1/* -*- C++ -*-
2 SPDX-FileCopyrightText: 1998 Netscape Communications Corporation <developer@mozilla.org>
3
4 SPDX-License-Identifier: MIT
5*/
6
7#ifndef nsHebrewProber_h__
8#define nsHebrewProber_h__
9
10#include "nsSBCharSetProber.h"
11namespace kencodingprober
12{
13// This prober doesn't actually recognize a language or a charset.
14// It is a helper prober for the use of the Hebrew model probers
15class KCODECS_NO_EXPORT nsHebrewProber : public nsCharSetProber
16{
17public:
18 nsHebrewProber(void)
19 : mLogicalProb(nullptr)
20 , mVisualProb(nullptr)
21 {
22 Reset();
23 }
24
25 ~nsHebrewProber(void) override
26 {
27 }
28 nsProbingState HandleData(const char *aBuf, unsigned int aLen) override;
29 const char *GetCharSetName() override;
30 void Reset(void) override;
31
32 nsProbingState GetState(void) override;
33
34 float GetConfidence(void) override
35 {
36 return (float)0.0;
37 }
38
39 void SetModelProbers(nsCharSetProber *logicalPrb, nsCharSetProber *visualPrb)
40 {
41 mLogicalProb = logicalPrb;
42 mVisualProb = visualPrb;
43 }
44
45#ifdef DEBUG_PROBE
46 void DumpStatus() override;
47#endif
48
49protected:
50 static bool isFinal(char c);
51 static bool isNonFinal(char c);
52
53 int mFinalCharLogicalScore, mFinalCharVisualScore;
54
55 // The two last characters seen in the previous buffer.
56 char mPrev, mBeforePrev;
57
58 // These probers are owned by the group prober.
59 nsCharSetProber *mLogicalProb, *mVisualProb;
60};
61}
62
63/**
64 * ** General ideas of the Hebrew charset recognition **
65 *
66 * Four main charsets exist in Hebrew:
67 * "ISO-8859-8" - Visual Hebrew
68 * "windows-1255" - Logical Hebrew
69 * "ISO-8859-8-I" - Logical Hebrew
70 * "x-mac-hebrew" - ?? Logical Hebrew ??
71 *
72 * Both "ISO" charsets use a completely identical set of code points, whereas
73 * "windows-1255" and "x-mac-hebrew" are two different proper supersets of
74 * these code points. windows-1255 defines additional characters in the range
75 * 0x80-0x9F as some misc punctuation marks as well as some Hebrew-specific
76 * diacritics and additional 'Yiddish' ligature letters in the range 0xc0-0xd6.
77 * x-mac-hebrew defines similar additional code points but with a different
78 * mapping.
79 *
80 * As far as an average Hebrew text with no diacritics is concerned, all four
81 * charsets are identical with respect to code points. Meaning that for the
82 * main Hebrew alphabet, all four map the same values to all 27 Hebrew letters
83 * (including final letters).
84 *
85 * The dominant difference between these charsets is their directionality.
86 * "Visual" directionality means that the text is ordered as if the renderer is
87 * not aware of a BIDI rendering algorithm. The renderer sees the text and
88 * draws it from left to right. The text itself when ordered naturally is read
89 * backwards. A buffer of Visual Hebrew generally looks like so:
90 * "[last word of first line spelled backwards] [whole line ordered backwards
91 * and spelled backwards] [first word of first line spelled backwards]
92 * [end of line] [last word of second line] ... etc' "
93 * adding punctuation marks, numbers and English text to visual text is
94 * naturally also "visual" and from left to right.
95 *
96 * "Logical" directionality means the text is ordered "naturally" according to
97 * the order it is read. It is the responsibility of the renderer to display
98 * the text from right to left. A BIDI algorithm is used to place general
99 * punctuation marks, numbers and English text in the text.
100 *
101 * Texts in x-mac-hebrew are almost impossible to find on the Internet. From
102 * what little evidence I could find, it seems that its general directionality
103 * is Logical.
104 *
105 * To sum up all of the above, the Hebrew probing mechanism knows about two
106 * charsets:
107 * Visual Hebrew - "ISO-8859-8" - backwards text - Words and sentences are
108 * backwards while line order is natural. For charset recognition purposes
109 * the line order is unimportant (In fact, for this implementation, even
110 * word order is unimportant).
111 * Logical Hebrew - "windows-1255" - normal, naturally ordered text.
112 *
113 * "ISO-8859-8-I" is a subset of windows-1255 and doesn't need to be
114 * specifically identified.
115 * "x-mac-hebrew" is also identified as windows-1255. A text in x-mac-hebrew
116 * that contain special punctuation marks or diacritics is displayed with
117 * some unconverted characters showing as question marks. This problem might
118 * be corrected using another model prober for x-mac-hebrew. Due to the fact
119 * that x-mac-hebrew texts are so rare, writing another model prober isn't
120 * worth the effort and performance hit.
121 *
122 * *** The Prober ***
123 *
124 * The prober is divided between two nsSBCharSetProbers and an nsHebrewProber,
125 * all of which are managed, created, fed data, inquired and deleted by the
126 * nsSBCSGroupProber. The two nsSBCharSetProbers identify that the text is in
127 * fact some kind of Hebrew, Logical or Visual. The final decision about which
128 * one is it is made by the nsHebrewProber by combining final-letter scores
129 * with the scores of the two nsSBCharSetProbers to produce a final answer.
130 *
131 * The nsSBCSGroupProber is responsible for stripping the original text of HTML
132 * tags, English characters, numbers, low-ASCII punctuation characters, spaces
133 * and new lines. It reduces any sequence of such characters to a single space.
134 * The buffer fed to each prober in the SBCS group prober is pure text in
135 * high-ASCII.
136 * The two nsSBCharSetProbers (model probers) share the same language model:
137 * Win1255Model.
138 * The first nsSBCharSetProber uses the model normally as any other
139 * nsSBCharSetProber does, to recognize windows-1255, upon which this model was
140 * built. The second nsSBCharSetProber is told to make the pair-of-letter
141 * lookup in the language model backwards. This in practice exactly simulates
142 * a visual Hebrew model using the windows-1255 logical Hebrew model.
143 *
144 * The nsHebrewProber is not using any language model. All it does is look for
145 * final-letter evidence suggesting the text is either logical Hebrew or visual
146 * Hebrew. Disjointed from the model probers, the results of the nsHebrewProber
147 * alone are meaningless. nsHebrewProber always returns 0.00 as confidence
148 * since it never identifies a charset by itself. Instead, the pointer to the
149 * nsHebrewProber is passed to the model probers as a helper "Name Prober".
150 * When the Group prober receives a positive identification from any prober,
151 * it asks for the name of the charset identified. If the prober queried is a
152 * Hebrew model prober, the model prober forwards the call to the
153 * nsHebrewProber to make the final decision. In the nsHebrewProber, the
154 * decision is made according to the final-letters scores maintained and Both
155 * model probers scores. The answer is returned in the form of the name of the
156 * charset identified, either "windows-1255" or "ISO-8859-8".
157 *
158 */
159#endif /* nsHebrewProber_h__ */
160

source code of kcodecs/src/probers/nsHebrewProber.h