Re: Fixes slope errors from incorrect X extents in Beam::print. (issue 5293060)

[mtsolo]

Reviewers: Keith, mike_apollinemike.com,

Message:
All problems fixed and ready for a review.

A few things:

1) please ignore the commented out printf statements.  i use them for
debugging, and i'll delete them before pushing.
2) the regtest may be a little excessive :)  unless people think this is
a good idea, is there a way to shorten it and still test it out in
situations that approximate real music?
3) there are a few regtests that change - all are intentional, and
represent either better handling of kneed beaming or quanted broken
beams (quanting in broken beaming used to be broken because the quant
point was taken from normal stems and not from the beam hangover - this
is fixed).
4) there's now a few warning messages in regtests because of a warning i
added for cases where lilypond tries to estimate the slope of a beam
with 1 stem, encouraging the user to use a different broken-beam-style.
if peters-prolongation becomes the default, this message will disappear
in all but the most absurd cases (beams with one stem, for example).

Description:
Fixes slope errors from incorrect X extents in Beam::print.

Please review this at http://codereview.appspot.com/5293060/

Affected files:
  A input/regression/beam-broken-scriabin-individual.ly
  A input/regression/beam-broken-scriabin-peters-prolongation.ly
  A input/regression/beam-broken-scriabin-strict-prolongation.ly
  D input/regression/beam-consistent-broken-slope.ly
  A input/regression/scriabin-defs.ly
  M lily/beam-quanting.cc
  M lily/beam.cc
  M lily/include/beam-scoring-problem.hh
  M lily/include/beam.hh
  M lily/spanner.cc
  M scm/define-grob-properties.scm
  M scm/define-grobs.scm