gsm_tools 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +4 -0
- data/.rvmrc +71 -0
- data/Gemfile +4 -0
- data/Guardfile +10 -0
- data/Rakefile +1 -0
- data/gsm_tools.gemspec +27 -0
- data/lib/gsm_tools/base.rb +235 -0
- data/lib/gsm_tools/string.rb +11 -0
- data/lib/gsm_tools/version.rb +3 -0
- data/lib/gsm_tools.rb +3 -0
- data/spec/lib/gsm_tools/base_spec.rb +81 -0
- data/spec/lib/gsm_tools/string_spec.rb +7 -0
- data/spec/lib/gsm_tools_spec.rb +11 -0
- data/spec/spec_helper.rb +146 -0
- metadata +91 -0
data/.gitignore
ADDED
data/.rvmrc
ADDED
@@ -0,0 +1,71 @@
|
|
1
|
+
#!/usr/bin/env bash
|
2
|
+
|
3
|
+
# This is an RVM Project .rvmrc file, used to automatically load the ruby
|
4
|
+
# development environment upon cd'ing into the directory
|
5
|
+
|
6
|
+
# First we specify our desired <ruby>[@<gemset>], the @gemset name is optional.
|
7
|
+
environment_id="ruby-1.9.2-p290@gsm_tools"
|
8
|
+
|
9
|
+
#
|
10
|
+
# Uncomment following line if you want options to be set only for given project.
|
11
|
+
#
|
12
|
+
# PROJECT_JRUBY_OPTS=( --1.9 )
|
13
|
+
#
|
14
|
+
# The variable PROJECT_JRUBY_OPTS requires the following to be run in shell:
|
15
|
+
#
|
16
|
+
# chmod +x ${rvm_path}/hooks/after_use_jruby_opts
|
17
|
+
#
|
18
|
+
|
19
|
+
#
|
20
|
+
# First we attempt to load the desired environment directly from the environment
|
21
|
+
# file. This is very fast and efficient compared to running through the entire
|
22
|
+
# CLI and selector. If you want feedback on which environment was used then
|
23
|
+
# insert the word 'use' after --create as this triggers verbose mode.
|
24
|
+
#
|
25
|
+
if [[ -d "${rvm_path:-$HOME/.rvm}/environments" \
|
26
|
+
&& -s "${rvm_path:-$HOME/.rvm}/environments/$environment_id" ]]
|
27
|
+
then
|
28
|
+
\. "${rvm_path:-$HOME/.rvm}/environments/$environment_id"
|
29
|
+
|
30
|
+
if [[ -s "${rvm_path:-$HOME/.rvm}/hooks/after_use" ]]
|
31
|
+
then
|
32
|
+
. "${rvm_path:-$HOME/.rvm}/hooks/after_use"
|
33
|
+
fi
|
34
|
+
else
|
35
|
+
# If the environment file has not yet been created, use the RVM CLI to select.
|
36
|
+
if ! rvm --create "$environment_id"
|
37
|
+
then
|
38
|
+
echo "Failed to create RVM environment '${environment_id}'."
|
39
|
+
return 1
|
40
|
+
fi
|
41
|
+
fi
|
42
|
+
|
43
|
+
#
|
44
|
+
# If you use an RVM gemset file to install a list of gems (*.gems), you can have
|
45
|
+
# it be automatically loaded. Uncomment the following and adjust the filename if
|
46
|
+
# necessary.
|
47
|
+
#
|
48
|
+
# filename=".gems"
|
49
|
+
# if [[ -s "$filename" ]]
|
50
|
+
# then
|
51
|
+
# rvm gemset import "$filename" | grep -v already | grep -v listed | grep -v complete | sed '/^$/d'
|
52
|
+
# fi
|
53
|
+
|
54
|
+
# If you use bundler, this might be useful to you:
|
55
|
+
# if [[ -s Gemfile ]] && ! command -v bundle >/dev/null
|
56
|
+
# then
|
57
|
+
# printf "The rubygem 'bundler' is not installed. Installing it now.\n"
|
58
|
+
# gem install bundler
|
59
|
+
# fi
|
60
|
+
# if [[ -s Gemfile ]] && command -v bundle
|
61
|
+
# then
|
62
|
+
# bundle install
|
63
|
+
# fi
|
64
|
+
|
65
|
+
if [[ $- == *i* ]] # check for interactive shells
|
66
|
+
then
|
67
|
+
echo "Using: $(tput setaf 2)$GEM_HOME$(tput sgr0)" # show the user the ruby and gemset they are using in green
|
68
|
+
else
|
69
|
+
echo "Using: $GEM_HOME" # don't use colors in interactive shells
|
70
|
+
fi
|
71
|
+
|
data/Gemfile
ADDED
data/Guardfile
ADDED
data/Rakefile
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
require "bundler/gem_tasks"
|
data/gsm_tools.gemspec
ADDED
@@ -0,0 +1,27 @@
|
|
1
|
+
# -*- encoding: utf-8 -*-
|
2
|
+
$:.push File.expand_path("../lib", __FILE__)
|
3
|
+
require "gsm_tools/version"
|
4
|
+
|
5
|
+
Gem::Specification.new do |s|
|
6
|
+
s.name = "gsm_tools"
|
7
|
+
s.version = GsmTools::VERSION
|
8
|
+
s.authors = ["Andrew Ang"]
|
9
|
+
s.email = ["andukz@gmail.com"]
|
10
|
+
s.homepage = ""
|
11
|
+
s.summary = %q{GSM 03.38 extension libraries}
|
12
|
+
s.description = %q{Adds various methods to check GSM 03.38 character encoding compatibility.}
|
13
|
+
|
14
|
+
s.rubyforge_project = "gsm_tools"
|
15
|
+
|
16
|
+
s.files = `git ls-files`.split("\n")
|
17
|
+
s.test_files = `git ls-files -- {test,spec,features}/*`.split("\n")
|
18
|
+
s.executables = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
|
19
|
+
s.require_paths = ["lib"]
|
20
|
+
|
21
|
+
# specify any dependencies here; for example:
|
22
|
+
# s.add_development_dependency "rspec"
|
23
|
+
# s.add_runtime_dependency "rest-client"
|
24
|
+
|
25
|
+
s.add_development_dependency "rspec"
|
26
|
+
s.add_development_dependency "guard-rspec"
|
27
|
+
end
|
@@ -0,0 +1,235 @@
|
|
1
|
+
#encoding: utf-8
|
2
|
+
#
|
3
|
+
# Unicode/GSM Reference: http://unicode.org/Public/MAPPINGS/ETSI/GSM0338.TXT
|
4
|
+
|
5
|
+
require 'iconv'
|
6
|
+
|
7
|
+
module GSMTools
|
8
|
+
NL = "\n"
|
9
|
+
CR = "\r"
|
10
|
+
BS = '\\'
|
11
|
+
|
12
|
+
NON_GSM_7BIT_MATCHER = /[^@£\$¥èéùìòÇ\nØø\rÅåΔ_ΦΓΛΩΠΨΣΘΞÆæßÉ !"#¤%&'\(\)\*\+,-\.\/0123456789:;<=>\?¡ABCDEFGHIJKLMNOPQRSTUVWXYZÄÖÑܧ¿abcdefghijklmnopqrstuvwxyzäöñüà\^{}\\\[~\]\|€]/
|
13
|
+
|
14
|
+
GSM_7BIT_MAP = [
|
15
|
+
0x00, '@'.ord,
|
16
|
+
0x01, '£'.ord,
|
17
|
+
0x02, '$'.ord,
|
18
|
+
0x03, '¥'.ord,
|
19
|
+
0x04, 'è'.ord,
|
20
|
+
0x05, 'é'.ord,
|
21
|
+
0x06, 'ù'.ord,
|
22
|
+
0x07, 'ì'.ord,
|
23
|
+
0x08, 'ò'.ord,
|
24
|
+
0x09, 'Ç'.ord,
|
25
|
+
0x0a, 0x000A,
|
26
|
+
0x0b, 'Ø'.ord,
|
27
|
+
0x0c, 'ø'.ord,
|
28
|
+
0x0d, 0x000D,
|
29
|
+
0x0e, 'Å'.ord,
|
30
|
+
0x0f, 'å'.ord,
|
31
|
+
0x10, 'Δ'.ord,
|
32
|
+
0x11, '_'.ord,
|
33
|
+
0x12, 'Φ'.ord,
|
34
|
+
0x13, 'Γ'.ord,
|
35
|
+
0x14, 'Λ'.ord,
|
36
|
+
0x15, 'Ω'.ord,
|
37
|
+
0x16, 'Π'.ord,
|
38
|
+
0x17, 'Ψ'.ord,
|
39
|
+
0x18, 'Σ'.ord,
|
40
|
+
0x19, 'Θ'.ord,
|
41
|
+
0x1a, 'Ξ'.ord,
|
42
|
+
0x1b, 0x00A0,
|
43
|
+
0x1c, 'Æ'.ord,
|
44
|
+
0x1d, 'æ'.ord,
|
45
|
+
0x1e, 'ß'.ord,
|
46
|
+
0x1f, 'É'.ord,
|
47
|
+
0x20, " ".ord,
|
48
|
+
0x21, '!'.ord,
|
49
|
+
0x22, '"'.ord,
|
50
|
+
0x23, '#'.ord,
|
51
|
+
0x24, '¤'.ord,
|
52
|
+
0x25, '%'.ord,
|
53
|
+
0x26, '&'.ord,
|
54
|
+
0x27, "'".ord,
|
55
|
+
0x28, '('.ord,
|
56
|
+
0x29, ')'.ord,
|
57
|
+
0x2a, '*'.ord,
|
58
|
+
0x2b, '+'.ord,
|
59
|
+
0x2c, ','.ord,
|
60
|
+
0x2d, '-'.ord,
|
61
|
+
0x2e, '.'.ord,
|
62
|
+
0x2f, '/'.ord,
|
63
|
+
0x30, '0'.ord,
|
64
|
+
0x31, '1'.ord,
|
65
|
+
0x32, '2'.ord,
|
66
|
+
0x33, '3'.ord,
|
67
|
+
0x34, '4'.ord,
|
68
|
+
0x35, '5'.ord,
|
69
|
+
0x36, '6'.ord,
|
70
|
+
0x37, '7'.ord,
|
71
|
+
0x38, '8'.ord,
|
72
|
+
0x39, '9'.ord,
|
73
|
+
0x3a, ':'.ord,
|
74
|
+
0x3b, ';'.ord,
|
75
|
+
0x3c, '<'.ord,
|
76
|
+
0x3d, '='.ord,
|
77
|
+
0x3e, '>'.ord,
|
78
|
+
0x3f, '?'.ord,
|
79
|
+
0x40, '¡'.ord,
|
80
|
+
0x41, 'A'.ord,
|
81
|
+
0x42, 'B'.ord,
|
82
|
+
0x43, 'C'.ord,
|
83
|
+
0x44, 'D'.ord,
|
84
|
+
0x45, 'E'.ord,
|
85
|
+
0x46, 'F'.ord,
|
86
|
+
0x47, 'G'.ord,
|
87
|
+
0x48, 'H'.ord,
|
88
|
+
0x49, 'I'.ord,
|
89
|
+
0x4a, 'J'.ord,
|
90
|
+
0x4b, 'K'.ord,
|
91
|
+
0x4c, 'L'.ord,
|
92
|
+
0x4d, 'M'.ord,
|
93
|
+
0x4e, 'N'.ord,
|
94
|
+
0x4f, 'O'.ord,
|
95
|
+
0x50, 'P'.ord,
|
96
|
+
0x51, 'Q'.ord,
|
97
|
+
0x52, 'R'.ord,
|
98
|
+
0x53, 'S'.ord,
|
99
|
+
0x54, 'T'.ord,
|
100
|
+
0x55, 'U'.ord,
|
101
|
+
0x56, 'V'.ord,
|
102
|
+
0x57, 'W'.ord,
|
103
|
+
0x58, 'X'.ord,
|
104
|
+
0x59, 'Y'.ord,
|
105
|
+
0x5a, 'Z'.ord,
|
106
|
+
0x5b, 'Ä'.ord,
|
107
|
+
0x5c, 'Ö'.ord,
|
108
|
+
0x5d, 'Ñ'.ord,
|
109
|
+
0x5e, 'Ü'.ord,
|
110
|
+
0x5f, '§'.ord,
|
111
|
+
0x60, '¿'.ord,
|
112
|
+
0x61, 'a'.ord,
|
113
|
+
0x62, 'b'.ord,
|
114
|
+
0x63, 'c'.ord,
|
115
|
+
0x64, 'd'.ord,
|
116
|
+
0x65, 'e'.ord,
|
117
|
+
0x66, 'f'.ord,
|
118
|
+
0x67, 'g'.ord,
|
119
|
+
0x68, 'h'.ord,
|
120
|
+
0x69, 'i'.ord,
|
121
|
+
0x6a, 'j'.ord,
|
122
|
+
0x6b, 'k'.ord,
|
123
|
+
0x6c, 'l'.ord,
|
124
|
+
0x6d, 'm'.ord,
|
125
|
+
0x6e, 'n'.ord,
|
126
|
+
0x6f, 'o'.ord,
|
127
|
+
0x70, 'p'.ord,
|
128
|
+
0x71, 'q'.ord,
|
129
|
+
0x72, 'r'.ord,
|
130
|
+
0x73, 's'.ord,
|
131
|
+
0x74, 't'.ord,
|
132
|
+
0x75, 'u'.ord,
|
133
|
+
0x76, 'v'.ord,
|
134
|
+
0x77, 'w'.ord,
|
135
|
+
0x78, 'x'.ord,
|
136
|
+
0x79, 'y'.ord,
|
137
|
+
0x7a, 'z'.ord,
|
138
|
+
0x7b, 'ä'.ord,
|
139
|
+
0x7c, 'ö'.ord,
|
140
|
+
0x7d, 'ñ'.ord,
|
141
|
+
0x7e, 'ü'.ord,
|
142
|
+
0x7f, 'à'.ord,
|
143
|
+
#[0x1B, 0x0A], 0x000C, # Removed for now, Need to figure out how to Regex match this character
|
144
|
+
[0x1B, 0x14], '^'.ord,
|
145
|
+
[0x1B, 0x28], '{'.ord,
|
146
|
+
[0x1B, 0x29], '}'.ord,
|
147
|
+
[0x1B, 0x2f], 0x005C,
|
148
|
+
[0x1B, 0x3c], '['.ord,
|
149
|
+
[0x1B, 0x3d], '~'.ord,
|
150
|
+
[0x1B, 0x3e], ']'.ord,
|
151
|
+
[0x1B, 0x40], '|'.ord,
|
152
|
+
[0x1B, 0x65], '€'.ord
|
153
|
+
]
|
154
|
+
|
155
|
+
@@uni_to_gsm = Hash.new
|
156
|
+
@@gsm_to_uni = Hash.new
|
157
|
+
GSM_7BIT_MAP.each_slice(2) { |gsm, uni| @@uni_to_gsm[uni] = gsm; @@gsm_to_uni[gsm] = uni; }
|
158
|
+
|
159
|
+
class << self
|
160
|
+
def is_gsm_7bit? str
|
161
|
+
str.match(NON_GSM_7BIT_MATCHER).nil?
|
162
|
+
end
|
163
|
+
|
164
|
+
def to_gsm str
|
165
|
+
if self.is_gsm_7bit?(str)
|
166
|
+
str.unpack('U*').collect { |utf8_char| @@uni_to_gsm[utf8_char] }.flatten.pack('c*')
|
167
|
+
else
|
168
|
+
Iconv.iconv('UCS-2', 'UTF-8', str).first
|
169
|
+
end
|
170
|
+
end
|
171
|
+
|
172
|
+
def gsm_string_counter str
|
173
|
+
hash = {}
|
174
|
+
|
175
|
+
if self.is_gsm_7bit?(str)
|
176
|
+
hash = { :length => 0, :segments => 1, :remaining_length_for_segment => 160 }
|
177
|
+
gsm_str = self.to_gsm(str)
|
178
|
+
hash[:length] = gsm_str.length
|
179
|
+
|
180
|
+
case gsm_str.length
|
181
|
+
when 0..160
|
182
|
+
hash[:segments] = 1
|
183
|
+
hash[:remaining_length_for_segment] = 160 - gsm_str.length
|
184
|
+
when 161..306
|
185
|
+
hash[:segments] = 2
|
186
|
+
hash[:remaining_length_for_segment] = 306 - gsm_str.length
|
187
|
+
when 307..459
|
188
|
+
hash[:segments] = 3
|
189
|
+
hash[:remaining_length_for_segment] = 459 - gsm_str.length
|
190
|
+
else
|
191
|
+
hash = { :invalid => true }
|
192
|
+
end
|
193
|
+
|
194
|
+
else
|
195
|
+
hash = { :length => 0, :segments => 1, :remaining_length_for_segment => 70 }
|
196
|
+
hash[:length] = str.length
|
197
|
+
|
198
|
+
case str.length
|
199
|
+
when 0..70
|
200
|
+
hash[:segments] = 1
|
201
|
+
hash[:remaining_length_for_segment] = 70 - str.length
|
202
|
+
when 71..134
|
203
|
+
hash[:segments] = 2
|
204
|
+
hash[:remaining_length_for_segment] = 134 - str.length
|
205
|
+
when 135..201
|
206
|
+
hash[:segments] = 3
|
207
|
+
hash[:remaining_length_for_segment] = 201 - str.length
|
208
|
+
else
|
209
|
+
hash = { :invalid => true }
|
210
|
+
end
|
211
|
+
end
|
212
|
+
|
213
|
+
return hash
|
214
|
+
end
|
215
|
+
|
216
|
+
def gsm_to_s gsm
|
217
|
+
last_is_escape = false
|
218
|
+
collect = gsm.unpack('c*').collect do |bin_char|
|
219
|
+
if bin_char == 27
|
220
|
+
last_is_escape = true
|
221
|
+
next
|
222
|
+
end
|
223
|
+
|
224
|
+
if last_is_escape
|
225
|
+
last_is_escape = false
|
226
|
+
@@gsm_to_uni[[27, bin_char]]
|
227
|
+
else
|
228
|
+
@@gsm_to_uni[bin_char]
|
229
|
+
end
|
230
|
+
end
|
231
|
+
collect.reject { |x| x.nil? }.flatten.pack('U*')
|
232
|
+
end
|
233
|
+
|
234
|
+
end
|
235
|
+
end
|
data/lib/gsm_tools.rb
ADDED
@@ -0,0 +1,81 @@
|
|
1
|
+
#encoding: utf-8
|
2
|
+
|
3
|
+
require 'spec_helper'
|
4
|
+
|
5
|
+
describe GSMTools, "base" do
|
6
|
+
it "must check for correct non-GSM 7-bit string" do
|
7
|
+
GSMTools.is_gsm_7bit?(GSM_7BIT_CHARS).should == true
|
8
|
+
GSMTools.is_gsm_7bit?("你好").should == false
|
9
|
+
end
|
10
|
+
|
11
|
+
it "must encode the correct unicode chars to GSM 7-bit" do
|
12
|
+
GSM_UNICODE_MAP.each_slice(2) do |gsm, unicode|
|
13
|
+
GSMTools.is_gsm_7bit?(unicode.chr('UTF-8')).should == true
|
14
|
+
result = GSMTools.to_gsm(unicode.chr('UTF-8'))
|
15
|
+
(result.length == 1 ? (result.ord.should == gsm) : result.bytes.collect {|x| x }.should == gsm)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
|
19
|
+
it "must be able to decode form GSM 7-bit" do
|
20
|
+
gsm_7bit = GSMTools.to_gsm(GSM_7BIT_CHARS)
|
21
|
+
reverted = GSMTools.gsm_to_s(gsm_7bit)
|
22
|
+
reverted.should == GSM_7BIT_CHARS
|
23
|
+
end
|
24
|
+
|
25
|
+
it "must calculate the length of GSM 7-bit string properly" do
|
26
|
+
gsm_7bit = GSMTools.to_gsm("0123456789abcde")
|
27
|
+
gsm_7bit.length.should == 15
|
28
|
+
gsm_7bit = GSMTools.to_gsm("0123456789abcde[]{}~")
|
29
|
+
gsm_7bit.length.should == 25
|
30
|
+
end
|
31
|
+
|
32
|
+
it "must encode non GSM 7-bit characters to UCS-2" do
|
33
|
+
GSMTools.to_gsm("你好").length.should == 4
|
34
|
+
end
|
35
|
+
|
36
|
+
it "must check the length and segments of the GSM string" do
|
37
|
+
GSMTools.gsm_string_counter("hello there").should == { :length => 11,
|
38
|
+
:segments => 1,
|
39
|
+
:remaining_length_for_segment => 149 }
|
40
|
+
|
41
|
+
GSMTools.gsm_string_counter("Andreasen gag high friggin Argyll Argyll artful Tyrol archon stock serial studbook Argyll Argyll archbishop drink sitcom restful Argyll stubborn objects scoffs actinolite quinta stump pigs electric").should == { :length => 197,
|
42
|
+
:segments => 2,
|
43
|
+
:remaining_length_for_segment => 109 }
|
44
|
+
|
45
|
+
GSMTools.gsm_string_counter("Axndreasen gag high friggin Argyll Argyll artful Tyrol archon stock serial studbook Argyll Argyll archbishop drink sitcom restful Argyll stubborn objects scoffs actinolite quinta stump pigs electric andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang").should == { :length => 306,
|
46
|
+
:segments => 2,
|
47
|
+
:remaining_length_for_segment => 0 }
|
48
|
+
|
49
|
+
GSMTools.gsm_string_counter("Axndreasen gag high friggin Argyll Argyll artful Tyrol archon stock serial studbook Argyll Argyll archbishop drink sitcom restful Argyll stubborn objects scoffs actinolite quinta stump pigs electric andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo an[").should == { :length => 307,
|
50
|
+
:segments => 3,
|
51
|
+
:remaining_length_for_segment => 152 }
|
52
|
+
|
53
|
+
GSMTools.gsm_string_counter("Andreasen gag high friggin Argyll Argyll artful Tyrol archon stock serial studbook Argyll Argyll archbishop drink sitcom restful Argyll stubborn objects scoffs actinolite quinta stump pigs electric andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang fdafda fdafa dsafsaf dsaf dsaf dsa fs af sa fsa fsdaf dasf dsa f dsaf dsa fds afd sa fdsa fdsafasfdsa fas fd saf da fd afdsafdasfdasf dsaf das f asf as[").should == { :length => 459,
|
54
|
+
:segments => 3,
|
55
|
+
:remaining_length_for_segment => 0 }
|
56
|
+
|
57
|
+
GSMTools.gsm_string_counter("Andreasen gag high friggin Argyll Argyll artful Tyrol archon stock serial studbook Argyll Argyll archbishop drink sitcom restful Argyll stubborn objects scoffs actinolite quinta stump pigs electric andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang andrew angelo ang fdafda fdafa dsafsaf dsaf dsaf dsa fs af sa fsa fsdaf dasf dsa f dsaf dsa fds afd sa fdsa fdsafasfdsa fas fd saf da fd afdsafdasfdasf dsaf das f asf as[x").should == { :invalid => true }
|
58
|
+
|
59
|
+
GSMTools.gsm_string_counter("你好我是洪一尧ANG").should == { :length => 10,
|
60
|
+
:segments => 1,
|
61
|
+
:remaining_length_for_segment => 60 }
|
62
|
+
|
63
|
+
GSMTools.gsm_string_counter("你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG").should == { :length => 70,
|
64
|
+
:segments => 1,
|
65
|
+
:remaining_length_for_segment => 0 }
|
66
|
+
|
67
|
+
GSMTools.gsm_string_counter("你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG!").should == { :length => 71,
|
68
|
+
:segments => 2,
|
69
|
+
:remaining_length_for_segment => 63 }
|
70
|
+
|
71
|
+
GSMTools.gsm_string_counter("你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG!!!!").should == { :length => 134,
|
72
|
+
:segments => 2,
|
73
|
+
:remaining_length_for_segment => 0 }
|
74
|
+
|
75
|
+
GSMTools.gsm_string_counter("你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG[").should == { :length => 201,
|
76
|
+
:segments => 3,
|
77
|
+
:remaining_length_for_segment => 0 }
|
78
|
+
|
79
|
+
GSMTools.gsm_string_counter("你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG你好我是洪一尧ANG!!").should == { :invalid => true }
|
80
|
+
end
|
81
|
+
end
|
data/spec/spec_helper.rb
ADDED
@@ -0,0 +1,146 @@
|
|
1
|
+
#encoding: utf-8
|
2
|
+
|
3
|
+
require 'gsm_tools'
|
4
|
+
|
5
|
+
GSM_7BIT_CHARS = "@£$¥èéùìòÇ\nØø\rÅåΔ_ΦΓΛΩΠΨΣΘΞÆæßÉ !\"#¤%&'()*+,-./0123456789:;<=>?¡ABCDEFGHIJKLMNOPQRSTUVWXYZÄÖÑܧ¿abcdefghijklmnopqrstuvwxyzäöñüà^{}\\[~]|€"
|
6
|
+
|
7
|
+
GSM_UNICODE_MAP = [
|
8
|
+
0x00, 0x0040, # COMMERCIAL AT
|
9
|
+
0x01, 0x00A3, # POUND SIGN
|
10
|
+
0x02, 0x0024, # DOLLAR SIGN
|
11
|
+
0x03, 0x00A5, # YEN SIGN
|
12
|
+
0x04, 0x00E8, # LATIN SMALL LETTER E WITH GRAVE
|
13
|
+
0x05, 0x00E9, # LATIN SMALL LETTER E WITH ACUTE
|
14
|
+
0x06, 0x00F9, # LATIN SMALL LETTER U WITH GRAVE
|
15
|
+
0x07, 0x00EC, # LATIN SMALL LETTER I WITH GRAVE
|
16
|
+
0x08, 0x00F2, # LATIN SMALL LETTER O WITH GRAVE
|
17
|
+
#0x09, 0x00E7, # LATIN SMALL LETTER C WITH CEDILLA
|
18
|
+
0x09, 0x00C7, # LATIN CAPITAL LETTER C WITH CEDILLA (see note above), THIS SHOULD BE THE CORRECT ONE
|
19
|
+
0x0A, 0x000A, # LINE FEED
|
20
|
+
0x0B, 0x00D8, # LATIN CAPITAL LETTER O WITH STROKE
|
21
|
+
0x0C, 0x00F8, # LATIN SMALL LETTER O WITH STROKE
|
22
|
+
0x0D, 0x000D, # CARRIAGE RETURN
|
23
|
+
0x0E, 0x00C5, # LATIN CAPITAL LETTER A WITH RING ABOVE
|
24
|
+
0x0F, 0x00E5, # LATIN SMALL LETTER A WITH RING ABOVE
|
25
|
+
0x10, 0x0394, # GREEK CAPITAL LETTER DELTA
|
26
|
+
0x11, 0x005F, # LOW LINE
|
27
|
+
0x12, 0x03A6, # GREEK CAPITAL LETTER PHI
|
28
|
+
0x13, 0x0393, # GREEK CAPITAL LETTER GAMMA
|
29
|
+
0x14, 0x039B, # GREEK CAPITAL LETTER LAMDA
|
30
|
+
0x15, 0x03A9, # GREEK CAPITAL LETTER OMEGA
|
31
|
+
0x16, 0x03A0, # GREEK CAPITAL LETTER PI
|
32
|
+
0x17, 0x03A8, # GREEK CAPITAL LETTER PSI
|
33
|
+
0x18, 0x03A3, # GREEK CAPITAL LETTER SIGMA
|
34
|
+
0x19, 0x0398, # GREEK CAPITAL LETTER THETA
|
35
|
+
0x1A, 0x039E, # GREEK CAPITAL LETTER XI
|
36
|
+
#[0x1B, 0x0A], 0x000C, # FORM FEED # Removed for now, Need to figure out how to Regex match this character
|
37
|
+
[0x1B, 0x14], 0x005E, # CIRCUMFLEX ACCENT
|
38
|
+
[0x1B, 0x28], 0x007B, # LEFT CURLY BRACKET
|
39
|
+
[0x1B, 0x29], 0x007D, # RIGHT CURLY BRACKET
|
40
|
+
[0x1B, 0x2F], 0x005C, # REVERSE SOLIDUS
|
41
|
+
[0x1B, 0x3C], 0x005B, # LEFT SQUARE BRACKET
|
42
|
+
[0x1B, 0x3D], 0x007E, # TILDE
|
43
|
+
[0x1B, 0x3E], 0x005D, # RIGHT SQUARE BRACKET
|
44
|
+
[0x1B, 0x40], 0x007C, # VERTICAL LINE
|
45
|
+
[0x1B, 0x65], 0x20AC, # EURO SIGN
|
46
|
+
0x1C, 0x00C6, # LATIN CAPITAL LETTER AE
|
47
|
+
0x1D, 0x00E6, # LATIN SMALL LETTER AE
|
48
|
+
0x1E, 0x00DF, # LATIN SMALL LETTER SHARP S (German)
|
49
|
+
0x1F, 0x00C9, # LATIN CAPITAL LETTER E WITH ACUTE
|
50
|
+
0x20, 0x0020, # SPACE
|
51
|
+
0x21, 0x0021, # EXCLAMATION MARK
|
52
|
+
0x22, 0x0022, # QUOTATION MARK
|
53
|
+
0x23, 0x0023, # NUMBER SIGN
|
54
|
+
0x24, 0x00A4, # CURRENCY SIGN
|
55
|
+
0x25, 0x0025, # PERCENT SIGN
|
56
|
+
0x26, 0x0026, # AMPERSAND
|
57
|
+
0x27, 0x0027, # APOSTROPHE
|
58
|
+
0x28, 0x0028, # LEFT PARENTHESIS
|
59
|
+
0x29, 0x0029, # RIGHT PARENTHESIS
|
60
|
+
0x2A, 0x002A, # ASTERISK
|
61
|
+
0x2B, 0x002B, # PLUS SIGN
|
62
|
+
0x2C, 0x002C, # COMMA
|
63
|
+
0x2D, 0x002D, # HYPHEN-MINUS
|
64
|
+
0x2E, 0x002E, # FULL STOP
|
65
|
+
0x2F, 0x002F, # SOLIDUS
|
66
|
+
0x30, 0x0030, # DIGIT ZERO
|
67
|
+
0x31, 0x0031, # DIGIT ONE
|
68
|
+
0x32, 0x0032, # DIGIT TWO
|
69
|
+
0x33, 0x0033, # DIGIT THREE
|
70
|
+
0x34, 0x0034, # DIGIT FOUR
|
71
|
+
0x35, 0x0035, # DIGIT FIVE
|
72
|
+
0x36, 0x0036, # DIGIT SIX
|
73
|
+
0x37, 0x0037, # DIGIT SEVEN
|
74
|
+
0x38, 0x0038, # DIGIT EIGHT
|
75
|
+
0x39, 0x0039, # DIGIT NINE
|
76
|
+
0x3A, 0x003A, # COLON
|
77
|
+
0x3B, 0x003B, # SEMICOLON
|
78
|
+
0x3C, 0x003C, # LESS-THAN SIGN
|
79
|
+
0x3D, 0x003D, # EQUALS SIGN
|
80
|
+
0x3E, 0x003E, # GREATER-THAN SIGN
|
81
|
+
0x3F, 0x003F, # QUESTION MARK
|
82
|
+
0x40, 0x00A1, # INVERTED EXCLAMATION MARK
|
83
|
+
0x41, 0x0041, # LATIN CAPITAL LETTER A
|
84
|
+
0x42, 0x0042, # LATIN CAPITAL LETTER B
|
85
|
+
0x43, 0x0043, # LATIN CAPITAL LETTER C
|
86
|
+
0x44, 0x0044, # LATIN CAPITAL LETTER D
|
87
|
+
0x45, 0x0045, # LATIN CAPITAL LETTER E
|
88
|
+
0x46, 0x0046, # LATIN CAPITAL LETTER F
|
89
|
+
0x47, 0x0047, # LATIN CAPITAL LETTER G
|
90
|
+
0x48, 0x0048, # LATIN CAPITAL LETTER H
|
91
|
+
0x49, 0x0049, # LATIN CAPITAL LETTER I
|
92
|
+
0x4A, 0x004A, # LATIN CAPITAL LETTER J
|
93
|
+
0x4B, 0x004B, # LATIN CAPITAL LETTER K
|
94
|
+
0x4C, 0x004C, # LATIN CAPITAL LETTER L
|
95
|
+
0x4D, 0x004D, # LATIN CAPITAL LETTER M
|
96
|
+
0x4E, 0x004E, # LATIN CAPITAL LETTER N
|
97
|
+
0x4F, 0x004F, # LATIN CAPITAL LETTER O
|
98
|
+
0x50, 0x0050, # LATIN CAPITAL LETTER P
|
99
|
+
0x51, 0x0051, # LATIN CAPITAL LETTER Q
|
100
|
+
0x52, 0x0052, # LATIN CAPITAL LETTER R
|
101
|
+
0x53, 0x0053, # LATIN CAPITAL LETTER S
|
102
|
+
0x54, 0x0054, # LATIN CAPITAL LETTER T
|
103
|
+
0x55, 0x0055, # LATIN CAPITAL LETTER U
|
104
|
+
0x56, 0x0056, # LATIN CAPITAL LETTER V
|
105
|
+
0x57, 0x0057, # LATIN CAPITAL LETTER W
|
106
|
+
0x58, 0x0058, # LATIN CAPITAL LETTER X
|
107
|
+
0x59, 0x0059, # LATIN CAPITAL LETTER Y
|
108
|
+
0x5A, 0x005A, # LATIN CAPITAL LETTER Z
|
109
|
+
0x5B, 0x00C4, # LATIN CAPITAL LETTER A WITH DIAERESIS
|
110
|
+
0x5C, 0x00D6, # LATIN CAPITAL LETTER O WITH DIAERESIS
|
111
|
+
0x5D, 0x00D1, # LATIN CAPITAL LETTER N WITH TILDE
|
112
|
+
0x5E, 0x00DC, # LATIN CAPITAL LETTER U WITH DIAERESIS
|
113
|
+
0x5F, 0x00A7, # SECTION SIGN
|
114
|
+
0x60, 0x00BF, # INVERTED QUESTION MARK
|
115
|
+
0x61, 0x0061, # LATIN SMALL LETTER A
|
116
|
+
0x62, 0x0062, # LATIN SMALL LETTER B
|
117
|
+
0x63, 0x0063, # LATIN SMALL LETTER C
|
118
|
+
0x64, 0x0064, # LATIN SMALL LETTER D
|
119
|
+
0x65, 0x0065, # LATIN SMALL LETTER E
|
120
|
+
0x66, 0x0066, # LATIN SMALL LETTER F
|
121
|
+
0x67, 0x0067, # LATIN SMALL LETTER G
|
122
|
+
0x68, 0x0068, # LATIN SMALL LETTER H
|
123
|
+
0x69, 0x0069, # LATIN SMALL LETTER I
|
124
|
+
0x6A, 0x006A, # LATIN SMALL LETTER J
|
125
|
+
0x6B, 0x006B, # LATIN SMALL LETTER K
|
126
|
+
0x6C, 0x006C, # LATIN SMALL LETTER L
|
127
|
+
0x6D, 0x006D, # LATIN SMALL LETTER M
|
128
|
+
0x6E, 0x006E, # LATIN SMALL LETTER N
|
129
|
+
0x6F, 0x006F, # LATIN SMALL LETTER O
|
130
|
+
0x70, 0x0070, # LATIN SMALL LETTER P
|
131
|
+
0x71, 0x0071, # LATIN SMALL LETTER Q
|
132
|
+
0x72, 0x0072, # LATIN SMALL LETTER R
|
133
|
+
0x73, 0x0073, # LATIN SMALL LETTER S
|
134
|
+
0x74, 0x0074, # LATIN SMALL LETTER T
|
135
|
+
0x75, 0x0075, # LATIN SMALL LETTER U
|
136
|
+
0x76, 0x0076, # LATIN SMALL LETTER V
|
137
|
+
0x77, 0x0077, # LATIN SMALL LETTER W
|
138
|
+
0x78, 0x0078, # LATIN SMALL LETTER X
|
139
|
+
0x79, 0x0079, # LATIN SMALL LETTER Y
|
140
|
+
0x7A, 0x007A, # LATIN SMALL LETTER Z
|
141
|
+
0x7B, 0x00E4, # LATIN SMALL LETTER A WITH DIAERESIS
|
142
|
+
0x7C, 0x00F6, # LATIN SMALL LETTER O WITH DIAERESIS
|
143
|
+
0x7D, 0x00F1, # LATIN SMALL LETTER N WITH TILDE
|
144
|
+
0x7E, 0x00FC, # LATIN SMALL LETTER U WITH DIAERESIS
|
145
|
+
0x7F, 0x00E0, # LATIN SMALL LETTER A WITH GRAVE
|
146
|
+
]
|
metadata
ADDED
@@ -0,0 +1,91 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: gsm_tools
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Andrew Ang
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2012-01-09 00:00:00.000000000Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: rspec
|
16
|
+
requirement: &70344682777260 !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: *70344682777260
|
25
|
+
- !ruby/object:Gem::Dependency
|
26
|
+
name: guard-rspec
|
27
|
+
requirement: &70344682781580 !ruby/object:Gem::Requirement
|
28
|
+
none: false
|
29
|
+
requirements:
|
30
|
+
- - ! '>='
|
31
|
+
- !ruby/object:Gem::Version
|
32
|
+
version: '0'
|
33
|
+
type: :development
|
34
|
+
prerelease: false
|
35
|
+
version_requirements: *70344682781580
|
36
|
+
description: Adds various methods to check GSM 03.38 character encoding compatibility.
|
37
|
+
email:
|
38
|
+
- andukz@gmail.com
|
39
|
+
executables: []
|
40
|
+
extensions: []
|
41
|
+
extra_rdoc_files: []
|
42
|
+
files:
|
43
|
+
- .gitignore
|
44
|
+
- .rvmrc
|
45
|
+
- Gemfile
|
46
|
+
- Guardfile
|
47
|
+
- Rakefile
|
48
|
+
- gsm_tools.gemspec
|
49
|
+
- lib/gsm_tools.rb
|
50
|
+
- lib/gsm_tools/base.rb
|
51
|
+
- lib/gsm_tools/string.rb
|
52
|
+
- lib/gsm_tools/version.rb
|
53
|
+
- spec/lib/gsm_tools/base_spec.rb
|
54
|
+
- spec/lib/gsm_tools/string_spec.rb
|
55
|
+
- spec/lib/gsm_tools_spec.rb
|
56
|
+
- spec/spec_helper.rb
|
57
|
+
homepage: ''
|
58
|
+
licenses: []
|
59
|
+
post_install_message:
|
60
|
+
rdoc_options: []
|
61
|
+
require_paths:
|
62
|
+
- lib
|
63
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
64
|
+
none: false
|
65
|
+
requirements:
|
66
|
+
- - ! '>='
|
67
|
+
- !ruby/object:Gem::Version
|
68
|
+
version: '0'
|
69
|
+
segments:
|
70
|
+
- 0
|
71
|
+
hash: -1888712993988250446
|
72
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
73
|
+
none: false
|
74
|
+
requirements:
|
75
|
+
- - ! '>='
|
76
|
+
- !ruby/object:Gem::Version
|
77
|
+
version: '0'
|
78
|
+
segments:
|
79
|
+
- 0
|
80
|
+
hash: -1888712993988250446
|
81
|
+
requirements: []
|
82
|
+
rubyforge_project: gsm_tools
|
83
|
+
rubygems_version: 1.8.10
|
84
|
+
signing_key:
|
85
|
+
specification_version: 3
|
86
|
+
summary: GSM 03.38 extension libraries
|
87
|
+
test_files:
|
88
|
+
- spec/lib/gsm_tools/base_spec.rb
|
89
|
+
- spec/lib/gsm_tools/string_spec.rb
|
90
|
+
- spec/lib/gsm_tools_spec.rb
|
91
|
+
- spec/spec_helper.rb
|