Rekt

Rekt is a python module for generically and dynamically generating rest clients at run-time. Rekt tries to be as flexible as possible by being a light and convenient wrapper around the venerable requests module. rekt also supports asynchronous calls by utilizing concurrent.futures.

Rekt Requires:

• Python 3.4+
• Requests
• PyYAML

Why Use Rekt

The impetus for rekt is that the rest pattern for generating clients is fairly standard. The one off libraries for various services have their inconsistencies and they often lack Python3 support.

• It is a light weight code generation wrapper around requests. So at its core it supports what requests support. The exceptions thrown for bad requests will be the same as the exceptions from requests.

• Asynchronous when needed. All calls are generated with a synchronous and asynchronous call handler. Client objects contain a configurable concurrent.futures.ThreadPoolExecutor.

• Thread Safe. Some service specific rest libraries store context within client instance attributes in a way that does not allow for concurrent calls for mysterious reasons. State should be bound to the call, rekt was designed from the beginning to allow for concurrent calls.

• Generic. Rekt generates the code for clients at runtime from yaml configurations. Only the request parameter specifications need to be specified before hand. Rekt utilizes a dict subclass called DynamicObject as both the base Response class and as the object hook when deserializing the json responses in order to capture all of the parameters of the response as dot accessible object. For those that prefer it, since DynamicObject is also a dict subclass all of the normal dictionary methods are also available.

Example Using Rekt:

  >>> from rekt import utils
  >>> from rekt import load_service

  >>> conf = utils.load_builtin_config('googleplaces')
  >>> googleplaces = load_service(conf)

  >>> client = googleplaces.Client()
  >>> response = client.get_places(key=YOUR_API_KEY, location='47.6097,-122.3331', radius=1000)
  >>> print(response.keys())
  dict_keys(['html_attributions', 'results', 'next_page_token', 'status'])
  >>> my_place = response.results[0] # Note dot access, like normal objects
  >>> same_places = response['results'] # using keys works too!
  >>> print(response.__class__.__name__)
  GetPlacesResponse

  >>> import concurrent.futures
  >>> f = client.async_get_details(key=YOUR_API_KEY, placeid=my_place.place_id)
  <Future at 0x7f6880a52978 state=running>
  >>> f = next(concurrent.futures.as_completed([f]))
  >>> print(f.result().keys())
  dict_keys(['result', 'html_attributions', 'status'])

Get Rekt!