Still: Beaming

[Joerg Anders]

On Thu, 18 Mar 2004, Jan Nieuwenhuizen wrote:

> Joerg Anders writes:
> 
> > Or perhaps a pointer into the LilyPond C code ?
> 

First of all: Thank you all who answered this question.

> LilyPond uses rule based autobeaming, but it's far from perfect.  It's
> one of those things we'll have to Get Right (TM) some time.  The rules
> are listed in
> 
>    scm/auto-beam.scm
> 
> and use time signature, position in measure and shortest note in beam.

Very interesting! 

What happens if 2 rules match ? 

For instance in scm/auto-beam.scm:

        ((end * * 4 4) . ,(ly:make-moment 1 2))
                ...
        ((end 1 16 4 4) . ,(ly:make-moment 1 4))

I made an attempt: The time signature is 4/4. I placed 16 16ths
in 1st measure and and 8 8ths in 2nd measure.

The 16ths are grouped:

   |--|--|--|    |--|--|--|    |--|--|--|    |--|--|--|       
   |--|--|--|    |--|--|--|    |--|--|--|    |--|--|--|       
   |  |  |  |    |  |  |  |    |  |  |  |    |  |  |  |       
   |  |  |  |    |  |  |  |    |  |  |  |    |  |  |  |       
   /  /  /  /    /  /  /  /    /  /  /  /    /  /  /  /       

i.e. according to 2nd rule (If I comprehended all right)

The 8th are grouped:

  |---|---|---|     |---|---|---| 
  |   |   |   |     |   |   |   |
  |   |   |   |     |   |   |   |
  /   /   /   /     /   /   /   /

i.e. according to 1st rule (If I comprehended all right)
Ok here is no problem because the 2nd rule does not match.
But if I apply the 1st rule to 1st measure the result could be:

   |--|--|--|--|--|--|--|  |--|--|--|--|--|--|--|       
   |--|--|--|--|--|--|--|  |--|--|--|--|--|--|--|       
   |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |       
   |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |       
   /  /  /  /  /  /  /  /  /  /  /  /  /  /  /  /       

What's the rule over rules ?

        a.) The last matching rules is applied ?

        b.) The most precise rule is applied ?


-- 
J.Anders, Chemnitz, GERMANY (address@hidden)