Re: [O] [patch, koma-letter] Change of subject behavior

[Nicolas Goaziou]

Rasmus <address@hidden> writes:

> Nicolas Goaziou <address@hidden> writes:

>> However, I don't think :opening setting belongs to
>> `org-koma-letter-headline', as it means it is potentially set each time
>> a headline is being processed.
>
> The property is at most set once cf. the second clause.

By "potentially set", I meant the (trivial) check would be done
nonetheless, which is a waste.

>> Another (smallish) advantage is that the following case would be
>> allowed:
>>
>>   ***** Dear myself,
>>   * Part 1
>
> Is the advantage that it won't fold?

No, since

  * Dear myself,
  * Part 1

won't fold either, but in "****** Dear myself," stands out and doesn't
look like a standard container.

> I updated the patch.  Perhaps the loop determination shod be moved to a
> separate function to keep the template clean.

Yes, it needs to be factored out (maybe
`org-koma-letter--special-headline-p') since it is also used in
`org-koma-letter-headline'. What about

  (defun org-koma-letter--special-headline-p (h info)
    (let ((special-tags (plist-get info :special-tags)))
      (org-some (lambda (tag) (assoc-string tag special-tags))
                (org-export-get-tags h info))))

> I thinking this cause we (or at least I) sometimes point to
> ox-koma-letter as an example of a derived backend. WDTY on that?

What do you mean?

> +    (:with-headline-opening nil nil 
> org-koma-letter-headline-is-opening-maybe)

I don't think this line is required.


Regards,