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
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
|
#ifndef SIMDUTF_UTF8_H
#define SIMDUTF_UTF8_H
namespace simdutf {
namespace scalar {
namespace {
namespace utf8 {
#if SIMDUTF_IMPLEMENTATION_FALLBACK || SIMDUTF_IMPLEMENTATION_RVV
// only used by the fallback kernel.
// credit: based on code from Google Fuchsia (Apache Licensed)
inline simdutf_warn_unused bool validate(const char *buf, size_t len) noexcept {
const uint8_t *data = reinterpret_cast<const uint8_t *>(buf);
uint64_t pos = 0;
uint32_t code_point = 0;
while (pos < len) {
// check of the next 16 bytes are ascii.
uint64_t next_pos = pos + 16;
if (next_pos <=
len) { // if it is safe to read 16 more bytes, check that they are ascii
uint64_t v1;
std::memcpy(&v1, data + pos, sizeof(uint64_t));
uint64_t v2;
std::memcpy(&v2, data + pos + sizeof(uint64_t), sizeof(uint64_t));
uint64_t v{v1 | v2};
if ((v & 0x8080808080808080) == 0) {
pos = next_pos;
continue;
}
}
unsigned char byte = data[pos];
while (byte < 0b10000000) {
if (++pos == len) {
return true;
}
byte = data[pos];
}
if ((byte & 0b11100000) == 0b11000000) {
next_pos = pos + 2;
if (next_pos > len) {
return false;
}
if ((data[pos + 1] & 0b11000000) != 0b10000000) {
return false;
}
// range check
code_point = (byte & 0b00011111) << 6 | (data[pos + 1] & 0b00111111);
if ((code_point < 0x80) || (0x7ff < code_point)) {
return false;
}
} else if ((byte & 0b11110000) == 0b11100000) {
next_pos = pos + 3;
if (next_pos > len) {
return false;
}
if ((data[pos + 1] & 0b11000000) != 0b10000000) {
return false;
}
if ((data[pos + 2] & 0b11000000) != 0b10000000) {
return false;
}
// range check
code_point = (byte & 0b00001111) << 12 |
(data[pos + 1] & 0b00111111) << 6 |
(data[pos + 2] & 0b00111111);
if ((code_point < 0x800) || (0xffff < code_point) ||
(0xd7ff < code_point && code_point < 0xe000)) {
return false;
}
} else if ((byte & 0b11111000) == 0b11110000) { // 0b11110000
next_pos = pos + 4;
if (next_pos > len) {
return false;
}
if ((data[pos + 1] & 0b11000000) != 0b10000000) {
return false;
}
if ((data[pos + 2] & 0b11000000) != 0b10000000) {
return false;
}
if ((data[pos + 3] & 0b11000000) != 0b10000000) {
return false;
}
// range check
code_point =
(byte & 0b00000111) << 18 | (data[pos + 1] & 0b00111111) << 12 |
(data[pos + 2] & 0b00111111) << 6 | (data[pos + 3] & 0b00111111);
if (code_point <= 0xffff || 0x10ffff < code_point) {
return false;
}
} else {
// we may have a continuation
return false;
}
pos = next_pos;
}
return true;
}
#endif
inline simdutf_warn_unused result validate_with_errors(const char *buf,
size_t len) noexcept {
const uint8_t *data = reinterpret_cast<const uint8_t *>(buf);
size_t pos = 0;
uint32_t code_point = 0;
while (pos < len) {
// check of the next 16 bytes are ascii.
size_t next_pos = pos + 16;
if (next_pos <=
len) { // if it is safe to read 16 more bytes, check that they are ascii
uint64_t v1;
std::memcpy(&v1, data + pos, sizeof(uint64_t));
uint64_t v2;
std::memcpy(&v2, data + pos + sizeof(uint64_t), sizeof(uint64_t));
uint64_t v{v1 | v2};
if ((v & 0x8080808080808080) == 0) {
pos = next_pos;
continue;
}
}
unsigned char byte = data[pos];
while (byte < 0b10000000) {
if (++pos == len) {
return result(error_code::SUCCESS, len);
}
byte = data[pos];
}
if ((byte & 0b11100000) == 0b11000000) {
next_pos = pos + 2;
if (next_pos > len) {
return result(error_code::TOO_SHORT, pos);
}
if ((data[pos + 1] & 0b11000000) != 0b10000000) {
return result(error_code::TOO_SHORT, pos);
}
// range check
code_point = (byte & 0b00011111) << 6 | (data[pos + 1] & 0b00111111);
if ((code_point < 0x80) || (0x7ff < code_point)) {
return result(error_code::OVERLONG, pos);
}
} else if ((byte & 0b11110000) == 0b11100000) {
next_pos = pos + 3;
if (next_pos > len) {
return result(error_code::TOO_SHORT, pos);
}
if ((data[pos + 1] & 0b11000000) != 0b10000000) {
return result(error_code::TOO_SHORT, pos);
}
if ((data[pos + 2] & 0b11000000) != 0b10000000) {
return result(error_code::TOO_SHORT, pos);
}
// range check
code_point = (byte & 0b00001111) << 12 |
(data[pos + 1] & 0b00111111) << 6 |
(data[pos + 2] & 0b00111111);
if ((code_point < 0x800) || (0xffff < code_point)) {
return result(error_code::OVERLONG, pos);
}
if (0xd7ff < code_point && code_point < 0xe000) {
return result(error_code::SURROGATE, pos);
}
} else if ((byte & 0b11111000) == 0b11110000) { // 0b11110000
next_pos = pos + 4;
if (next_pos > len) {
return result(error_code::TOO_SHORT, pos);
}
if ((data[pos + 1] & 0b11000000) != 0b10000000) {
return result(error_code::TOO_SHORT, pos);
}
if ((data[pos + 2] & 0b11000000) != 0b10000000) {
return result(error_code::TOO_SHORT, pos);
}
if ((data[pos + 3] & 0b11000000) != 0b10000000) {
return result(error_code::TOO_SHORT, pos);
}
// range check
code_point =
(byte & 0b00000111) << 18 | (data[pos + 1] & 0b00111111) << 12 |
(data[pos + 2] & 0b00111111) << 6 | (data[pos + 3] & 0b00111111);
if (code_point <= 0xffff) {
return result(error_code::OVERLONG, pos);
}
if (0x10ffff < code_point) {
return result(error_code::TOO_LARGE, pos);
}
} else {
// we either have too many continuation bytes or an invalid leading byte
if ((byte & 0b11000000) == 0b10000000) {
return result(error_code::TOO_LONG, pos);
} else {
return result(error_code::HEADER_BITS, pos);
}
}
pos = next_pos;
}
return result(error_code::SUCCESS, len);
}
// Finds the previous leading byte starting backward from buf and validates with
// errors from there Used to pinpoint the location of an error when an invalid
// chunk is detected We assume that the stream starts with a leading byte, and
// to check that it is the case, we ask that you pass a pointer to the start of
// the stream (start).
inline simdutf_warn_unused result rewind_and_validate_with_errors(
const char *start, const char *buf, size_t len) noexcept {
// First check that we start with a leading byte
if ((*start & 0b11000000) == 0b10000000) {
return result(error_code::TOO_LONG, 0);
}
size_t extra_len{0};
// A leading byte cannot be further than 4 bytes away
for (int i = 0; i < 5; i++) {
unsigned char byte = *buf;
if ((byte & 0b11000000) != 0b10000000) {
break;
} else {
buf--;
extra_len++;
}
}
result res = validate_with_errors(buf, len + extra_len);
res.count -= extra_len;
return res;
}
inline size_t count_code_points(const char *buf, size_t len) {
const int8_t *p = reinterpret_cast<const int8_t *>(buf);
size_t counter{0};
for (size_t i = 0; i < len; i++) {
// -65 is 0b10111111, anything larger in two-complement's should start a new
// code point.
if (p[i] > -65) {
counter++;
}
}
return counter;
}
inline size_t utf16_length_from_utf8(const char *buf, size_t len) {
const int8_t *p = reinterpret_cast<const int8_t *>(buf);
size_t counter{0};
for (size_t i = 0; i < len; i++) {
if (p[i] > -65) {
counter++;
}
if (uint8_t(p[i]) >= 240) {
counter++;
}
}
return counter;
}
simdutf_warn_unused inline size_t trim_partial_utf8(const char *input,
size_t length) {
if (length < 3) {
switch (length) {
case 2:
if (uint8_t(input[length - 1]) >= 0xc0) {
return length - 1;
} // 2-, 3- and 4-byte characters with only 1 byte left
if (uint8_t(input[length - 2]) >= 0xe0) {
return length - 2;
} // 3- and 4-byte characters with only 2 bytes left
return length;
case 1:
if (uint8_t(input[length - 1]) >= 0xc0) {
return length - 1;
} // 2-, 3- and 4-byte characters with only 1 byte left
return length;
case 0:
return length;
}
}
if (uint8_t(input[length - 1]) >= 0xc0) {
return length - 1;
} // 2-, 3- and 4-byte characters with only 1 byte left
if (uint8_t(input[length - 2]) >= 0xe0) {
return length - 2;
} // 3- and 4-byte characters with only 1 byte left
if (uint8_t(input[length - 3]) >= 0xf0) {
return length - 3;
} // 4-byte characters with only 3 bytes left
return length;
}
} // namespace utf8
} // unnamed namespace
} // namespace scalar
} // namespace simdutf
#endif
|
