|
| 1 | +######################## BEGIN LICENSE BLOCK ######################## |
| 2 | +# This code was modified from latin1prober.py by Rob Speer <rob@lumino.so>. |
| 3 | +# The Original Code is Mozilla Universal charset detector code. |
| 4 | +# |
| 5 | +# The Initial Developer of the Original Code is |
| 6 | +# Netscape Communications Corporation. |
| 7 | +# Portions created by the Initial Developer are Copyright (C) 2001 |
| 8 | +# the Initial Developer. All Rights Reserved. |
| 9 | +# |
| 10 | +# Contributor(s): |
| 11 | +# Rob Speer - adapt to MacRoman encoding |
| 12 | +# Mark Pilgrim - port to Python |
| 13 | +# Shy Shalom - original C code |
| 14 | +# |
| 15 | +# This library is free software; you can redistribute it and/or |
| 16 | +# modify it under the terms of the GNU Lesser General Public |
| 17 | +# License as published by the Free Software Foundation; either |
| 18 | +# version 2.1 of the License, or (at your option) any later version. |
| 19 | +# |
| 20 | +# This library is distributed in the hope that it will be useful, |
| 21 | +# but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 22 | +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 23 | +# Lesser General Public License for more details. |
| 24 | +# |
| 25 | +# You should have received a copy of the GNU Lesser General Public |
| 26 | +# License along with this library; if not, write to the Free Software |
| 27 | +# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA |
| 28 | +# 02110-1301 USA |
| 29 | +######################### END LICENSE BLOCK ######################### |
| 30 | + |
| 31 | +from typing import List, Union |
| 32 | + |
| 33 | +from .charsetprober import CharSetProber |
| 34 | +from .enums import ProbingState |
| 35 | + |
| 36 | +FREQ_CAT_NUM = 4 |
| 37 | + |
| 38 | +UDF = 0 # undefined |
| 39 | +OTH = 1 # other |
| 40 | +ASC = 2 # ascii capital letter |
| 41 | +ASS = 3 # ascii small letter |
| 42 | +ACV = 4 # accent capital vowel |
| 43 | +ACO = 5 # accent capital other |
| 44 | +ASV = 6 # accent small vowel |
| 45 | +ASO = 7 # accent small other |
| 46 | +ODD = 8 # character that is unlikely to appear |
| 47 | +CLASS_NUM = 9 # total classes |
| 48 | + |
| 49 | +# The change from Latin1 is that we explicitly look for extended characters |
| 50 | +# that are infrequently-occurring symbols, and consider them to always be |
| 51 | +# improbable. This should let MacRoman get out of the way of more likely |
| 52 | +# encodings in most situations. |
| 53 | + |
| 54 | +# fmt: off |
| 55 | +MacRoman_CharToClass = ( |
| 56 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 00 - 07 |
| 57 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 08 - 0F |
| 58 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 10 - 17 |
| 59 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 18 - 1F |
| 60 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 20 - 27 |
| 61 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 28 - 2F |
| 62 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 30 - 37 |
| 63 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # 38 - 3F |
| 64 | + OTH, ASC, ASC, ASC, ASC, ASC, ASC, ASC, # 40 - 47 |
| 65 | + ASC, ASC, ASC, ASC, ASC, ASC, ASC, ASC, # 48 - 4F |
| 66 | + ASC, ASC, ASC, ASC, ASC, ASC, ASC, ASC, # 50 - 57 |
| 67 | + ASC, ASC, ASC, OTH, OTH, OTH, OTH, OTH, # 58 - 5F |
| 68 | + OTH, ASS, ASS, ASS, ASS, ASS, ASS, ASS, # 60 - 67 |
| 69 | + ASS, ASS, ASS, ASS, ASS, ASS, ASS, ASS, # 68 - 6F |
| 70 | + ASS, ASS, ASS, ASS, ASS, ASS, ASS, ASS, # 70 - 77 |
| 71 | + ASS, ASS, ASS, OTH, OTH, OTH, OTH, OTH, # 78 - 7F |
| 72 | + ACV, ACV, ACO, ACV, ACO, ACV, ACV, ASV, # 80 - 87 |
| 73 | + ASV, ASV, ASV, ASV, ASV, ASO, ASV, ASV, # 88 - 8F |
| 74 | + ASV, ASV, ASV, ASV, ASV, ASV, ASO, ASV, # 90 - 97 |
| 75 | + ASV, ASV, ASV, ASV, ASV, ASV, ASV, ASV, # 98 - 9F |
| 76 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, ASO, # A0 - A7 |
| 77 | + OTH, OTH, ODD, ODD, OTH, OTH, ACV, ACV, # A8 - AF |
| 78 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, OTH, # B0 - B7 |
| 79 | + OTH, OTH, OTH, OTH, OTH, OTH, ASV, ASV, # B8 - BF |
| 80 | + OTH, OTH, ODD, OTH, ODD, OTH, OTH, OTH, # C0 - C7 |
| 81 | + OTH, OTH, OTH, ACV, ACV, ACV, ACV, ASV, # C8 - CF |
| 82 | + OTH, OTH, OTH, OTH, OTH, OTH, OTH, ODD, # D0 - D7 |
| 83 | + ASV, ACV, ODD, OTH, OTH, OTH, OTH, OTH, # D8 - DF |
| 84 | + OTH, OTH, OTH, OTH, OTH, ACV, ACV, ACV, # E0 - E7 |
| 85 | + ACV, ACV, ACV, ACV, ACV, ACV, ACV, ACV, # E8 - EF |
| 86 | + ODD, ACV, ACV, ACV, ACV, ASV, ODD, ODD, # F0 - F7 |
| 87 | + ODD, ODD, ODD, ODD, ODD, ODD, ODD, ODD, # F8 - FF |
| 88 | +) |
| 89 | + |
| 90 | +# 0 : illegal |
| 91 | +# 1 : very unlikely |
| 92 | +# 2 : normal |
| 93 | +# 3 : very likely |
| 94 | +MacRomanClassModel = ( |
| 95 | +# UDF OTH ASC ASS ACV ACO ASV ASO ODD |
| 96 | + 0, 0, 0, 0, 0, 0, 0, 0, 0, # UDF |
| 97 | + 0, 3, 3, 3, 3, 3, 3, 3, 1, # OTH |
| 98 | + 0, 3, 3, 3, 3, 3, 3, 3, 1, # ASC |
| 99 | + 0, 3, 3, 3, 1, 1, 3, 3, 1, # ASS |
| 100 | + 0, 3, 3, 3, 1, 2, 1, 2, 1, # ACV |
| 101 | + 0, 3, 3, 3, 3, 3, 3, 3, 1, # ACO |
| 102 | + 0, 3, 1, 3, 1, 1, 1, 3, 1, # ASV |
| 103 | + 0, 3, 1, 3, 1, 1, 3, 3, 1, # ASO |
| 104 | + 0, 1, 1, 1, 1, 1, 1, 1, 1, # ODD |
| 105 | +) |
| 106 | +# fmt: on |
| 107 | + |
| 108 | + |
| 109 | +class MacRomanProber(CharSetProber): |
| 110 | + def __init__(self) -> None: |
| 111 | + super().__init__() |
| 112 | + self._last_char_class = OTH |
| 113 | + self._freq_counter: List[int] = [] |
| 114 | + self.reset() |
| 115 | + |
| 116 | + def reset(self) -> None: |
| 117 | + self._last_char_class = OTH |
| 118 | + self._freq_counter = [0] * FREQ_CAT_NUM |
| 119 | + |
| 120 | + # express the prior that MacRoman is a somewhat rare encoding; |
| 121 | + # this can be done by starting out in a slightly improbable state |
| 122 | + # that must be overcome |
| 123 | + self._freq_counter[2] = 10 |
| 124 | + |
| 125 | + super().reset() |
| 126 | + |
| 127 | + @property |
| 128 | + def charset_name(self) -> str: |
| 129 | + return "MacRoman" |
| 130 | + |
| 131 | + @property |
| 132 | + def language(self) -> str: |
| 133 | + return "" |
| 134 | + |
| 135 | + def feed(self, byte_str: Union[bytes, bytearray]) -> ProbingState: |
| 136 | + byte_str = self.remove_xml_tags(byte_str) |
| 137 | + for c in byte_str: |
| 138 | + char_class = MacRoman_CharToClass[c] |
| 139 | + freq = MacRomanClassModel[(self._last_char_class * CLASS_NUM) + char_class] |
| 140 | + if freq == 0: |
| 141 | + self._state = ProbingState.NOT_ME |
| 142 | + break |
| 143 | + self._freq_counter[freq] += 1 |
| 144 | + self._last_char_class = char_class |
| 145 | + |
| 146 | + return self.state |
| 147 | + |
| 148 | + def get_confidence(self) -> float: |
| 149 | + if self.state == ProbingState.NOT_ME: |
| 150 | + return 0.01 |
| 151 | + |
| 152 | + total = sum(self._freq_counter) |
| 153 | + confidence = ( |
| 154 | + 0.0 |
| 155 | + if total < 0.01 |
| 156 | + else (self._freq_counter[3] - self._freq_counter[1] * 20.0) / total |
| 157 | + ) |
| 158 | + confidence = max(confidence, 0.0) |
| 159 | + # lower the confidence of MacRoman so that other more accurate |
| 160 | + # detector can take priority. |
| 161 | + confidence *= 0.73 |
| 162 | + return confidence |
0 commit comments