Match url patterns as regular expressions with named groups.

gistfile1.py

import re


class NoURLMatchException(Exception):
    pass


class URLParser(object):

    def __init__(self, regex_definition_handle):
        """
        Accepts an iterator of strings that define potential regex URLs.

        In django, a way to produce these URLs is as follows:
        >> from django.core import urlresolvers
        >> resolver = urlresolvers.get_resolver(None)
        >> url_re_list = ['^/' + url[1] for url in resolver.reverse_dict.itervalues()]
        """

        # Match names fields in python.
        # http://regex101.com/r/sJ1wY0
        self.re_replace_fields = re.compile(r'\(\?P\<[^\>]+\>[^)]+\)')
        # If there is a non-capturing wrapper, it will be eliminated here.
        self.re_replace_non_capture = re.compile(r'\(\?\:#/?\)\?')
        self.re_replace_wrapped_capture = re.compile(r'\(#\?\/\?\)?\??')

        self.re_list = [re.compile(s.strip()) for s in regex_definition_handle]

    def parse_url(self, path):
        """
        Try regular expressions until one matches.

        Return tuple containing all named groups and a URL with fields stripped.
        """
        for regex in self.re_list:
            match = regex.search(path)
            if match:
                pattern = self.re_replace_fields.sub('#', regex.pattern)
                pattern = self.re_replace_non_capture.sub('#', pattern)
                pattern = self.re_replace_wrapped_capture.sub('#', pattern)
                return match.groupdict(), pattern
        raise NoURLMatchException()