ssm_param_api_key: str):
    assert ssm_param_api_key, 'expecting non empty api_key param'

    def _api_key_loader():
        # very simplified code
        return session.client('ssm').get_parameter(Name=ssm_param_api_key)["Parameter"]["Value"]

    return _api_key_loader

def sdk_client_factory(load_api_key: api_key_loader):
    def _sdk_client_factory(client_type):
        # very simplified code
        api_key = load_api_key()
        if client_type == 'User':
            return some_sdk.user.Client(api_key)
        raise ValueError(f"unknown client type: {client_type}")
    return sdk_client_factory

def sdk_data_transfer(client_factory: sdk_client_factory, formatter: some_sdk.Formatter, context: object):
    def _sdk_data_transfer(payload: dict)
        # very simplified code
        client = client_factory(payload['type'])
        client.publish(formatter.format(payload['body']), handler=context.invoked_function_arn)
    return _sdk_data_transfer

# ...

Configure and run in file:

import os
from . import ioc

init_error = None

    container = ioc.build_container(
        format_type=os.environ.get("FORMAT_TYPE", "YAML"),
except Exception as ex:
    init_error = ex

def lambda_handler(event: dict, context: object):
    if init_error:
        raise init_error


    return container.event_handler(event)

In this example container would be compiled once per Lambda Function life cycle, while lambda_handler processes incoming requests sequentially, until it dies after idle timeout.

Project details

Release history Release notifications | RSS feed

This version


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pydioc-0.1.tar.gz (4.1 kB view hashes)

Uploaded source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page