XPath-friendlier ElementTree namespace helper

By Filip Salo; published on February 21, 2007.

The effbot's namespace helper is nice, but doesn't handle the simple xpath expressions that ElementTree supports. This variant of it does:

class NS:
    def __init__(self, uri):
        self.uri = uri
    def __getattr__(self, tag):
        return self.uri + tag
    def __call__(self, path):
        return "/".join((tag not in ("", ".", "*"))
                        and getattr(self, tag)
                        or tag
                        for tag in path.split("/"))

You can then do stuff like this:

>>> XHTML = NS("{http://www.w3.org/1999/xhtml}")
>>> tree.findall(XHTML(".//li"))
[...]

It helps.