feat(reflect-cpp): Switched from glaze -> reflect cpp

A bug was discovered in glaze which prevented valid toml output. We have
switched to toml++ and reflect-cpp. The interface has remained the same
so this should not break any code

build-config/reflect-cpp/include/rfl/internal/wrap_in_fields.hpp

@@ -0,0 +1,47 @@
+#ifndef RFL_INTERNAL_WRAP_IN_FIELDS_HPP_
+#define RFL_INTERNAL_WRAP_IN_FIELDS_HPP_
+
+#include <type_traits>
+
+#include "../Field.hpp"
+#include "../Tuple.hpp"
+#include "flattened_ptr_tuple_t.hpp"
+#include "is_flatten_field.hpp"
+#include "lit_name.hpp"
+
+namespace rfl {
+namespace internal {
+
+// TODO: Non-recursive implementation
+template <class FieldNames, int j = 0, class... Fields>
+auto wrap_in_fields(auto&& _tuple, Fields&&... _fields) {
+  constexpr auto size =
+      rfl::tuple_size_v<std::remove_cvref_t<decltype(_tuple)>>;
+  constexpr auto i = sizeof...(_fields);
+  if constexpr (i == size) {
+    return rfl::make_tuple(std::move(_fields)...);
+  } else {
+    auto value = std::move(std::get<i>(_tuple));
+    using Type = std::remove_cvref_t<std::remove_pointer_t<decltype(value)>>;
+    if constexpr (is_flatten_field_v<Type>) {
+      // The problem here is that the FieldNames are already flattened, but this
+      // is not, so we need to determine how many field names to skip.
+      constexpr auto n_skip = rfl::tuple_size_v<
+          std::remove_cvref_t<flattened_ptr_tuple_t<typename Type::Type>>>;
+      return wrap_in_fields<FieldNames, j + n_skip>(
+          std::move(_tuple), std::move(_fields)..., std::move(value));
+    } else {
+      const auto name_literal = FieldNames::template name_of<j>();
+      auto new_field = rfl::make_field<
+          lit_name_v<std::remove_cvref_t<decltype(name_literal)>>>(
+          std::move(value));
+      return wrap_in_fields<FieldNames, j + 1>(
+          std::move(_tuple), std::move(_fields)..., std::move(new_field));
+    }
+  }
+}
+
+}  // namespace internal
+}  // namespace rfl
+
+#endif