chardet 0.9.0
Sign up to get free protection for your applications and to get access to all the features.
- data/COPYING +504 -0
- data/README +12 -0
- data/lib/Big5Freq.rb +913 -0
- data/lib/Big5Prober.rb +48 -0
- data/lib/CharDistributionAnalysis.rb +245 -0
- data/lib/CharSetGroupProber.rb +114 -0
- data/lib/CharSetProber.rb +70 -0
- data/lib/CodingStateMachine.rb +74 -0
- data/lib/ESCSM.rb +242 -0
- data/lib/EUCJPProber.rb +97 -0
- data/lib/EUCKRFreq.rb +600 -0
- data/lib/EUCKRProber.rb +48 -0
- data/lib/EUCTWFreq.rb +432 -0
- data/lib/EUCTWProber.rb +48 -0
- data/lib/EscCharSetProber.rb +94 -0
- data/lib/GB2312Freq.rb +475 -0
- data/lib/GB2312Prober.rb +48 -0
- data/lib/HebrewProber.rb +292 -0
- data/lib/JISFreq.rb +573 -0
- data/lib/JapaneseContextAnalysis.rb +234 -0
- data/lib/LangBulgarianModel.rb +231 -0
- data/lib/LangCyrillicModel.rb +332 -0
- data/lib/LangGreekModel.rb +229 -0
- data/lib/LangHebrewModel.rb +202 -0
- data/lib/LangHungarianModel.rb +228 -0
- data/lib/LangThaiModel.rb +203 -0
- data/lib/Latin1Prober.rb +160 -0
- data/lib/MBCSGroupProber.rb +57 -0
- data/lib/MBCSSM.rb +513 -0
- data/lib/MultiByteCharSetProber.rb +94 -0
- data/lib/SBCSGroupProber.rb +71 -0
- data/lib/SJISProber.rb +99 -0
- data/lib/SingleByteCharSetProber.rb +131 -0
- data/lib/UTF8Prober.rb +91 -0
- data/lib/UniversalDetector.rb +209 -0
- data/python-docs/css/chardet.css +299 -0
- data/python-docs/faq.html +107 -0
- data/python-docs/how-it-works.html +113 -0
- data/python-docs/images/caution.png +0 -0
- data/python-docs/images/important.png +0 -0
- data/python-docs/images/note.png +0 -0
- data/python-docs/images/permalink.gif +0 -0
- data/python-docs/images/tip.png +0 -0
- data/python-docs/images/warning.png +0 -0
- data/python-docs/index.html +73 -0
- data/python-docs/license.html +62 -0
- data/python-docs/supported-encodings.html +86 -0
- data/python-docs/usage.html +107 -0
- metadata +86 -0
data/lib/Big5Prober.rb
ADDED
@@ -0,0 +1,48 @@
|
|
1
|
+
######################## BEGIN LICENSE BLOCK ########################
|
2
|
+
# The Original Code is mozilla.org code.
|
3
|
+
#
|
4
|
+
# The Initial Developer of the Original Code is
|
5
|
+
# Netscape Communications Corporation.
|
6
|
+
# Portions created by the Initial Developer are Copyright (C) 1998
|
7
|
+
# the Initial Developer. All Rights Reserved.
|
8
|
+
#
|
9
|
+
# Contributor(s):
|
10
|
+
# Hui (zhengzhengzheng@gmail.com) - port to Ruby
|
11
|
+
# Mark Pilgrim - first port to Python
|
12
|
+
#
|
13
|
+
# This library is free software; you can redistribute it and/or
|
14
|
+
# modify it under the terms of the GNU Lesser General Public
|
15
|
+
# License as published by the Free Software Foundation; either
|
16
|
+
# version 2.1 of the License, or (at your option) any later version.
|
17
|
+
#
|
18
|
+
# This library is distributed in the hope that it will be useful,
|
19
|
+
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
20
|
+
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
21
|
+
# Lesser General Public License for more details.
|
22
|
+
#
|
23
|
+
# You should have received a copy of the GNU Lesser General Public
|
24
|
+
# License along with this library; if not, write to the Free Software
|
25
|
+
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
|
26
|
+
# 02110-1301 USA
|
27
|
+
######################### END LICENSE BLOCK #########################
|
28
|
+
|
29
|
+
require 'UniversalDetector'
|
30
|
+
require 'MultiByteCharSetProber'
|
31
|
+
require 'CodingStateMachine'
|
32
|
+
require 'CharDistributionAnalysis'
|
33
|
+
require 'MBCSSM'
|
34
|
+
|
35
|
+
module UniversalDetector
|
36
|
+
class Big5Prober < MultiByteCharSetProber
|
37
|
+
def initialize
|
38
|
+
super
|
39
|
+
@_mCodingSM = CodingStateMachine.new(Big5SMModel)
|
40
|
+
@_mDistributionAnalyzer = Big5DistributionAnalysis.new
|
41
|
+
reset()
|
42
|
+
end
|
43
|
+
|
44
|
+
def get_charset_name
|
45
|
+
return "Big5"
|
46
|
+
end
|
47
|
+
end
|
48
|
+
end
|
@@ -0,0 +1,245 @@
|
|
1
|
+
######################## BEGIN LICENSE BLOCK ########################
|
2
|
+
# The Original Code is mozilla.org code.
|
3
|
+
#
|
4
|
+
# The Initial Developer of the Original Code is
|
5
|
+
# Netscape Communications Corporation.
|
6
|
+
# Portions created by the Initial Developer are Copyright (C) 1998
|
7
|
+
# the Initial Developer. All Rights Reserved.
|
8
|
+
#
|
9
|
+
# Contributor(s):
|
10
|
+
# Hui (zhengzhengzheng@gmail.com) - port to Ruby
|
11
|
+
# Mark Pilgrim - first port to Python
|
12
|
+
#
|
13
|
+
# This library is free software; you can redistribute it and/or
|
14
|
+
# modify it under the terms of the GNU Lesser General Public
|
15
|
+
# License as published by the Free Software Foundation; either
|
16
|
+
# version 2.1 of the License, or (at your option) any later version.
|
17
|
+
#
|
18
|
+
# This library is distributed in the hope that it will be useful,
|
19
|
+
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
20
|
+
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
21
|
+
# Lesser General Public License for more details.
|
22
|
+
#
|
23
|
+
# You should have received a copy of the GNU Lesser General Public
|
24
|
+
# License along with this library; if not, write to the Free Software
|
25
|
+
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
|
26
|
+
# 02110-1301 USA
|
27
|
+
######################### END LICENSE BLOCK #########################
|
28
|
+
|
29
|
+
require "UniversalDetector"
|
30
|
+
require "EUCTWFreq"
|
31
|
+
require "EUCKRFreq"
|
32
|
+
require "GB2312Freq"
|
33
|
+
require "Big5Freq"
|
34
|
+
require "JISFreq"
|
35
|
+
|
36
|
+
module UniversalDetector
|
37
|
+
class CharDistributionAnalysis
|
38
|
+
|
39
|
+
ENOUGH_DATA_THRESHOLD = 1024
|
40
|
+
SURE_YES = 0.99
|
41
|
+
SURE_NO = 0.01
|
42
|
+
|
43
|
+
def initialize
|
44
|
+
@_mCharToFreqOrder = nil # Mapping table to get frequency order from char order (get from GetOrder())
|
45
|
+
@_mTableSize = nil # Size of above table
|
46
|
+
@_mTypicalDistributionRatio = nil # This is a constant value which varies from language to language, used in calculating confidence. See http://www.mozilla.org/projects/intl/UniversalCharsetDetection.html for further detail.
|
47
|
+
reset()
|
48
|
+
end
|
49
|
+
|
50
|
+
def reset
|
51
|
+
#"""reset analyser, clear any state"""
|
52
|
+
@_mDone = false # If this flag is set to constants.True, detection is done and conclusion has been made
|
53
|
+
@_mTotalChars = 0 # Total characters encountered
|
54
|
+
@_mFreqChars = 0 # The number of characters whose frequency order is less than 512
|
55
|
+
end
|
56
|
+
|
57
|
+
def feed(aStr, aCharLen)
|
58
|
+
#"""feed a character with known length"""
|
59
|
+
if aCharLen == 2
|
60
|
+
# we only care about 2-bytes character in our distribution analysis
|
61
|
+
order = get_order(aStr)
|
62
|
+
else
|
63
|
+
order = -1
|
64
|
+
end
|
65
|
+
if order >= 0
|
66
|
+
@_mTotalChars += 1
|
67
|
+
# order is valid
|
68
|
+
if order < @_mTableSize
|
69
|
+
if 512 > @_mCharToFreqOrder[order]
|
70
|
+
@_mFreqChars += 1
|
71
|
+
end
|
72
|
+
end
|
73
|
+
end
|
74
|
+
end
|
75
|
+
|
76
|
+
def get_confidence
|
77
|
+
#"""return confidence based on existing data"""
|
78
|
+
# if we didn"t receive any character in our consideration range, return negative answer
|
79
|
+
if @_mTotalChars <= 0
|
80
|
+
return SURE_NO
|
81
|
+
end
|
82
|
+
|
83
|
+
if @_mTotalChars != @_mFreqChars
|
84
|
+
r = @_mFreqChars / ((@_mTotalChars - @_mFreqChars) * @_mTypicalDistributionRatio)
|
85
|
+
if r < SURE_YES
|
86
|
+
return r
|
87
|
+
end
|
88
|
+
end
|
89
|
+
|
90
|
+
# normalize confidence (we don"t want to be 100% sure)
|
91
|
+
return SURE_YES
|
92
|
+
end
|
93
|
+
|
94
|
+
def got_enough_data
|
95
|
+
# It is not necessary to receive all data to draw conclusion. For charset detection,
|
96
|
+
# certain amount of data is enough
|
97
|
+
return @_mTotalChars > ENOUGH_DATA_THRESHOLD
|
98
|
+
end
|
99
|
+
|
100
|
+
def get_order(aStr)
|
101
|
+
# We do not handle characters based on the original encoding string, but
|
102
|
+
# convert this encoding string to a number, here called order.
|
103
|
+
# This allows multiple encodings of a language to share one frequency table.
|
104
|
+
return -1
|
105
|
+
end
|
106
|
+
end
|
107
|
+
|
108
|
+
class EUCTWDistributionAnalysis < CharDistributionAnalysis
|
109
|
+
def initialize
|
110
|
+
super
|
111
|
+
@_mCharToFreqOrder = EUCTWCharToFreqOrder
|
112
|
+
@_mTableSize = EUCTW_TABLE_SIZE
|
113
|
+
@_mTypicalDistributionRatio = EUCTW_TYPICAL_DISTRIBUTION_RATIO
|
114
|
+
end
|
115
|
+
|
116
|
+
def get_order(aStr)
|
117
|
+
# for euc-TW encoding, we are interested
|
118
|
+
# first byte range: 0xc4 -- 0xfe
|
119
|
+
# second byte range: 0xa1 -- 0xfe
|
120
|
+
# no validation needed here. State machine has done that
|
121
|
+
if aStr[0] >= 0xC4
|
122
|
+
return 94 * (aStr[0] - 0xC4) + aStr[1] - 0xA1
|
123
|
+
else
|
124
|
+
return -1
|
125
|
+
end
|
126
|
+
end
|
127
|
+
end
|
128
|
+
|
129
|
+
class EUCKRDistributionAnalysis < CharDistributionAnalysis
|
130
|
+
def initialize
|
131
|
+
super
|
132
|
+
@_mCharToFreqOrder = EUCKRCharToFreqOrder
|
133
|
+
@_mTableSize = EUCKR_TABLE_SIZE
|
134
|
+
@_mTypicalDistributionRatio = EUCKR_TYPICAL_DISTRIBUTION_RATIO
|
135
|
+
end
|
136
|
+
|
137
|
+
def get_order(aStr)
|
138
|
+
# for euc-KR encoding, we are interested
|
139
|
+
# first byte range: 0xb0 -- 0xfe
|
140
|
+
# second byte range: 0xa1 -- 0xfe
|
141
|
+
# no validation needed here. State machine has done that
|
142
|
+
if aStr[0] >= 0xB0
|
143
|
+
return 94 * (aStr[0] - 0xB0) + aStr[1] - 0xA1
|
144
|
+
else
|
145
|
+
return -1;
|
146
|
+
end
|
147
|
+
end
|
148
|
+
end
|
149
|
+
|
150
|
+
class GB2312DistributionAnalysis < CharDistributionAnalysis
|
151
|
+
def initialize
|
152
|
+
super
|
153
|
+
@_mCharToFreqOrder = GB2312CharToFreqOrder
|
154
|
+
@_mTableSize = GB2312_TABLE_SIZE
|
155
|
+
@_mTypicalDistributionRatio = GB2312_TYPICAL_DISTRIBUTION_RATIO
|
156
|
+
end
|
157
|
+
|
158
|
+
def get_order(aStr)
|
159
|
+
# for GB2312 encoding, we are interested
|
160
|
+
# first byte range: 0xb0 -- 0xfe
|
161
|
+
# second byte range: 0xa1 -- 0xfe
|
162
|
+
# no validation needed here. State machine has done that
|
163
|
+
if (aStr[0] >= 0xB0) and (aStr[1] >= 0xA1)
|
164
|
+
return 94 * (aStr[0] - 0xB0) + aStr[1] - 0xA1
|
165
|
+
else
|
166
|
+
return -1;
|
167
|
+
end
|
168
|
+
end
|
169
|
+
end
|
170
|
+
|
171
|
+
class Big5DistributionAnalysis < CharDistributionAnalysis
|
172
|
+
def initialize
|
173
|
+
super
|
174
|
+
@_mCharToFreqOrder = Big5CharToFreqOrder
|
175
|
+
@_mTableSize = BIG5_TABLE_SIZE
|
176
|
+
@_mTypicalDistributionRatio = BIG5_TYPICAL_DISTRIBUTION_RATIO
|
177
|
+
end
|
178
|
+
|
179
|
+
def get_order(aStr)
|
180
|
+
# for big5 encoding, we are interested
|
181
|
+
# first byte range: 0xa4 -- 0xfe
|
182
|
+
# second byte range: 0x40 -- 0x7e , 0xa1 -- 0xfe
|
183
|
+
# no validation needed here. State machine has done that
|
184
|
+
if aStr[0] >= 0xA4
|
185
|
+
if aStr[1] >= 0xA1
|
186
|
+
return 157 * (aStr[0] - 0xA4) + aStr[1] - 0xA1 + 63
|
187
|
+
else
|
188
|
+
return 157 * (aStr[0] - 0xA4) + aStr[1] - 0x40
|
189
|
+
end
|
190
|
+
else
|
191
|
+
return -1
|
192
|
+
end
|
193
|
+
end
|
194
|
+
end
|
195
|
+
|
196
|
+
class SJISDistributionAnalysis < CharDistributionAnalysis
|
197
|
+
def initialize
|
198
|
+
super
|
199
|
+
@_mCharToFreqOrder = JISCharToFreqOrder
|
200
|
+
@_mTableSize = JIS_TABLE_SIZE
|
201
|
+
@_mTypicalDistributionRatio = JIS_TYPICAL_DISTRIBUTION_RATIO
|
202
|
+
end
|
203
|
+
|
204
|
+
def get_order(aStr)
|
205
|
+
# for sjis encoding, we are interested
|
206
|
+
# first byte range: 0x81 -- 0x9f , 0xe0 -- 0xfe
|
207
|
+
# second byte range: 0x40 -- 0x7e, 0x81 -- oxfe
|
208
|
+
# no validation needed here. State machine has done that
|
209
|
+
if (aStr[0] >= 0x81) && (aStr[0] <= 0x9F)
|
210
|
+
order = 188 * (aStr[0] - 0x81)
|
211
|
+
elsif (aStr[0] >= 0xE0) and (aStr[0] <= 0xEF)
|
212
|
+
order = 188 * (aStr[0] - 0xE0 + 31)
|
213
|
+
else
|
214
|
+
return -1;
|
215
|
+
end
|
216
|
+
order = order + aStr[1] - 0x40
|
217
|
+
if aStr[1] > 0x7F
|
218
|
+
order =- 1
|
219
|
+
end
|
220
|
+
return order
|
221
|
+
end
|
222
|
+
end
|
223
|
+
|
224
|
+
class EUCJPDistributionAnalysis < CharDistributionAnalysis
|
225
|
+
def initialize
|
226
|
+
super
|
227
|
+
@_mCharToFreqOrder = JISCharToFreqOrder
|
228
|
+
@_mTableSize = JIS_TABLE_SIZE
|
229
|
+
@_mTypicalDistributionRatio = JIS_TYPICAL_DISTRIBUTION_RATIO
|
230
|
+
end
|
231
|
+
|
232
|
+
def get_order(aStr)
|
233
|
+
# for euc-JP encoding, we are interested
|
234
|
+
# first byte range: 0xa0 -- 0xfe
|
235
|
+
# second byte range: 0xa1 -- 0xfe
|
236
|
+
# no validation needed here. State machine has done that
|
237
|
+
if aStr[0] >= 0xA0
|
238
|
+
return 94 * (aStr[0] - 0xA1) + aStr[1] - 0xa1
|
239
|
+
else
|
240
|
+
return -1
|
241
|
+
end
|
242
|
+
end
|
243
|
+
end
|
244
|
+
|
245
|
+
end
|
@@ -0,0 +1,114 @@
|
|
1
|
+
######################## BEGIN LICENSE BLOCK ########################
|
2
|
+
# The Original Code is mozilla.org code.
|
3
|
+
#
|
4
|
+
# The Initial Developer of the Original Code is
|
5
|
+
# Netscape Communications Corporation.
|
6
|
+
# Portions created by the Initial Developer are Copyright (C) 1998
|
7
|
+
# the Initial Developer. All Rights Reserved.
|
8
|
+
#
|
9
|
+
# Contributor(s):
|
10
|
+
# Hui (zhengzhengzheng@gmail.com) - port to Ruby
|
11
|
+
# Mark Pilgrim - first port to Python
|
12
|
+
#
|
13
|
+
# This library is free software; you can redistribute it and/or
|
14
|
+
# modify it under the terms of the GNU Lesser General Public
|
15
|
+
# License as published by the Free Software Foundation; either
|
16
|
+
# version 2.1 of the License, or (at your option) any later version.
|
17
|
+
#
|
18
|
+
# This library is distributed in the hope that it will be useful,
|
19
|
+
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
20
|
+
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
21
|
+
# Lesser General Public License for more details.
|
22
|
+
#
|
23
|
+
# You should have received a copy of the GNU Lesser General Public
|
24
|
+
# License along with this library; if not, write to the Free Software
|
25
|
+
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
|
26
|
+
# 02110-1301 USA
|
27
|
+
######################### END LICENSE BLOCK #########################
|
28
|
+
|
29
|
+
require 'UniversalDetector'
|
30
|
+
require 'CharSetProber'
|
31
|
+
|
32
|
+
module UniversalDetector
|
33
|
+
class CharSetGroupProber < CharSetProber
|
34
|
+
def initialize
|
35
|
+
@_mActiveNum = 0
|
36
|
+
@_mProbers = []
|
37
|
+
@_mBestGuessProber = nil
|
38
|
+
end
|
39
|
+
|
40
|
+
def reset
|
41
|
+
super
|
42
|
+
@_mActiveNum = 0
|
43
|
+
for prober in @_mProbers
|
44
|
+
if prober
|
45
|
+
prober.reset()
|
46
|
+
prober.active = true
|
47
|
+
@_mActiveNum += 1
|
48
|
+
end
|
49
|
+
end
|
50
|
+
@_mBestGuessProber = nil
|
51
|
+
end
|
52
|
+
|
53
|
+
def get_charset_name
|
54
|
+
unless @_mBestGuessProber
|
55
|
+
get_confidence()
|
56
|
+
unless @_mBestGuessProber then return nil end
|
57
|
+
# @_mBestGuessProber = @_mProbers[0]
|
58
|
+
end
|
59
|
+
return @_mBestGuessProber.get_charset_name()
|
60
|
+
end
|
61
|
+
|
62
|
+
def feed(aBuf)
|
63
|
+
for prober in @_mProbers
|
64
|
+
unless prober then next end
|
65
|
+
unless prober.active then next end
|
66
|
+
st = prober.feed(aBuf)
|
67
|
+
unless st then next end
|
68
|
+
if st == :FoundIt
|
69
|
+
@_mBestGuessProber = prober
|
70
|
+
return get_state()
|
71
|
+
elsif st == :NotMe
|
72
|
+
prober.active = false
|
73
|
+
@_mActiveNum -= 1
|
74
|
+
if @_mActiveNum <= 0
|
75
|
+
@_mState = :NotMe
|
76
|
+
return get_state()
|
77
|
+
end
|
78
|
+
end
|
79
|
+
end
|
80
|
+
return get_state()
|
81
|
+
end
|
82
|
+
|
83
|
+
def get_confidence
|
84
|
+
st = get_state()
|
85
|
+
if st == :FoundIt
|
86
|
+
return 0.99
|
87
|
+
elsif st == :NotMe
|
88
|
+
return 0.01
|
89
|
+
end
|
90
|
+
|
91
|
+
bestConf = 0.0
|
92
|
+
@_mBestGuessProber = nil
|
93
|
+
for prober in @_mProbers
|
94
|
+
unless prober then next end
|
95
|
+
unless prober.active
|
96
|
+
if UniversalDetector::DEBUG
|
97
|
+
p(prober.get_charset_name() + ' not active\n')
|
98
|
+
end
|
99
|
+
next
|
100
|
+
end
|
101
|
+
cf = prober.get_confidence()
|
102
|
+
if UniversalDetector::DEBUG
|
103
|
+
p('%s confidence = %s\n' % [prober.get_charset_name(), cf])
|
104
|
+
end
|
105
|
+
if bestConf < cf
|
106
|
+
bestConf = cf
|
107
|
+
@_mBestGuessProber = prober
|
108
|
+
end
|
109
|
+
end
|
110
|
+
unless @_mBestGuessProber then return 0.0 end
|
111
|
+
return bestConf
|
112
|
+
end
|
113
|
+
end
|
114
|
+
end
|
@@ -0,0 +1,70 @@
|
|
1
|
+
######################## BEGIN LICENSE BLOCK ########################
|
2
|
+
# The Original Code is mozilla.org code.
|
3
|
+
#
|
4
|
+
# The Initial Developer of the Original Code is
|
5
|
+
# Netscape Communications Corporation.
|
6
|
+
# Portions created by the Initial Developer are Copyright (C) 1998
|
7
|
+
# the Initial Developer. All Rights Reserved.
|
8
|
+
#
|
9
|
+
# Contributor(s):
|
10
|
+
# Hui (zhengzhengzheng@gmail.com) - port to Ruby
|
11
|
+
# Mark Pilgrim - first port to Python
|
12
|
+
#
|
13
|
+
# This library is free software; you can redistribute it and/or
|
14
|
+
# modify it under the terms of the GNU Lesser General Public
|
15
|
+
# License as published by the Free Software Foundation; either
|
16
|
+
# version 2.1 of the License, or (at your option) any later version.
|
17
|
+
#
|
18
|
+
# This library is distributed in the hope that it will be useful,
|
19
|
+
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
20
|
+
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
21
|
+
# Lesser General Public License for more details.
|
22
|
+
#
|
23
|
+
# You should have received a copy of the GNU Lesser General Public
|
24
|
+
# License along with this library; if not, write to the Free Software
|
25
|
+
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
|
26
|
+
# 02110-1301 USA
|
27
|
+
######################### END LICENSE BLOCK #########################
|
28
|
+
|
29
|
+
require 'UniversalDetector'
|
30
|
+
|
31
|
+
module UniversalDetector
|
32
|
+
|
33
|
+
class CharSetProber
|
34
|
+
def reset
|
35
|
+
@_mState = :Detecting
|
36
|
+
end
|
37
|
+
|
38
|
+
def get_charset_name
|
39
|
+
return nil
|
40
|
+
end
|
41
|
+
|
42
|
+
def feed(data)
|
43
|
+
end
|
44
|
+
|
45
|
+
def get_state
|
46
|
+
return @_mState
|
47
|
+
end
|
48
|
+
|
49
|
+
def get_confidence
|
50
|
+
return 0.0
|
51
|
+
end
|
52
|
+
|
53
|
+
def filter_high_bit_only(aBuf)
|
54
|
+
aBuf = aBuf.gsub(/([\x00-\x7F])+/, '')
|
55
|
+
return aBuf
|
56
|
+
end
|
57
|
+
|
58
|
+
def filter_without_english_letters(aBuf)
|
59
|
+
aBuf = aBuf.gsub(/([A-Za-z])+/, '')
|
60
|
+
return aBuf
|
61
|
+
end
|
62
|
+
|
63
|
+
def filter_with_english_letters(aBuf)
|
64
|
+
# TODO
|
65
|
+
return aBuf
|
66
|
+
end
|
67
|
+
|
68
|
+
end #class
|
69
|
+
|
70
|
+
end #module
|