Regexp-based conversion from Roman numerals to decimal

roman.py

#!/bin/python3
__author__ = 'ncoop'

from re import search, VERBOSE, IGNORECASE

roman_pattern = '''
    ^
    # M: thousands
    ((?P<m1000>M)|(?P<m2000>MM)|(?P<m3000>MMM))?

    # C,D: hundreds
    ((?P<c100>C)|(?P<c200>CC)|(?P<c300>CCC)|(?P<c400>CD)|(?P<c500>D)|
    (?P<c600>DC)|(?P<c700>DCC)|(?P<c800>DCCC)|(?P<c900>CM))?

    # X,L: tens
    ((?P<x10>X)|(?P<x20>XX)|(?P<x30>XXX)|(?P<x40>XL)|(?P<x50>L)|
    (?P<x60>LX)|(?P<x70>LXX)|(?P<x80>LXXX)|(?P<x90>XC))?

    # I,V: ones
    ((?P<i1>I)|(?P<i2>II)|(?P<i3>III)|(?P<i4>IV)|(?P<i5>V)|
    (?P<i6>VI)|(?P<i7>VII)|(?P<i8>VIII)|(?P<i9>IX))?
    $
'''


def rom2dec(roman_string, ignore_case=False) -> int:
    """
    Convert a string of Roman numerals to its int equivalent.
    Minimum 0 for empty string, maximum 3999 for MMMCMXCIX
    Returns None for invalid string.

    :rtype : int
    :param roman_string: the string to be converted
    :param ignore_case: True if case-insensitivity is required
    :return: the int value of the string
    """
    match = search(roman_pattern, roman_string, VERBOSE | (IGNORECASE if ignore_case else 0))
    if not match:
        return None
    match_dict = match.groupdict()
    # filter out unmatched groups, interpret the key as an int value, and sum the results
    return sum([int(k[1:]) for k in match_dict if match_dict[k]])

if __name__ == '__main__':
    for r in ['', 'I', 'iv', 'VIII', 'XXI', 'LXXX', 'XC', 'MMM', 'MMI', 'MMCMIII', 'mmmCMXCIX']:
        print(r + ' => ' + str(rom2dec(r, True)))