#!/usr/bin/python
"""
ircsoundex.py - A quick IRC Soundex implementation
Usage: ./ircsoundex.py <nickname>
Author: Sean B. Palmer, inamidst.com
License:
   "I really couldn't care less what anyone does with this, but 
   claiming that they wrote it would be just sad and in asserting 
   my right of intellectual property I assert my right to publicly 
   ridicule anyone who does."
   - http://www.hackcraft.net/rssvalid.xsl
"""

import sys, re

# People often use |status modifiers, so we only use the first set of 
# [\w_-]+ characters, stripping any trailing [^\w_-]+ characters.
r_soundex = re.compile('^(?:[^A-Za-z0-9_-]+)?([A-Za-z0-9_-]+)')

# Cf. http://www.archives.gov/research_room/genealogy/census/soundex.html
# This implementation is too greedy on the side-by-side rules, and doesn't 
# apply it for the first character. Consonantal separators aren't checked 
# either.
soundexMap = { 
   'b': 1, 'f': 1, 'p': 1, 'v': 1, 
   'c': 2, 'g': 2, 'j': 2, 'k': 2, 'q': 2, 's': 2, 'x': 2, 'z': 2, 
   'd': 3, 't': 3, 
   'l': 4, 
   'm': 5, 'n': 5, 
   'r': 6
}

def ircSoundex(nick): 
   m = r_soundex.match(nick.lower())
   if not m: return '0-000'
   shortnick = m.group(1)
   result = shortnick[0].upper() + '-'
   for char in shortnick[1:]: 
      if soundexMap.has_key(char): 
         code = str(soundexMap[char])
         if not result.endswith(code): 
            result += code
      if len(result) > 4: break
   if len(result) < 5: 
      result += '0' * (5 - len(result))
   return result

def main(args=None): 
   if args is None: 
      args = sys.argv[1:]

   if args: 
      for arg in args: 
         print arg, ircSoundex(arg)
   else: print __doc__

if __name__=="__main__": 
   main()