[ruby-list:26400] Re: Ruby/KAKASI bug?

前田です。

At Thu, 7 Dec 2000 00:56:06 +0900,
[e-mail] (Yukihiro Matsumoto) wrote:
> |希望ですけど，kconv でも *いわゆる* 半角カナをサポートできませんか?
> |# Jcode.pm でいう z2h 
> 
> ちゅーか、半角かなが登場すると自動的に全角にしますよ(自動判
> 別はできませんが、それはしょうがない)。そうしたくない場合が
> あれば、nkfを直接使ってください。NKF.nkf("-xSe", str) とか。

Jcode.pm でいう z2h、ということなので全角から半角への変換がしたい
んじゃないですかね?

Jcode.pmのぱくりですけど、こんなんでどうでしょう?
# 文字コードはEUCを仮定


=begin

= hzconv.rb

== module HZconv

=== Module Functions

: h2z(str, keep_dakuten = false)
    Converts X201 kana (Hankazu) to X208 kana (Zenkaku).
    When keep_dakuten is true, it leaves dakuten as is
    (That is, "ka + dakuten" is left as is instead of
    being converted to "ga")

: z2h(str)
    Converts X208 kana (Zenkaku) to X201 kana (Hankazu).

=end

module HZconv
  H2Z = {
    "\x8e\xa1"	=>	"\xa1\xa3",	#。
    "\x8e\xa2"	=>	"\xa1\xd6",	#「
    "\x8e\xa3"	=>	"\xa1\xd7",	#」
    "\x8e\xa4"	=>	"\xa1\xa2",	#、
    "\x8e\xa5"	=>	"\xa1\xa6",	#・
    "\x8e\xa6"	=>	"\xa5\xf2",	#ヲ
    "\x8e\xa7"	=>	"\xa5\xa1",	#ァ
    "\x8e\xa8"	=>	"\xa5\xa3",	#ィ
    "\x8e\xa9"	=>	"\xa5\xa5",	#ゥ
    "\x8e\xaa"	=>	"\xa5\xa7",	#ェ
    "\x8e\xab"	=>	"\xa5\xa9",	#ォ
    "\x8e\xac"	=>	"\xa5\xe3",	#ャ
    "\x8e\xad"	=>	"\xa5\xe5",	#ュ
    "\x8e\xae"	=>	"\xa5\xe7",	#ョ
    "\x8e\xaf"	=>	"\xa5\xc3",	#ッ
    "\x8e\xb0"	=>	"\xa1\xbc",	#ー
    "\x8e\xb1"	=>	"\xa5\xa2",	#ア
    "\x8e\xb2"	=>	"\xa5\xa4",	#イ
    "\x8e\xb3"	=>	"\xa5\xa6",	#ウ
    "\x8e\xb4"	=>	"\xa5\xa8",	#エ
    "\x8e\xb5"	=>	"\xa5\xaa",	#オ
    "\x8e\xb6"	=>	"\xa5\xab",	#カ
    "\x8e\xb7"	=>	"\xa5\xad",	#キ
    "\x8e\xb8"	=>	"\xa5\xaf",	#ク
    "\x8e\xb9"	=>	"\xa5\xb1",	#ケ
    "\x8e\xba"	=>	"\xa5\xb3",	#コ
    "\x8e\xbb"	=>	"\xa5\xb5",	#サ
    "\x8e\xbc"	=>	"\xa5\xb7",	#シ
    "\x8e\xbd"	=>	"\xa5\xb9",	#ス
    "\x8e\xbe"	=>	"\xa5\xbb",	#セ
    "\x8e\xbf"	=>	"\xa5\xbd",	#ソ
    "\x8e\xc0"	=>	"\xa5\xbf",	#タ
    "\x8e\xc1"	=>	"\xa5\xc1",	#チ
    "\x8e\xc2"	=>	"\xa5\xc4",	#ツ
    "\x8e\xc3"	=>	"\xa5\xc6",	#テ
    "\x8e\xc4"	=>	"\xa5\xc8",	#ト
    "\x8e\xc5"	=>	"\xa5\xca",	#ナ
    "\x8e\xc6"	=>	"\xa5\xcb",	#ニ
    "\x8e\xc7"	=>	"\xa5\xcc",	#ヌ
    "\x8e\xc8"	=>	"\xa5\xcd",	#ネ
    "\x8e\xc9"	=>	"\xa5\xce",	#ノ
    "\x8e\xca"	=>	"\xa5\xcf",	#ハ
    "\x8e\xcb"	=>	"\xa5\xd2",	#ヒ
    "\x8e\xcc"	=>	"\xa5\xd5",	#フ
    "\x8e\xcd"	=>	"\xa5\xd8",	#ヘ
    "\x8e\xce"	=>	"\xa5\xdb",	#ホ
    "\x8e\xcf"	=>	"\xa5\xde",	#マ
    "\x8e\xd0"	=>	"\xa5\xdf",	#ミ
    "\x8e\xd1"	=>	"\xa5\xe0",	#ム
    "\x8e\xd2"	=>	"\xa5\xe1",	#メ
    "\x8e\xd3"	=>	"\xa5\xe2",	#モ
    "\x8e\xd4"	=>	"\xa5\xe4",	#ヤ
    "\x8e\xd5"	=>	"\xa5\xe6",	#ユ
    "\x8e\xd6"	=>	"\xa5\xe8",	#ヨ
    "\x8e\xd7"	=>	"\xa5\xe9",	#ラ
    "\x8e\xd8"	=>	"\xa5\xea",	#リ
    "\x8e\xd9"	=>	"\xa5\xeb",	#ル
    "\x8e\xda"	=>	"\xa5\xec",	#レ
    "\x8e\xdb"	=>	"\xa5\xed",	#ロ
    "\x8e\xdc"	=>	"\xa5\xef",	#ワ
    "\x8e\xdd"	=>	"\xa5\xf3",	#ン
    "\x8e\xde"	=>	"\xa1\xab",	#゛
    "\x8e\xdf"	=>	"\xa1\xac",	#゜
  }

  D2Z = {
    "\x8e\xb6\x8e\xde"	=>	"\xa5\xac",	#ガ
    "\x8e\xb7\x8e\xde"	=>	"\xa5\xae",	#ギ
    "\x8e\xb8\x8e\xde"	=>	"\xa5\xb0",	#グ
    "\x8e\xb9\x8e\xde"	=>	"\xa5\xb2",	#ゲ
    "\x8e\xba\x8e\xde"	=>	"\xa5\xb4",	#ゴ
    "\x8e\xbb\x8e\xde"	=>	"\xa5\xb6",	#ザ
    "\x8e\xbc\x8e\xde"	=>	"\xa5\xb8",	#ジ
    "\x8e\xbd\x8e\xde"	=>	"\xa5\xba",	#ズ
    "\x8e\xbe\x8e\xde"	=>	"\xa5\xbc",	#ゼ
    "\x8e\xbf\x8e\xde"	=>	"\xa5\xbe",	#ゾ
    "\x8e\xc0\x8e\xde"	=>	"\xa5\xc0",	#ダ
    "\x8e\xc1\x8e\xde"	=>	"\xa5\xc2",	#ヂ
    "\x8e\xc2\x8e\xde"	=>	"\xa5\xc5",	#ヅ
    "\x8e\xc3\x8e\xde"	=>	"\xa5\xc7",	#デ
    "\x8e\xc4\x8e\xde"	=>	"\xa5\xc9",	#ド
    "\x8e\xca\x8e\xde"	=>	"\xa5\xd0",	#バ
    "\x8e\xcb\x8e\xde"	=>	"\xa5\xd3",	#ビ
    "\x8e\xcc\x8e\xde"	=>	"\xa5\xd6",	#ブ
    "\x8e\xcd\x8e\xde"	=>	"\xa5\xd9",	#ベ
    "\x8e\xce\x8e\xde"	=>	"\xa5\xdc",	#ボ
    "\x8e\xca\x8e\xdf"	=>	"\xa5\xd1",	#パ
    "\x8e\xcb\x8e\xdf"	=>	"\xa5\xd4",	#ピ
    "\x8e\xcc\x8e\xdf"	=>	"\xa5\xd7",	#プ
    "\x8e\xcd\x8e\xdf"	=>	"\xa5\xda",	#ペ
    "\x8e\xce\x8e\xdf"	=>	"\xa5\xdd",	#ポ
    "\x8e\xb3\x8e\xde"	=>      "\xa5\xf4",     #ヴ
  }

  Z2H = H2Z.invert
  Z2D = D2Z.invert

  EUC_C     = '[\xa1-\xfe][\xa1-\xfe]'
  EUC_KANA  = '\x8e[\xa1-\xdf]'

  def h2z(str, keep_dakuten = false)
    if keep_dakuten
      return str.gsub(/(\x8e[\xa1-\xdf])/on) { H2Z[$1] }
    else
      return str.gsub(/(\x8e[\xa1-\xdf](:?\x8e[\xde\xdf])?)/on) {
	D2Z[$1] || H2Z[$1] || 
	  # in case dakuten and handakuten are side-by-side!
	  H2Z[$1[0,2]] + H2Z[$1[2,2]]
      }
    end
  end
  module_function :h2z

  def z2h(str)
    return str.gsub(/([\xa1-\xfe][\xa1-\xfe]|\x8e[\xa1-\xdf])/on) {
      Z2D[$1] || Z2H[$1] || $1
    }
  end
  module_function :z2h
end


-- 
前田 修吾