vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

grammars: handle x{n} and fix x{n,n}

Browse source
This commit is contained in:
Olivier Chafik 2024-04-12 16:11:10 +01:00
parent 01604690c1
commit f2030e3210
1 changed files with 26 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

21
common/grammar-parser.cpp
View file

@ -199,7 +199,8 @@ namespace grammar_parser {
sub_rule.push_back({LLAMA_GRETYPE_RULE_REF, star_rule_id}); sub_rule.push_back({LLAMA_GRETYPE_RULE_REF, star_rule_id});
} else { } else {
uint32_t last_rec_rule_id = 0; uint32_t last_rec_rule_id = 0;
for (int i = 0, n = max_times - min_times; i < n; i++) { auto n_opt = max_times - min_times;
for (int i = 0; i < n_opt; i++) {
uint32_t rec_rule_id = generate_symbol_id(state, rule_name + "_" + std::to_string(i + 1)); uint32_t rec_rule_id = generate_symbol_id(state, rule_name + "_" + std::to_string(i + 1));
if (i == 0) { if (i == 0) {
add_rule(state, rec_rule_id, { add_rule(state, rec_rule_id, {
@ -217,8 +218,10 @@ namespace grammar_parser {
} }
last_rec_rule_id = rec_rule_id; last_rec_rule_id = rec_rule_id;
} }
if (n_opt > 0) {
sub_rule.push_back({LLAMA_GRETYPE_RULE_REF, last_rec_rule_id}); sub_rule.push_back({LLAMA_GRETYPE_RULE_REF, last_rec_rule_id});
} }
}
sub_rule.push_back({LLAMA_GRETYPE_END, 0}); sub_rule.push_back({LLAMA_GRETYPE_END, 0});
add_rule(state, sub_rule_id, sub_rule); add_rule(state, sub_rule_id, sub_rule);
@ -291,24 +294,34 @@ namespace grammar_parser {
pos = parse_space(pos + 1, is_nested); pos = parse_space(pos + 1, is_nested);
size_t min_times = 0; size_t min_times = 0;
int max_times = -1; int max_times = -1;
if (is_digit_char(*pos)) { if (is_digit_char(*pos)) {
const char * int_end = parse_int(pos); const char * int_end = parse_int(pos);
min_times = std::stoul(std::string(pos, int_end - pos)); min_times = std::stoul(std::string(pos, int_end - pos));
pos = parse_space(int_end, is_nested); pos = parse_space(int_end, is_nested);
} else if (*pos != ',') {
throw std::runtime_error(std::string("expecting an int or ',' at ") + pos);
} }
if (*pos != ',') {
throw std::runtime_error(std::string("expecting ',' at ") + pos); if (*pos == '}') {
} max_times = min_times;
pos = parse_space(pos + 1, is_nested); pos = parse_space(pos + 1, is_nested);
} else if (*pos == ',') {
pos = parse_space(pos + 1, is_nested);
if (is_digit_char(*pos)) { if (is_digit_char(*pos)) {
const char * int_end = parse_int(pos); const char * int_end = parse_int(pos);
max_times = std::stoul(std::string(pos, int_end - pos)); max_times = std::stoul(std::string(pos, int_end - pos));
pos = parse_space(int_end, is_nested); pos = parse_space(int_end, is_nested);
} }
if (*pos != '}') { if (*pos != '}') {
throw std::runtime_error(std::string("expecting '}' at ") + pos); throw std::runtime_error(std::string("expecting '}' at ") + pos);
} }
pos = parse_space(pos + 1, is_nested); pos = parse_space(pos + 1, is_nested);
} else {
throw std::runtime_error(std::string("expecting ',' at ") + pos);
}
handle_repetitions(min_times, max_times); handle_repetitions(min_times, max_times);
} else { } else {
break; break;