Clipping instability #42
Description
Still finding missing pieces. First attempt at a fix was ensuring pieces are in the correct winding order at various stages in the algorithm. Beyond that, more attention is needed to get to the bottom of this.
EDIT: The best way to go about this would probably be to create visual examples of degeneracies. I did this the first time around when I was building the whole darn thing, but I've since deleted them now that the code is basically at version 1.0 or something like that.
Test
- Line intersection.
- Polygon intersection with a line.
- Triangle to rectangle.
- Rectangle to rectangle (changing the dimensions when a rectangle is too wide, by halving).
- Stacking rectangles to form a square.
- Sutherland Hodgman clipping of subject and source polygons.
- SH clipping of collections of polygons.
- Polygon orientation, winding order.