#!/usr/bin/env python

# Rank is based on: 
# a) recency
# b) frequency
# c) quality (CQ)

import sys, re, operator, itertools
# from decimal import Decimal as dec

r_word = re.compile(r"[A-Za-z']+")
r_comment = re.compile(r'\[[^\]]+\]')

def parse(s): 
   s = r_comment.sub(' ', s)
   return [word.lower().strip("'") for word in r_word.findall(s)]

def main(): 
   words = {}
   f = open(sys.argv[1])
   for line in f: 
      line = line.strip(' \t\r\n')
      if line.count('\t') != 3: continue

      num, kal, eng, rat = line.split('\t')
      num, rat = float(num), float(rat)
      for word in parse(kal): 
         if words.has_key(word): 
            freq, qual, rec = words[word]
            words[word] = (freq + 1, qual + [rat], num)
         else: words[word] = (1, [rat], num)
   f.close()

   result = {}
   for word, (freq, qual, rec) in words.iteritems(): 
      rating = (sum(qual) / len(qual)) * ((freq / 20) + 0.25)
      if rec < 750: rating = rating * 0.9
      result[word] = round(rating, 1)

   results = []
   ri = sorted(result.iteritems(), key=operator.itemgetter(1))
   for k, g in itertools.groupby(ri, key=operator.itemgetter(1)): 
      results.append(('%s: ' % k, ', '.join(map(operator.itemgetter(0), g))))
   for p, q in reversed(results): 
      print p, q


if __name__ == '__main__': 
   main()