gimchi 0.1.9 → 0.2.0

data/config/default.yml

@@ -1,174 +1,174 @@
 ---
 structure:
-    chosung:  [ㄱ, ㄲ, ㄴ, ㄷ, ㄸ, ㄹ, ㅁ, ㅂ, ㅃ, ㅅ, ㅆ, ㅇ, ㅈ, ㅉ, ㅊ, ㅋ, ㅌ, ㅍ, ㅎ]
-    jungsung: [ㅏ, ㅐ, ㅑ, ㅒ, ㅓ, ㅔ, ㅕ, ㅖ, ㅗ, ㅘ, ㅙ, ㅚ, ㅛ, ㅜ, ㅝ, ㅞ, ㅟ, ㅠ, ㅡ, ㅢ, ㅣ]
-    jongsung: [ㄱ, ㄲ, ㄳ, ㄴ, ㄵ, ㄶ, ㄷ, ㄹ, ㄺ, ㄻ, ㄼ, ㄽ, ㄾ, ㄿ, ㅀ, ㅁ, ㅂ, ㅄ, ㅅ, 
-                ㅆ, ㅇ, ㅈ, ㅊ, ㅋ, ㅌ, ㅍ, ㅎ]
+  chosung:  [ㄱ, ㄲ, ㄴ, ㄷ, ㄸ, ㄹ, ㅁ, ㅂ, ㅃ, ㅅ, ㅆ, ㅇ, ㅈ, ㅉ, ㅊ, ㅋ, ㅌ, ㅍ, ㅎ]
+  jungsung: [ㅏ, ㅐ, ㅑ, ㅒ, ㅓ, ㅔ, ㅕ, ㅖ, ㅗ, ㅘ, ㅙ, ㅚ, ㅛ, ㅜ, ㅝ, ㅞ, ㅟ, ㅠ, ㅡ, ㅢ, ㅣ]
+  jongsung: [ㄱ, ㄲ, ㄳ, ㄴ, ㄵ, ㄶ, ㄷ, ㄹ, ㄺ, ㄻ, ㄼ, ㄽ, ㄾ, ㄿ, ㅀ, ㅁ, ㅂ, ㅄ, ㅅ,
+        ㅆ, ㅇ, ㅈ, ㅊ, ㅋ, ㅌ, ㅍ, ㅎ]
 
-    fortis map:
-        ㄱ: ㄲ
-        ㄷ: ㄸ
-        ㅂ: ㅃ
-        ㅅ: ㅆ
-        ㅈ: ㅉ
+  fortis map:
+    ㄱ: ㄲ
+    ㄷ: ㄸ
+    ㅂ: ㅃ
+    ㅅ: ㅆ
+    ㅈ: ㅉ
 
-    double consonant map:
-        ㄳ: [ㄱ, ㅅ]
-        ㄵ: [ㄴ, ㅈ]
-        ㄶ: [ㄴ, ㅎ]
-        ㄺ: [ㄹ, ㄱ]
-        ㄻ: [ㄹ, ㅁ]
-        ㄼ: [ㄹ, ㅂ]
-        ㄽ: [ㄹ, ㅅ]
-        ㄾ: [ㄹ, ㅌ]
-        ㄿ: [ㄹ, ㅍ]
-        ㅀ: [ㄹ, ㅎ]
-        ㅄ: [ㅂ, ㅅ]
+  double consonant map:
+    ㄳ: [ㄱ, ㅅ]
+    ㄵ: [ㄴ, ㅈ]
+    ㄶ: [ㄴ, ㅎ]
+    ㄺ: [ㄹ, ㄱ]
+    ㄻ: [ㄹ, ㅁ]
+    ㄼ: [ㄹ, ㅂ]
+    ㄽ: [ㄹ, ㅅ]
+    ㄾ: [ㄹ, ㅌ]
+    ㄿ: [ㄹ, ㅍ]
+    ㅀ: [ㄹ, ㅎ]
+    ㅄ: [ㅂ, ㅅ]
 
 pronouncer:
-    jongsung sound:
-        ㄱ: ㄱ
-        ㄲ: ㄱ
-        ㄳ: ㄱ
-        ㄴ: ㄴ
-        ㄵ: ㄴ
-        ㄶ: ㄴ
-        ㄷ: ㄷ
-        ㄹ: ㄹ
-        ㄺ: ㄱ
-        ㄻ: ㅁ
-        ㄼ: ㄹ
-        ㄽ: ㄹ
-        ㄾ: ㅌ
-        ㄿ: ㅂ
-        ㅀ: ㄹ
-        ㅁ: ㅁ
-        ㅂ: ㅂ
-        ㅄ: ㅂ
-        ㅅ: ㄷ
-        ㅆ: ㄷ
-        ㅇ: ㅇ
-        ㅈ: ㄷ
-        ㅊ: ㄷ
-        ㅋ: ㄱ
-        ㅌ: ㄷ
-        ㅍ: ㅂ
-        ㅎ:
-    transformation:
-        # changing the order affects the quality of the transformation
-        sequence for 1:
-            - rule_5_1
-            - rule_5_3
+  jongsung sound:
+    ㄱ: ㄱ
+    ㄲ: ㄱ
+    ㄳ: ㄱ
+    ㄴ: ㄴ
+    ㄵ: ㄴ
+    ㄶ: ㄴ
+    ㄷ: ㄷ
+    ㄹ: ㄹ
+    ㄺ: ㄱ
+    ㄻ: ㅁ
+    ㄼ: ㄹ
+    ㄽ: ㄹ
+    ㄾ: ㅌ
+    ㄿ: ㅂ
+    ㅀ: ㄹ
+    ㅁ: ㅁ
+    ㅂ: ㅂ
+    ㅄ: ㅂ
+    ㅅ: ㄷ
+    ㅆ: ㄷ
+    ㅇ: ㅇ
+    ㅈ: ㄷ
+    ㅊ: ㄷ
+    ㅋ: ㄱ
+    ㅌ: ㄷ
+    ㅍ: ㅂ
+    ㅎ:
+  transformation:
+    # changing the order affects the quality of the transformation
+    sequence for 1:
+      - rule_5_1
+      - rule_5_3
 
-        sequence for 2:
-            - rule_16
-            - rule_17
-            - rule_18
-            - rule_19
-            - rule_5_1
-            - rule_5_3
-            - rule_30
-            - rule_23
-            - rule_24
-            - rule_25
-            - rule_12
-            - rule_20
-            - rule_10
-            - rule_27
-            - rule_9
-            - rule_11
-            - rule_14
-            - rule_13
-            - rule_15
-        blocking rule:
-            rule_16: [rule_30]
+    sequence for 2:
+      - rule_16
+      - rule_17
+      - rule_18
+      - rule_19
+      - rule_5_1
+      - rule_5_3
+      - rule_30
+      - rule_23
+      - rule_24
+      - rule_25
+      - rule_12
+      - rule_20
+      - rule_10
+      - rule_27
+      - rule_9
+      - rule_11
+      - rule_14
+      - rule_13
+      - rule_15
+    blocking rule:
+      rule_16: [rule_30]
 
 number:
-    positive: 플러스
-    negative: 마이너스
-    decimal point: 점
-    units: ["", 만, 억, 조, 경, 해, 자, 양, 구, 간, 정, 재, 극, 항하사, 아승기, 나유타, 불가사의, 무량대수]
-    digits: [영, 일, 이, 삼, 사, 오, 육, 칠, 팔, 구]
-    post substitution:
-        "^일만": 만
+  positive: 플러스
+  negative: 마이너스
+  decimal point: 점
+  units: ["", 만, 억, 조, 경, 해, 자, 양, 구, 간, 정, 재, 극, 항하사, 아승기, 나유타, 불가사의, 무량대수]
+  digits: [영, 일, 이, 삼, 사, 오, 육, 칠, 팔, 구]
+  post substitution:
+    "^일만": 만
 
-    # 정수형일 때 또다른 표현법 (나이, 시간)
-    alt notation:
-        when suffix:
-            개:
-                max: 
-            명:
-                max: 
-            살:
-                max: 
-            시:
-                max: 12
-        tenfolds: [열, 스물, 서른, 마흔, 쉰, 예순, 일흔, 여든, 아흔, 백]
-        digits: ["", 한, 두, 세, 네, 다섯, 여섯, 일곱, 여덟, 아홉]
-        post substitution:
-            "스물$": 스무
+  # 정수형일 때 또다른 표현법 (나이, 시간)
+  alt notation:
+    when suffix:
+      개:
+        max:
+      명:
+        max:
+      살:
+        max:
+      시:
+        max: 12
+    tenfolds: [열, 스물, 서른, 마흔, 쉰, 예순, 일흔, 여든, 아흔, 백]
+    digits: ["", 한, 두, 세, 네, 다섯, 여섯, 일곱, 여덟, 아홉]
+    post substitution:
+      "스물$": 스무
 
 romanization:
-    chosung:
-        ㄱ: g
-        ㄲ: kk
-        ㅋ: k
-        ㄷ: d
-        ㄸ: tt
-        ㅌ: t
-        ㅂ: b
-        ㅃ: pp
-        ㅍ: p
-        ㅈ: j
-        ㅉ: jj
-        ㅊ: ch
-        ㅅ: s
-        ㅆ: ss
-        ㅎ: h
-        ㄴ: n
-        ㅁ: m
-        ㄹ: r
-        ㅇ: "-"
-    jungsung:
-        ㅏ: a
-        ㅓ: eo
-        ㅗ: o
-        ㅜ: u
-        ㅡ: eu
-        ㅣ: i
-        ㅐ: ae
-        ㅔ: e
-        ㅚ: oe
-        ㅟ: wi
-        ㅑ: ya
-        ㅕ: yeo
-        ㅛ: yo
-        ㅠ: yu
-        ㅒ: yae
-        ㅖ: ye
-        ㅘ: wa
-        ㅙ: wae
-        ㅝ: wo
-        ㅞ: we
-        ㅢ: ui
-    jongsung:
-        ㄱ: k
-        ㄴ: n-
-        ㄷ: t
-        ㄹ: l
-        ㅁ: m
-        ㅂ: p
-        ㅇ: ng
-    post substitution:
-        # 제2항 [붙임 2]‘ㄹ’은 모음 앞에서는 ‘ｒ’로, 자음 앞이나 어말에서는
-        # ‘ｌ’로 적는다. 단, ‘ㄹㄹ’은 ‘ll’로 적는다.
-        lr: ll
-        "-w": w
-        "-y": y
-        kkk: k-kk
-        ttt: t-tt
-        ppp: p-pp
-        "--": "-"
-        "n-([^gaeiou])": "n\\1"
-        "-(\\s)": "\\1"
-        "-$": ""
+  chosung:
+    ㄱ: g
+    ㄲ: kk
+    ㅋ: k
+    ㄷ: d
+    ㄸ: tt
+    ㅌ: t
+    ㅂ: b
+    ㅃ: pp
+    ㅍ: p
+    ㅈ: j
+    ㅉ: jj
+    ㅊ: ch
+    ㅅ: s
+    ㅆ: ss
+    ㅎ: h
+    ㄴ: n
+    ㅁ: m
+    ㄹ: r
+    ㅇ: "-"
+  jungsung:
+    ㅏ: a
+    ㅓ: eo
+    ㅗ: o
+    ㅜ: u
+    ㅡ: eu
+    ㅣ: i
+    ㅐ: ae
+    ㅔ: e
+    ㅚ: oe
+    ㅟ: wi
+    ㅑ: ya
+    ㅕ: yeo
+    ㅛ: yo
+    ㅠ: yu
+    ㅒ: yae
+    ㅖ: ye
+    ㅘ: wa
+    ㅙ: wae
+    ㅝ: wo
+    ㅞ: we
+    ㅢ: ui
+  jongsung:
+    ㄱ: k
+    ㄴ: n-
+    ㄷ: t
+    ㄹ: l
+    ㅁ: m
+    ㅂ: p
+    ㅇ: ng
+  post substitution:
+    # 제2항 [붙임 2]‘ㄹ’은 모음 앞에서는 ‘ｒ’로, 자음 앞이나 어말에서는
+    # ‘ｌ’로 적는다. 단, ‘ㄹㄹ’은 ‘ll’로 적는다.
+    lr: ll
+    "-w": w
+    "-y": y
+    kkk: k-kk
+    ttt: t-tt
+    ppp: p-pp
+    "--": "-"
+    "n-([^gaeiou])": "n\\1"
+    "-(\\s)": "\\1"
+    "-$": ""

data/crawler/crawler.rb

@@ -0,0 +1,49 @@
+#!/usr/bin/env ruby
+# encoding: UTF-8
+# Junegunn Choi (junegunn.c@gmail.com)
+# 2011/04/02-
+
+# A dirty little script to fetch test sets from http://www.korean.go.kr
+
+require 'open-uri'
+require 'yaml'
+
+# Crawl romanization test set
+rdata = open('http://www.korean.go.kr/09_new/dic/rule/rule_roman_0101.jsp').read.
+	scan(%r{th>(.*?)</td}m).flatten.map { |e| e.split %r{<.*>}m }.
+	select { |e| e.length == 2 }
+
+File.open(File.dirname(__FILE__) + '/../test/romanization.yml', 'w') do | f |
+	f.puts "---"
+
+	rdata.each do | arr |
+		f.puts "\"#{arr.first}\": \"#{arr.last}\""
+	end
+end
+
+exit
+
+# Crawl pronunciation test set
+m = {}
+%w[
+	http://www.korean.go.kr/09_new/dic/rule/rule02_0202.jsp
+	http://www.korean.go.kr/09_new/dic/rule/rule02_0204.jsp
+	http://www.korean.go.kr/09_new/dic/rule/rule02_0205.jsp
+	http://www.korean.go.kr/09_new/dic/rule/rule02_0206.jsp
+	http://www.korean.go.kr/09_new/dic/rule/rule02_0207.jsp
+].each do | url |
+	open(url).read.scan(/>([^0-9<>);?]+?)\[(.*?)\]</).each do | match |
+		puts match[0, 2].join(' => ')
+		m[match[0]] = match[1]
+	end
+end
+
+File.open(File.dirname(__FILE__) + '/../test/pronunciation.yml', 'w') do | f |
+	f.puts "---"
+	m.each do | k, v |
+		k = k.sub(/.*→/, '').gsub(/-/, '')
+		v = v.sub(/.*→/, '').gsub(/[\(:ː\)]/, '').split(%r{[/∼]})
+		f.puts "\"#{k}\": [#{v.join(', ')}]"
+	end
+end
+

data/gimchi.gemspec

@@ -0,0 +1,21 @@
+# -*- encoding: utf-8 -*-
+lib = File.expand_path('../lib', __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+
+Gem::Specification.new do |gem|
+  gem.name          = %q{gimchi}
+  gem.version       = "0.2.0"
+  gem.authors       = ["Junegunn Choi"]
+  gem.email         = ["junegunn.c@gmail.com"]
+  gem.description   = %q{A Ruby gem for Korean characters}
+  gem.summary       = %q{A Ruby gem for Korean characters}
+  gem.homepage      = "https://github.com/junegunn/gimchi"
+
+  gem.files         = `git ls-files`.split($/).reject { |f| f =~ %r[^viz/] }
+  gem.executables   = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
+  gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})
+  gem.require_paths = ["lib"]
+  gem.license       = "MIT"
+
+  gem.add_development_dependency 'ansi'
+end

data/lib/gimchi.rb

@@ -2,10 +2,380 @@
 # encoding: UTF-8
 # Junegunn Choi (junegunn.c@gmail.com)
 
-require 'gimchi/korean'
+require 'yaml'
+require 'set'
 require 'gimchi/char'
 require 'gimchi/pronouncer'
 
-if RUBY_VERSION =~ /^1\.8\./
-  require 'gimchi/patch_1.8' 
-end
+class Gimchi
+  class << self
+    def setup
+      @@default ||= Gimchi.new
+    end
+
+    def Char ch
+      @@default.kchar ch
+    end
+
+    [
+      :decompose,
+      :compose,
+      :korean_char?,
+      :complete_korean_char?,
+      :kchar,
+      :kchar?,
+      :chosung?,
+      :jungsung?,
+      :jongsung?,
+      :read_number,
+      :pronounce,
+      :romanize
+    ].each do |sym|
+      define_method(sym) do |*arg, &b|
+        @@default.send sym, *arg, &b
+      end
+    end
+  end
+
+  CONFIG_FILE_PATH = File.expand_path('../../config/default.yml', __FILE__)
+  attr_reader :config, :chosungs, :jungsungs, :jongsungs
+
+  # Initialize Gimchi::Korean.
+  def initialize
+    symbolize_keys = lambda do |val|
+      case val
+      when Hash
+        {}.tap do |h|
+          val.each do |k, v|
+            k = k.gsub(' ', '_').to_sym if k =~ /[a-z0-9 ]/
+            h[k] = symbolize_keys.call v
+          end
+        end
+      when Array
+        val.map { |v| symbolize_keys.call v }
+      else
+        val
+      end
+    end
+    @config = symbolize_keys.call YAML.load(File.read CONFIG_FILE_PATH)
+
+    [
+      @config[:romanization][:post_substitution],
+      @config[:number][:post_substitution],
+      @config[:number][:alt_notation][:post_substitution]
+    ].each do |r|
+      r.keys.each do |k|
+        r[Regexp.compile k.to_s] = r.delete k
+      end
+    end
+    @config.freeze
+
+    @pronouncer = Gimchi::Pronouncer.send :new, self
+
+    @chosungs  = config[:structure][:chosung]
+    @jungsungs = config[:structure][:jungsung]
+    @jongsungs = config[:structure][:jongsung]
+    @chosung_set  = Set[*@chosungs]
+    @jungsung_set = Set[*@jungsungs]
+    @jongsung_set = Set[*@jongsungs]
+    @all          = @chosung_set + @jungsung_set + @jongsung_set
+  end
+
+  # Decompose a Korean character into 3 components
+  # @param [String] ch Korean character
+  # @return [Array]
+  def decompose ch
+    kchar(ch).to_a
+  end
+
+  # Compose 3 elements into a Korean character String
+  # @param [String] chosung
+  # @param [String] jungsung
+  # @param [String] jongsung
+  # @return [String]
+  def compose chosung, jungsung = nil, jongsung = nil
+    if chosung.nil? && jungsung.nil?
+      ""
+    elsif chosung && jungsung
+      n1, n2, n3 =
+      n1 = chosungs.index(chosung) || 0
+      n2 = jungsungs.index(jungsung) || 0
+      n3 = ([nil] + jongsungs).index(jongsung) || 0
+      [ 0xAC00 + n1 * (21 * 28) + n2 * 28 + n3 ].pack('U')
+    else
+      chosung || jungsung
+    end
+  end
+
+  # @param [String] ch
+  # @return [Boolean]
+  def chosung? ch
+    @chosung_set.include? ch
+  end
+
+  # @param [String] ch
+  # @return [Boolean]
+  def jungsung? ch
+    @jungsung_set.include? ch
+  end
+
+  # @param [String] ch
+  # @return [Boolean]
+  def jongsung? ch
+    @jongsung_set.include? ch
+  end
+
+  # Checks if the given character is a korean character.
+  # @param [String] ch A string of size 1
+  def korean_char? ch
+    raise ArgumentError.new('Lengthy input') if str_length(ch) > 1
+
+    complete_korean_char?(ch) || @all.include?(ch)
+  end
+  alias kchar? korean_char?
+
+  # Checks if the given character is a "complete" korean character.
+  # "Complete" Korean character must have chosung and jungsung, with optional jongsung.
+  # @param [String] ch A string of size 1
+  def complete_korean_char? ch
+    raise ArgumentError.new('Lengthy input') if str_length(ch) > 1
+
+    # Range of Korean chracters in Unicode 2.0: AC00(가) ~ D7A3(힣)
+    ch.unpack('U').all? { | c | c >= 0xAC00 && c <= 0xD7A3 }
+  end
+
+  # Returns a Gimchi::Char object for the given Korean character.
+  # @param [String] ch Korean character in String
+  # @return [Gimchi::Char] Gimchi::Char instance
+  def kchar ch
+    Gimchi::Char.new(self, ch)
+  end
+
+  # Reads numeric expressions in Korean way.
+  # @param [String, Number] str Numeric type or String containing numeric expressions
+  # @return [String] Output string
+  def read_number str
+    str.to_s.gsub(/(([+-]\s*)?[0-9,]*,*[0-9]+(\.[0-9]+(e[+-][0-9]+)?)?)(\s*.)?/) {
+      read_number_sub($1, $5)
+    }
+  end
+
+  # Returns the pronunciation of the given string containing Korean characters.
+  # Takes optional options hash.
+  #
+  # @param [String] Input string
+  # @param [Hash] options Options
+  # @option options [Boolean] each_char Each character of the string is pronounced respectively.
+  # @option options [Boolean] slur      Strings separated by whitespaces are processed again as if they were contiguous.
+  # @option options [Boolean] number    Numberic parts of the string is also pronounced in Korean.
+  # @option options [Array]   except    Allows you to skip certain transformations.
+  # @return [String] Output string
+  def pronounce str, options = {}
+    options = {
+      :each_char => false,
+      :slur      => false,
+      :number    => true,
+      :except    => [],
+      :debug     => false
+    }.merge options
+
+    str = read_number(str) if options[:number]
+
+    result, transforms = @pronouncer.send :pronounce!, str, options
+
+    if options[:debug]
+      return result, transforms
+    else
+      return result
+    end
+  end
+
+  # Returns the romanization (alphabetical notation) of the given Korean string.
+  # http://en.wikipedia.org/wiki/Korean_romanization
+  # @param [String] str Input Korean string
+  # @param [Hash] options Options
+  # @option options [Boolean] as_pronounced If true, #pronounce is internally called before romanize
+  # @option options [Boolean] number        Whether to read numeric expressions in the string
+  # @option options [Boolean] slur          Same as :slur in #pronounce
+  # @return [String] Output string in Roman Alphabet
+  # @see Korean#pronounce
+  def romanize str, options = {}
+    options = {
+      :as_pronounced => true,
+      :number        => true,
+      :slur          => false
+    }.merge options
+
+    rdata = config[:romanization]
+    post_subs = rdata[:post_substitution]
+    rdata = [rdata[:chosung], rdata[:jungsung], rdata[:jongsung]]
+
+    str = pronounce str,
+      :each_char => !options[:as_pronounced],
+      :number    => options[:number],
+      :slur      => options[:slur],
+      # 제1항 [붙임 1] ‘ㅢ’는 ‘ㅣ’로 소리 나더라도 ‘ui’로 적는다.
+      :except => %w[rule_5_3]
+    dash = rdata[0]["ㅇ"]
+    romanization = ""
+
+    romanize_chunk = lambda do |chunk|
+      chunk.each_char.map { |ch| kchar(ch) rescue ch }.each do |kc|
+        kc.to_a.each_with_index do |comp, idx|
+          next if comp.nil?
+          comp = rdata[idx][comp] || comp
+          comp = comp[1..-1] if comp[0, 1] == dash &&
+              (romanization.empty? || romanization[-1, 1] =~ /\s/)
+          romanization += comp
+        end
+      end
+
+      return post_subs.keys.inject(romanization) { | output, pattern |
+        output.gsub(pattern, post_subs[pattern])
+      }
+    end
+
+    k_chunk = ""
+    str.each_char do | c |
+      if korean_char? c
+        k_chunk += c
+      else
+        unless k_chunk.empty?
+          romanization = romanize_chunk.call k_chunk
+          k_chunk = ""
+        end
+        romanization += c
+      end
+    end
+    romanization = romanize_chunk.call k_chunk unless k_chunk.empty?
+    romanization
+  end
+
+private
+  def str_length str
+    str.length
+  end
+
+  def read_number_sub num, next_char
+    nconfig = config[:number]
+
+    if num == '0'
+      return nconfig[:digits].first
+    end
+
+    num = num.gsub(',', '')
+    next_char = next_char.to_s
+    is_float = num.match(/[\.e]/) != nil
+
+    # Alternative notation for integers with proper suffix
+    alt = false
+    if is_float == false &&
+        nconfig[:alt_notation][:when_suffix].keys.include?(next_char.strip)
+      max = nconfig[:alt_notation][:when_suffix][next_char.strip][:max]
+
+      if max.nil? || num.to_i <= max
+        alt = true
+      end
+    end
+
+    # Sign
+    sign = []
+    negative = false
+    if num =~ /^-/
+      num = num.sub(/^-\s*/, '')
+      sign << nconfig[:negative]
+      negative = true
+    elsif num =~ /^\+/
+      num = num.sub(/^\+\s*/, '')
+      sign << nconfig[:positive]
+    end
+
+    if is_float
+      below = nconfig[:decimal_point]
+      below = nconfig[:digits][0] + below if num.to_f < 1
+
+      if md = num.match(/(.*)e(.*)/)
+        dp = md[1].index('.')
+        num = md[1].tr '.', ''
+        exp = md[2].to_i
+
+        dp += exp
+        if dp > num.length
+          num = num.ljust(dp, '0')
+          num = num.sub(/^0+([1-9])/, "\\1")
+
+          below = ""
+        elsif dp < 0
+          num = '0.' + '0' * (-dp) + num
+        else
+          num[dp, 1] = '.' + num[dp, 1]
+        end
+      end
+      num.sub(/.*\./, '').each_char do | char |
+        below += nconfig[:digits][char.to_i]
+      end if num.include? '.'
+      num = num.sub(/\..*/, '')
+    else
+      below = ""
+    end
+
+    tokens = []
+    unit_idx = -1
+    num = num.to_i
+    while num > 0
+      v = num % 10000
+
+      unit_idx += 1
+      if v > 0
+        if alt == false || unit_idx >= 1
+          str = ""
+          # Cannot use hash as they're unordered in 1.8
+          [[1000, '천'],
+           [100,  '백'],
+           [10,   '십']].each do | arr |
+            u, sub_unit = arr
+            str += (nconfig[:digits][v/u] if v/u != 1).to_s + sub_unit + ' ' if v / u > 0
+            v %= u
+          end
+          str += nconfig[:digits][v] if v > 0
+
+          raise RangeError, "number too large" unless nconfig[:units][unit_idx]
+          tokens << str.sub(/ $/, '') + nconfig[:units][unit_idx]
+        else
+          str = ""
+          tenfolds = nconfig[:alt_notation][:tenfolds]
+          digits = nconfig[:alt_notation][:digits]
+          alt_post_subs = nconfig[:alt_notation][:post_substitution]
+
+          # Likewise.
+          [[1000, '천'],
+           [100,  '백']].each do |u, sub_unit|
+            str += (nconfig[:digits][v/u] if v/u != 1).to_s + sub_unit + ' ' if v / u > 0
+            v %= u
+          end
+
+          str += tenfolds[(v / 10) - 1] if v / 10 > 0
+          v %= 10
+          str += digits[v] if v > 0
+
+          alt_post_subs.each do |p, s|
+            str.gsub!(p, s)
+          end if alt
+          tokens << str.sub(/ $/, '') + nconfig[:units][unit_idx]
+        end
+      end
+      num /= 10000
+    end
+
+    tokens += sign unless sign.empty?
+    ret = tokens.reverse.join(' ') + below + next_char
+    nconfig[:post_substitution].each do |p, s|
+      ret.gsub!(p, s)
+    end
+    ret
+  end
+end#Gimchi
+
+require 'gimchi/patch_1.8'
+
+Gimchi.setup