1 | ######################## BEGIN LICENSE BLOCK ######################## |
---|
2 | # The Original Code is Mozilla Communicator client 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 | # Mark Pilgrim - port to Python |
---|
11 | # |
---|
12 | # This library is free software; you can redistribute it and/or |
---|
13 | # modify it under the terms of the GNU Lesser General Public |
---|
14 | # License as published by the Free Software Foundation; either |
---|
15 | # version 2.1 of the License, or (at your option) any later version. |
---|
16 | # |
---|
17 | # This library is distributed in the hope that it will be useful, |
---|
18 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
---|
20 | # Lesser General Public License for more details. |
---|
21 | # |
---|
22 | # You should have received a copy of the GNU Lesser General Public |
---|
23 | # License along with this library; if not, write to the Free Software |
---|
24 | # Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA |
---|
25 | # 02110-1301 USA |
---|
26 | ######################### END LICENSE BLOCK ######################### |
---|
27 | |
---|
28 | from .euctwfreq import (EUCTW_CHAR_TO_FREQ_ORDER, EUCTW_TABLE_SIZE, |
---|
29 | EUCTW_TYPICAL_DISTRIBUTION_RATIO) |
---|
30 | from .euckrfreq import (EUCKR_CHAR_TO_FREQ_ORDER, EUCKR_TABLE_SIZE, |
---|
31 | EUCKR_TYPICAL_DISTRIBUTION_RATIO) |
---|
32 | from .gb2312freq import (GB2312_CHAR_TO_FREQ_ORDER, GB2312_TABLE_SIZE, |
---|
33 | GB2312_TYPICAL_DISTRIBUTION_RATIO) |
---|
34 | from .big5freq import (BIG5_CHAR_TO_FREQ_ORDER, BIG5_TABLE_SIZE, |
---|
35 | BIG5_TYPICAL_DISTRIBUTION_RATIO) |
---|
36 | from .jisfreq import (JIS_CHAR_TO_FREQ_ORDER, JIS_TABLE_SIZE, |
---|
37 | JIS_TYPICAL_DISTRIBUTION_RATIO) |
---|
38 | |
---|
39 | |
---|
40 | class CharDistributionAnalysis(object): |
---|
41 | ENOUGH_DATA_THRESHOLD = 1024 |
---|
42 | SURE_YES = 0.99 |
---|
43 | SURE_NO = 0.01 |
---|
44 | MINIMUM_DATA_THRESHOLD = 3 |
---|
45 | |
---|
46 | def __init__(self): |
---|
47 | # Mapping table to get frequency order from char order (get from |
---|
48 | # GetOrder()) |
---|
49 | self._char_to_freq_order = None |
---|
50 | self._table_size = None # Size of above table |
---|
51 | # This is a constant value which varies from language to language, |
---|
52 | # used in calculating confidence. See |
---|
53 | # http://www.mozilla.org/projects/intl/UniversalCharsetDetection.html |
---|
54 | # for further detail. |
---|
55 | self.typical_distribution_ratio = None |
---|
56 | self._done = None |
---|
57 | self._total_chars = None |
---|
58 | self._freq_chars = None |
---|
59 | self.reset() |
---|
60 | |
---|
61 | def reset(self): |
---|
62 | """reset analyser, clear any state""" |
---|
63 | # If this flag is set to True, detection is done and conclusion has |
---|
64 | # been made |
---|
65 | self._done = False |
---|
66 | self._total_chars = 0 # Total characters encountered |
---|
67 | # The number of characters whose frequency order is less than 512 |
---|
68 | self._freq_chars = 0 |
---|
69 | |
---|
70 | def feed(self, char, char_len): |
---|
71 | """feed a character with known length""" |
---|
72 | if char_len == 2: |
---|
73 | # we only care about 2-bytes character in our distribution analysis |
---|
74 | order = self.get_order(char) |
---|
75 | else: |
---|
76 | order = -1 |
---|
77 | if order >= 0: |
---|
78 | self._total_chars += 1 |
---|
79 | # order is valid |
---|
80 | if order < self._table_size: |
---|
81 | if 512 > self._char_to_freq_order[order]: |
---|
82 | self._freq_chars += 1 |
---|
83 | |
---|
84 | def get_confidence(self): |
---|
85 | """return confidence based on existing data""" |
---|
86 | # if we didn't receive any character in our consideration range, |
---|
87 | # return negative answer |
---|
88 | if self._total_chars <= 0 or self._freq_chars <= self.MINIMUM_DATA_THRESHOLD: |
---|
89 | return self.SURE_NO |
---|
90 | |
---|
91 | if self._total_chars != self._freq_chars: |
---|
92 | r = (self._freq_chars / ((self._total_chars - self._freq_chars) |
---|
93 | * self.typical_distribution_ratio)) |
---|
94 | if r < self.SURE_YES: |
---|
95 | return r |
---|
96 | |
---|
97 | # normalize confidence (we don't want to be 100% sure) |
---|
98 | return self.SURE_YES |
---|
99 | |
---|
100 | def got_enough_data(self): |
---|
101 | # It is not necessary to receive all data to draw conclusion. |
---|
102 | # For charset detection, certain amount of data is enough |
---|
103 | return self._total_chars > self.ENOUGH_DATA_THRESHOLD |
---|
104 | |
---|
105 | def get_order(self, byte_str): |
---|
106 | # We do not handle characters based on the original encoding string, |
---|
107 | # but convert this encoding string to a number, here called order. |
---|
108 | # This allows multiple encodings of a language to share one frequency |
---|
109 | # table. |
---|
110 | return -1 |
---|
111 | |
---|
112 | |
---|
113 | class EUCTWDistributionAnalysis(CharDistributionAnalysis): |
---|
114 | def __init__(self): |
---|
115 | super(EUCTWDistributionAnalysis, self).__init__() |
---|
116 | self._char_to_freq_order = EUCTW_CHAR_TO_FREQ_ORDER |
---|
117 | self._table_size = EUCTW_TABLE_SIZE |
---|
118 | self.typical_distribution_ratio = EUCTW_TYPICAL_DISTRIBUTION_RATIO |
---|
119 | |
---|
120 | def get_order(self, byte_str): |
---|
121 | # for euc-TW encoding, we are interested |
---|
122 | # first byte range: 0xc4 -- 0xfe |
---|
123 | # second byte range: 0xa1 -- 0xfe |
---|
124 | # no validation needed here. State machine has done that |
---|
125 | first_char = byte_str[0] |
---|
126 | if first_char >= 0xC4: |
---|
127 | return 94 * (first_char - 0xC4) + byte_str[1] - 0xA1 |
---|
128 | else: |
---|
129 | return -1 |
---|
130 | |
---|
131 | |
---|
132 | class EUCKRDistributionAnalysis(CharDistributionAnalysis): |
---|
133 | def __init__(self): |
---|
134 | super(EUCKRDistributionAnalysis, self).__init__() |
---|
135 | self._char_to_freq_order = EUCKR_CHAR_TO_FREQ_ORDER |
---|
136 | self._table_size = EUCKR_TABLE_SIZE |
---|
137 | self.typical_distribution_ratio = EUCKR_TYPICAL_DISTRIBUTION_RATIO |
---|
138 | |
---|
139 | def get_order(self, byte_str): |
---|
140 | # for euc-KR encoding, we are interested |
---|
141 | # first byte range: 0xb0 -- 0xfe |
---|
142 | # second byte range: 0xa1 -- 0xfe |
---|
143 | # no validation needed here. State machine has done that |
---|
144 | first_char = byte_str[0] |
---|
145 | if first_char >= 0xB0: |
---|
146 | return 94 * (first_char - 0xB0) + byte_str[1] - 0xA1 |
---|
147 | else: |
---|
148 | return -1 |
---|
149 | |
---|
150 | |
---|
151 | class GB2312DistributionAnalysis(CharDistributionAnalysis): |
---|
152 | def __init__(self): |
---|
153 | super(GB2312DistributionAnalysis, self).__init__() |
---|
154 | self._char_to_freq_order = GB2312_CHAR_TO_FREQ_ORDER |
---|
155 | self._table_size = GB2312_TABLE_SIZE |
---|
156 | self.typical_distribution_ratio = GB2312_TYPICAL_DISTRIBUTION_RATIO |
---|
157 | |
---|
158 | def get_order(self, byte_str): |
---|
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 | first_char, second_char = byte_str[0], byte_str[1] |
---|
164 | if (first_char >= 0xB0) and (second_char >= 0xA1): |
---|
165 | return 94 * (first_char - 0xB0) + second_char - 0xA1 |
---|
166 | else: |
---|
167 | return -1 |
---|
168 | |
---|
169 | |
---|
170 | class Big5DistributionAnalysis(CharDistributionAnalysis): |
---|
171 | def __init__(self): |
---|
172 | super(Big5DistributionAnalysis, self).__init__() |
---|
173 | self._char_to_freq_order = BIG5_CHAR_TO_FREQ_ORDER |
---|
174 | self._table_size = BIG5_TABLE_SIZE |
---|
175 | self.typical_distribution_ratio = BIG5_TYPICAL_DISTRIBUTION_RATIO |
---|
176 | |
---|
177 | def get_order(self, byte_str): |
---|
178 | # for big5 encoding, we are interested |
---|
179 | # first byte range: 0xa4 -- 0xfe |
---|
180 | # second byte range: 0x40 -- 0x7e , 0xa1 -- 0xfe |
---|
181 | # no validation needed here. State machine has done that |
---|
182 | first_char, second_char = byte_str[0], byte_str[1] |
---|
183 | if first_char >= 0xA4: |
---|
184 | if second_char >= 0xA1: |
---|
185 | return 157 * (first_char - 0xA4) + second_char - 0xA1 + 63 |
---|
186 | else: |
---|
187 | return 157 * (first_char - 0xA4) + second_char - 0x40 |
---|
188 | else: |
---|
189 | return -1 |
---|
190 | |
---|
191 | |
---|
192 | class SJISDistributionAnalysis(CharDistributionAnalysis): |
---|
193 | def __init__(self): |
---|
194 | super(SJISDistributionAnalysis, self).__init__() |
---|
195 | self._char_to_freq_order = JIS_CHAR_TO_FREQ_ORDER |
---|
196 | self._table_size = JIS_TABLE_SIZE |
---|
197 | self.typical_distribution_ratio = JIS_TYPICAL_DISTRIBUTION_RATIO |
---|
198 | |
---|
199 | def get_order(self, byte_str): |
---|
200 | # for sjis encoding, we are interested |
---|
201 | # first byte range: 0x81 -- 0x9f , 0xe0 -- 0xfe |
---|
202 | # second byte range: 0x40 -- 0x7e, 0x81 -- oxfe |
---|
203 | # no validation needed here. State machine has done that |
---|
204 | first_char, second_char = byte_str[0], byte_str[1] |
---|
205 | if (first_char >= 0x81) and (first_char <= 0x9F): |
---|
206 | order = 188 * (first_char - 0x81) |
---|
207 | elif (first_char >= 0xE0) and (first_char <= 0xEF): |
---|
208 | order = 188 * (first_char - 0xE0 + 31) |
---|
209 | else: |
---|
210 | return -1 |
---|
211 | order = order + second_char - 0x40 |
---|
212 | if second_char > 0x7F: |
---|
213 | order = -1 |
---|
214 | return order |
---|
215 | |
---|
216 | |
---|
217 | class EUCJPDistributionAnalysis(CharDistributionAnalysis): |
---|
218 | def __init__(self): |
---|
219 | super(EUCJPDistributionAnalysis, self).__init__() |
---|
220 | self._char_to_freq_order = JIS_CHAR_TO_FREQ_ORDER |
---|
221 | self._table_size = JIS_TABLE_SIZE |
---|
222 | self.typical_distribution_ratio = JIS_TYPICAL_DISTRIBUTION_RATIO |
---|
223 | |
---|
224 | def get_order(self, byte_str): |
---|
225 | # for euc-JP encoding, we are interested |
---|
226 | # first byte range: 0xa0 -- 0xfe |
---|
227 | # second byte range: 0xa1 -- 0xfe |
---|
228 | # no validation needed here. State machine has done that |
---|
229 | char = byte_str[0] |
---|
230 | if char >= 0xA0: |
---|
231 | return 94 * (char - 0xA1) + byte_str[1] - 0xa1 |
---|
232 | else: |
---|
233 | return -1 |
---|