smart_translitter 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/smart_translitter.rb +113 -0
- metadata +56 -0
@@ -0,0 +1,113 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
class SmartTranslitter
|
3
|
+
LOWER_CASE = {
|
4
|
+
"а" => ["a"],
|
5
|
+
"б" => ["b"],
|
6
|
+
"в" => ["v"],
|
7
|
+
"г" => ["g"],
|
8
|
+
"д" => ["d"],
|
9
|
+
"е" => ["e"],
|
10
|
+
"ё" => ["yo", "jo"],
|
11
|
+
"ж" => ["zh", "j"],
|
12
|
+
"з" => ["z"],
|
13
|
+
"и" => ["i"],
|
14
|
+
"й" => ["j", "jj", "y"],
|
15
|
+
"к" => ["k"],
|
16
|
+
"л" => ["l"],
|
17
|
+
"м" => ["m"],
|
18
|
+
"н" => ["n"],
|
19
|
+
"о" => ["o"],
|
20
|
+
"п" => ["p"],
|
21
|
+
"р" => ["r"],
|
22
|
+
"с" => ["s"],
|
23
|
+
"т" => ["t"],
|
24
|
+
"у" => ["u"],
|
25
|
+
"ф" => ["f"],
|
26
|
+
"х" => ["ch", "h", "kh"],
|
27
|
+
"ц" => ["c", "ts", "cz"],
|
28
|
+
"ч" => ["ch"],
|
29
|
+
"ш" => ["sh"],
|
30
|
+
"щ" => ["shh", "shch", "sc"],
|
31
|
+
"ъ" => ["\""],
|
32
|
+
"ы" => ["y"],
|
33
|
+
"ь" => ["\'"],
|
34
|
+
"э" => ["e", "eh"],
|
35
|
+
"ю" => ["ju", "yu", "iu"],
|
36
|
+
"я" => ["ja", "ya", "ia"]
|
37
|
+
}
|
38
|
+
|
39
|
+
UPPER_CASE = {
|
40
|
+
"А" => ["A"],
|
41
|
+
"Б" => ["B"],
|
42
|
+
"В" => ["V"],
|
43
|
+
"Г" => ["G"],
|
44
|
+
"Д" => ["D"],
|
45
|
+
"Е" => ["E", "Je", "Ye"],
|
46
|
+
"Ё" => ["Yo", "Jo"],
|
47
|
+
"Ж" => ["Zh", "J"],
|
48
|
+
"З" => ["Z"],
|
49
|
+
"И" => ["I"],
|
50
|
+
"Й" => ["J", "Jj", "Y"],
|
51
|
+
"К" => ["K"],
|
52
|
+
"Л" => ["L"],
|
53
|
+
"М" => ["M"],
|
54
|
+
"Н" => ["N"],
|
55
|
+
"О" => ["O"],
|
56
|
+
"П" => ["P"],
|
57
|
+
"Р" => ["R"],
|
58
|
+
"С" => ["S"],
|
59
|
+
"Т" => ["T"],
|
60
|
+
"У" => ["U"],
|
61
|
+
"Ф" => ["F"],
|
62
|
+
"Х" => ["Ch", "H", "Kh"],
|
63
|
+
"Ц" => ["C", "Ts", "Cz"],
|
64
|
+
"Ч" => ["Ch"],
|
65
|
+
"Ш" => ["Sh"],
|
66
|
+
"Щ" => ["Shh", "Shch", "Sc"],
|
67
|
+
"Ъ" => ["\""],
|
68
|
+
"Ы" => ["Y"],
|
69
|
+
"Ь" => ["\'"],
|
70
|
+
"Э" => ["E", "Eh"],
|
71
|
+
"Ю" => ["Ju", "Yu", "Iu"],
|
72
|
+
"Я" => ["Ja", "Ya", "Ia"]
|
73
|
+
}
|
74
|
+
|
75
|
+
def self.t(word)
|
76
|
+
|
77
|
+
# Набор финальных вариантов транслитерации строки
|
78
|
+
results = [""]
|
79
|
+
word.chars do |ch|
|
80
|
+
# Определяем регистр и выбираем набор возможных интерпреаций символа
|
81
|
+
if UPPER_CASE[ch].nil? && LOWER_CASE[ch].nil?
|
82
|
+
variants = [ch]
|
83
|
+
else
|
84
|
+
variants = !UPPER_CASE[ch].nil? ? UPPER_CASE[ch] : LOWER_CASE[ch]
|
85
|
+
end
|
86
|
+
|
87
|
+
# Для каждой возможной интерпретации символа копируем текущий набор вариантов интерпретации всей строки
|
88
|
+
additional_results = []
|
89
|
+
results.each do |result|
|
90
|
+
(variants.size - 1).times do
|
91
|
+
str_copy = result.dup
|
92
|
+
additional_results << str_copy
|
93
|
+
end
|
94
|
+
end
|
95
|
+
results += additional_results
|
96
|
+
|
97
|
+
# Набор обновленных вариантов интепретации
|
98
|
+
new_results = []
|
99
|
+
# Добавляем к каждому варианту возможную интпретацию символа
|
100
|
+
index = 0
|
101
|
+
results.each_slice(results.size / variants.size) do |set|
|
102
|
+
set.each do |result|
|
103
|
+
result += variants[index]
|
104
|
+
new_results << result
|
105
|
+
end
|
106
|
+
index += 1
|
107
|
+
end
|
108
|
+
results = new_results
|
109
|
+
end
|
110
|
+
results
|
111
|
+
end
|
112
|
+
|
113
|
+
end
|
metadata
ADDED
@@ -0,0 +1,56 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: smart_translitter
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Stanislav Mekhonoshin
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2012-07-10 00:00:00.000000000 Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: rspec
|
16
|
+
requirement: &2151909640 !ruby/object:Gem::Requirement
|
17
|
+
none: false
|
18
|
+
requirements:
|
19
|
+
- - ! '>='
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: '0'
|
22
|
+
type: :development
|
23
|
+
prerelease: false
|
24
|
+
version_requirements: *2151909640
|
25
|
+
description: Translits doubly translitting russian chars
|
26
|
+
email: ejabberd@gmail.com
|
27
|
+
executables: []
|
28
|
+
extensions: []
|
29
|
+
extra_rdoc_files: []
|
30
|
+
files:
|
31
|
+
- lib/smart_translitter.rb
|
32
|
+
homepage: http://rubygems.org/gems/smart_translitter
|
33
|
+
licenses: []
|
34
|
+
post_install_message:
|
35
|
+
rdoc_options: []
|
36
|
+
require_paths:
|
37
|
+
- lib
|
38
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
39
|
+
none: false
|
40
|
+
requirements:
|
41
|
+
- - ! '>='
|
42
|
+
- !ruby/object:Gem::Version
|
43
|
+
version: '0'
|
44
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
45
|
+
none: false
|
46
|
+
requirements:
|
47
|
+
- - ! '>='
|
48
|
+
- !ruby/object:Gem::Version
|
49
|
+
version: '0'
|
50
|
+
requirements: []
|
51
|
+
rubyforge_project:
|
52
|
+
rubygems_version: 1.8.17
|
53
|
+
signing_key:
|
54
|
+
specification_version: 3
|
55
|
+
summary: Smart cyrillic2latin translitter!
|
56
|
+
test_files: []
|