Hi forum! Is there a way to replace topology of adjacencies from heteroptera using native components?
Avoid wasting time. Walk the right walk. Time flies.
What happens to adjacent rectangles that don’t share the full edge?
for the indices of the initial curves thing, you can still use closest point on area centroids, and get the answer from its “i” output
for the being slow part… what is slow in your definition? Brep Join? Brep Topology looks pretty fast to me:
Region Union? No geometry posted though… 
If the question is related with the C# above … appears that an EnryLevel_V1A (for the 2 fast options) is required. A Recursive search acc a depth user var is also required.
Omg thank you guys for all the replies:) I will reply to all of you, thank you
they should be treated as full length, like in joined breps
No. Thomas he’s - obviously - right, just think the general case (partial [no full edge/edge] contact so to speak). But I confess that I have removed some sensitive very fast (real-time in fact) Methods for the general case.
I.e. check first Point3dList proximity matters AND if it fails … go to 2nd check > find if 2 lines [i.e. the poly edges] are colinear and have some sort of overlap.
BTW: the above C# is labeled as EntryLevel (meaning a bit of a crap occasionally).
Anyway see a preview of the V1A update (plus spot the depth Recursive search user var).
I hear you: but … for the conn “Graph” you are using the Polyline center point (a bit confusing in irregular cases like the one shown) instead of the so called visual center (life sucks).
Free bonus: remember your recent case with Graph islands? Well … if you add some minor (so to speak) stuff to the SearchAdjacent Method … you’ll get what you want.
