Re: Reconciling org-mode idiosyncrasies with Emacs core

[Kévin Le Gouguec]

Nicolas Goaziou <address@hidden> writes:

> Tests for `org-return' (named "test-org/return") are in the
> "test-org.el" file in the "testing/lisp" directory. We only need to test
> if electric-indent-mode has an effect, but only in regular cases.

Thanks for the pointer!

(I forgot to mention that AFAICT, all current tests pass with this
patch.)

>> - INTERACTIVE is what makes 'newline' run 'post-self-insert-hook' (thus
>>   triggering indentation through electric-indent-mode),
>
> OK. I thought it was necessary to call
> `electric-newline-and-maybe-indent'.

Nope; took me some time to piece everything together, but the logic (as
of 24.4) seems to be

- RET runs `newline': if electric-indent-mode is disabled, then it's a
  dumb newline, otherwise electric-indent-post-self-insert-function
  kicks in *if run interactively*;

- C-j runs `electric-newline-and-maybe-indent': it's meant to be the
  "smart newline" command when electric-indent-mode is disabled,
  otherwise it shrugs and just inserts a newline.

>> I took the liberty of using
>> this function in the "list item" case too, otherwise there's no way to
>> indent the trailing text.
>
> I'm not sure what you mean. It would be a regression if you didn't use
> the function there, too, wouldn't it?

The "list-item" case currently calls `newline-and-indent'
unconditionally, because the condition for that cond branch starts with
(and indent …).  Therefore, to make this case work with
electric-indent-mode, I had to tweak the condition; I just wanted to
bring attention to this change, since it was not as "mechanical" as for
the "at-headline" and "default" branches.

> I cannot speak for the Emacs side, but it should land in Org 9.4, not
> Org 9.3.6.
>
> It is a very visible change, one that every Org user is going to face.
> This requires a new ORG-NEWS entry. Those only appear in new minor+
> releases. Therefore, if you apply it in Emacs 27.1, the change will be
> announced nowhere.

Right.  Org master it is, then.

>> Now for C-j, in order to minimize breakage (for anyone calling
>> org-return-indent from Lisp code) and simplify disabling the new
>> behaviour (by simply turning off electric-indent-mode in Org), should we
>> bind C-j to a new function?  E.g.:
>>
>> (defun org-return-and-maybe-indent ()
>>   (interactive)
>>   (org-return (not electric-indent-mode)))
>
> I think so. Then we can mark `org-return-indent' as obsolete and suggest
> to call `org-return' instead.

Alright.


I'll try to follow-up with this additional command, tests, and
changelog/news entries in the next few days.

Thank you for the review!