#!/usr/bin/env python # -*- coding: utf-8 -*- import re regex = u'^[ぁ-ゞ・ー]+$' # ひらがな+・+ーだけ p = re.compile(regex) for line in open('ruby.txt'): file, kanji, yomi = line.strip().split('\t') if p.match(yomi.decode('utf-8')): print "%s\t%s\t%s" % (file, kanji, yomi)