/**
* @brief Get previous cell from string
*
* @param s : the string
* @param pos : the position in string to get cell previous to
* @param cell : the output buffer
* @param is_decomp_am : whether SARA AM is to be decomposed into
* NIKHANIT and SARA AA and to be in separate cells
*
* @returns total chars consumed by the cell
*
* Gets last cell from the string bounded by @a s and @a pos, and if @a cell
* is not null, stores the cell data in it.
*/
size_t
th_prev_cell (const thchar_t *s, size_t pos,
struct thcell_t *cell, int is_decomp_am)
{
size_t n = 0;
struct thcell_t acell;
acell.base = acell.hilo = acell.top = 0;
if (pos > 0) {
do {
thchar_t c = s[pos-1];
switch (th_chlevel (c)) {
case 0:
if (is_decomp_am && c == TIS_SARA_AM) {
acell.hilo = c;
} else {
acell.base = c;
}
break;
case 1:
if (acell.hilo && th_chlevel(acell.hilo) == 3) {
acell.top = acell.hilo;
}
/* fall through */
case -1:
acell.hilo = c;
break;
case 2:
acell.top = c;
break;
case 3:
if (!acell.hilo) {
acell.hilo = c;
} else {
acell.top = c;
}
break;
}
++n; --pos;
} while (
pos > 0 && (
is_composible (s[pos-1], s[pos]) || (
acell.hilo == TIS_SARA_AM &&
!acell.base && (
(!acell.top && th_isthtone (s[pos-1])) ||
th_isthcons (s[pos-1])
)
)
)
);
}
if (cell) {
*cell = acell;
}
return n;
}
/**
* @brief Get first cell from string
*
* @param s : the string
* @param len : the length of string
* @param cell : the output buffer
* @param is_decomp_am : whether SARA AM is to be decomposed into
* NIKHANIT and SARA AA and to be in separate cells
*
* @returns total chars consumed by the cell
*
* Gets first cell from the string bounded by @a s and @a len, and, if @a cell
* is not null, stores the cell data in it.
*/
size_t
th_next_cell (const thchar_t *s, size_t len,
struct thcell_t *cell, int is_decomp_am)
{
size_t n = 0;
struct thcell_t acell;
acell.base = acell.hilo = acell.top = 0;
if (len > 0) {
do {
switch (th_chlevel (*s)) {
case 0:
if (is_decomp_am && *s == TIS_SARA_AM) {
acell.hilo = *s++;
} else {
acell.base = *s++;
}
break;
case -1:
case 1:
acell.hilo = *s++;
break;
case 2:
acell.top = *s++;
break;
case 3:
if (!acell.hilo) {
acell.hilo = *s++;
} else {
acell.top = *s++;
}
break;
}
++n; --len;
} while (
len > 0 && (
is_composible (s[-1], s[0]) || (
is_decomp_am && s[0] == TIS_SARA_AM
&& th_isthcons (acell.base) && acell.hilo == 0
)
)
);
}
if (cell) {
*cell = acell;
}
return n;
}
static void
th_brkpos_hints (const thchar_t *str, int len, char *hints)
{
int i;
if (len < 0)
len = strlen ((const char *)str);
memset (hints, 0, len);
for (i = 0; i < len; /* nop */) {
if (th_isthcons (str[i])) {
if (i+1 < len && str[i+1] == TIS_THANTHAKHAT) {
i += 2; /* the cons + THANTHAKHAT */
} else if (i+2 < len && str[i+2] == TIS_THANTHAKHAT) {
i += 3; /* the cons + intermediate char + THANTHAKHAT */
} else if (i+2 < len
&& str[i] != TIS_KO_KAI && str[i+1] == TIS_MAITAIKHU
&& (str[i+2] == TIS_O_ANG || str[i+2] == TIS_WO_WAEN))
{
hints[i] = 1;
i += 4; /* the cons + MAITAIKHU + OANG/WOWAEN + cons */
} else if ((i > 0
&& (str[i-1] == TIS_MAI_HAN_AKAT
|| str[i-1] == TIS_SARA_UEE))
|| (i > 1 && th_isthtone (str[i-1])
&& (str[i-2] == TIS_MAI_HAN_AKAT
|| str[i-2] == TIS_SARA_UEE)))
{
i++;
} else {
hints[i++] = 1;
}
} else if (str[i] == TIS_SARA_E || str[i] == TIS_SARA_AE) {
hints[i] = 1; /* sara e/ae */
i += 2; /* sara e/ae + the supposedly cons */
if (i >= len)
break;
if (str[i] == TIS_MAITAIKHU) {
i += 2; /* MAITAIKHU + the supposedly cons */
} else if (th_isupvowel (str[i])) {
i++; /* the upper vowel, as part of composite vowel */
if (i < len && th_isthtone (str[i]))
i++;
i++; /* the supposedly cons */
} else if (i+2 < len
&& ((str[i+1] == TIS_SARA_AA && str[i+2] == TIS_SARA_A)
|| (str[i] != TIS_KO_KAI
&& str[i+1] == TIS_MAITAIKHU
&& str[i+2] != TIS_O_ANG
&& str[i+2] != TIS_WO_WAEN)))
{
i += 3; /* 2nd cons + SARA_AA + SARA_A, or
* 2nd cons + MAITAIKHU + final cons
*/
}
} else if (th_isldvowel (str[i])) {
hints[i] = 1; /* the ldvowel */
i += 2; /* the ldvowel + the supposedly cons */
} else if (str[i] == TIS_RU || str[i] == TIS_LU) {
hints[i++] = 1;
} else {
i++;
}
}
}