I'm not really in favor of this change as making the iteration over the heap nodes non-const can have severe side effects. e.g. by changing the node values that are used for comparison, which would result in corruption of the entire datastructure.

I thought about this for a while and I think we should use a dedicated erasure method for clear and not use the iter_all method. A simple breadth-first or depth-first search would be enough for this. Then we don't have to worry about invalid pointer accesses too.

e as making the iteration over the heap nodes non-const can have severe side effects. e.g. by changing the node values that are used for

Do you think it might be worth implementing an iterator for the tree?

Do you think it might be worth implementing an iterator for the tree?

That would be cool! Currently we can get a problem with the recursion depth if the tree is too large. An iterator should solve that.

Do you think it might be worth implementing an iterator for the tree?

That would be cool! Currently we can get a problem with the recursion depth if the tree is too large. An iterator should solve that.

iterator added