Package: emacs;
Reported by: James Cherti <contact <at> jamescherti.com>
Date: Thu, 21 Aug 2025 18:10:02 UTC
Severity: normal
View this message in rfc822 format
From: James Cherti <contact <at> jamescherti.com> To: Rahguzar <rahguzar <at> mailbox.org> Cc: 79286 <at> debbugs.gnu.org Subject: bug#79286: outline-mode: attempting to open folded sections does not work as expected inside a deeply nested header Date: Thu, 21 Aug 2025 15:09:19 -0400
On 2025-08-21 14:34, Rahguzar wrote:
> Hi James,
Hi Rahguzar,
> James Cherti <contact <at> jamescherti.com> writes:
>
>> When using outline-mode or outline-minor-mode, attempting to
>> open folded sections (outline-show-entry) does not work as
>> expected when the cursor is positioned inside a deeply
>> nested header.
>>
>> Steps to Reproduce:
>> -------------------
>> 1. Open an outline-mode buffer with multiple
>> nested headers, for example:
>>
>> Level 1
>> ** Level 2
>> *** Level 3
>> **** Level 4: A
>> Some content here
>> Some content here
>> Some content here
>> Some content here
>> **** Level 4: B
>> Some content here
>> Some content here
>> Some content here
>> Some content here
>> **** Level 4: C
>> Some content here
>> Some content here< PUT THE CURSOR HERE
>> Some content here
>> Some content here
>> **** Level 4: D
>> Some content here
>> Some content here
>> Some content here
>> Some content here
>> Some content here
>> **** Level 4: E
>> Some content here
>> Some content here
>> Some content here
>>
>> 2. Put the cursor before `< PUT THE CURSOR HERE`
>>
>> 3. Collapse all folds with: (hide-sublevels 1)
>>
>> 6. Show entry: (outline-show-entry)
>>
>> The folds fail to expand fully in nested hierarchies, leaving all
>> content hidden except for `level 4: C`.
>>
>> Screenshot:
>> https://github.com/user-attachments/assets/2ea6c3bd-0284-40a7-90e5-1619bdf3acc5
>>
>> Expected Behavior:
>> ------------------
>> The expected behavior is that only the heading at point (and
>> its associated content) should be expanded, while all
>> ancestor headings remain collapsed. (Similar to other editors.)
>>
>> Actual Behavior:
>> ----------------
>> Only the heading at point expands, while other headings
>> remain invisible, which is confusing. In some cases, these
>> hidden headings cannot be revealed until the parent heading
>> is first collapsed and then expanded again.
>
> I have encountered this with isearch and was able to fix this
> with in my init file.
>
> ```emacs-lisp
> (defun +outline-reveal (pos)
> (let ((heading-pos nil))
> (while (not (eq heading-pos (progn (outline-back-to-heading) (point))))
> (setq heading-pos (point))
> (outline-show-children)
> (goto-char pos))
> (outline-show-entry)
> (goto-char pos)))
>
> (defun +outline-invsible-open (&optional _)
> (when (and outline-minor-mode
> (outline-invisible-p))
> (+outline-reveal (point))))
>
> (setq outline-isearch-open-invisible-function #'+outline-invsible-open)
> ```
> With an interactive spec the function +outline-reveal can be used as an
> alternative to outline-show-entry without this problem but I can't
> imagine a scenario where that might be needed.
>
Unfortunately, neither `+outline-reveal` nor `+outline-invisible-open`
resolves this issue.
I wrote the following Elisp code to fix this issue:
--8<---------------cut here---------------start------------->8---
;; Author James Cherti
;; GitHub: https://github.com/jamescherti
(defun my-outline-show-entry ()
"Show the body directly following this heading.
Show the heading too, if it is currently invisible."
(interactive)
;; Show previous headers
(unless (derived-mode-p 'outline-indent-minor-mode)
(save-excursion
(let ((func-invisible-p
(cond
((and (derived-mode-p 'org-mode)
(fboundp 'org-invisible-p))
'org-invisible-p)
((and (or (derived-mode-p 'outline-mode)
(derived-mode-p 'outline-minor-mode))
(fboundp 'outline-invisible-p))
'outline-invisible-p)
(t
'invisible-p)))
(initial-point (point))
(visual-point nil)
(visual-init-point nil))
;; Unfolding loop
(while (and
;; Folded?
(save-excursion
(end-of-line)
(funcall func-invisible-p (point)))
;; When the real point differs from the visual point,
it indicates
;; that the cursor is still located within a hidden header.
;;
;; This prevents infinite loops if the unfolding stops
changing
;; the cursor position.
(or
(not visual-init-point)
(not visual-point)
(/= visual-init-point visual-point)))
(save-excursion
(outline-back-to-heading)
(beginning-of-line)
(setq visual-point (point))
(with-no-warnings
(show-children))
;; Go back to point
(goto-char initial-point)
(outline-back-to-heading)
(beginning-of-line)
(setq visual-init-point (point))))
;; Previous version of `outline-show-entry'
(save-excursion
(outline-back-to-heading t)
(outline-flag-region (1- (point))
(progn
(outline-next-preface)
(if (= 1 (- (point-max) (point)))
(point-max)
(point)))
nil))))))
(advice-add 'outline-show-entry :override #'my-outline-show-entry)
--8<---------------cut here---------------start------------->8---
This could likely be simplified using built-in functions.
I welcome any suggestions for improvement.
>> Environment:
>> ------------
>> * Emacs version: 30.2 and master branch
>> (ade6608e2587452c8ea565ce3057879379ebd0b5)
>>
>> --
>> James Cherti
>> GitHub: https://github.com/jamescherti
>> Website: https://www.jamescherti.com/
>
> Rahguzar
--
James Cherti
GitHub: https://github.com/jamescherti
Website: https://www.jamescherti.com/
GNU bug tracking system
Copyright (C) 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson.