#!/usr/bin/python

import sys, re

r_comment = re.compile(r'<!--(?:[^-]+|-(?!-))+-->')
r_tag = re.compile(r'<[^>]+>')
r_em = re.compile(r'(?ims)(?:<em>.*?</em>)|(?:<i>.*?</i>)')
r_paren = re.compile(r'\([^\)]+\)')
r_tr = re.compile(r'(?ims)<tr[^>]*>.*?</tr>')
r_td = re.compile(r'(?ims)<td[^>]+>.*?</td>')
r_ws = re.compile(r'[ \t\r\n]+')

def stripComments(s): 
   return r_comment.sub(' ', s)

def resolveEntities(s): 
   replacements = {'&lt;': '<', '&gt;': '>', '&nbsp;': ''}
   for (before, after) in replacements.iteritems(): 
      s = s.replace(before, after)
   return s.replace('&amp;', '&')

def getWord(s): 
   s = r_tag.sub(' ', s)
   s = r_ws.sub(' ', s)
   return s.strip(' \t\r\n')

def getTranslations(s): 
   s = r_em.sub(' ', s)
   s = r_tag.sub(' ', s)
   s = r_paren.sub(' ', s)
   s = r_ws.sub(' ', s)
   s = s.strip(' \t\r\n')
   if s: return s.split(', ')
   else: return []

def main(): 
   fn = sys.argv[1]
   # get all the tr blocks
   f = open(fn)
   data = f.read()
   f.close()
   for tr in r_tr.findall(data): 
      tr = stripComments(tr)
      try: dt, dd = tuple(r_td.findall(tr))
      except ValueError: continue
      dt, dd = resolveEntities(dt), resolveEntities(dd)
      word = getWord(dt)
      translations = getTranslations(dd)
      if word and translations: 
         print '%s: %s' % (word, ', '.join(translations[:2]))

if __name__=="__main__": 
   main()