Re: page breaking cleanups

[Joe Neeman]

On Sat, 2006-08-19 at 20:00 +0200, Han-Wen Nienhuys wrote:
> Joe Neeman wrote:
> > +vector<SCM>
> > +Page_breaking::systems ()
> 
> this looks like a potential memory error.  The SCM vector is allocated 
> on the heap, and hence, is not GC scanned, and may be freed prematurely. 
> Please double check

OK, this is now just SCM list.

(and the other 2 issues are fixed, too)

In addition, I've added better support for ragged-bottom and
ragged-last-bottom in both of the new page-breakers.

2006-08-23  Joe Neeman  <address@hidden>

        * input/regression/optimal-page-breaking-hstretch.ly: test for
        ragged-last-bottom also

        * lily/paper-column-engraver.cc (finalize): make the end of a score
        breakable by default. This is to balance out a change in behaviour
        of the page-turn-breaker which no longer makes the end of a score
        breakable.

        * lily/paper-book.cc (pages): set the systems_ once the pages are
        broken

        * lily/page-turn-page-breaking.cc (calc_subproblem): use the new
        Page_breaking interface.

        * lily/page-breaking.cc (class Page_breaking): make the interface
        more consistent and provide abstractions for dealing with
        Line_divisions.

        * lily/optimal-page-breaking.cc (solve): use a more straightforward
        algorithm. Use the new interface to Page_breaking.

        * lily/page-spacing.cc: better support for ragged-bottom and
        ragged-last-bottom

page_breaking_cleanup.patch
Description: Text Data