First step towards fixing the cross staff problem

First step towards fixing cross staff problem

[MikeSol, 2011-04-24 12:14:16]

This patch seems to get everything right in the beam-collision engraver.

If the typesetting in beam.cc can accommodate it, I think we'll have hit beam
collision nirvana.

In:


<<
  \new Staff = up \relative c' {
    c8 c c c
  }
  \new Staff = down \relative c' {
    s8 c c c
    \change Staff = up
  }
>>

<<
  \new Staff = up \relative c' {
    c8 c c c
  }
  \new Staff = down \relative c' {
    s8 c c
    \change Staff = up
    c8
  }
>>

The first example now clears, whereas the last one still has problems.  These
problems stem from the first beam trying to avoid the second beam (not stems,
noteheads, etc.).

If anyone can figure out a way to make this work in beam.cc, then it seems like
a good solution, as it genuinely handles cross-staff beams.

[hanwenn, 2011-04-24 13:41:49]

http://codereview.appspot.com/4430063/diff/1/lily/beam-collision-engraver.cc
File lily/beam-collision-engraver.cc (right):

http://codereview.appspot.com/4430063/diff/1/lily/beam-collision-engraver.cc#...
lily/beam-collision-engraver.cc:95: // First, find the staves to which any
covered grob may belong
this is weird; why dont you extract the list of staves directly from stems ?

http://codereview.appspot.com/4430063/diff/1/lily/beam-collision-engraver.cc#...
lily/beam-collision-engraver.cc:102: staff_symbols.insert
(Staff_symbol_referencer::get_staff_symbol (covered_grob));
This is unidiomatic: it's better to use the originating staff context for this,
as that is translation information rather than backend info . This
implementation will stop working if somebody removes Staff_symbol_engraver.

maybe add a TODO about this.

http://codereview.appspot.com/4430063/diff/1/lily/beam-collision-engraver.cc#...
lily/beam-collision-engraver.cc:119: for (it = staff_symbols.begin (); it !=
staff_symbols.end (); it++)
why a loop?

you can do 

 staff_symbols.find(Staff_referencer::get_staff_sym(covered))