Re: Find Voice/etc. context from Staff/etc. context

[Urs Liska]

Am 12.07.2018 um 00:13 schrieb David Kastrup:
> Urs Liska <address@hidden> writes:
>
> > Am 11.07.2018 um 10:33 schrieb David Kastrup:
> > > Urs Liska <address@hidden> writes:
> > >
> > > > One more understanding question regarding contexts and engravers.
> > > >
> > > > If I \consist an engraver in a bottom context like Voice I can travel
> > > > up the chain of parent contexts, retrieving their name and id
> > > > properties. But if I \consist it in Staff level contexts I don't see
> > > > if/how I can get to the bottom-level context a grob is possibly in.
> > > To be in any meaningful way "in a bottom-level context", a grob needs to
> > > be made by an engraver at bottom level.  In that case you can call
> > > ly:translator-context on the third argument of the grob acknowledger
> > > which is the _source_ engraver rather than the current engraver.
> > OK, I have tested how this works, and I see how I can retrieve the
> > lowest context of any given grob. But if I'm not mistaken this is only
> > possible in acknowledgers, isn't it?
> Sure.
>
> > But I have to get to that context in process-acknowledged or
> > stop-translation-timestep.
> So store it until then.
>
> > By now I would know how to link that information to the grob in
> > acknowledgers and read it later - but that would mean I have to look
> > up the source-engraver context for *all* grobs and not only for those
> > (the annotated grobs) where I actually need them.
> So?  How many millions per time step do you expect?

per time step and context I only expect a few dozens. But I was 
reluctant to do that (store but mainly retrieve the context) for *every* 
grob in the score when I only have very few actual annotations.

But it appears to work that way, and I assume this is just the typical 
tendency of trying to optimize on the wrong end ;-)