Re: [Denemo-devel] scheme_change_chord_notes

[Jeremiah Benham]

On Jul 14, 2009, at 10:53 AM, Richard Shann <address@hidden>  
wrote:

> On Tue, 2009-07-14 at 09:37 -0500, Jeremiah Benham wrote:
> > On Tue, Jul 14, 2009 at 01:06:25PM +0100, Richard Shann wrote:
> > > On Tue, 2009-07-14 at 13:02 +0100, Richard Shann wrote:
> > > > (We really *should* have a goto-note-position command for the  
> > > > cursor
> > > > if
> > > > we don't already).
> > > In fact it is madness not to, that is what happens when you click  
> > > with
> > > the mouse, more or less (there is also the enshift). So the
> > > functionality must be there, we just need to wire a scheme_xxx  
> > > function
> > > to it.
> >
> > Ok. I will write that and remove my latest function from view.c.  
> > Would enshift receive an int argument? +1 adds sharp and -2 double  
> > flat?
> Sorry I was getting it muddled, there is no enshift, only an octave  
> spec
> needed. The cursor is in the same place regardless of whether the note
> is sharp or flat etc.
> What I have in mind is
>
> (d-CursorToNote "f,,")
>
> would move the cursor to the note position on the staff that is f,,
> and so on for other possible cursor y positions. I guess we should  
> allow
> (but ignore) accidentals, so that
>
> (d-CursorToNote "ees'")
>
> does exactly the same thing as
>
> (d-CursorToNote "e'")

I wrote a function that does this. I pushed it to git. It is currently  
giving the wrong octave because of a bug in name2mid_c_offset. My  
guess is that on line 1309 gint octave should be -1 not -2.  
scheme_change_chord_notes calls the same function and has the same  
problem. I have not tested my hypothesis yet though.

> I have looked thru the code now and I am fairly sure this function has
> not already been written. It looks like the following function (from
> pitchentry.c) is the nearest thing to the desired functionality
>
> static void enter_note_in_score (DenemoGUI *gui, notepitch * found,  
> gint
> octave) {
>  gui->si->cursor_y = gui->si->staffletter_y = found->spec.step;
>  gui->si->cursor_y += 7*octave;
>  shiftcursor(gui, found->spec.step);
>  setenshift(gui->si, found->spec.alteration);
>  displayhelper (gui);
> }
>
> from this I would guess that you need
>
> gui->si->cursor_y = gui->si->staffletter_y =  the step of the scale
> gui->si->cursor_y += 7*octave;
> gui->mode=CLASSIC_MODE
> shiftcursor(gui, the step of the scale);
> gui->mode=the mode on entry

I am not using or setting mode. I am assigning a value to cursor_y  
using name2mid_c_offset.

> roughly speaking that is. The point being that shiftcursor, despite  
> its
> name is mode sensitive, in the original demo, pre-modes, it did just
> shift the cursor, but only relative to its current position, and so  
> you
> have to set the octave first.

That is why I want to avoid using something like shiftcursor.

> It is crying out for a tidy up of
> course...
>
> I've just done some checking on the (d-PutNoteName ...) and I find  
> there
> is a problem. If you track it through it ends up at modify_note() in

Why don't we remove this code and create a macro for it that does  
something like:

(d-CursorToNote lilyname)
(d-Insert2)

This would not be for appendind notes to a chord though.
> chordops.c which states that you cannot use it on multi-note chords.
> I think you could do
>
> (d-CursorToNote notename)
> (d-RemoveNoteFromChord)
> (d-CursorToNote newnotename)
> (d-AddNoteToChord)
>
> for each notename in (d-GetNotes), with newnotename stepping thru the
> replacement notes.

A user could sting append (d-GetNotes) with a desired new note. This  
output could be given as argument to (d-ChangeChordNotes).

I have modified (d-ChangeChordNotes) to accept string input. I will  
next save directives.

> This would lose fingerings etc attached to notes. It would be possible
> to write the remove-then-add part of this as a single C-function, then
> you could detach the directives attached to the note before deleting  
> it,
> and re-attach them to the new note after creating it, viz
>
> GList *g = note->directives;
> note->directives = NULL;
> delete_note_from_chord(note);
> add_note_to_chord(newnote);
> newnote->directives = g;

Sounds good. I will implement this next.

Jeremiah
> where I have just invented names for the add/removing of notes from a
> chord but I hope the meaning is obvious.
>
> Richard
>
> > Jeremiah
> >
> > > Richard
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Denemo-devel mailing list
> > > address@hidden
> > > http://lists.gnu.org/mailman/listinfo/denemo-devel
>
>
>
> _______________________________________________
> Denemo-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/denemo-devel