Your reply is not relevant to my reply. The original poster is asking for this functionality and appears to believe it is something other than an iterator and requires some sort of special language support. However, it is completely implementable as an iterator, in a reasonably usable manner, with no additional language support. My specific code is written only to show that fact off.

Anyone who copies and pastes it is welcome to both pieces when it breaks. Others have already alluded to possible improvements that could be made, and I already have my own analysis in a grandchild reply as to why I don't think this is a terribly pressing need or necessarily even a good idea.

The reason I provide code is that it gets past the "oh, you say it's just an iterator, but I still don't believe you, since you haven't spelled it out to the n'th degree". When code is provided, belief ceases to be an issue. It is clearly something an iterator can implement, in existing languages, with existing iterator support.

Unless you're going to claim it is somehow impossible to provide this functionality in a tested manner, you're completely changing the topic in an uninteresting direction, since it is always true that functionality generally needs testing and bits of code slammed into an HN conversation just to make a particular point probably shouldn't be copied wholesale into your production code.

Sure but all pre-made, battle-tested tree datastructures you'd use in production in all languages already come with some form of iterator that you can just for-loop over, so the original articles point is still moot.

Sure, but it can be done by a library. There's no reason it needs to be built into the language.