gimchi 0.2.0 → 0.2.1
Sign up to get free protection for your applications and to get access to all the features.
- data/gimchi.gemspec +1 -1
- data/lib/gimchi.rb +58 -77
- data/lib/gimchi/char.rb +16 -18
- data/lib/gimchi/patch_1.8.rb +3 -1
- data/lib/gimchi/pronouncer.rb +7 -7
- data/test/test_gimchi.rb +4 -0
- metadata +2 -2
data/gimchi.gemspec
CHANGED
@@ -4,7 +4,7 @@ $LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
|
4
4
|
|
5
5
|
Gem::Specification.new do |gem|
|
6
6
|
gem.name = %q{gimchi}
|
7
|
-
gem.version = "0.2.
|
7
|
+
gem.version = "0.2.1"
|
8
8
|
gem.authors = ["Junegunn Choi"]
|
9
9
|
gem.email = ["junegunn.c@gmail.com"]
|
10
10
|
gem.description = %q{A Ruby gem for Korean characters}
|
data/lib/gimchi.rb
CHANGED
@@ -8,77 +8,11 @@ require 'gimchi/char'
|
|
8
8
|
require 'gimchi/pronouncer'
|
9
9
|
|
10
10
|
class Gimchi
|
11
|
-
|
12
|
-
|
13
|
-
@@default ||= Gimchi.new
|
14
|
-
end
|
15
|
-
|
16
|
-
def Char ch
|
17
|
-
@@default.kchar ch
|
18
|
-
end
|
19
|
-
|
20
|
-
[
|
21
|
-
:decompose,
|
22
|
-
:compose,
|
23
|
-
:korean_char?,
|
24
|
-
:complete_korean_char?,
|
25
|
-
:kchar,
|
26
|
-
:kchar?,
|
27
|
-
:chosung?,
|
28
|
-
:jungsung?,
|
29
|
-
:jongsung?,
|
30
|
-
:read_number,
|
31
|
-
:pronounce,
|
32
|
-
:romanize
|
33
|
-
].each do |sym|
|
34
|
-
define_method(sym) do |*arg, &b|
|
35
|
-
@@default.send sym, *arg, &b
|
36
|
-
end
|
37
|
-
end
|
38
|
-
end
|
39
|
-
|
40
|
-
CONFIG_FILE_PATH = File.expand_path('../../config/default.yml', __FILE__)
|
41
|
-
attr_reader :config, :chosungs, :jungsungs, :jongsungs
|
42
|
-
|
43
|
-
# Initialize Gimchi::Korean.
|
44
|
-
def initialize
|
45
|
-
symbolize_keys = lambda do |val|
|
46
|
-
case val
|
47
|
-
when Hash
|
48
|
-
{}.tap do |h|
|
49
|
-
val.each do |k, v|
|
50
|
-
k = k.gsub(' ', '_').to_sym if k =~ /[a-z0-9 ]/
|
51
|
-
h[k] = symbolize_keys.call v
|
52
|
-
end
|
53
|
-
end
|
54
|
-
when Array
|
55
|
-
val.map { |v| symbolize_keys.call v }
|
56
|
-
else
|
57
|
-
val
|
58
|
-
end
|
59
|
-
end
|
60
|
-
@config = symbolize_keys.call YAML.load(File.read CONFIG_FILE_PATH)
|
11
|
+
class << self
|
12
|
+
attr_reader :chosungs, :jungsungs, :jongsungs
|
61
13
|
|
62
|
-
|
63
|
-
|
64
|
-
@config[:number][:post_substitution],
|
65
|
-
@config[:number][:alt_notation][:post_substitution]
|
66
|
-
].each do |r|
|
67
|
-
r.keys.each do |k|
|
68
|
-
r[Regexp.compile k.to_s] = r.delete k
|
69
|
-
end
|
70
|
-
end
|
71
|
-
@config.freeze
|
72
|
-
|
73
|
-
@pronouncer = Gimchi::Pronouncer.send :new, self
|
74
|
-
|
75
|
-
@chosungs = config[:structure][:chosung]
|
76
|
-
@jungsungs = config[:structure][:jungsung]
|
77
|
-
@jongsungs = config[:structure][:jongsung]
|
78
|
-
@chosung_set = Set[*@chosungs]
|
79
|
-
@jungsung_set = Set[*@jungsungs]
|
80
|
-
@jongsung_set = Set[*@jongsungs]
|
81
|
-
@all = @chosung_set + @jungsung_set + @jongsung_set
|
14
|
+
def Char ch
|
15
|
+
kchar ch
|
82
16
|
end
|
83
17
|
|
84
18
|
# Decompose a Korean character into 3 components
|
@@ -144,11 +78,10 @@ class Gimchi
|
|
144
78
|
ch.unpack('U').all? { | c | c >= 0xAC00 && c <= 0xD7A3 }
|
145
79
|
end
|
146
80
|
|
147
|
-
#
|
148
|
-
# @
|
149
|
-
# @return [Gimchi::Char] Gimchi::Char instance
|
81
|
+
# @deprecated
|
82
|
+
# @private
|
150
83
|
def kchar ch
|
151
|
-
Gimchi::Char.new(
|
84
|
+
Gimchi::Char.new(ch)
|
152
85
|
end
|
153
86
|
|
154
87
|
# Reads numeric expressions in Korean way.
|
@@ -206,7 +139,7 @@ class Gimchi
|
|
206
139
|
:slur => false
|
207
140
|
}.merge options
|
208
141
|
|
209
|
-
rdata = config[:romanization]
|
142
|
+
rdata = @config[:romanization]
|
210
143
|
post_subs = rdata[:post_substitution]
|
211
144
|
rdata = [rdata[:chosung], rdata[:jungsung], rdata[:jongsung]]
|
212
145
|
|
@@ -252,12 +185,14 @@ class Gimchi
|
|
252
185
|
end
|
253
186
|
|
254
187
|
private
|
188
|
+
CONFIG_FILE_PATH = File.expand_path('../../config/default.yml', __FILE__)
|
189
|
+
|
255
190
|
def str_length str
|
256
191
|
str.length
|
257
192
|
end
|
258
193
|
|
259
194
|
def read_number_sub num, next_char
|
260
|
-
nconfig = config[:number]
|
195
|
+
nconfig = @config[:number]
|
261
196
|
|
262
197
|
if num == '0'
|
263
198
|
return nconfig[:digits].first
|
@@ -374,8 +309,54 @@ private
|
|
374
309
|
end
|
375
310
|
ret
|
376
311
|
end
|
312
|
+
|
313
|
+
# @private
|
314
|
+
def setup
|
315
|
+
symbolize_keys = lambda do |val|
|
316
|
+
case val
|
317
|
+
when Hash
|
318
|
+
{}.tap do |h|
|
319
|
+
val.each do |k, v|
|
320
|
+
k = k.gsub(' ', '_').to_sym if k =~ /[a-z0-9 ]/
|
321
|
+
h[k] = symbolize_keys.call v
|
322
|
+
end
|
323
|
+
end
|
324
|
+
when Array
|
325
|
+
val.map { |v| symbolize_keys.call v }
|
326
|
+
else
|
327
|
+
val
|
328
|
+
end
|
329
|
+
end
|
330
|
+
@config = symbolize_keys.call YAML.load(File.read CONFIG_FILE_PATH)
|
331
|
+
|
332
|
+
[
|
333
|
+
@config[:romanization][:post_substitution],
|
334
|
+
@config[:number][:post_substitution],
|
335
|
+
@config[:number][:alt_notation][:post_substitution]
|
336
|
+
].each do |r|
|
337
|
+
r.keys.each do |k|
|
338
|
+
r[Regexp.compile k.to_s] = r.delete k
|
339
|
+
end
|
340
|
+
end
|
341
|
+
@config.freeze
|
342
|
+
|
343
|
+
@pronouncer = Gimchi::Pronouncer.send :new, @config[:pronouncer], @config[:structure]
|
344
|
+
|
345
|
+
@chosungs = @config[:structure][:chosung]
|
346
|
+
@jungsungs = @config[:structure][:jungsung]
|
347
|
+
@jongsungs = @config[:structure][:jongsung]
|
348
|
+
@chosung_set = Set[*@chosungs]
|
349
|
+
@jungsung_set = Set[*@jungsungs]
|
350
|
+
@jongsung_set = Set[*@jongsungs]
|
351
|
+
@all = @chosung_set + @jungsung_set + @jongsung_set
|
352
|
+
end
|
353
|
+
end
|
354
|
+
private
|
355
|
+
def initialize
|
356
|
+
raise NoMethodError, "Gimchi is a singleton class"
|
357
|
+
end
|
377
358
|
end#Gimchi
|
378
359
|
|
379
360
|
require 'gimchi/patch_1.8'
|
380
361
|
|
381
|
-
Gimchi.setup
|
362
|
+
Gimchi.send :setup
|
data/lib/gimchi/char.rb
CHANGED
@@ -13,13 +13,11 @@ class Gimchi
|
|
13
13
|
# @return [String] Jongsung component of this character.
|
14
14
|
attr_reader :jongsung
|
15
15
|
|
16
|
-
# @param [Gimchi] gimchi Gimchi instance
|
17
16
|
# @param [String] kchar Korean character string
|
18
|
-
def initialize
|
19
|
-
raise ArgumentError.new('Not a korean character') unless
|
17
|
+
def initialize kchar
|
18
|
+
raise ArgumentError.new('Not a korean character') unless Gimchi.korean_char? kchar
|
20
19
|
|
21
|
-
|
22
|
-
if @gimchi.complete_korean_char? kchar
|
20
|
+
if Gimchi.complete_korean_char? kchar
|
23
21
|
c = kchar.unpack('U').first
|
24
22
|
n = c - 0xAC00
|
25
23
|
# '가' ~ '깋' -> 'ㄱ'
|
@@ -28,14 +26,14 @@ class Gimchi
|
|
28
26
|
n = n % (21 * 28)
|
29
27
|
n2 = n / 28;
|
30
28
|
n3 = n % 28;
|
31
|
-
self.chosung =
|
32
|
-
self.jungsung =
|
33
|
-
self.jongsung = ([nil] +
|
34
|
-
elsif
|
29
|
+
self.chosung = Gimchi.chosungs[n1]
|
30
|
+
self.jungsung = Gimchi.jungsungs[n2]
|
31
|
+
self.jongsung = ([nil] + Gimchi.jongsungs)[n3]
|
32
|
+
elsif Gimchi.chosung? kchar
|
35
33
|
self.chosung = kchar
|
36
|
-
elsif
|
34
|
+
elsif Gimchi.jungsung? kchar
|
37
35
|
self.jungsung = kchar
|
38
|
-
elsif
|
36
|
+
elsif Gimchi.jongsung? kchar
|
39
37
|
self.jongsung = kchar
|
40
38
|
end
|
41
39
|
end
|
@@ -43,23 +41,23 @@ class Gimchi
|
|
43
41
|
# Recombines components into a korean character.
|
44
42
|
# @return [String] Combined korean character
|
45
43
|
def to_s
|
46
|
-
|
44
|
+
Gimchi.compose chosung, jungsung, jongsung
|
47
45
|
end
|
48
46
|
|
49
47
|
# Sets the chosung component.
|
50
48
|
# @param [String]
|
51
49
|
def chosung= c
|
52
50
|
raise ArgumentError.new('Invalid chosung component') if
|
53
|
-
c &&
|
54
|
-
@chosung = c && c.dup.extend(Component).tap { |e| e.kor =
|
51
|
+
c && Gimchi.chosung?(c) == false
|
52
|
+
@chosung = c && c.dup.extend(Component).tap { |e| e.kor = Gimchi }
|
55
53
|
end
|
56
54
|
|
57
55
|
# Sets the jungsung component
|
58
56
|
# @param [String]
|
59
57
|
def jungsung= c
|
60
58
|
raise ArgumentError.new('Invalid jungsung component') if
|
61
|
-
c &&
|
62
|
-
@jungsung = c && c.dup.extend(Component).tap { |e| e.kor =
|
59
|
+
c && Gimchi.jungsung?(c) == false
|
60
|
+
@jungsung = c && c.dup.extend(Component).tap { |e| e.kor = Gimchi }
|
63
61
|
end
|
64
62
|
|
65
63
|
# Sets the jongsung component
|
@@ -67,8 +65,8 @@ class Gimchi
|
|
67
65
|
# @param [String]
|
68
66
|
def jongsung= c
|
69
67
|
raise ArgumentError.new('Invalid jongsung component') if
|
70
|
-
c &&
|
71
|
-
@jongsung = c && c.dup.extend(Component).tap { |e| e.kor =
|
68
|
+
c && Gimchi.jongsung?(c) == false
|
69
|
+
@jongsung = c && c.dup.extend(Component).tap { |e| e.kor = Gimchi }
|
72
70
|
end
|
73
71
|
|
74
72
|
# Returns Array of three components.
|
data/lib/gimchi/patch_1.8.rb
CHANGED
data/lib/gimchi/pronouncer.rb
CHANGED
@@ -7,9 +7,9 @@ class Gimchi
|
|
7
7
|
# @private
|
8
8
|
class Pronouncer
|
9
9
|
private
|
10
|
-
def initialize
|
11
|
-
@
|
12
|
-
@
|
10
|
+
def initialize pconfig, structure
|
11
|
+
@pconfig = pconfig
|
12
|
+
@structure = structure
|
13
13
|
end
|
14
14
|
|
15
15
|
def pronounce! str, options = {}
|
@@ -17,7 +17,7 @@ class Gimchi
|
|
17
17
|
"sequence_for_#{options[:each_char] ? '1' : '2'}".to_sym] - options[:except]
|
18
18
|
|
19
19
|
# Dissecting
|
20
|
-
@chars = str.each_char.map { |c|
|
20
|
+
@chars = str.each_char.map { |c| Gimchi.kchar(c) rescue c }
|
21
21
|
@orig_chars = @chars.dup
|
22
22
|
|
23
23
|
# Padding
|
@@ -86,12 +86,12 @@ class Gimchi
|
|
86
86
|
|
87
87
|
# shortcut
|
88
88
|
def fortis_map
|
89
|
-
@
|
89
|
+
@structure[:fortis_map]
|
90
90
|
end
|
91
91
|
|
92
92
|
# shortcut
|
93
93
|
def double_consonant_map
|
94
|
-
@
|
94
|
+
@structure[:double_consonant_map]
|
95
95
|
end
|
96
96
|
|
97
97
|
# 제5항: ‘ㅑ ㅒ ㅕ ㅖ ㅘ ㅙ ㅛ ㅝ ㅞ ㅠ ㅢ’는 이중 모음으로 발음한다.
|
@@ -299,7 +299,7 @@ class Gimchi
|
|
299
299
|
|
300
300
|
word = @kc.to_s + @next_kc.to_s
|
301
301
|
if map.keys.include? word
|
302
|
-
new_char =
|
302
|
+
new_char = Gimchi.kchar(map[word].scan(/./mu)[1])
|
303
303
|
@next_kc.chosung = new_char.chosung
|
304
304
|
@next_kc.jongsung = new_char.jongsung
|
305
305
|
|
data/test/test_gimchi.rb
CHANGED
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: gimchi
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.2.
|
4
|
+
version: 0.2.1
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,7 +9,7 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2013-03-
|
12
|
+
date: 2013-03-29 00:00:00.000000000 Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: ansi
|