It isn't actually necessary to use insert_points...?

[sampotter]

I spent a little more time looking into how to add points to the mesh, and appears that using insert_points (see #60, #68, and #69) may be unnecessary.

In section 5.1.1 of the TetGen manual, it mentions that polygons in the PLC may be degenerate (only consist of one or two nodes), in which case they can be used to specify constrained points or edges. I tried this out, and it appears that while assembling the points and facets of a meshpy.tet.MeshInfo, just adding an extra point that isn't included in a facet has the desired effect. I guess I should have read the section titled "Useful Things to Know" earlier. 😨

Maybe the -i switch makes more sense when using TetGen from the command-line... But it would be helpful to indicate that there are different ways to go about adding constrained features to a mesh in meshpy's docs.

[sampotter]

I should also mention that I've run into at least one case where trying to insert a point using insert_points (as implemented in #69) leads to a segmentation fault, while just adding the extra point to MeshInfo's list of points works just fine. I can add a MWE for this if there's interest.