1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
|
######################## BEGIN LICENSE BLOCK ########################
# The Original Code is Mozilla Universal charset detector code.
#
# The Initial Developer of the Original Code is
# Netscape Communications Corporation.
# Portions created by the Initial Developer are Copyright (C) 2001
# the Initial Developer. All Rights Reserved.
#
# Contributor(s):
# Mark Pilgrim - port to Python
# Shy Shalom - original C code
#
# This library is free software; you can redistribute it and/or
# modify it under the terms of the GNU Lesser General Public
# License as published by the Free Software Foundation; either
# version 2.1 of the License, or (at your option) any later version.
#
# This library is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with this library; if not, write to the Free Software
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
# 02110-1301 USA
######################### END LICENSE BLOCK #########################
from collections import namedtuple
from .charsetprober import CharSetProber
from .enums import CharacterCategory, ProbingState, SequenceLikelihood
SingleByteCharSetModel = namedtuple(
"SingleByteCharSetModel",
[
"charset_name",
"language",
"char_to_order_map",
"language_model",
"typical_positive_ratio",
"keep_ascii_letters",
"alphabet",
],
)
class SingleByteCharSetProber(CharSetProber):
SAMPLE_SIZE = 64
SB_ENOUGH_REL_THRESHOLD = 1024 # 0.25 * SAMPLE_SIZE^2
POSITIVE_SHORTCUT_THRESHOLD = 0.95
NEGATIVE_SHORTCUT_THRESHOLD = 0.05
def __init__(self, model, is_reversed=False, name_prober=None):
super().__init__()
self._model = model
# TRUE if we need to reverse every pair in the model lookup
self._reversed = is_reversed
# Optional auxiliary prober for name decision
self._name_prober = name_prober
self._last_order = None
self._seq_counters = None
self._total_seqs = None
self._total_char = None
self._control_char = None
self._freq_char = None
self.reset()
def reset(self):
super().reset()
# char order of last character
self._last_order = 255
self._seq_counters = [0] * SequenceLikelihood.get_num_categories()
self._total_seqs = 0
self._total_char = 0
self._control_char = 0
# characters that fall in our sampling range
self._freq_char = 0
@property
def charset_name(self):
if self._name_prober:
return self._name_prober.charset_name
return self._model.charset_name
@property
def language(self):
if self._name_prober:
return self._name_prober.language
return self._model.language
def feed(self, byte_str):
# TODO: Make filter_international_words keep things in self.alphabet
if not self._model.keep_ascii_letters:
byte_str = self.filter_international_words(byte_str)
else:
byte_str = self.remove_xml_tags(byte_str)
if not byte_str:
return self.state
char_to_order_map = self._model.char_to_order_map
language_model = self._model.language_model
for char in byte_str:
order = char_to_order_map.get(char, CharacterCategory.UNDEFINED)
# XXX: This was SYMBOL_CAT_ORDER before, with a value of 250, but
# CharacterCategory.SYMBOL is actually 253, so we use CONTROL
# to make it closer to the original intent. The only difference
# is whether or not we count digits and control characters for
# _total_char purposes.
if order < CharacterCategory.CONTROL:
self._total_char += 1
if order < self.SAMPLE_SIZE:
self._freq_char += 1
if self._last_order < self.SAMPLE_SIZE:
self._total_seqs += 1
if not self._reversed:
lm_cat = language_model[self._last_order][order]
else:
lm_cat = language_model[order][self._last_order]
self._seq_counters[lm_cat] += 1
self._last_order = order
charset_name = self._model.charset_name
if self.state == ProbingState.DETECTING:
if self._total_seqs > self.SB_ENOUGH_REL_THRESHOLD:
confidence = self.get_confidence()
if confidence > self.POSITIVE_SHORTCUT_THRESHOLD:
self.logger.debug(
"%s confidence = %s, we have a winner", charset_name, confidence
)
self._state = ProbingState.FOUND_IT
elif confidence < self.NEGATIVE_SHORTCUT_THRESHOLD:
self.logger.debug(
"%s confidence = %s, below negative shortcut threshold %s",
charset_name,
confidence,
self.NEGATIVE_SHORTCUT_THRESHOLD,
)
self._state = ProbingState.NOT_ME
return self.state
def get_confidence(self):
r = 0.01
if self._total_seqs > 0:
r = (
(
self._seq_counters[SequenceLikelihood.POSITIVE]
+ 0.25 * self._seq_counters[SequenceLikelihood.LIKELY]
)
/ self._total_seqs
/ self._model.typical_positive_ratio
)
# The more control characters (proportionnaly to the size
# of the text), the less confident we become in the current
# charset.
r = r * (self._total_char - self._control_char) / self._total_char
r = r * self._freq_char / self._total_char
if r >= 1.0:
r = 0.99
return r
|
