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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
|
// © 2016 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html
/*
**********************************************************************
* Copyright (C) 2005-2013, International Business Machines
* Corporation and others. All Rights Reserved.
**********************************************************************
*/
#include "unicode/utypes.h"
#if !UCONFIG_NO_CONVERSION
#include "csrucode.h"
#include "csmatch.h"
U_NAMESPACE_BEGIN
CharsetRecog_Unicode::~CharsetRecog_Unicode()
{
// nothing to do
}
CharsetRecog_UTF_16_BE::~CharsetRecog_UTF_16_BE()
{
// nothing to do
}
const char *CharsetRecog_UTF_16_BE::getName() const
{
return "UTF-16BE";
}
// UTF-16 confidence calculation. Very simple minded, but better than nothing.
// Any 8 bit non-control characters bump the confidence up. These have a zero high byte,
// and are very likely to be UTF-16, although they could also be part of a UTF-32 code.
// NULs are a contra-indication, they will appear commonly if the actual encoding is UTF-32.
// NULs should be rare in actual text.
static int32_t adjustConfidence(char16_t codeUnit, int32_t confidence) {
if (codeUnit == 0) {
confidence -= 10;
} else if ((codeUnit >= 0x20 && codeUnit <= 0xff) || codeUnit == 0x0a) {
confidence += 10;
}
if (confidence < 0) {
confidence = 0;
} else if (confidence > 100) {
confidence = 100;
}
return confidence;
}
UBool CharsetRecog_UTF_16_BE::match(InputText* textIn, CharsetMatch *results) const
{
const uint8_t *input = textIn->fRawInput;
int32_t confidence = 10;
int32_t length = textIn->fRawLength;
int32_t bytesToCheck = (length > 30) ? 30 : length;
for (int32_t charIndex=0; charIndex<bytesToCheck-1; charIndex+=2) {
char16_t codeUnit = (input[charIndex] << 8) | input[charIndex + 1];
if (charIndex == 0 && codeUnit == 0xFEFF) {
confidence = 100;
break;
}
confidence = adjustConfidence(codeUnit, confidence);
if (confidence == 0 || confidence == 100) {
break;
}
}
if (bytesToCheck < 4 && confidence < 100) {
confidence = 0;
}
results->set(textIn, this, confidence);
return (confidence > 0);
}
CharsetRecog_UTF_16_LE::~CharsetRecog_UTF_16_LE()
{
// nothing to do
}
const char *CharsetRecog_UTF_16_LE::getName() const
{
return "UTF-16LE";
}
UBool CharsetRecog_UTF_16_LE::match(InputText* textIn, CharsetMatch *results) const
{
const uint8_t *input = textIn->fRawInput;
int32_t confidence = 10;
int32_t length = textIn->fRawLength;
int32_t bytesToCheck = (length > 30) ? 30 : length;
for (int32_t charIndex=0; charIndex<bytesToCheck-1; charIndex+=2) {
char16_t codeUnit = input[charIndex] | (input[charIndex + 1] << 8);
if (charIndex == 0 && codeUnit == 0xFEFF) {
confidence = 100; // UTF-16 BOM
if (length >= 4 && input[2] == 0 && input[3] == 0) {
confidence = 0; // UTF-32 BOM
}
break;
}
confidence = adjustConfidence(codeUnit, confidence);
if (confidence == 0 || confidence == 100) {
break;
}
}
if (bytesToCheck < 4 && confidence < 100) {
confidence = 0;
}
results->set(textIn, this, confidence);
return (confidence > 0);
}
CharsetRecog_UTF_32::~CharsetRecog_UTF_32()
{
// nothing to do
}
UBool CharsetRecog_UTF_32::match(InputText* textIn, CharsetMatch *results) const
{
const uint8_t *input = textIn->fRawInput;
int32_t limit = (textIn->fRawLength / 4) * 4;
int32_t numValid = 0;
int32_t numInvalid = 0;
bool hasBOM = false;
int32_t confidence = 0;
if (limit > 0 && getChar(input, 0) == 0x0000FEFFUL) {
hasBOM = true;
}
for(int32_t i = 0; i < limit; i += 4) {
int32_t ch = getChar(input, i);
if (ch < 0 || ch >= 0x10FFFF || (ch >= 0xD800 && ch <= 0xDFFF)) {
numInvalid += 1;
} else {
numValid += 1;
}
}
// Cook up some sort of confidence score, based on presence of a BOM
// and the existence of valid and/or invalid multi-byte sequences.
if (hasBOM && numInvalid==0) {
confidence = 100;
} else if (hasBOM && numValid > numInvalid*10) {
confidence = 80;
} else if (numValid > 3 && numInvalid == 0) {
confidence = 100;
} else if (numValid > 0 && numInvalid == 0) {
confidence = 80;
} else if (numValid > numInvalid*10) {
// Probably corrupt UTF-32BE data. Valid sequences aren't likely by chance.
confidence = 25;
}
results->set(textIn, this, confidence);
return (confidence > 0);
}
CharsetRecog_UTF_32_BE::~CharsetRecog_UTF_32_BE()
{
// nothing to do
}
const char *CharsetRecog_UTF_32_BE::getName() const
{
return "UTF-32BE";
}
int32_t CharsetRecog_UTF_32_BE::getChar(const uint8_t *input, int32_t index) const
{
return input[index + 0] << 24 | input[index + 1] << 16 |
input[index + 2] << 8 | input[index + 3];
}
CharsetRecog_UTF_32_LE::~CharsetRecog_UTF_32_LE()
{
// nothing to do
}
const char *CharsetRecog_UTF_32_LE::getName() const
{
return "UTF-32LE";
}
int32_t CharsetRecog_UTF_32_LE::getChar(const uint8_t *input, int32_t index) const
{
return input[index + 3] << 24 | input[index + 2] << 16 |
input[index + 1] << 8 | input[index + 0];
}
U_NAMESPACE_END
#endif
|
