About

A library for simple URI handling. It has no external dependancies, and works with Python 2.3+.

Currently contains only an implementation of URI-Templates (http://bitworking.org/projects/URI-Templates/) -- but see the TODO_ below.

Some bits are inspired by or based on:

* Joe Gregorio's example implementation
  (http://code.google.com/p/uri-templates/)

* Addressable (http://addressable.rubyforge.org/)

Examples

Simple usage::

>>> import uri

>>> args = {'foo': 'it worked'}
>>> uri.expand_template("http://example.com/{foo}", args)
'http://example.com/it%20worked'

>>> args = {'a':'foo', 'b':'bar', 'a_b':'baz'}
>>> uri.expand_template("http://example.org/{a}{b}/{a_b}", args)
'http://example.org/foobar/baz'

You can also use keyword arguments for a more pythonic style::

>>> uri.expand_template("http://example.org/?q={a}", a="foo")
'http://example.org/?q=foo'

Contributing

I use GitHub; the canonical repository lives at http://github.com/jacobian/uri/tree. Please feel free to send pull requests or patches to [email protected]

TODO

Over time, I'd like to add the following features to this library:

* ``uri.extract(template, uri)``: extract a dict of info given a template
  and a URI.
  
* ``uri.parse(uri)``: thin wrapper around ``urlparse.urlparse()`` that
  returns a class (so you can do e.g. ``some_uri.fragment`` or whathaveyou.)
  This URI class should have expand/extract methods.
  
* Add methods/functions to do sane URI joining -- ``urlparse.urljoin`` 
  *never* does what I think it's going to do, especially when faced
  with relative URIs.