BinarySearchTreeExercises VS MGCLib

Trippy and kind of cool. Don't remember if I knew that. I did make this red black tree map last May though. The only problem is I wrote the code without the key being const, but because a major part of my tree changing methodology was using std::swap which at the time I couldn't figure out how to replicate involving a const key, I left it that way. Partially it was due to the basis coming from following a tutorial in Java I think. But since it was only affecting me at the time with the iterator, I did a hack with some lambdas to make the iterator operators work like this:

This is theDS project that I have spent 2 weeks on and have not commented almost at all on. A red black tree map, I did not have anything well defined concerning how much I was going to end up doing with it, so it has evolved into what it is now. If I do anything more with it, it will probably be just to add in the rest of the special members and maybe 2 unary operators (+ and -) to the Iterator.

Regarding the formatting, it took me a little while but I installed clang-format and went through the options, it was close between google and gnu but I went with google. Does this look better? This is my current version of this class from my repository. RedBlackTree

The ordeal I have had with it makes me feel like it's full of duct tape and zip ties TBH. I am not in the know of benchmarking but using some timer class I found by googling; the run times for insertion and deletion of elements seem comparable between my raw version and the std::set. But like I indicated earlier, it is not a template class and we are only working with ints here. What I am wondering though is if an ideal erase() method is supposed really to be anywhere as long as mine turned out to be? Because seriously I felt like a dutch boy plugging holes on this thing for hours hahaha. Methods are below and entire class is available here https://github.com/sethvan/BinarySearchTreeExercise .

I did not go with the left leaning version you attached because the paper intimidated me as soon as it mentioned 2-3 trees which I was not familiar with, however I have been slightly glancing at the repository from 2022 (which is where I based my node off of in part). Two observations I have made of that class and another piece of code https://pastebin.com/7GGFiaEc I had looked at for an AVL tree in C is that first I think it is very cool that they have so many functions that return a node value back to it's own variable like node = doSomething(node) so to speak instead of all the void methods that I have been doing (not only on my own but that this professor also showed for my present version of a red black tree here https://pastebin.com/m1aG7w4s ). So that's cool and suspect I need to try and think more in that mannner. BUT.... in neither of those versions do I see any memory being deleted or free'd nywhere. So i am confused, Why is there no destructor in this guy https://github.com/maxgoren/MGCLib/blob/main/mgc_redblack.hpp and why is nothing being deleted anywhere that I can see?