/* utf32 index => original byte offset */
int UTF8Converter::getOriginOffset(int origIdx) {
uint offs = 0;
while (origIdx > 0 && offs < _str.size()) {
// increment if it's not the start of a utf8 sequence
(void)( (++offs < _str.size() && isutf(_str[offs])) ||
(++offs < _str.size() && isutf(_str[offs])) ||
(++offs < _str.size() && isutf(_str[offs])) ||
++offs);
origIdx--;
}
return offs;
}
/* conversions without error checking
only works for valid UTF-8, i.e. no 5- or 6-byte sequences
srcsz = source size in bytes
sz = dest size in # of wide characters
returns # characters converted
if sz == srcsz+1 (i.e. 4*srcsz+4 bytes), there will always be enough space.
*/
size_t u8_toucs(uint32_t *dest, size_t sz, const char *src, size_t srcsz)
{
uint32_t ch;
const char *src_end = src + srcsz;
size_t nb;
size_t i=0;
if (sz == 0 || srcsz == 0)
return 0;
while (i < sz) {
if (!isutf(*src)) { // invalid sequence
dest[i++] = 0xFFFD;
src++;
if (src >= src_end) break;
continue;
}
nb = trailingBytesForUTF8[(unsigned char)*src];
if (src + nb >= src_end)
break;
ch = 0;
switch (nb) {
/* these fall through deliberately */
case 5: ch += (unsigned char)*src++; ch <<= 6;
case 4: ch += (unsigned char)*src++; ch <<= 6;
case 3: ch += (unsigned char)*src++; ch <<= 6;
case 2: ch += (unsigned char)*src++; ch <<= 6;
case 1: ch += (unsigned char)*src++; ch <<= 6;
case 0: ch += (unsigned char)*src++;
}
ch -= offsetsFromUTF8[nb];
dest[i++] = ch;
}
return i;
}
static int right_callback(int count, int key) {
reset_indent();
do {
rl_point += (rl_line_buffer[rl_point] == '\n') ? prompt_length+1 : 1;
} while (locale_is_utf8 && !isutf(rl_line_buffer[rl_point]));
if (rl_end < rl_point) rl_point = rl_end;
return 0;
}
static int left_callback(int count, int key) {
reset_indent();
if (rl_point > 0) {
int i = line_start(rl_point);
do {
rl_point = (i == 0 || rl_point-i > prompt_length) ? rl_point-1 : i-1;
} while (locale_is_utf8 && !isutf(rl_line_buffer[rl_point]) && rl_point > i-1);
}
return 0;
}
static int delete_callback(int count, int key) {
reset_indent();
int j = rl_point;
do {
j += (rl_line_buffer[j] == '\n') ? prompt_length+1 : 1;
} while (locale_is_utf8 && !isutf(rl_line_buffer[j]));
if (rl_end < j) j = rl_end;
rl_delete_text(rl_point, j);
return 0;
}
/* reads the next utf-8 sequence out of a string, updating an index */
uint32 UTF8Converter::nextchar(const char *s, int *i) {
uint32 ch = 0;
int sz = 0;
do {
ch <<= 6;
ch += (byte)s[(*i)++];
sz++;
} while (s[*i] && !isutf(s[*i]));
ch -= offsetsFromUTF8[sz - 1];
return ch;
}
static int backspace_callback(int count, int key) {
reset_indent();
if (!rl_point) return 0;
int i = line_start(rl_point), j = rl_point, k;
if (!i || rl_point <= i + prompt_length) goto backspace;
for (k = i; k < rl_point; k++)
if (rl_line_buffer[k] != ' ') goto backspace;
//unindent:
k = i + prompt_length;
do { rl_point--; } while ((rl_point - k) % tab_width);
goto finish;
backspace:
do {
rl_point = (i == 0 || rl_point-i > prompt_length) ? rl_point-1 : i-1;
} while (locale_is_utf8 && !isutf(rl_line_buffer[rl_point]) && rl_point > i-1);
finish:
rl_delete_text(rl_point, j);
return 0;
}
/*
* Decrements i to point to the previous unicode glyph
*
*/
void u8_dec(char *s, int *i) {
(void)(isutf(s[--(*i)]) || isutf(s[--(*i)]) || isutf(s[--(*i)]) || --(*i));
}