#!/usr/bin/env python
# -*- coding: utf-8 -*-

import re

apple = {}
for line in open('RomajiRule_Default_original.txt', 'r'):
    line = line.strip()
    if len(line) > 0 and line[0] == '#':
        continue
    data = line.split('\t')
    if len(data) < 2:
        continue
    # print data
    apple[data[0]] = data

diff = []
for line in open('dvorakjp-edited.txt', 'r'):
    data = line.strip().split('\t')
    if data[0] not in apple:
        diff.append([data[0], ''.join(data[1:]).decode('utf-8')])
        # print data[0] + '\t' + ''.join(data[1:])

def compare(s1, s2):
    for i,j in zip(s1, s2):
        if ord(i) < ord(j):
            return -1
        elif ord(i) > ord(j):
            return 1
    if len(s1) < len(s2):
        return -1
    elif len(s1) > len(s2):
        return 1
    else:
        return 0

def strcmp(x, y):
    n = len(x[1]) - len(y[1])
    if n == 0:
        return compare(x[1], y[1])
    else:
        return n

p = re.compile(r'[,\.;]')
for t in sorted(diff, cmp=strcmp):
    n = len(t[1])
    if n > 1:
        t.append(('0,' * (n - 1) + '%d') % len(t[0]))
    if p.search(t[0]):
        t[0] = '#' + t[0]
        t.append(u'# 無効')
    print '\t'.join(t).encode('utf-8')