Bug in subtree narrowing

Bug in subtree narrowing

[Samuel Wales]

Hi Carsten,

org-tree-to-indirect-buffer does not include the final
newline.  This causes sorting to run two lines together.
For example, a headline can get joined to an :END: line.
This can cause a task to not be recognized.

Sorting makes a lot of changes, so some users will not
notice the bug until much later.

org-narrow-to-subtree appears similar.

A lot of user code might assume, as perhaps org-sort does,
that the final line has a newline.

The choice of solution I will leave to you :).

My preference is for the final newline to exist.  This is
because it feels normal to me as a user.  If I go to the end
of the buffer, then up, I usually expect to be on the last
line.  If I manually kill a subtree (or a paragraph) I kill
to the beginning of the next line.  m-x mark-paragraph
appears to do this.  So does m-x kill-paragraph.

However, I will not file a bug on outline.
outline-mark-subtree (which org users use also) does not include the
final newline.
outline-end-of-subtree does (forward-char -1), which means
that it is deliberate.  There are likely to be factors that
I have not considered.

"Org-mode version 6.30trans (release_6.30d.814.gdacd)" .

Thanks.


Samuel


-- 
Myalgic encephalomyelitis causes death (Jason et al. 2006)
and severe suffering.  Conflicts of interest are destroying
research.  What people "know" is wrong.  Silence = death.
http://www.meactionuk.org.uk/What_Is_ME_What_Is_CFS.htm

Re: Bug in subtree narrowing

[Carsten Dominik]

On Sep 18, 2009, at 12:09 AM, Samuel Wales wrote:

> Hi Carsten,
>
> org-tree-to-indirect-buffer does not include the final
> newline.  This causes sorting to run two lines together.
> For example, a headline can get joined to an :END: line.
> This can cause a task to not be recognized.

Fixed for this case.

>
> Sorting makes a lot of changes, so some users will not
> notice the bug until much later.

I do not understand what this has to do with narrowing to a subtree?

>
> org-narrow-to-subtree appears similar.

Fixed for this case as well.

>
> A lot of user code might assume, as perhaps org-sort does,
> that the final line has a newline.
>
> The choice of solution I will leave to you :).
>
> My preference is for the final newline to exist.

Are you suggesting that I should add the final newline?
Under what circumstances?
Are you aware of the variable `require-final-newline'?  It might be  
for you?
I have turned it on.

- Carsten


> This is
> because it feels normal to me as a user.  If I go to the end
> of the buffer, then up, I usually expect to be on the last
> line.  If I manually kill a subtree (or a paragraph) I kill
> to the beginning of the next line.  m-x mark-paragraph
> appears to do this.  So does m-x kill-paragraph.
>
> However, I will not file a bug on outline.
> outline-mark-subtree (which org users use also) does not include the
> final newline.
> outline-end-of-subtree does (forward-char -1), which means
> that it is deliberate.  There are likely to be factors that
> I have not considered.
>
> "Org-mode version 6.30trans (release_6.30d.814.gdacd)" .
>
> Thanks.
>
>
> Samuel
>
>
> -- 
> Myalgic encephalomyelitis causes death (Jason et al. 2006)
> and severe suffering.  Conflicts of interest are destroying
> research.  What people "know" is wrong.  Silence = death.
> http://www.meactionuk.org.uk/What_Is_ME_What_Is_CFS.htm
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Remember: use `Reply All' to send replies to the list.
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode

Re: Bug in subtree narrowing

[Samuel Wales]

On Thu, Sep 17, 2009 at 23:22, Carsten Dominik
<carsten.dominik@gmail.com> wrote:
>> Sorting makes a lot of changes, so some users will not
>> notice the bug until much later.
>
> I do not understand what this has to do with narrowing to a subtree?

Sorting obscures the bug.

> Are you suggesting that I should add the final newline?

No.  Move the end forward by one char.

> Under what circumstances?

Perhaps when it won't break anything.

> Are you aware of the variable `require-final-newline'?  It might be for you?
> I have turned it on.

Isn't that for files?

I am referring to narrowed regions.  Point at end should be after the
newline, to match my intuition.  But this conflicts with outline's
intuition.

Adding a newline is not necessary.

Thanks for fixing the bug.