bug#50139: 27.2; An issue with outline-move-subtree-down

bug#50139: 27.2; An issue with outline-move-subtree-down

[Alex]

[-- Attachment #1: Type: text/plain, Size: 1747 bytes --]


Consider a buffer in |outline-mode| having the following text (it is 
important that the buffer ends in a newline):

* first heading
the body of the first heading
* second heading
the body of the second heading


Now position point on the second heading and invoke |M-x 
outline-hide-subtree|, after which go to the first heading and invoke 
|M-x outline-move-subtree-down|. On my computer, this resulted in the 
first heading appearing to be on the same visual line as the second.

I've figured out why this happens. It is because 
|outline-end-of-subtree| works a bit differently depending on whether 
the subtree is the last one in the buffer. In this case, point will be 
positioned at the end of the buffer, but when the subtree is not the 
last one, it will be positioned on the last position of the last line of 
the subtree, before the newline which separates it from the next 
heading. I think this implementation is more consistent, as it works the 
same in both cases:


|(defun outline-end-of-subtree ()
   "Move to the end of the current subtree.
More precisely, move to the last position of the last line of the 
subtree: the
position before the newline character of that line."
   (outline-back-to-heading)
   (let ((level (funcall outline-level)))
     ;; this loop stops at the end of the buffer or on the first heading 
whose
     ;; level is not less than that of the current subtree
     (while (and (not (eobp))
                 (> (progn (outline-next-heading)
                           (funcall outline-level))
                    level)))
     (if (eobp)
         (if (= (char-before) ?\n) (backward-char))
       (backward-char)))) |


[-- Attachment #2: Type: text/html, Size: 2224 bytes --]

bug#50139: 27.2; An issue with outline-move-subtree-down

[Lars Ingebrigtsen]

[-- Attachment #1: Type: text/plain, Size: 639 bytes --]

Alex <alex.giorev@gmail.com> writes:

> Consider a buffer in outline-mode having the following text (it is important
> that the buffer ends in a newline):
> * first heading
> the body of the first heading
> * second heading
> the body of the second heading 
>
> Now position point on the second heading and invoke M-x
> outline-hide-subtree, after which go to the first heading and invoke M-x
> outline-move-subtree-down. On my computer, this resulted in the first heading
> appearing to be on the same visual line as the second. 

I'm unable to reproduce this in Emacs 27.2.  After following this
recipe, I get a buffer that looks like:


[-- Attachment #2: Type: image/png, Size: 13491 bytes --]

[-- Attachment #3: Type: text/plain, Size: 210 bytes --]


Which seems correct to me.

Is this a complete recipe for reproducing the bug, starting from "emacs -Q"?

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

bug#50139: 27.2; An issue with outline-move-subtree-down

[Stephen Berman]

On Sat, 21 Aug 2021 14:59:59 +0200 Lars Ingebrigtsen <larsi@gnus.org> wrote:

> Alex <alex.giorev@gmail.com> writes:
>
>> Consider a buffer in outline-mode having the following text (it is important
>> that the buffer ends in a newline):
>> * first heading
>> the body of the first heading
>> * second heading
>> the body of the second heading
>>
>> Now position point on the second heading and invoke M-x
>> outline-hide-subtree, after which go to the first heading and invoke M-x
>> outline-move-subtree-down. On my computer, this resulted in the first heading
>> appearing to be on the same visual line as the second.
>
> I'm unable to reproduce this in Emacs 27.2.  After following this
> recipe, I get a buffer that looks like:
[...]
> Which seems correct to me.
>
> Is this a complete recipe for reproducing the bug, starting from "emacs -Q"?

Your image looks like you hid the subtree of the first heading, not the
second as in the recipe.  Following the recipe exactly, I can reproduce
the problem in Emacs 27 and 28.

Steve Berman

bug#50139: 27.2; An issue with outline-move-subtree-down

[Lars Ingebrigtsen]

Stephen Berman <stephen.berman@gmx.net> writes:

> Your image looks like you hid the subtree of the first heading, not the
> second as in the recipe.  Following the recipe exactly, I can reproduce
> the problem in Emacs 27 and 28.

Oh, right; misread that.  Yes, when I follow the recipe correctly, I can
reproduce the problem here, too.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no