Corrected and optimized DFS function and optimized BFS

[Shu-AFK]

• Fixed the DFS function(for weighted graphs): it was previously performing a BFS instead of a DFS due to a mistake in the traversal algorithm. Corrected the issue to ensure the function now performs a proper Depth-first Search.
• Optimized the BFS and DFS functions to improve its memory usage and or its performance:
  • by checking whether a node has been visited before it's added to the queue, rather than after. The previous implementation pushed every node into the queue, then checked for visited status upon popping the node, which can result in unnecessary queue operations. By checking for visited status beforehand, unnecessary push operations into the queue are avoided, leading to performance optimization.
  • changed std::unordered_map<T, bool> to std::unordered_set, due to the bools not actually being used, optimizing memory usage

[spirosmaggioros]

you are correct about the bfs, i was probably tired and missed it when i was writting it. I don't agree with the unordered_set instead of unordered_map. It's the same thing and there's no need to change it right now, cause i have unordered_maps everywhere in they graph header, and it'll consume a lot of time to just change those things.So let me just fix the bfs myself and we'll see about the unordered_map later. I don't really think it's faster, it is just a matter of implementation.Both are equally fast.

[Shu-AFK]

The change is minor but switching to unordered_sets reduces memory usage, due to the bools not getting used anyways but it doesn't really make a big difference if the graph is not massive

[spirosmaggioros]

there is not enough documentation to check about the memory, but it makes sense that hashmaps have one more operator [] for their values. Let me worry about that, i'll change them to unordered_sets when i have the chance in this commit, i'll close it for now and re-open it when i'll make the change.Thanks for the comments.