Re: Help for \beamExceptions

[David Kastrup]

Richard Shann <address@hidden> writes:

> On Tue, 2015-09-01 at 21:04 +0200, David Kastrup wrote:
>> Richard Shann <address@hidden> writes:
>> 
>> > I've been trying to use the music function \beamExceptions
>> 
>> "scheme function"
>
> Is that to say that this should be called a scheme function not a music
> function? I found it documented at
>
> http://www.lilypond.org/doc/v2.19/Documentation/notation/available-music-functions
>
> as Available music functions ...

Well.  Most of them are, but this one has a documented different return
type.  Should we call the superset of them LilyPond functions, or rather
bite the bullet and create multiple categories?

>> > which is in version 2.19.25, I can't find anything much to document it
>> > though.  For example, trying this
>> >
>> > \version "2.19.25"
>> >
>> > MvmntIVoiceI = {
>> >          e'8[ f'8 g'] a'8\noBeam b'8 a'[ g'8 f'] |
>> >          e'8 f' g' a' b' a' g' f' 
>> > }
>> >
>> > \score { 
>> >         \new Staff {\time 4/4 \MvmntIVoiceI } 
>> >         \layout {
>> >            \set beamExceptions = \beamExceptions { \time 4/4  e' 8[ f' 8
>> > g' ] a' 8\noBeam b' 8 a' [ g' 8 f' ] }
>> >         }
>> >       }
>> >
>> > the second bar is not beamed like the first, despite the music argument
>> > to \beamExceptions being a copy of the first bar.
>> 
>> \beamExceptions only takes explicit beams.  \noBeam is useless here.
>> 
>> > Is the next stop some source file for finding out what can be done, or
>> > am I missing some documentation?
>> 
>> It's just notes and [ and ].
>
> hmm, well I have tried \tuplet (I think it was, or the older \times
> perhaps) and it seemed to understand that - I haven't yet tried \grace,
> it would be nice if that could be autobeamed, though perhaps that is not
> possible.

It isn't.

> And there is a comment somewhere about using | to separate something,
> I guess bars with the same \time signature (or, indeed different
> ones).

No, it's always for exactly one time signature, but at the time it is
called, LilyPond has no idea just how long a bar is and can't count it
off itself.  So you need to separate rule patterns by |.

> I did have it barfing on empty measures (s1*4/4 perhaps) where the
> underlying scheme errored without any pleasant LilyPond error message
> resulting, I'm not sure if this is something that should be pursued...

Possibly, but it does not exactly make much sense to beam over that.

-- 
David Kastrup