plumb 1.2.5

Connect systems via many brokers such as Kafka, AWS SQS, RabbitMQ and more.

Author:

Spectro

Date:

2018-10-01

Version:

1.2.5

License:

MIT License

Provides higher level interfaces to work with data producers and consumers. plumb supports five backends: RabbitMQ, Redis, Amazon SQS and SNS services, Kafka, and a memory implementation (designed to be a testing artifact).

The focus is on python 3 and boto3 (for AWS support).

The API

plumb uses the concepts of Source and Sink to abstract the backend. Data is transfered encoded in JSON, and by default compressed using zlib.

To fetch raw packages use a “Package Source” object, which handles a backend connection and its configuration and exposes a “get()” method. Similarly, a “Package Sink” object exposes a “put(pkg)” method.

You can either create the Sources and Sinks directly by instantiating them from their packages or use the convenience factory functions exposed in the plumb package. This functions receive a JSON with the configuration for the Sources or Sinks and returns a list of such configured devices.

Tests

The library is provided with unit tests and integration tests for Redis and AWS. To run the unit tests:

python setup.py test

or using nose:

nosetests tests/unit

The integration tests can be run using nose:

nosetests tests/integration

Keep in mind that boto3 will fetch your AWS credentials. It currently tries the environment variables AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY, then tries the ~/.aws directory. For details, see Boto3 Credentials Configuration.