Django-esque URL Resolution in Nevow

---

So, from the previous post on object publishing in Nevow, we had the following resources:

• http://localhost:8080/
• http://localhost:8080/m ystuff
• http://localhost:8080/yourstuff

We added child_* attributes and methods to the "root" class, thus defining the "mystuff" and "yourstuff" resources. But what if we want to do it like Django does it? In other words, how do we map paths in the URL to resources via regular expressions?

It's actually fairly straight-forward. All we need to do is the following:

1. Create a tuple of patterns.
2. Override the locateChild() method in our "root" class.

Working with and extending the examples from the previous post, we might create a patterns tuple like this:

urlPatterns = (
 (r'/mystuff(.*)', MyStuffResource),
 (r'/yourstuff/blog(.)', BlogResource),
 (r'/yourstuff(.)', YourStuffResource),
)

Then, in SiteRoot, we could do something like the following:

 def locateChild(self, context, segments):
 path = '/'.join(('',) + segments)
 for regex, resource in urlPatterns:
 match = re.match(regex, path)
 if match:
 newPath = match.groups()[0]
 r = resource()
 if newPath in ['', '/']:
 return r, ()
 else:
 newSegments = newPath.split('/')[1:]
 return r.locateChild(context, newSegments)
 return super(SiteRoot, self).locateChild(context, segments)

What we're doing here is interrupting the "natural" flow of Nevow's path processing. If there are more segments once we've found a match, we pass the additional segments on to the child resource's locateChild() method. If not, we have a final destination and can return the resource itself.

Here are some screenshots of this in action:







You can browse the code for this at the following links:

• The example as a single .tac file
• The same example, split up into separate files

Technorati Tags: django, nevow, python, twisted, web