interview.py

"""
Provide a module which parses URLs into useful features (protocol, hostname, path, etc).
"""

import re

class url:
  def parse(self, feature, url):
      """
      a feature may be any of "protocol", "hostname",
      or "path", and looks like this: 
          {"protocol": "ftp"}
          {"hostname": "google"}
          {"path": "news/google"}
          
      The hostname is fully qualified domain name,
      including both domain and tld, e.g.: google.com

      An example call to this function:
          url_parse({"protocol": "http"}, "http://www.google.com")
      """
      
      # check for valid feature
      valid_feature = [f for f in feature if f in ["protocol", "hostname", "path"]]
      if valid_feature == []:
          try:
              raise NameError("Please use 'protocol', 'hostname', or 'path' when listing the feature you want to parse")
          except NameError as e:
              return e

      protocol = re.findall('^(\w+://)', url)

      if protocol != []:
          self.is_fqdn(True, url)
          hostname = re.findall('^\w+\:\/\/(\w+)', url)
      else:
          hostname = re.findall('^(\w+\.\w+)', url)
          path = re.findall('^\w+\:\/\/\w+.*(/*)', url)
      
      if "protocol" in feature:
          if protocol != [] and \
             (protocol[0] == feature["protocol"] or \
              protocol[0] == feature["protocol"] + str("://")):
                 return (True, protocol, url) 
          else:
              return (False, feature, protocol)

      if "hostname" in feature and \
         hostname != [] and \
         hostname == feature["hostname"]:
             return (True, hostname)

      if "path" in feature and path != []:
          return (True, path)

  def is_fqdn(self, has_protocol, url):
      if has_protocol:
          # check for lack of tld
          if re.findall('^\w+\:\/\/(\w+\.\w+)', url) == []:
            try:
                raise NameError("url doesn't include a tld.")
            except NameError as e:
                return e