[frogs] Stencil combination [was: Re: How can I obtain articulation informations?]

[Marc Hohl <marc@xxxxxxxxxx>]

Neil Puttock schrieb:
> [...]
> OK, but surely you need to work with tie::handle-tab-note-head?  IIRC,
> the parentheses on the tied notes come via a stencil override, not
> from ParenthesesItem.
>   
You are right. I included my "harmonic detector" now in 
parenthesize-tab-note-head;
this seems the appropriate place to me, because this function is called 
several times
(for ties and repeat ties), so it would be convenient if this function 
handles the
harmonics properly.

I have now

(define-public (parenthesize-tab-note-head grob)

 ;; helper function
 (define (is-harmonic grob)
   (let* ((articulations (ly:event-property (event-cause grob) 
'articulations))
          (harmonic-found #f))

     (for-each (lambda (art)
                 (if (eq? 'harmonic-event (ly:event-property art 'class))
                     (set! harmonic-found #t)))
               articulations)
     harmonic-found))

 (let* ((font (ly:grob-default-font grob))
        (open (stencil-whiteout
               (ly:font-get-glyph font "accidentals.leftparen")))
        (close (stencil-whiteout
                (ly:font-get-glyph font "accidentals.rightparen")))
        (me (ly:text-interface::print grob))
        (padding (if (is-harmonic grob) 10 0)))

   (ly:stencil-combine-at-edge
    (ly:stencil-combine-at-edge me X LEFT open padding) X RIGHT close 
padding)))

The value of 10 for the padding is ridicously, but so I can spot the 
effect at first glance.
Now it seems that the padding itself has a whiteout effect, because the 
string lines
are erased and sometimes the harmonic brackets, too (I think the latter 
is due to the
internal order in which the grobs are placed) - see the attached pictures.

Why does the whitespace occur? Is there another possibility to combine 
the stencils appropriately
so that they "leave space" for the harmonic brackets?


> For the articulations check, try looking at note-finger in
> scm/translation-functions.scm.  Something similar to this, using
> for-each instead of map, should work.
>
>   
Yes, works great - thanks for the tip, I was sure that I saw some 
functions handling with
the articulation informations before while digging through the code some 
time ago,
but I forgot where.

Marc
> Cheers,
> Neil
>
>