Module: ZhongwenTools::Regex

Extended by:
Regex
Included in:
Regex
Defined in:
lib/zhongwen_tools/regex.rb,
lib/zhongwen_tools/regex/ruby18.rb

Instance Method Summary collapse

Instance Method Details

#bopomofo ⇒ Object

Public: A Regex for bopomofo, a.k.a. Zhuyin Fuhao 注音符号.

Examples

bopomofo #=> <Regex>

Returns a Regex.



55
56
57
 
# File 'lib/zhongwen_tools/regex.rb', line 55

def bopomofo
  /[ㄅㄆㄇㄈㄉㄊㄋㄌㄍㄎㄏㄐㄑㄒㄓㄔㄕㄖㄗㄘㄙㄚㄛㄜㄝㄞㄟㄠㄡㄢㄣㄤㄥㄦㄧㄨㄩ]/
end

#capital_letters ⇒ Object 



20
21
22
 
# File 'lib/zhongwen_tools/regex.rb', line 20

def capital_letters
  /(#{Regexp.union(ZhongwenTools::UNICODE_CAPS.keys)})/
end

#fullwidth ⇒ Object 



16
17
18
 
# File 'lib/zhongwen_tools/regex.rb', line 16

def fullwidth
  /[０-９Ａ-Ｚａ-ｚ％．：＃＄＆＋－／＼＝；＜＞]/
end

#lowercase_letters ⇒ Object 



24
25
26
 
# File 'lib/zhongwen_tools/regex.rb', line 24

def lowercase_letters
  /(#{Regexp.union(ZhongwenTools::UNICODE_CAPS.values)})/
end

#punc ⇒ Object 



32
33
34
 
# File 'lib/zhongwen_tools/regex.rb', line 32

def punc
  /[\u0021-\u0023]|[\u0025-\u002A]|[\u002C-\u002F]|[\u003A\u003B\u003F\u0040]|[\u005B-\u005D\u005F\u007B\u007D\u00A1\u00A7\u00AB\u00B6\u00B7\u00BB\u00BF\u037E\u0387]/
end

#py ⇒ Object 



10
11
12
13
14
 
# File 'lib/zhongwen_tools/regex.rb', line 10

def py
  # FIXME: need to detect Ālābó
  # ([ĀÁǍÀA][io]?|[io]?|[][āáǎàaēéěèeūúǔùu]?o?|[ĒÉĚÈE]i?|[]i?|[ŌÓǑÒO]u?|[]u?|u[āáǎàaēoēéěèe]?i?|[]e?)(n?g?r?)){1,}
  /(#{pyn_regexes.map{|k,v| v.to_s[7..-2].gsub_with_hash(/[aeiouv]/,py_tones)}.join('|')}([\s\-])?)/
end

#py_tones ⇒ Object 



80
81
82
83
84
85
86
87
88
89
 
# File 'lib/zhongwen_tools/regex.rb', line 80

def py_tones
  py_tones = {
    'a' => '[āáǎàa]',
    'e' => '[ēéěèe]',
    'i' => '[īíǐìi]',
    'o' => '[ōóǒòo]',
    'u' => '[ūúǔùu]',
    'v' => '[ǖǘǚǜü]'
  }
end

#pyn ⇒ Object 



6
7
8
 
# File 'lib/zhongwen_tools/regex.rb', line 6

def pyn
  /(#{pyn_regexes.values.join('|')}|r)([1-5])?([\s\-]+)?/
end

#zh ⇒ Object 



28
29
30
 
# File 'lib/zhongwen_tools/regex.rb', line 28

def zh
  /[\u2E80-\u2E99]|[\u2E9B-\u2EF3]|[\u2F00-\u2FD5]|[\u3005|\u3007]|[\u3021-\u3029]|[\u3038-\u303B]|[\u3400-\u4DB5]|[\u4E00-\u9FCC]|[\uF900-\uFA6D]|[\uFA70-\uFAD9]/
end

#zh_numbers ⇒ Object 



41
42
43
44
45
 
# File 'lib/zhongwen_tools/regex.rb', line 41

def zh_numbers
  # TODO: include numbers like yotta, etc.
  # 垓	秭	穰	溝	澗	正	載 --> beyond 100,000,000!
  /[〇零一壹幺二贰貳两兩三弎叁參四肆䦉五伍六陆陸七柒八捌九玖十拾廿百佰千仟万萬亿億]/
end

#zh_punc ⇒ Object 



36
37
38
39
 
# File 'lib/zhongwen_tools/regex.rb', line 36

def zh_punc
  # TODO: includes non-zh punctuation codes. Should only include punctuation in CJK ranges.
  /[\u2E00-\u2E2E]|[\u2E30-\u2E3B]|[\u3001-\u3003]|[\u3008-\u3011]|[\u3014-\u301F]|[\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF]|[\uA60D-\uA60F]|[\uA673\uA67E]|[\uA6F2-\uA6F7]|[\uA874-\uA877]|[\uA8CE\uA8CF]|[\uA8F8-\uA8FA]|[\uA92E\uA92F\uA95F]|[\uA9C1-\uA9CD]|[\uA9DE\uA9DF]|[\uAA5C-\uAA5F]|[\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F]|[\uFE10-\uFE19]|[\uFE30-\uFE52]|[\uFE54-\uFE61]|[\uFE63\uFE68\uFE6A\uFE6B]|[\uFF01-\uFF03]|[\uFF05-\uFF0A]|[\uFF0C-\uFF0F]|[\uFF1A\uFF1B\uFF1F\uFF20]|[\uFF3B-\uFF3D]|[\uFF3F\uFF5B\uFF5D]|[\uFF5F-\uFF65]/
end