rpc4django 0.1.2

Handles JSONRPC and XMLRPC requests easily with Django

Installation

Prerequisites

RPC4Django has been tested on Mac OS, Linux and Windows.

• Python 2.4 - 2.6

• Django 1.0+

• Docutils (optional)

Source Install Method

This is the preferred method for installing RPC4Django.

$> tar xvfz rpc4django-x.y.z.tar.gz
$> cd rpc4django-x.y.z
$> python setup.py install

Easy Install Method

RPC4Django can be installed this way if setuptools is installed and easy_install is in your path.

$> easy_install rpc4django

No Installation Method

This method installs RPC4Django only for one specific django project but does not require any special system access.

$> tar xvfz rpc4django-x.y.z.tar.gz
$> cd rpc4django-x.y.z
$> cp -r rpc4django YOUR_DJANGO_PROJECT_DIRECTORY

---

Configuration

1. First, you need to add new url pattern to your root urls.py file. This file should be the one pointed to by ROOT_URLCONF in settings.py. You can replace r'^RPC2$' with anything you like.

   # urls.py
   #...
   urlpatterns = patterns('',
       # if installed via no install method
       #(r'^RPC2$', 'YOURPROJECT.rpc4django.views.serve_rpc_request'),
   
       # if installed via source or easy_install
       (r'^RPC2$', 'rpc4django.views.serve_rpc_request'),
   )

2. Second, add RPC4Django to the list of installed applications in your settings.py.

   # settings.py
   #... INSTALLED_APPS = (
       # if installed via no install
       #'YOURPROJECT.rpc4django',
   
       # if installed via source or easy_install
       'rpc4django',
   )

3. Lastly, you need to let RPC4Django know which methods to make available. This is done with the decorator @rpcmethod. RPC4Django imports all the apps in INSTALLED_APPS and makes any methods importable via __init__.py with the @rpcmethod decorator available as RPC methods. You can always write your RPC methods in another module and simply import it in __init__.py.

   # testapp/__init__.py
   from rpc4django import rpcmethod
   
   # This imports another method to be made available as an RPC method
   # This method should also have the @rpcmethod decorator
   # from mymodule import myrpcmethod
   
   # The doc string supports reST if docutils is installed
   @rpcmethod(name='mynamespace.add', signature=['int', 'int', 'int'])
   def add(a, b):
       '''Adds two numbers together
       >>> add(1, 2)
       3
       '''
   
       return a+b

   The decorator @rpcmethod accepts two optional parameters:

   • name - is the name under which the method will be made externally available. It defaults to the method’s real name.

   • signature - is the signature of the method and will be returned by calls to the XMLRPC introspection method system.methodSignature. It is of the form: [return_value, arg1, arg2, arg3, ...]. All of the types should be XMLRPC types (eg. struct, int, array, etc. – see the XMLRPC spec for details). Since python is relatively loosely typed, all types are set to object by default.