I'm just re-stating this to make sure I'm getting it right: Definitions are
lazy, and the expression in front of # is lazy too, so it gets a chance to
see the definition once it's done? Am I right?
That's good to know!
On Sat, Oct 2, 2021 at 4:00 PM David Kastrup <d...@gnu.org> wrote:
> "Omid Mo'menzadeh" <omid.mnza...@gmail.com> writes:
>
> > Of course!I just attached two files that demonstrate the difference.
> >
> > On Sat, Oct 2, 2021 at 3:30 PM David Kastrup <d...@gnu.org> wrote:
> >
> >> "Omid Mo'menzadeh" <omid.mnza...@gmail.com> writes:
> >>
> >> > I didn't know about $ at all! Where in the documentation can I find
> more
> >> > about it? It was hard to search for!
> >> >
> >> > I tried it, but it doesn't seem to find my variable "vocal" defined in
> >> the
> >> > same .ily file. I tried using #(top-repl) and $(top-repl) too, in the
> #
> >> > version it knows what \vocal is, but in the $ version it doesn't.
> >>
> >> How about a minimal example exhibiting the problem?
> >>
> >> --
> >> David Kastrup
> >>
> >
> > \version "2.22.1"
> >
> > vocal = \relative c'' {
> > a4 b c d |
> > }
> >
> > $(if (not (defined? 'is-loaded-file)) #{ \score { \vocal } #})
>
> Well, that one's a timing problem. You couldn't write
>
> vocal = \relative c'' {
> a4 b c d |
> }
>
> \vocal
>
> either. The reason is that the assignment is not complete before
> LilyPond has a chance to check whether something like \addlyrics { Ah }
> follows, and so when it looks at the next thing that comes up to make
> that decision, \vocal is not yet defined. So there really needs to be
> anything after that assignment that does not immediately reference
> \vocal .
>
> With #, LilyPond does not allow different syntactic functionality, so
> when it sees # it does not bother evaluating it until after it has
> completed the assignment, being sure that it won't have to add something
> like \addlyrics ... to the expression to be assigned.
>
> This is one case where minimal examples tend to fail more often than
> actual use of assigned variables in practice, because usually when you
> employ variables seriously, you don't reference them immediately right
> afterwards.
>
> --
> David Kastrup
>
