> I present a free to use/modify C++ "tree" container that is so named
> because it can represent a general hierarchical collection of elements
> using a tree-like structure. Examples of things which use such general
> tree-like structures include file systems, XML elements in an XML
> document or the items in a GUI tree control or menu. The C++ Standard
> Library includes containers which use a binary (2-ary) search tree as
> their underlying data structure (std::set, std::multiset, std::map and
> std::multimap) but their interfaces are designed for accessing the
> elements as an ordered sequence of elements with no hierarchy and do not
> provide direct access to the underlying tree data structure; tree on the
> other hand provides an interface for accessing the (non-fixed-ary) tree
> directly allowing siblings to be iterated, parent elements to be
> determined etc.
Standard containers usually need a comparison function or a functor to sort the contained elements. The type of the functor is a template parameter and is std::less<T> by default, which only requires a suitable operator<() to be defined for T, where T is the type of the elements.
With regard to your tree, not only it does not specify any template parameter as "comparator", but, if I am not wrong, somewhere it internally uses operator==() to compare the elements, which turns out to be a superfluous (and maybe erroneous from the std point of view) requirement for T, since given two elements e1, e2, then instead of (e1 == e2) you can (or should) write !(e1 < e2) && !(e2 < e1). There are less evident reasons why the use of operator<() has been preferred to operator==(), but is another issue.
I think you should also document what kind of guarantees the tree iterators give to the user after the tree has been modified with one of its operations.
