Origin
======

This package was originally written as a sort of proof-of-concept github gist
on December 25, 2011 by Josiah Carlson. This updated package is copyright
2012, and is released under the GNU LGPL v2.1 license.

Description
===========

This package offers the ability to connect to http servers via the asyncore
and asynchat modules, in an attempt to download files, or otherwise integrate
connecting to other services from within asyncore-derived services.


Download Files
==============

The included async_http.get module, which can be used on the command line:
    $ python -m async_http.get <list of urls>

... will fetch the provided urls in parallel. If you check the source for the
get.py module, you can see how you can override the http_body(), http_done(),
and http_close() method to (for example) fetch content remotely and possibly
redirect it somewhere else.

By properly subclassing from DownloadFile you can serve remote files, similar
to the requested feature here:
http://code.google.com/p/pyftpdlib/issues/detail?id=197


Make OAuth Requests
===================

The included async_http.oauth module, which can be used on the command line:
    $ python -m async_http.oauth ckey,csecret [tkey,tsecret] <url>

... will perform an Oauth 1.0a request against the provided url with the given
client key and secret, with optional token key and secret (your url will
determine whether you need a some sort of access or request token).


Features
========

* HTTP/HTTPS
* redirect support
* chunked transfer encoding
* timeouts
* transparent gzip handling
  (derived from http://effbot.org/zone/consumer-gzip.htm)
* OAuth 1.0a client support (no other 3rd party libraries required)


API
===

If you want to write your own handlers for events, subclass from
_http.AsyncHTTPRequest, and override one or more of the following methods::

    def http_setup(self):
        '''
        Called just before the connection is set up. You can manipulate
        headers, request body, etc.
        '''

    def http_response(self):
        '''
        Called after the response is read. You can handle redirects, perform
        additional logging, start a reply in a proxy, etc.
        '''

    def http_body(self):
        '''
        Called at the end of every chunk with chunked transfer encoding, and
        any time data is read for the body otherwise.
        '''

    def http_chunk(self):
        '''
        Called after every chunk with the chunked transfer encoding,
        immediately after the body callback.
        '''

    def http_done(self):
        '''
        Called when the body has finshed being transferred. This will not be
        called when there is an error.
        '''

    def http_close(self):
        '''
        Called when the connection has been closed.
        '''