llama : refactor sampling v2 (#9294)

- Add `struct llama_sampler` and `struct llama_sampler_i`
- Add `llama_sampler_` API
- Add `llama_sampler_chain_` API for chaining multiple samplers
- Remove `LLAMA_API_INTERNAL`
- Add `llama_perf_` API and remove old `llama_print_timings` and `llama_reset_timings`

tests/test-json-schema-to-grammar.cpp

@@ -2,14 +2,15 @@
 #undef NDEBUG
 #endif
 
+#include "json-schema-to-grammar.h"
+
+#include "llama-grammar.h"
+
 #include <cassert>
 #include <fstream>
 #include <sstream>
 #include <regex>
 
-#include "json-schema-to-grammar.h"
-#include "grammar-parser.h"
-
 static std::string trim(const std::string & source) {
     std::string s(source);
     s.erase(0,s.find_first_not_of(" \n\r\t"));
@@ -40,7 +41,8 @@ struct TestCase {
     }
     void verify_expectation_parseable() const {
         try {
-            auto state = grammar_parser::parse(expected_grammar.c_str());
+            llama_grammar_parser state;
+            state.parse(expected_grammar.c_str());
             if (state.symbol_ids.find("root") == state.symbol_ids.end()) {
                 throw std::runtime_error("Grammar failed to parse:\n" + expected_grammar);
             }