New incorrect case about cavaliercontours

[jbuckmccready]

This is fixed in the Rust code, try putting those polylines in to the interactive demo here: https://jbuckmccready.github.io/cavalier_contours_web_demo_page/#/pline_boolean

I recommend using the C FFI from the Rust code.

[timoria21]

Can you help me to find the changeset that resolves this issue? I'm currently unable to switch to Rust code.

Thank you.

[jbuckmccready]

I have made a lot of fixes so I really don't know which would apply, best bet if you must use the C++ code is to follow the code in both projects and track down what has changed (that is what I would do - or re-implement the C++ code according to the Rust code). If you'd like to help fix it/maintain the C++ side of things I can accept pull requests or also look into making you a maintainer of the repo so you can commit to it directly.

I made a lot of small fixes with the intersect handling (the primitive segment intersect functions for line-line, line-arc, and arc-arc), so that would probably be the easiest place to start as those functions are small. But I know I made other changes as well for finding all intersects between two polylines and the boolean operation code.

Note you can both dynamically and statically link to the Rust C FFI from C++ so there should be no reason you can't use it other than some sort of company policy.