Re: Beam_collision_engraver sometimes fails on demisemiquavers or shorter notes

[Mike Solomon]

On May 4, 2011, at 9:58 AM, Bertrand Bordage wrote:

> This isn't a regression. I suggest a medium or high prority.
> Here is an example :
> 
> \version "2.15"
> <<
>   { a'32[ a' a' a'] }
>   \\
>   { c''[ d'' d'' d''] }
> >>
> 
> Regards,
> Bertrand

The issue would need to be fixed in beam-quanting.cc .  
beam-collision-engraver.cc registers the covered grobs correctly, and beam.cc 
doesn't treat these grobs (I think), as they are too close to to the beam's 
placement to be scoped by shift_region_to_valid.  There'd need to be something 
in beam-quanting.cc (in Beam_scoring_problem::init_collisions or 
Beam_scoring_problem::score_collisions) that registers these correctly.  I 
don't have time to get on this in the near future, but if you feel like 
tackling it, go for it!

~Mike