Re: [Denemo-devel] gsoc work - Tue morning

[Richard Shann]

On Tue, 2010-08-10 at 09:09 +0100, Richard Shann wrote:
> Also, my reading was not close enough to see what was happening to the
> octave indicators - the steno_pitch has extra bits for them which you
> haven't yet done:
I have had a few more minutes to look at this. What you have done is to
change the rule for pitch to include the quotes. Unless there was a very
good reason don't do this, instead leave them in the rule for
steno_pitch.
I notice you have done a string-append for the value. This would be fine
if we were doing the direct creation of Denemo commands - you would have
(d-A)(d-UpOctave) as a result and could execute that when you got to
score block.

 (There is a wrinkle here BTW, because you can't actually do up octave
until you have done the duration, which would mean either
      * performing a macro engineered reversal
or
      * implementing the d-DeferredOctaveUp command that a user
        requested 
- so as to be able to type note name accidental duration in that order).


But as you are going for creating a parse tree for the input rather than
a string of commands you will want to replace that string-append with a
cons of an octave-shift object onto your tree. But that is for later. I
think it is good to keep the first target as a complete parse of a
simple musicxml2ly converted .ly file, without taking any particular
actions at all. As you say, it will be difficult to be sure it is right
until we do take actions, but if we stick very closely to the lilypond
one it will likely be ok.

Richard