# coding: utf-8 require 'nkf' require 'romaji/string_extension' require 'romaji/constants' require 'romaji/version' module Romaji def self.romaji2kana(text, options = {}) text = hira2kata(self.normalize(text)) pos = 0 k = nil kana = '' chars = text.split(//u) while true # ン if chars[pos] == 'm' && ['p', 'b', 'm'].include?(chars[pos + 1]) kana += 'ン' pos += 1 next end # ッ if chars[pos] == chars[pos + 1] && !['a', 'i', 'u', 'e', 'o', 'n'].include?(chars[pos]) && chars[pos] =~ /[a-z]/ kana += 'ッ' pos += 1 next end ROMAJI_MAX_LENGTH.downto(1) do |t| substr = chars.slice(pos, t).join k = ROMAJI2KANA[substr] if k kana += k pos += t break end end unless k kana += chars.slice(pos, 1).join pos += 1 end break if pos >= chars.size end kana_type = options[:kana_type] || :katakana kana = kata2hira(kana) if :hiragana == kana_type.to_sym return kana end def self.kana2romaji(text) text = hira2kata(self.normalize(text)) pos = 0 k = nil romaji = '' chars = text.split(//u) while true # ン if chars[pos] == 'ン' next_char_romaji = KANA2ROMAJI[chars[pos + 1]] if next_char_romaji && ['p', 'b', 'm'].include?(next_char_romaji[0].slice(0,1)) romaji += 'm' else romaji += 'n' end pos += 1 next end # ッ if chars[pos] == 'ッ' next_char_romaji = KANA2ROMAJI[chars[pos + 1]] if ['a', 'i', 'u', 'e', 'o', 'n', nil].include?(chars[pos + 1]) || next_char_romaji.nil? romaji += 'xtsu' else romaji += (next_char_romaji[0].slice(0,1)) end pos += 1 next end ROMAJI_MAX_LENGTH.downto(1) do |t| substr = chars.slice(pos, t).join k = KANA2ROMAJI[substr] if k romaji += k[0] pos += t break end end unless k romaji += chars.slice(pos, 1).join pos += 1 end break if pos >= chars.size end romaji end def self.hira2kata(text) NKF.nkf("--katakana -Ww", text) end def self.kata2hira(text) NKF.nkf("--hiragana -Ww", text) end def self.normalize(text) NKF.nkf('-mZ0Wwh0', text).downcase end end