bug#50630: [PATCH] Add tests for insert-directory

[John Cummings]

Eli Zaretskii <eliz@gnu.org> wrote:

> > > With your recipe, indeed the free space is not always correct. But
> > > why is that use case interesting? insert-directory is not a normally
> > > a user-visible function.
> >
> > If being a user-visible function is necessary to consider it interesting,
> > then I reported this as a bug in the verbose `list-directory' output,
> > which is documented in the manual.
>
> Like I said before: let's pick this up when your test is installed,
> and we can run it and debug it if needed.

I'd appreciate it if you continued this discussion until we can agree there
is actually a bug, or at least something that could be improved.  I don't
think we need the tests to be merged to be able to do that. In fact, I'd say
it's the other way around: if there's a chance that a maintainer might tell
me there is no bug, or that whatever I found is not worth changing, I would
decide not to invest any more time into it.

So are we in agreement about these 3 things below?

1. insert-directory in files.el used to be able to show the incorrect free
   space for its FILE arg. The same function in ls-lisp.el and tramp-sh.el
   continues to be able to show the incorrect free space today.

2. The recipes/instructions I sent will be able to confirm this

3. This should be changed because insert-directory should not be able to
   insert free space info that is not correct for its FILE arg.