fix(cpp): add regression tests for functor and type conversion operators (#1759)

gitgalaxy/core/detector.py

@@ -4445,7 +4445,7 @@ def _extract_name(self, raw_match: str) -> str:
             # Group 1 now grabs the optional `(Ident::)+` chain immediately
             # before the `operator` keyword and it's prefixed back on below.
             op_match = re.search(
-                r"((?:[a-zA-Z_]\w*::)*)\b(operator\s*(?:\[\s*\]|\(\s*\)|[^a-zA-Z0-9_\s({]+|[a-zA-Z_]\w*(?:\s*\*+)?))",
+                r"((?:[a-zA-Z_]\w*::)*)\b(operator\s*(?:\[\s*\]|\(\s*\)|(?:::)?[a-zA-Z_]\w*(?:::[a-zA-Z_]\w*)*(?:\s*\*+)?|[^a-zA-Z0-9_\s({]+))",
                 match_strip,
             )
             if op_match:

gitgalaxy/standards/language_standards.py

@@ -36,7 +36,7 @@
 | -------- | ----------- | -------------- | ------------ | --------------- |
 | Apex | 100.0% | 95.0% | 100.0% | 100.0% |
 | C | 93.3% | 99.5% | 100.0% | 100.0% |
-| Cpp | 92.3% | 96.2% | 98.6% | 92.6% |
+| Cpp | 93.4% | 95.7% | 98.6% | 92.6% |
 | Csharp | 99.2% | 99.8% | 100.0% | 100.0% |
 | Css | 100.0% | 100.0% | N/A | N/A |
 | Dart | 96.0% | 97.8% | 100.0% | 100.0% |
@@ -3011,7 +3011,7 @@ class PrismConfigSchema(TypedDict):
                 # literal `[*&]`, the other forbids consuming past the first
                 # non-whitespace char), so this doesn't reopen the Rule 14
                 # backtracking gap the surrounding bounds were built to close.
-                r"(?:(?![ \t]*#)[a-zA-Z_]\w*(?:::[a-zA-Z_]\w*)*"
+                r"(?:(?![ \t]*#)(?!(?:[a-zA-Z_]\w*::)*operator\b)[a-zA-Z_]\w*(?:::[a-zA-Z_]\w*)*"
                 r"(?:<(?:[^<>]|<(?:[^<>]|<[^<>]*>)*>)*>)?"
                 r"(?:[ \t]{0,20}[*&]{1,5}[ \t\n]{0,200}|[ \t\n]{1,200})){0,5}"
                 # 5. THE "NOT A FUNCTION" SHIELD
@@ -3025,7 +3025,7 @@ class PrismConfigSchema(TypedDict):
                 # `(?:[a-zA-Z_]\w*::)*`. Out-of-line operator overload definitions (defined in a
                 # .cpp file, declared in the header) are mainstream, common C++ -- completely
                 # invisible to func_start before this fix.
-                r"(?![ \t]*#)((?:[a-zA-Z_]\w*::)*[~a-zA-Z_]\w*|(?:[a-zA-Z_]\w*::)*operator[ \t]*[^a-zA-Z_\s(]+|(?:[a-zA-Z_]\w*::)*operator[ \t]+(?:new|delete)(?:\[\])?)"
+                r"(?![ \t]*#)((?:[a-zA-Z_]\w*::)*operator[ \t]*\(\)|(?:[a-zA-Z_]\w*::)*operator[ \t]+(?:::)?[a-zA-Z_]\w*(?:::[a-zA-Z_]\w*)*(?:[ \t]*[*&]+)?|(?:[a-zA-Z_]\w*::)*operator[ \t]*[^a-zA-Z_\s(]+|(?:[a-zA-Z_]\w*::)*operator[ \t]+(?:new|delete)(?:\[\])?|(?:[a-zA-Z_]\w*::)*[~a-zA-Z_]\w*)"
                 # 7. THE PARAMETER BLOCK (Supports vertical gap)
                 # [NESTED PARENTHESIS FIX]: Uses 1-Level Nesting Trick to swallow function pointers without ReDoS.
                 r"[ \t\n]{0,200}(?:ARGS\d+\s*\([^)]*\)|\((?:[^)(]|\([^)]*\))*\)|NOARGS)"

tests/extraction/languages/test_cpp.py

@@ -60,6 +60,22 @@
             "TargetClass& TargetClass::operator=(const TargetClass& other) {",
             "TargetClass::operator=",
         ),  # out-of-line operator= -- was a real bug, now fixed
+        (
+            "MyClass::operator()() const {",
+            "MyClass::operator()",
+        ),  # functor operator()
+        (
+            "MyClass::operator bool() const {",
+            "MyClass::operator bool",
+        ),  # primitive type conversion operator
+        (
+            "MyClass::operator std::string() const {",
+            "MyClass::operator std::string",
+        ),  # namespace-qualified type conversion operator
+        (
+            "MyClass::operator Foo() const {",
+            "MyClass::operator Foo",
+        ),  # custom type conversion operator
         (
             "MyClass::MyClass(int x) : field_(x), other_(0) {",
             "MyClass::MyClass",

tests/tree_sitter_accuracy_baseline_cpp.json

@@ -1,12 +1,12 @@
 {
-  "args_comparable": 1376,
-  "args_exact_match": 1202,
+  "args_comparable": 1392,
+  "args_exact_match": 1207,
   "corpus_path": "language-crucible/data/cpp",
   "extra_classes": 11,
-  "extra_functions": 55,
+  "extra_functions": 62,
   "files_scanned": 29,
   "found_classes": 138,
-  "found_functions": 1376,
+  "found_functions": 1392,
   "real_classes": 140,
   "real_functions": 1491
 }