On Sat. (Jun 11, 2022) at 04:21:59 AM -04, Eli Zaretskii <eliz@gnu.org> wrote:
>> From: "Wamm K. D." <jaft.r@outlook.com>
>> Date: Sat, 11 Jun 2022 01:52:13 -0500
>> 
>> Pretty much as the subject describes; this should allow for users to
>> lazily load hierarchy branches when loading everything at once would
>> otherwise be too consuming, up-front (or if dealing with an infinite
>> tree).
>
> Thanks.
>
>> * lisp/emacs-lisp/hierarchy.el (hierarchy-add-tree) (hierarchy-add-trees):
>> Add parameter 'delay-children-p'.
>
> Please use our style of commit log messages.  The above should be
>
> * lisp/emacs-lisp/hierarchy.el (hierarchy-add-tree)
> (hierarchy-add-trees): Add parameter 'delay-children-p'.

Mmm; gotcha. This hadn't been clear, to me; I thought the example given
in the CONTRIBUTE file was that way because the line was long and the
main point of its demonstration was what to do with a case of more than
one function. I've adjusted the commit message appropriately, now.

>> +(defun hierarchy--create-delayed-tree-widget (elem labelfn indent fn)
>> +  "Return a list of tree-widgets for a hierarchy created from the childrenfn.
>
> The first line of the doc string should ideally mention the
> arguments.  This one doesn't, and it also seems to misname one
> argument: it's FN, not childrenfn, right?

Yeah; I was referencing the fact that this function is called childrenfn
in a bunch of other functions but, for whatever reason, I didn't call it
that here so that's why it isn't all-capitalized (while the reference to
it, at the end, is and says FN). But that's weird; I don't know why I
went with that name when I could've just…used the same name as every
other function. I've corrected that, standardized.

> Can you add some tests of this new functionality?

I've added some; I didn't created tests for children functionality since
the delayed computation means there aren't children (until someone
activates the tree-widget and the children are computed with the
provided function) but I create some tests to check that the root and
parent elements were still functioning as expected, when using delayed
children.

> Also, I think this warrants a NEWS entry to describe the new feature.

I've added that, as well.