#!/usr/bin/env python3

#oneoffs = set(["D", "CD", "L", "XL", "V", "IV"])
roman_numerals  = {"M" : 1000, "CM" : 900,
                   "D" : 500,  "CD" : 400,
                   "C" : 100,  "XC" : 90,
                   "L" : 50,   "XL" : 40,
                   "X" : 10,   "IX" : 9,
                   "V" : 5,    "IV" : 4,
                   "I" : 1}

def parse(path):
    with open(path) as f:
        return f.read().split("\n")

def roman2int(num):
    rn = roman_numerals
    r = 0
    for dig, i in zip(num, num[1:]+"I"):
        r += rn[dig] * (-1 if rn[dig] < rn[i] else 1)
    return r

def int2roman(num):
    rnum = str()
    pairs = list(roman_numerals.items())
    pairs.sort(key = lambda x: -x[1])
    for dig, val in pairs:
        while val <= num:
            rnum += dig
            num  -= val
            #if rnum in oneoffs:
                #break
    return rnum

def main(path):
    nums = parse(path)
    min_nums = map(int2roman, map(roman2int, nums))
    return sum(map(len, nums)) - sum(map(len, min_nums))

if __name__ == '__main__':
    print(main("data/089_roman_numerals.txt"))