PolyTreeD Precision

[hillaryan]

I have been playing around with this wonderful library and ran into something that puzzled me with the PolyTreeD:
Suppose I build a tree with the following path in the subject:

u: -67.86204528808594, -60.34636306762695
u: -67.86204528808594, 0
u: 0, 0
u: 0, -60.34636306762695

After execution, I can pull the path back out from Child(0) of the tree:

v: 0, -60.34375
v: 0, 0
v: -67.859375, 0
v: -67.859375, -60.34375

It looks to me like the path got reversed and lost precision. From reading the docs, I know that Clipper2 converts doubles to 64 bit integers and then back again, but I am surprised that the precision loss is this great. Is this expected behavior or am I likely making a mistake somewhere?

Many Thanks!

[hillaryan]

To answer my own question, my issue was that I initialized ClipperD like so ClipperD c. This sets clipper to the default precision of 2. Using ClipperD c(6) solves my problem.

[AngusJohnson]

It looks to me like the path got reversed

"Clipping operations wills always return Positive oriented solutions (unless the Clipper object's ReverseSolution property has been enabled). This means that outer polygon contours will wind anti-clockwise and inner 'hole' contours will wind clockwise (in Cartesian coordinates)."

See: http://www.angusj.com/clipper2/Docs/Overview.htm#closed_paths