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
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
|
/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* vim: set ts=8 sts=2 et sw=2 tw=80: */
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
#ifndef nsUTF8Utils_h_
#define nsUTF8Utils_h_
// NB: This code may be used from non-XPCOM code, in particular, the
// standalone updater executable. That is, this file may be used in
// two ways: if MOZILLA_INTERNAL_API is defined, this file will
// provide signatures for the Mozilla abstract string types. It will
// use XPCOM assertion/debugging macros, etc.
#include <type_traits>
#include "mozilla/Assertions.h"
#include "mozilla/EndianUtils.h"
#include "nsCharTraits.h"
#ifdef MOZILLA_INTERNAL_API
# define UTF8UTILS_WARNING(msg) NS_WARNING(msg)
#else
# define UTF8UTILS_WARNING(msg)
#endif
class UTF8traits {
public:
static bool isASCII(char aChar) { return (aChar & 0x80) == 0x00; }
static bool isInSeq(char aChar) { return (aChar & 0xC0) == 0x80; }
static bool is2byte(char aChar) { return (aChar & 0xE0) == 0xC0; }
static bool is3byte(char aChar) { return (aChar & 0xF0) == 0xE0; }
static bool is4byte(char aChar) { return (aChar & 0xF8) == 0xF0; }
static bool is5byte(char aChar) { return (aChar & 0xFC) == 0xF8; }
static bool is6byte(char aChar) { return (aChar & 0xFE) == 0xFC; }
// return the number of bytes in a sequence beginning with aChar
static int bytes(char aChar) {
if (isASCII(aChar)) {
return 1;
}
if (is2byte(aChar)) {
return 2;
}
if (is3byte(aChar)) {
return 3;
}
if (is4byte(aChar)) {
return 4;
}
MOZ_ASSERT_UNREACHABLE("should not be used for in-sequence characters");
return 1;
}
};
/**
* Extract the next Unicode scalar value from the buffer and return it. The
* pointer passed in is advanced to the start of the next character in the
* buffer. Upon error, the return value is 0xFFFD, *aBuffer is advanced
* over the maximal valid prefix and *aErr is set to true (if aErr is not
* null).
*
* Note: This method never sets *aErr to false to allow error accumulation
* across multiple calls.
*
* Precondition: *aBuffer < aEnd
*/
class UTF8CharEnumerator {
public:
static inline char32_t NextChar(const char** aBuffer, const char* aEnd,
bool* aErr = nullptr) {
MOZ_ASSERT(aBuffer, "null buffer pointer pointer");
MOZ_ASSERT(aEnd, "null end pointer");
const unsigned char* p = reinterpret_cast<const unsigned char*>(*aBuffer);
const unsigned char* end = reinterpret_cast<const unsigned char*>(aEnd);
MOZ_ASSERT(p, "null buffer");
MOZ_ASSERT(p < end, "Bogus range");
unsigned char first = *p;
++p;
if (MOZ_LIKELY(first < 0x80U)) {
*aBuffer = reinterpret_cast<const char*>(p);
return first;
}
// Unsigned underflow is defined behavior
if (MOZ_UNLIKELY((p == end) || ((first - 0xC2U) >= (0xF5U - 0xC2U)))) {
*aBuffer = reinterpret_cast<const char*>(p);
if (aErr) {
*aErr = true;
}
return 0xFFFDU;
}
unsigned char second = *p;
if (first < 0xE0U) {
// Two-byte
if (MOZ_LIKELY((second & 0xC0U) == 0x80U)) {
++p;
*aBuffer = reinterpret_cast<const char*>(p);
return ((uint32_t(first) & 0x1FU) << 6) | (uint32_t(second) & 0x3FU);
}
*aBuffer = reinterpret_cast<const char*>(p);
if (aErr) {
*aErr = true;
}
return 0xFFFDU;
}
if (MOZ_LIKELY(first < 0xF0U)) {
// Three-byte
unsigned char lower = 0x80U;
unsigned char upper = 0xBFU;
if (first == 0xE0U) {
lower = 0xA0U;
} else if (first == 0xEDU) {
upper = 0x9FU;
}
if (MOZ_LIKELY(second >= lower && second <= upper)) {
++p;
if (MOZ_LIKELY(p != end)) {
unsigned char third = *p;
if (MOZ_LIKELY((third & 0xC0U) == 0x80U)) {
++p;
*aBuffer = reinterpret_cast<const char*>(p);
return ((uint32_t(first) & 0xFU) << 12) |
((uint32_t(second) & 0x3FU) << 6) |
(uint32_t(third) & 0x3FU);
}
}
}
*aBuffer = reinterpret_cast<const char*>(p);
if (aErr) {
*aErr = true;
}
return 0xFFFDU;
}
// Four-byte
unsigned char lower = 0x80U;
unsigned char upper = 0xBFU;
if (first == 0xF0U) {
lower = 0x90U;
} else if (first == 0xF4U) {
upper = 0x8FU;
}
if (MOZ_LIKELY(second >= lower && second <= upper)) {
++p;
if (MOZ_LIKELY(p != end)) {
unsigned char third = *p;
if (MOZ_LIKELY((third & 0xC0U) == 0x80U)) {
++p;
if (MOZ_LIKELY(p != end)) {
unsigned char fourth = *p;
if (MOZ_LIKELY((fourth & 0xC0U) == 0x80U)) {
++p;
*aBuffer = reinterpret_cast<const char*>(p);
return ((uint32_t(first) & 0x7U) << 18) |
((uint32_t(second) & 0x3FU) << 12) |
((uint32_t(third) & 0x3FU) << 6) |
(uint32_t(fourth) & 0x3FU);
}
}
}
}
}
*aBuffer = reinterpret_cast<const char*>(p);
if (aErr) {
*aErr = true;
}
return 0xFFFDU;
}
};
/**
* Extract the next Unicode scalar value from the buffer and return it. The
* pointer passed in is advanced to the start of the next character in the
* buffer. Upon error, the return value is 0xFFFD, *aBuffer is advanced over
* the unpaired surrogate and *aErr is set to true (if aErr is not null).
*
* Note: This method never sets *aErr to false to allow error accumulation
* across multiple calls.
*
* Precondition: *aBuffer < aEnd
*/
class UTF16CharEnumerator {
public:
static inline char32_t NextChar(const char16_t** aBuffer,
const char16_t* aEnd, bool* aErr = nullptr) {
MOZ_ASSERT(aBuffer, "null buffer pointer pointer");
MOZ_ASSERT(aEnd, "null end pointer");
const char16_t* p = *aBuffer;
MOZ_ASSERT(p, "null buffer");
MOZ_ASSERT(p < aEnd, "Bogus range");
char16_t c = *p++;
// Let's use encoding_rs-style code golf here.
// Unsigned underflow is defined behavior
char16_t cMinusSurrogateStart = c - 0xD800U;
if (MOZ_LIKELY(cMinusSurrogateStart > (0xDFFFU - 0xD800U))) {
*aBuffer = p;
return c;
}
if (MOZ_LIKELY(cMinusSurrogateStart <= (0xDBFFU - 0xD800U))) {
// High surrogate
if (MOZ_LIKELY(p != aEnd)) {
char16_t second = *p;
// Unsigned underflow is defined behavior
if (MOZ_LIKELY((second - 0xDC00U) <= (0xDFFFU - 0xDC00U))) {
*aBuffer = ++p;
return (uint32_t(c) << 10) + uint32_t(second) -
(((0xD800U << 10) - 0x10000U) + 0xDC00U);
}
}
}
// Unpaired surrogate
*aBuffer = p;
if (aErr) {
*aErr = true;
}
return 0xFFFDU;
}
};
template <typename Char, typename UnsignedT>
inline UnsignedT RewindToPriorUTF8Codepoint(const Char* utf8Chars,
UnsignedT index) {
static_assert(std::is_same_v<Char, char> ||
std::is_same_v<Char, unsigned char> ||
std::is_same_v<Char, signed char>,
"UTF-8 data must be in 8-bit units");
static_assert(std::is_unsigned_v<UnsignedT>, "index type must be unsigned");
while (index > 0 && (utf8Chars[index] & 0xC0) == 0x80) --index;
return index;
}
#undef UTF8UTILS_WARNING
#endif /* !defined(nsUTF8Utils_h_) */
|
