flask-botox

Flask extension that ties boto3 connectors to the application context. To be used with Python 3.7+.

Install

• Via pip:

  $ pip install flask-botox

• Locally with Poetry for development purposes:

  $ git clone https://github.com/jperras/flask-botox.git
  $ cd flask-botox
  $ poetry install

How-to

The main class flask_botox.Boto3 takes a Flask application as its contructor's parameter:

from flask import Flask
from flask_botox import Boto3
app = Flask(__name__)
app.config["BOTOX_SERVICES"] = ["s3", "ses", "sqs"]

botox = Boto3(app)

The application factory pattern for extensions is also valid:

from flask import Flask
from flask_botox import Boto3

botox = Boto3()

app = Flask(__name__)
app.config["BOTOX_SERVICES"] = ["s3", "ses", "sqs"]

botox.init_app(app)

Then boto3's clients and resources will be available as properties within the application context:

>>> with app.app_context():
        print(botox.clients)
        print(botox.resources)
{'s3': <botocore.client.S3 object at 0x..>}
{'s3': s3.ServiceResource()}

Configuration

Flask-botox uses several keys from a Flask configuration objects to customize its behaviour. Any of the AWS_* keys are not required; if they are not specified, then the usual boto3 configuration parameter rules will apply.

• AWS_ACCESS_KEY_ID & AWS_SECRET_ACCESS_KEY : The AWS credentials. Note that it's not a good idea to put your secret access key in a configuration file, but it can be useful for e.g. testing purposes.
• AWS_DEFAULT_REGION : The region, e.g. us-east-1, for the boto3 AWS services.
• AWS_PROFILE : The AWS nanmed profile to use, if one is desired.
• BOTOX_SERVICES : The name of the AWS resources you want to use, e.g. ['sqs', 's3', 'ses'].
• BOTOX_OPTIONAL_PARAMS : Useful if you need to pass additional parameters to the client/resource connections, e.g. a custom endpoint_url for a particular service. The format is a dict where the top-level keys are the name of the services you're using and for each the value is a dict containing to keys args (contains the parameters as tuple) and kwargs (contains the parameters as a dict when they should be passed as keyword arguments).