master	coverage	PyPI	Python	Licence

SSMEnv allows you to read parameters from AWS Parameter Store and operate on results as on dictionary.

Only requirement is to have boto3 installed.

pip install ssmenv

Let's assume we have two parameters token and url under /service/my-service namespace. Reading both parameters is as simple as initialising class object.

from ssmenv import SSMEnv

params = SSMEnv("/service/my-service")

Done! Now we can access /service/my-service/token and /service/my-service/url in params variable!

Now params can be accesses as python dict type.

As you know by now, instance of SSMEnv can be accessed as any dict in python which means you can do things like:

from ssmenv import SSMEnv

params = SSMEnv("/service/my-service")

# 1. Access value directly
token = params["SERVICE_MY_SERVICE_TOKEN"]

# 2. Get list of all loaded parameter's names
list(params.keys())

# 3. Get list of all loaded parameter's values
list(params.values())

# and so on...

In real world most often you will access parameters from different namespaces, you can easily do that with SSMEnv by passing tuple

from ssmenv import SSMEnv

params = SSMEnv("/service/my-service", "/resource/mysql")

Now params will have all parameters from both /service/my-service and /resource/mysql.

If you use AWS lambda, you might find handy ssmenv decorator. It behaves same as if you would initialise SSMEnv by hand, but additionally it injects instance of SSMEnv into context.params attribute.

from ssmenv import ssmenv

@ssmenv("/service/my-service")
def handler(event, context):
    return context.params

You can hide use of SSMEnv by using os.environ dict.

import os
from ssmenv import SSMEnv

os.environ = {**os.environ, **SSMEnv("/service/my-service")}

Accessing your parameters through whole namespaces can sometimes be not convenient especially if you have very long names.

Hence why you can use prefixes parameter, to make your code cleaner.

from ssmenv import SSMEnv

params = SSMEnv("/service/my-service", prefixes=("/service/my-service",))
params["TOKEN"]

You might want to run your application without AWS, e.g. through docker on your local machine and mock parameters. For that you can use no_aws_default attribute.

import os
from ssmenv import SSMEnv

os.environ["SERVICE_MY_SERVICE_TOKEN"] = "mocked-token" # that might be set in docker-compose

params = SSMEnv("/service/my-service", no_aws_default=os.environ)

You can pass your own boto3 client as well.

import boto3
from ssmenv import SSMEnv

ssm_client = boto3.client("ssm")
params = SSMEnv("/service/my-service", ssm_client=ssm_client)