Add version guardrail.

src/python.cpp

@@ -3485,7 +3485,18 @@ PyObject* r_to_py_pandas_nullable_series (const RObject& column, const bool conv
     Rcpp::stop("R type not handled. Please supply one of int, double, logical");
   }
 
-  if (!constructor) throw PythonException(py_fetch_error());
+  if (!constructor) {
+    // if the constructor is not available it means that the user doesn't have
+    // the minimum pandas version.
+    // we show a warning and force the numpy construction.
+    Rcpp::warning(
+      "Nullable data types require pandas version >= 1.5.0."
+      "Forcing numpy cast. Use `options(reticulate.pandas_force_numpy = TRUE)`"
+      "to disable this warning."
+    );
+
+    return r_to_py_numpy(column, convert);
+  }
 
   // tuples own the objects - thus we don't leak the value and mask
   PyObjectPtr args(PyTuple_New(3));