require 'nkf'

module JapanShippingCSV
  module Encoding
    # Eg. HALFWIDTH + ROMAJI + NUMBERS can be represented by:
    # valid_encodings = Encoding::HALFWIDTH | Encoding::ROMAJI | Encoding::NUMBERS
    # and to figure out whether a flag is set:
    # romaji_allowed? = valid_encodings & Encoding::ROMAJI
    # 0 = false, 1 = true (which is not true in ruby)
    HALFWIDTH = 1 # Hankaku
    FULLWIDTH = 2 # Zenkaku
    ROMAJI = 4
    NUMBERS = 8
    KATAKANA = 16
    HYPHEN = 32

    module CodePoints
      # Character set lower/upper bound definitions
      ## Bounds for Hiragana
      ZENKAKU_HIRAGANA_FIRST = 'ぁ' # U+3041
      ZENKAKU_HIRAGANA_LAST_FOR_CONVERT = 'ん' # U+3093
      ZENKAKU_HIRAGANA_LAST = 'ゖ' # U+3096

      ## Bounds for Katakana
      HANKAKU_KATAKANA_FIRST  = 'ｦ'              # U+FF66
      HANKAKU_KATAKANA_LAST   = 'ﾝ'              # U+FF9D

      ZENKAKU_KATAKANA_FIRST  = 'ァ' # U+30A1
      ZENKAKU_KATAKANA_LAST_FOR_CONVERT = 'ン' # U+30F3
      ZENKAKU_KATAKANA_LAST = 'ヺ'             # U+30FA

      ## Bounds for Punctuation (kutoten)
      HANKAKU_PUNCTUATION_FIRST   = '｡'          # U+FF61
      HANKAKU_PUNCTUATION_LAST    = 'ﾟ'          # U+FF9F
      HANKAKU_PUNCTUATION_ONBIKI  = 'ｰ' # U+FF70

      ZENKAKU_PUNCTUATION_FIRST     = '、'       # U+3001
      ZENKAKU_PUNCTUATION_LAST      = '〜'       # U+301C
      ZENKAKU_PUNCTUATION_HG_FIRST  = '゛'       # U+309B
      ZENKAKU_PUNCTUATION_HG_LAST   = 'ゞ'       # U+309E
      ZENKAKU_PUNCTUATION_KK_FIRST  = '・'       # U+30FB
      ZENKAKU_PUNCTUATION_KK_LAST   = 'ヾ'       # U+30FE

      ## Bounds for Numeric
      HANKAKU_NUMBER_FIRST  = '0'                # U+0030
      HANKAKU_NUMBER_LAST   = '9'                # U+0039

      ZENKAKU_NUMBER_FIRST  = '０'               # U+FF10
      ZENKAKU_NUMBER_LAST   = '９'               # U+FF19

      ## Bounds for Alphabetic
      HANKAKU_LETTER_UPPER_FIRST  = 'A'          # U+0041
      HANKAKU_LETTER_UPPER_LAST   = 'Z'          # U+005A
      HANKAKU_LETTER_LOWER_FIRST  = 'a'          # U+0061
      HANKAKU_LETTER_LOWER_LAST   = 'z'          # U+007A

      ZENKAKU_LETTER_UPPER_FIRST  = 'Ａ'         # U+FF21
      ZENKAKU_LETTER_UPPER_LAST   = 'Ｚ'         # U+FF3A
      ZENKAKU_LETTER_LOWER_FIRST  = 'ａ'         # U+FF41
      ZENKAKU_LETTER_LOWER_LAST   = 'ｚ'         # U+FF5A

      # Bounds for All Alphanumeric and Symbol ASCII
      HANKAKU_SPACE = ' ' # U+0020
      HANKAKU_ASCII_FIRST  = '!'                 # U+0021
      HANKAKU_ASCII_LAST   = '~'                 # U+007E

      ZENKAKU_SPACE = '　' # U+3000
      ZENKAKU_ASCII_FIRST  = '！'                # U+FF01
      ZENKAKU_ASCII_LAST   = '～'                # U+FF5E
    end

    class << self
      def reencode(value, encoding_flags)
        return value if value.nil?

        result = value
        # if half-width not allowed
        half_width_allowed = (encoding_flags & Encoding::HALFWIDTH) == Encoding::HALFWIDTH
        if half_width_allowed
          result = NKF.nkf('-X -w', value).tr('０-９ａ-ｚＡ-Ｚ　－', '0-9a-zA-Z -')
        end
        # if full-width not allowed
        fullwidth_allowed = (encoding_flags & Encoding::FULLWIDTH) == Encoding::FULLWIDTH
        if fullwidth_allowed
          result = NKF.nkf('-X -w', value).tr('0-9a-zA-Z -', '０-９ａ-ｚＡ-Ｚ　－').tr('＇', '’')
        end
        result
      end

      def is_romaji(value)
        return false if value.empty?

        value.split("").each do |c|
          # return !(::kana::isHankakuLetter(u) || ::kana::isZenkakuLetter(u));
          is_hankaku_upper = c >= CodePoints::HANKAKU_LETTER_UPPER_FIRST && c <= CodePoints::HANKAKU_LETTER_UPPER_LAST
          is_hankaku_lower = c >= CodePoints::HANKAKU_LETTER_LOWER_FIRST && c <= CodePoints::HANKAKU_LETTER_LOWER_LAST
          is_zenkaku_upper = c >= CodePoints::ZENKAKU_LETTER_UPPER_FIRST && c <= CodePoints::ZENKAKU_LETTER_UPPER_LAST
          is_zenkaku_lower = c >= CodePoints::ZENKAKU_LETTER_LOWER_FIRST && c <= CodePoints::ZENKAKU_LETTER_LOWER_LAST
          return false if !is_hankaku_upper && !is_hankaku_lower && !is_zenkaku_upper && !is_zenkaku_lower
        end
        true
      end

      def is_number(value)
        return false if value.empty?

        value.split("").each do |c|
          is_hankaku_number = c >= CodePoints::HANKAKU_NUMBER_FIRST && c <= CodePoints::HANKAKU_NUMBER_LAST
          is_zenkaku_number = c >= CodePoints::ZENKAKU_NUMBER_FIRST && c <= CodePoints::ZENKAKU_NUMBER_LAST
          return false if !is_hankaku_number && !is_zenkaku_number
        end
        true
      end

      def is_katakana(value)
        return false if value.empty?

        value.split("").each do |c|
          is_hankaku_katakana = c >= CodePoints::HANKAKU_KATAKANA_FIRST && c <= CodePoints::HANKAKU_KATAKANA_LAST && c != CodePoints::HANKAKU_PUNCTUATION_ONBIKI
          is_zenkaku_katakana = c >= CodePoints::ZENKAKU_KATAKANA_FIRST && c <= CodePoints::ZENKAKU_KATAKANA_LAST
          return false if !is_hankaku_katakana && !is_zenkaku_katakana
        end
        true
      end

      def is_hyphen(value)
        return false if value.empty?

        value.split("").each do |c|
          if c != '-' && c != '－' && c != 'ｰ' && c != 'ー'
            return false
          end
        end
        true
      end

      def is_valid_string(value, encoding_flags)
        value.split("").each do |c|
          # if is_half_width(c) && !(encoding_flags & Encoding::HALFWIDTH)
          #   return false
          # end
          # if is_full_width(c) && !(encoding_flags & Encoding::FULLWIDTH)
          #   return false
          # end
          if is_romaji(c) && !(encoding_flags & Encoding::ROMAJI)
            return false
          end
          if is_number(c) && !(encoding_flags & Encoding::NUMBERS)
            return false
          end
          if is_katakana(c) && !(encoding_flags & Encoding::KATAKANA)
            return false
          end
          if is_hyphen(c) && !(encoding_flags & Encoding::HYPHEN)
            return false
          end
        end
        return true
      end
    end
  end
end