comparison env/lib/python3.7/site-packages/chardet/hebrewprober.py @ 0:26e78fe6e8c4 draft

"planemo upload commit c699937486c35866861690329de38ec1a5d9f783"
author shellac
date Sat, 02 May 2020 07:14:21 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:26e78fe6e8c4
1 ######################## BEGIN LICENSE BLOCK ########################
2 # The Original Code is Mozilla Universal charset detector code.
3 #
4 # The Initial Developer of the Original Code is
5 # Shy Shalom
6 # Portions created by the Initial Developer are Copyright (C) 2005
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 .charsetprober import CharSetProber
29 from .enums import ProbingState
30
31 # This prober doesn't actually recognize a language or a charset.
32 # It is a helper prober for the use of the Hebrew model probers
33
34 ### General ideas of the Hebrew charset recognition ###
35 #
36 # Four main charsets exist in Hebrew:
37 # "ISO-8859-8" - Visual Hebrew
38 # "windows-1255" - Logical Hebrew
39 # "ISO-8859-8-I" - Logical Hebrew
40 # "x-mac-hebrew" - ?? Logical Hebrew ??
41 #
42 # Both "ISO" charsets use a completely identical set of code points, whereas
43 # "windows-1255" and "x-mac-hebrew" are two different proper supersets of
44 # these code points. windows-1255 defines additional characters in the range
45 # 0x80-0x9F as some misc punctuation marks as well as some Hebrew-specific
46 # diacritics and additional 'Yiddish' ligature letters in the range 0xc0-0xd6.
47 # x-mac-hebrew defines similar additional code points but with a different
48 # mapping.
49 #
50 # As far as an average Hebrew text with no diacritics is concerned, all four
51 # charsets are identical with respect to code points. Meaning that for the
52 # main Hebrew alphabet, all four map the same values to all 27 Hebrew letters
53 # (including final letters).
54 #
55 # The dominant difference between these charsets is their directionality.
56 # "Visual" directionality means that the text is ordered as if the renderer is
57 # not aware of a BIDI rendering algorithm. The renderer sees the text and
58 # draws it from left to right. The text itself when ordered naturally is read
59 # backwards. A buffer of Visual Hebrew generally looks like so:
60 # "[last word of first line spelled backwards] [whole line ordered backwards
61 # and spelled backwards] [first word of first line spelled backwards]
62 # [end of line] [last word of second line] ... etc' "
63 # adding punctuation marks, numbers and English text to visual text is
64 # naturally also "visual" and from left to right.
65 #
66 # "Logical" directionality means the text is ordered "naturally" according to
67 # the order it is read. It is the responsibility of the renderer to display
68 # the text from right to left. A BIDI algorithm is used to place general
69 # punctuation marks, numbers and English text in the text.
70 #
71 # Texts in x-mac-hebrew are almost impossible to find on the Internet. From
72 # what little evidence I could find, it seems that its general directionality
73 # is Logical.
74 #
75 # To sum up all of the above, the Hebrew probing mechanism knows about two
76 # charsets:
77 # Visual Hebrew - "ISO-8859-8" - backwards text - Words and sentences are
78 # backwards while line order is natural. For charset recognition purposes
79 # the line order is unimportant (In fact, for this implementation, even
80 # word order is unimportant).
81 # Logical Hebrew - "windows-1255" - normal, naturally ordered text.
82 #
83 # "ISO-8859-8-I" is a subset of windows-1255 and doesn't need to be
84 # specifically identified.
85 # "x-mac-hebrew" is also identified as windows-1255. A text in x-mac-hebrew
86 # that contain special punctuation marks or diacritics is displayed with
87 # some unconverted characters showing as question marks. This problem might
88 # be corrected using another model prober for x-mac-hebrew. Due to the fact
89 # that x-mac-hebrew texts are so rare, writing another model prober isn't
90 # worth the effort and performance hit.
91 #
92 #### The Prober ####
93 #
94 # The prober is divided between two SBCharSetProbers and a HebrewProber,
95 # all of which are managed, created, fed data, inquired and deleted by the
96 # SBCSGroupProber. The two SBCharSetProbers identify that the text is in
97 # fact some kind of Hebrew, Logical or Visual. The final decision about which
98 # one is it is made by the HebrewProber by combining final-letter scores
99 # with the scores of the two SBCharSetProbers to produce a final answer.
100 #
101 # The SBCSGroupProber is responsible for stripping the original text of HTML
102 # tags, English characters, numbers, low-ASCII punctuation characters, spaces
103 # and new lines. It reduces any sequence of such characters to a single space.
104 # The buffer fed to each prober in the SBCS group prober is pure text in
105 # high-ASCII.
106 # The two SBCharSetProbers (model probers) share the same language model:
107 # Win1255Model.
108 # The first SBCharSetProber uses the model normally as any other
109 # SBCharSetProber does, to recognize windows-1255, upon which this model was
110 # built. The second SBCharSetProber is told to make the pair-of-letter
111 # lookup in the language model backwards. This in practice exactly simulates
112 # a visual Hebrew model using the windows-1255 logical Hebrew model.
113 #
114 # The HebrewProber is not using any language model. All it does is look for
115 # final-letter evidence suggesting the text is either logical Hebrew or visual
116 # Hebrew. Disjointed from the model probers, the results of the HebrewProber
117 # alone are meaningless. HebrewProber always returns 0.00 as confidence
118 # since it never identifies a charset by itself. Instead, the pointer to the
119 # HebrewProber is passed to the model probers as a helper "Name Prober".
120 # When the Group prober receives a positive identification from any prober,
121 # it asks for the name of the charset identified. If the prober queried is a
122 # Hebrew model prober, the model prober forwards the call to the
123 # HebrewProber to make the final decision. In the HebrewProber, the
124 # decision is made according to the final-letters scores maintained and Both
125 # model probers scores. The answer is returned in the form of the name of the
126 # charset identified, either "windows-1255" or "ISO-8859-8".
127
128 class HebrewProber(CharSetProber):
129 # windows-1255 / ISO-8859-8 code points of interest
130 FINAL_KAF = 0xea
131 NORMAL_KAF = 0xeb
132 FINAL_MEM = 0xed
133 NORMAL_MEM = 0xee
134 FINAL_NUN = 0xef
135 NORMAL_NUN = 0xf0
136 FINAL_PE = 0xf3
137 NORMAL_PE = 0xf4
138 FINAL_TSADI = 0xf5
139 NORMAL_TSADI = 0xf6
140
141 # Minimum Visual vs Logical final letter score difference.
142 # If the difference is below this, don't rely solely on the final letter score
143 # distance.
144 MIN_FINAL_CHAR_DISTANCE = 5
145
146 # Minimum Visual vs Logical model score difference.
147 # If the difference is below this, don't rely at all on the model score
148 # distance.
149 MIN_MODEL_DISTANCE = 0.01
150
151 VISUAL_HEBREW_NAME = "ISO-8859-8"
152 LOGICAL_HEBREW_NAME = "windows-1255"
153
154 def __init__(self):
155 super(HebrewProber, self).__init__()
156 self._final_char_logical_score = None
157 self._final_char_visual_score = None
158 self._prev = None
159 self._before_prev = None
160 self._logical_prober = None
161 self._visual_prober = None
162 self.reset()
163
164 def reset(self):
165 self._final_char_logical_score = 0
166 self._final_char_visual_score = 0
167 # The two last characters seen in the previous buffer,
168 # mPrev and mBeforePrev are initialized to space in order to simulate
169 # a word delimiter at the beginning of the data
170 self._prev = ' '
171 self._before_prev = ' '
172 # These probers are owned by the group prober.
173
174 def set_model_probers(self, logicalProber, visualProber):
175 self._logical_prober = logicalProber
176 self._visual_prober = visualProber
177
178 def is_final(self, c):
179 return c in [self.FINAL_KAF, self.FINAL_MEM, self.FINAL_NUN,
180 self.FINAL_PE, self.FINAL_TSADI]
181
182 def is_non_final(self, c):
183 # The normal Tsadi is not a good Non-Final letter due to words like
184 # 'lechotet' (to chat) containing an apostrophe after the tsadi. This
185 # apostrophe is converted to a space in FilterWithoutEnglishLetters
186 # causing the Non-Final tsadi to appear at an end of a word even
187 # though this is not the case in the original text.
188 # The letters Pe and Kaf rarely display a related behavior of not being
189 # a good Non-Final letter. Words like 'Pop', 'Winamp' and 'Mubarak'
190 # for example legally end with a Non-Final Pe or Kaf. However, the
191 # benefit of these letters as Non-Final letters outweighs the damage
192 # since these words are quite rare.
193 return c in [self.NORMAL_KAF, self.NORMAL_MEM,
194 self.NORMAL_NUN, self.NORMAL_PE]
195
196 def feed(self, byte_str):
197 # Final letter analysis for logical-visual decision.
198 # Look for evidence that the received buffer is either logical Hebrew
199 # or visual Hebrew.
200 # The following cases are checked:
201 # 1) A word longer than 1 letter, ending with a final letter. This is
202 # an indication that the text is laid out "naturally" since the
203 # final letter really appears at the end. +1 for logical score.
204 # 2) A word longer than 1 letter, ending with a Non-Final letter. In
205 # normal Hebrew, words ending with Kaf, Mem, Nun, Pe or Tsadi,
206 # should not end with the Non-Final form of that letter. Exceptions
207 # to this rule are mentioned above in isNonFinal(). This is an
208 # indication that the text is laid out backwards. +1 for visual
209 # score
210 # 3) A word longer than 1 letter, starting with a final letter. Final
211 # letters should not appear at the beginning of a word. This is an
212 # indication that the text is laid out backwards. +1 for visual
213 # score.
214 #
215 # The visual score and logical score are accumulated throughout the
216 # text and are finally checked against each other in GetCharSetName().
217 # No checking for final letters in the middle of words is done since
218 # that case is not an indication for either Logical or Visual text.
219 #
220 # We automatically filter out all 7-bit characters (replace them with
221 # spaces) so the word boundary detection works properly. [MAP]
222
223 if self.state == ProbingState.NOT_ME:
224 # Both model probers say it's not them. No reason to continue.
225 return ProbingState.NOT_ME
226
227 byte_str = self.filter_high_byte_only(byte_str)
228
229 for cur in byte_str:
230 if cur == ' ':
231 # We stand on a space - a word just ended
232 if self._before_prev != ' ':
233 # next-to-last char was not a space so self._prev is not a
234 # 1 letter word
235 if self.is_final(self._prev):
236 # case (1) [-2:not space][-1:final letter][cur:space]
237 self._final_char_logical_score += 1
238 elif self.is_non_final(self._prev):
239 # case (2) [-2:not space][-1:Non-Final letter][
240 # cur:space]
241 self._final_char_visual_score += 1
242 else:
243 # Not standing on a space
244 if ((self._before_prev == ' ') and
245 (self.is_final(self._prev)) and (cur != ' ')):
246 # case (3) [-2:space][-1:final letter][cur:not space]
247 self._final_char_visual_score += 1
248 self._before_prev = self._prev
249 self._prev = cur
250
251 # Forever detecting, till the end or until both model probers return
252 # ProbingState.NOT_ME (handled above)
253 return ProbingState.DETECTING
254
255 @property
256 def charset_name(self):
257 # Make the decision: is it Logical or Visual?
258 # If the final letter score distance is dominant enough, rely on it.
259 finalsub = self._final_char_logical_score - self._final_char_visual_score
260 if finalsub >= self.MIN_FINAL_CHAR_DISTANCE:
261 return self.LOGICAL_HEBREW_NAME
262 if finalsub <= -self.MIN_FINAL_CHAR_DISTANCE:
263 return self.VISUAL_HEBREW_NAME
264
265 # It's not dominant enough, try to rely on the model scores instead.
266 modelsub = (self._logical_prober.get_confidence()
267 - self._visual_prober.get_confidence())
268 if modelsub > self.MIN_MODEL_DISTANCE:
269 return self.LOGICAL_HEBREW_NAME
270 if modelsub < -self.MIN_MODEL_DISTANCE:
271 return self.VISUAL_HEBREW_NAME
272
273 # Still no good, back to final letter distance, maybe it'll save the
274 # day.
275 if finalsub < 0.0:
276 return self.VISUAL_HEBREW_NAME
277
278 # (finalsub > 0 - Logical) or (don't know what to do) default to
279 # Logical.
280 return self.LOGICAL_HEBREW_NAME
281
282 @property
283 def language(self):
284 return 'Hebrew'
285
286 @property
287 def state(self):
288 # Remain active as long as any of the model probers are active.
289 if (self._logical_prober.state == ProbingState.NOT_ME) and \
290 (self._visual_prober.state == ProbingState.NOT_ME):
291 return ProbingState.NOT_ME
292 return ProbingState.DETECTING