scope_guard VS refl-cpp

E.g. see this implementation: https://ricab.github.io/scope_guard/

But also I should make the point that you generally don't need to write defer-like statements at all when using RAII because it's implemented in the actual type. So it's only more code overall for types that are used only once or maybe twice.

Meanwhile I will check and use https://github.com/ricab/scope_guard

> low level details leak into your high level description of code, making the two coupled. You can’t make them invisible.

It's your job to make it not to leak. You have to write Modern C++ wrappers around C libs.

Similarity, The same can be said for Java. You can do low level in Java.

C++ is not C. C++ has backward compatibility with C.

Look at Boost folks, they wrote a Modern C++ wrapper around a C HTTP parser.

> And as I said, I’m familiar with RAII, it’s really great when the given object is scope-based, but can’t do anything otherwise.

Nothing is impossible.

You can use Scope Exit Guard with QT Widget.

https://github.com/ricab/scope_guard

> And if the new subclass has some non-standard object life cycle you HAVE to handle that case somewhere else, modifying another aspect of the code. It is not invisible, unless you want leaking code/memory corruption.

Again, Scope Exit Guards solve your problem!

You'd be surprised how much friction you'd have for C++ reflection. First, since it's a custom build step, you can do a mix of custom code gen and C++ constexpr/consteval for static reflection. Here's a header-only implementation for adding compile time reflection purely within the language [1]. And v8 already does dynamic code gen as part of its build process (to generate the snapshot to speedup instantiation of the isolate). Dynamic reflection is a must since JS is a dynamic language with reflection support.

Now of course, I don't know the specific details of reflection needed for the abstractions you reference and clearly V8 is still doing some amount of manual IR generation, so it's possible it would be a substantial investment to actually retrofit those techniques into v8. One would have to do a careful analysis of historical security exploits & specific techniques and their ability to prevent to figure out if it's worth adding those abstractions (especially since there is a potential performance tradeoff as you mention). As I said, I think there's insufficient research in this area to establish a compelling body of best practices (not to take away from the contributions of the GraalJS team to this space).

[1] https://github.com/veselink1/refl-cpp

I’ve been using https://github.com/veselink1/refl-cpp for my recent project where i needed reflections, especially for serialization. Cant wait for c++48 and reflections in standard