[PPL-devel] [GIT] ppl/ppl(floating_point): Little correction in two_variables_affine_image method.

[Roberto Amadini]

Module: ppl/ppl
Branch: floating_point
Commit: 1508fe8fbb7be516e7cebabb60d8fe30c1399407
URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=1508fe8fbb7be516e7cebabb60d8fe30c1399407

Author: Roberto Amadini <r.amadini at virgilio.it>
Date:   Tue Sep 22 10:55:49 2009 +0200

Little correction in two_variables_affine_image method.

---

 src/BD_Shape.templates.hh |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/src/BD_Shape.templates.hh b/src/BD_Shape.templates.hh
index 08b1c5d..8b5c49d 100644
--- a/src/BD_Shape.templates.hh
+++ b/src/BD_Shape.templates.hh
@@ -4237,7 +4237,10 @@ void BD_Shape<T>
       const Linear_Form< Interval<T, Interval_Info> >& lf,
                           const dimension_type& space_dim) {
 
-  reset_shortest_path_closed();
+  // Shortest-path closure is preserved, but not reduction.
+  if (marked_shortest_path_reduced())
+    reset_shortest_path_reduced();
+
 
   Linear_Form< Interval<T, Interval_Info> > minus_lf(lf);
   minus_lf.negate();