Commit Graph

180 Commits

Author SHA1 Message Date
Marc Glisse 39ba4946b4 Split points and vectors some more. Might remerge them later...
Iterators are not objects, handle that everywhere. Still some not-lazy-enough issues I'll have to deal with.
2012-03-27 16:05:32 +00:00
Marc Glisse 2f6dbee262 Small fix to lazy_cartesian now that iterators are objects (will probably change that soon, only caused trouble).
Provide an easy way to have K::Point (shorter than K::Type<Point_tag>::type...). Could do the same for functors, maybe.
Now that I have an Object_list, it would be possible to use directly K::Point and get rid of that Type<Point_tag>::type horror, which sure looks simpler (although it is actually more complicated behind), but it has some drawbacks. In particular, in a CRTP context, typedef std::pair<Derived::Point> Segment; won't work (it worked inside Type<...>), one has to use typedef MySegment<Derived> Segment; instead (C++ can be a pain...).
2012-03-26 11:15:14 +00:00
Marc Glisse f3304589c0 More conversion to iterators as object. Still not sure about that, they end up special cased every time so they probably should have a different category. 2012-03-24 21:08:33 +00:00
Marc Glisse 98440b3bd2 hush noisy warnings with g++47 and eigen 2012-03-22 11:12:10 +00:00
Marc Glisse 212ca488ea CGAL_CFG macros 2011-12-07 11:21:00 +00:00
Marc Glisse 2e88bc0daf Move Segment to a module that can be included. Segmentd.h should disappear, the functionality should be in Wrapper/Segment_d.h. 2011-11-08 11:53:21 +00:00
Marc Glisse d90d515da4 Fix untested code.
Add more CRTP.
Work around too much CRTP...
2011-11-05 14:33:54 +00:00
Marc Glisse e12e4f0705 Reuse existing static filter 2011-10-28 17:38:26 +00:00
Marc Glisse 1d9e7b9858 try and make the converter more flexible 2011-10-25 11:57:11 +00:00
Marc Glisse 3fd4068328 Remove non-Type<tag> stuff 2011-10-17 11:34:11 +00:00
Marc Glisse 21228b30e6 Continue removing LA.
Start obfuscating^Wusing an indirection for object types.
2011-10-12 19:48:23 +00:00
Marc Glisse 43d50228f2 Side_of_oriented_sphere 2011-06-20 13:58:26 +00:00
Marc Glisse 3369e6120c rename Orientation 2011-05-30 10:08:31 +00:00
Marc Glisse 1347e068cd Fix Orientation when Point!=Vector. I don't think there is much point having a single functor for both. 2011-05-25 09:06:05 +00:00
Marc Glisse 9c655afded More stateful kernels. 2011-05-24 15:28:27 +00:00
Marc Glisse 32fecea31f Lazy_cartesian is an independent layer 2011-05-20 15:58:28 +00:00
Marc Glisse 363b57cbd9 finish refcounting wrapper 2011-05-17 20:02:04 +00:00
Marc Glisse 8aaf9c465a Wrap(Null_functor)==Null_functor 2011-05-17 17:30:33 +00:00
Marc Glisse b75920fff5 make Construct_*_cartesian_const_iterator a true functor 2011-05-17 16:40:30 +00:00
Marc Glisse 3690072964 Use a single Functor template regrouping Predicate, Construct and Compute, and a map to know the category of a functor. 2011-05-17 15:21:49 +00:00
Marc Glisse 12fe1be5bb Merge from next to get the is_iterator fix 2011-05-14 10:34:10 +00:00
Marc Glisse 7a58fd43b4 More wrapper ugliness 2011-05-13 20:23:19 +00:00
Marc Glisse a873c26099 Continue with Wrapper 2011-05-13 16:05:15 +00:00
Marc Glisse cb55b147eb d 2011-05-12 16:15:26 +00:00
Marc Glisse 5fa563096e c 2011-05-12 13:23:54 +00:00
Marc Glisse 0e0374300f b 2011-05-12 13:15:40 +00:00
Marc Glisse dc29aebfd2 ploum (but obscenities may come to fill in for empty commit messages) 2011-05-11 20:15:08 +00:00
Marc Glisse 217ed569f2 pouët 2011-05-11 15:14:26 +00:00
Marc Glisse b63f32ccdd Add possibility to filter on kernels instead of number-types. 2011-05-11 10:10:48 +00:00
Marc Glisse 1cd3febf83 A beginning. 2011-05-09 14:08:39 +00:00