STATIC mp_obj_t re_exec(bool is_anchored, uint n_args, const mp_obj_t *args) {
mp_obj_re_t *self = args[0];
Subject subj;
mp_uint_t len;
subj.begin = mp_obj_str_get_data(args[1], &len);
subj.end = subj.begin + len;
int caps_num = (self->re.sub + 1) * 2;
mp_obj_match_t *match = m_new_obj_var(mp_obj_match_t, char*, caps_num);
int res = re1_5_recursiveloopprog(&self->re, &subj, match->caps, caps_num, is_anchored);
if (res == 0) {
m_del_var(mp_obj_match_t, char*, caps_num, match);
return mp_const_none;
}
match->base.type = &match_type;
match->num_matches = caps_num;
match->str = args[1];
return match;
}
STATIC mp_obj_t re_split(mp_uint_t n_args, const mp_obj_t *args) {
mp_obj_re_t *self = MP_OBJ_TO_PTR(args[0]);
Subject subj;
mp_uint_t len;
subj.begin = mp_obj_str_get_data(args[1], &len);
subj.end = subj.begin + len;
int caps_num = (self->re.sub + 1) * 2;
int maxsplit = 0;
if (n_args > 2) {
maxsplit = mp_obj_get_int(args[2]);
}
mp_obj_t retval = mp_obj_new_list(0, NULL);
const char **caps = alloca(caps_num * sizeof(char*));
while (true) {
// cast is a workaround for a bug in msvc: it treats const char** as a const pointer instead of a pointer to pointer to const char
memset((char**)caps, 0, caps_num * sizeof(char*));
int res = re1_5_recursiveloopprog(&self->re, &subj, caps, caps_num, false);
// if we didn't have a match, or had an empty match, it's time to stop
if (!res || caps[0] == caps[1]) {
break;
}
mp_obj_t s = mp_obj_new_str(subj.begin, caps[0] - subj.begin, false);
mp_obj_list_append(retval, s);
if (self->re.sub > 0) {
mp_not_implemented("Splitting with sub-captures");
}
subj.begin = caps[1];
if (maxsplit > 0 && --maxsplit == 0) {
break;
}
}
mp_obj_t s = mp_obj_new_str(subj.begin, subj.end - subj.begin, false);
mp_obj_list_append(retval, s);
return retval;
}
STATIC mp_obj_t re_exec(bool is_anchored, uint n_args, const mp_obj_t *args) {
(void)n_args;
mp_obj_re_t *self = MP_OBJ_TO_PTR(args[0]);
Subject subj;
mp_uint_t len;
subj.begin = mp_obj_str_get_data(args[1], &len);
subj.end = subj.begin + len;
int caps_num = (self->re.sub + 1) * 2;
mp_obj_match_t *match = m_new_obj_var(mp_obj_match_t, char*, caps_num);
// cast is a workaround for a bug in msvc: it treats const char** as a const pointer instead of a pointer to pointer to const char
memset((char*)match->caps, 0, caps_num * sizeof(char*));
int res = re1_5_recursiveloopprog(&self->re, &subj, match->caps, caps_num, is_anchored);
if (res == 0) {
m_del_var(mp_obj_match_t, char*, caps_num, match);
return mp_const_none;
}
match->base.type = &match_type;
match->num_matches = caps_num / 2; // caps_num counts start and end pointers
match->str = args[1];
return MP_OBJ_FROM_PTR(match);
}
STATIC mp_obj_t re_split(uint n_args, const mp_obj_t *args) {
mp_obj_re_t *self = args[0];
Subject subj;
mp_uint_t len;
subj.begin = mp_obj_str_get_data(args[1], &len);
subj.end = subj.begin + len;
int caps_num = (self->re.sub + 1) * 2;
int maxsplit = 0;
if (n_args > 2) {
maxsplit = mp_obj_int_get_truncated(args[2]);
}
mp_obj_t retval = mp_obj_new_list(0, NULL);
const char **caps = alloca(caps_num * sizeof(char*));
while (true) {
int res = re1_5_recursiveloopprog(&self->re, &subj, caps, caps_num, false);
// if we didn't have a match, or had an empty match, it's time to stop
if (!res || caps[0] == caps[1]) {
break;
}
mp_obj_t s = mp_obj_new_str(subj.begin, caps[0] - subj.begin, false);
mp_obj_list_append(retval, s);
if (self->re.sub > 0) {
mp_not_implemented("Splitting with sub-captures");
}
subj.begin = caps[1];
if (maxsplit > 0 && --maxsplit == 0) {
break;
}
}
mp_obj_t s = mp_obj_new_str(subj.begin, subj.end - subj.begin, false);
mp_obj_list_append(retval, s);
return retval;
}
