hulu/fry

Library for maintaining tracked requests sessions

fry

All right! I'm a delivery boy!

Description

Fry is an extension of the python requests library built to better manage service-level http interactions.

• Manages dependency settings within the Session object, lessening the amount of overhead boilerplate code
  • Dependency settings include retry count and timeout; split per host
• Given a statsd object, will also record and report stats per request; split per endpoint

StatsD libraries currently supported:

• datadog

Usage

import fry

adapter_settings = {
    'http://www.example.com': {
        'retry': {
            'total': 3,
            'read': 3
        },
        'adapter': {
            'pool_maxsize': 4,
        },
        'adapter_config': {
            'timeout': 0.5
        }
    }
}

fsession = fry.FrySession(stats_client=DogStatsd('example'), adapter_settings=adapter_settings)

signature = 'hostname.endpoint'
request_params = {'test': 'value'}

response = fsession.make_request('GET', 'http://www.example.com', signature, params=request_params)

Development

Update the version when updating the library in: fry/__init__.py

Build the library locally: python setup.py sdist

Change Log

0.2.0

• Support for datadog statsd api