grammar : improve regex when naming ref derived rules

Author: aldehir

common/json-schema-to-grammar.cpp

@@ -603,7 +603,8 @@ class SchemaConverter {
     std::string _resolve_ref(const std::string & ref) {
         auto it = ref.find('#');
         std::string ref_fragment = it != std::string::npos ? ref.substr(it + 1) : ref;
-        std::string ref_name = "ref" + std::regex_replace(ref_fragment, std::regex(R"([^a-zA-Z0-9-])"), "-");
+        static const std::regex nonalphanumeric_regex(R"([^a-zA-Z0-9-]+)");
+        std::string ref_name = "ref" + std::regex_replace(ref_fragment, nonalphanumeric_regex, "-");
         if (_rules.find(ref_name) == _rules.end() && _refs_being_resolved.find(ref) == _refs_being_resolved.end()) {
             _refs_being_resolved.insert(ref);
             json resolved = _refs[ref];

examples/json_schema_to_grammar.py

@@ -557,7 +557,7 @@ def join_seq():
 
     def _resolve_ref(self, ref):
         ref_fragment = ref.split('#')[-1]
-        ref_name = 'ref' + re.sub(r'[^a-zA-Z0-9-]', '-', ref_fragment)
+        ref_name = 'ref' + re.sub(r'[^a-zA-Z0-9-]+', '-', ref_fragment)
         if ref_name not in self._rules and ref not in self._refs_being_resolved:
             self._refs_being_resolved.add(ref)
             resolved = self._refs[ref]

tools/server/public_legacy/json-schema-to-grammar.mjs

@@ -599,7 +599,7 @@ export class SchemaConverter {
 
   _resolveRef(ref) {
     let refFragment = ref.split('#').pop();
-    let refName = 'ref' + refFragment.replace(/[^a-zA-Z0-9-]/, '-');
+    let refName = 'ref' + refFragment.replace(/[^a-zA-Z0-9-]+/g, '-');
     if (!(refName in this._rules) && !this._refsBeingResolved.has(ref)) {
       this._refsBeingResolved.add(ref);
       const resolved = this._refs[ref];