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
|
/* We're missing a start byte here */
if (OF_UNLIKELY(!(str[i] & 0x40))) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/* We have at minimum a 2 byte character -> check next byte */
if (OF_UNLIKELY(len < i + 1 || (str[i + 1] & 0xC0) != 0x80)) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/* Check if we have at minimum a 3 byte character */
if (OF_LIKELY(!(str[i] & 0x20))) {
i++;
continue;
}
/* We have at minimum a 3 byte char -> check second next byte */
if (OF_UNLIKELY(len < i + 2 || (str[i + 2] & 0xC0) != 0x80)) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/* Check if we have a 4 byte character */
if (OF_LIKELY(!(str[i] & 0x10))) {
i += 2;
continue;
}
/* We have a 4 byte character -> check third next byte */
if (OF_UNLIKELY(len < i + 3 || (str[i + 3] & 0xC0) != 0x80)) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/*
* Just in case, check if there's a 5th character, which is
* forbidden by UTF-8
|
|
|
|
|
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
|
/* We're missing a start byte here */
if (OF_UNLIKELY(!(str[i] & 0x40))) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/* We have at minimum a 2 byte character -> check next byte */
if (OF_UNLIKELY(len <= i + 1 || (str[i + 1] & 0xC0) != 0x80)) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/* Check if we have at minimum a 3 byte character */
if (OF_LIKELY(!(str[i] & 0x20))) {
i++;
continue;
}
/* We have at minimum a 3 byte char -> check second next byte */
if (OF_UNLIKELY(len <= i + 2 || (str[i + 2] & 0xC0) != 0x80)) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/* Check if we have a 4 byte character */
if (OF_LIKELY(!(str[i] & 0x10))) {
i += 2;
continue;
}
/* We have a 4 byte character -> check third next byte */
if (OF_UNLIKELY(len <= i + 3 || (str[i + 3] & 0xC0) != 0x80)) {
madvise((void*)str, len, MADV_NORMAL);
return -1;
}
/*
* Just in case, check if there's a 5th character, which is
* forbidden by UTF-8
|