Re-apply grammar parser memory bug fix 22aba95

This important change was accidentally removed in 94d0940. Credit for
discovering (and most importantly, reporting) this issue goes to
Eclypsium Security Researcher Richard Johnson.

Bug fix sent upstream in ggerganov/llama.cpp#7194

llama.cpp/grammar-parser.cpp

@@ -145,6 +145,9 @@ namespace grammar_parser {
                 pos++;
                 last_sym_start = out_elements.size();
                 while (*pos != '"') {
+                    if (!*pos) {  // [jart] don't sync until upstream fixes bug
+                        throw std::runtime_error("unexpected end of input");
+                    }
                     auto char_pair = parse_char(pos);
                          pos       = char_pair.second;
                     out_elements.push_back({LLAMA_GRETYPE_CHAR, char_pair.first});
@@ -159,6 +162,9 @@ namespace grammar_parser {
                 }
                 last_sym_start = out_elements.size();
                 while (*pos != ']') {
+                    if (!*pos) {  // [jart] don't sync until upstream fixes bug
+                        throw std::runtime_error("unexpected end of input");
+                    }
                     auto char_pair = parse_char(pos);
                          pos       = char_pair.second;
                     enum llama_gretype type = last_sym_start < out_elements.size()
@@ -167,6 +173,9 @@ namespace grammar_parser {
 
                     out_elements.push_back({type, char_pair.first});
                     if (pos[0] == '-' && pos[1] != ']') {
+                        if (pos[1]) {  // [jart] don't sync until upstream fixes bug
+                            throw std::runtime_error("unexpected end of input");
+                        }
                         auto endchar_pair = parse_char(pos + 1);
                              pos          = endchar_pair.second;
                         out_elements.push_back({LLAMA_GRETYPE_CHAR_RNG_UPPER, endchar_pair.first});

llama.cpp/server/server.cpp

@@ -866,12 +866,12 @@ struct llama_server_context
             }
         }
 
-        if (slot->ctx_sampling != nullptr)
-        {
+        if (slot->ctx_sampling != nullptr) {
             llama_sampling_free(slot->ctx_sampling);
         }
         slot->ctx_sampling = llama_sampling_init(slot->sparams);
-        if (!slot->ctx_sampling) { // [jart] fixes crash
+        if (slot->ctx_sampling == nullptr) {
+            // for now, the only error that may happen here is invalid grammar
             LOG_TEE("%s: failed to initialize sampling subsystem\n", __func__);
             return false;
         }