mangum 0.5.1

ASGI serverless adapters

Mangum

Documentation: https://erm.github.io/mangum/

Mangum is a library for using ASGI applications with FaaS platforms.

Important: This project is under active development and in an experimental/unstable state.

Requirements

Python 3.6+

Installation

$ pip3 install mangum

Dependencies

Currently there are two optional dependencies.

• azure-functions - Required for Azure.
• boto3 - Required for the AWS CLI commands.

This can be installed with:

$ pip3 install mangum[full]

Supported Platforms

Only two platforms are currently supported, but if you'd like to see others, please open an issue.

AWS Lambda / API Gateway

Example

Below is a basic ASGI application example using the AWS run method:

from mangum.platforms.aws.adapter import run_asgi

class App:
    def __init__(self, scope) -> None:
        self.scope = scope

    async def __call__(self, receive, send) -> None:
        message = await receive()
        if message["type"] == "http.request":
            await send(
                {
                    "type": "http.response.start",
                    "status": 200,
                    "headers": [[b"content-type", b"text/plain"]],
                }
            )
            await send({"type": "http.response.body", "body": b"Hello, world!"})

def lambda_handler(event, context):
    return run_asgi(App, event, context)

Mangum CLI (experimental)

Experimental AWS packaging/deployment support to generate, package, and deploy an application. This requires installation of the optional dependencies for AWS:

$ pip install mangum[full]

It also requires the AWS CLI and AWS credentials.

The available commands are briefly outlined below, but there is also a quickstart guide here:

• mangum init - Create a new application and deployment configuration.

• mangum package - Package the local project to prepare for deployment.

• mangum deploy - Deploy the packaged application to AWS.

• mangum tail - Tail the last 10 minutes of CloudWatch for the function.

• mangum describe - Retrieve the API endpoints for the function.

Azure Functions

Example

The same example application as above may be used with the Azure run method:

from mangum.platforms.azure.adapter import run_asgi


class App:
    def __init__(self, scope) -> None:
        self.scope = scope

    async def __call__(self, receive, send) -> None:
        message = await receive()
        if message["type"] == "http.request":
            await send(
                {
                    "type": "http.response.start",
                    "status": 200,
                    "headers": [[b"content-type", b"text/plain"]],
                }
            )
            await send({"type": "http.response.body", "body": b"Hello, world!"})

def lambda_handler(req):
    return run_asgi(App, req)

The command-line tools for Azure Functions can do pretty much everything you need. A basic quickstart guide for using it with Mangum is outlined here.