| rev |
line source |
|
jbe@10
|
1 # Copyright (c) 2009 Public Software Group e. V., Berlin, Germany
|
|
jbe@7
|
2 #
|
|
jbe@7
|
3 # Permission is hereby granted, free of charge, to any person obtaining a
|
|
jbe@7
|
4 # copy of this software and associated documentation files (the "Software"),
|
|
jbe@7
|
5 # to deal in the Software without restriction, including without limitation
|
|
jbe@7
|
6 # the rights to use, copy, modify, merge, publish, distribute, sublicense,
|
|
jbe@7
|
7 # and/or sell copies of the Software, and to permit persons to whom the
|
|
jbe@7
|
8 # Software is furnished to do so, subject to the following conditions:
|
|
jbe@7
|
9 #
|
|
jbe@7
|
10 # The above copyright notice and this permission notice shall be included in
|
|
jbe@7
|
11 # all copies or substantial portions of the Software.
|
|
jbe@7
|
12 #
|
|
jbe@7
|
13 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
jbe@7
|
14 # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
jbe@7
|
15 # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
jbe@7
|
16 # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
jbe@7
|
17 # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
|
jbe@7
|
18 # FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
|
|
jbe@7
|
19 # DEALINGS IN THE SOFTWARE.
|
|
jbe@0
|
20
|
|
jbe@7
|
21
|
|
jbe@7
|
22 #
|
|
jbe@7
|
23 # File name: ruby/utf8proc.rb
|
|
jbe@7
|
24 #
|
|
jbe@7
|
25 # Description:
|
|
jbe@7
|
26 # Part of the ruby wrapper for libutf8proc, which is written in ruby.
|
|
jbe@7
|
27 #
|
|
jbe@0
|
28
|
|
jbe@0
|
29
|
|
jbe@0
|
30 require 'utf8proc_native'
|
|
jbe@0
|
31
|
|
jbe@2
|
32
|
|
jbe@0
|
33 module Utf8Proc
|
|
jbe@2
|
34
|
|
jbe@0
|
35 SpecialChars = {
|
|
jbe@0
|
36 :HT => "\x09",
|
|
jbe@0
|
37 :LF => "\x0A",
|
|
jbe@0
|
38 :VT => "\x0B",
|
|
jbe@0
|
39 :FF => "\x0C",
|
|
jbe@0
|
40 :CR => "\x0D",
|
|
jbe@0
|
41 :FS => "\x1C",
|
|
jbe@0
|
42 :GS => "\x1D",
|
|
jbe@0
|
43 :RS => "\x1E",
|
|
jbe@0
|
44 :US => "\x1F",
|
|
jbe@0
|
45 :LS => "\xE2\x80\xA8",
|
|
jbe@0
|
46 :PS => "\xE2\x80\xA9",
|
|
jbe@0
|
47 }
|
|
jbe@2
|
48
|
|
jbe@2
|
49 module StringExtensions
|
|
jbe@2
|
50 def utf8map(*option_array)
|
|
jbe@2
|
51 options = 0
|
|
jbe@2
|
52 option_array.each do |option|
|
|
jbe@2
|
53 flag = Utf8Proc::Options[option]
|
|
jbe@2
|
54 raise ArgumentError, "Unknown argument given to String#utf8map." unless
|
|
jbe@2
|
55 flag
|
|
jbe@2
|
56 options |= flag
|
|
jbe@2
|
57 end
|
|
jbe@2
|
58 return Utf8Proc::utf8map(self, options)
|
|
jbe@2
|
59 end
|
|
jbe@2
|
60 def utf8map!(*option_array)
|
|
jbe@2
|
61 self.replace(self.utf8map(*option_array))
|
|
jbe@2
|
62 end
|
|
jbe@2
|
63 def utf8nfd; utf8map( :stable, :decompose); end
|
|
jbe@2
|
64 def utf8nfd!; utf8map!(:stable, :decompose); end
|
|
jbe@2
|
65 def utf8nfc; utf8map( :stable, :compose); end
|
|
jbe@2
|
66 def utf8nfc!; utf8map!(:stable, :compose); end
|
|
jbe@2
|
67 def utf8nfkd; utf8map( :stable, :decompose, :compat); end
|
|
jbe@2
|
68 def utf8nfkd!; utf8map!(:stable, :decompose, :compat); end
|
|
jbe@2
|
69 def utf8nfkc; utf8map( :stable, :compose, :compat); end
|
|
jbe@2
|
70 def utf8nfkc!; utf8map!(:stable, :compose, :compat); end
|
|
jbe@3
|
71 def utf8chars
|
|
jbe@3
|
72 result = self.utf8map(:charbound).split("\377")
|
|
jbe@6
|
73 result.shift if result.first == ""
|
|
jbe@3
|
74 result
|
|
jbe@3
|
75 end
|
|
jbe@2
|
76 def char_ary
|
|
jbe@3
|
77 # depecated, use String#utf8chars instead
|
|
jbe@3
|
78 utf8chars
|
|
jbe@2
|
79 end
|
|
jbe@2
|
80 end
|
|
jbe@2
|
81
|
|
jbe@2
|
82 module IntegerExtensions
|
|
jbe@2
|
83 def utf8
|
|
jbe@2
|
84 return Utf8Proc::utf8char(self)
|
|
jbe@2
|
85 end
|
|
jbe@2
|
86 end
|
|
jbe@2
|
87
|
|
jbe@0
|
88 end
|
|
jbe@0
|
89
|
|
jbe@2
|
90
|
|
jbe@0
|
91 class String
|
|
jbe@2
|
92 include(Utf8Proc::StringExtensions)
|
|
jbe@0
|
93 end
|
|
jbe@0
|
94
|
|
jbe@0
|
95 class Integer
|
|
jbe@2
|
96 include(Utf8Proc::IntegerExtensions)
|
|
jbe@0
|
97 end
|
|
jbe@0
|
98
|