don't permit arcs to have decreasing angle domains

[andrewheumann]

BACKGROUND:

• Arc.Polyline doesn't work if its domain has negative length. #994
• We want to allow users to construct Arcs with apparently decreasing angle domains, in order to have intuitive control over arc direction.

DESCRIPTION:

• Updates Arc constructors to ensure that EndAngle is always greater than StartAngle.

TESTING:

• I added a test for every modified constructor, to verify that for any pair of angles A,B, arc(A,B) looked just like arc(B,A), but parameterized in the opposite direction.
• I added a test to construct an IndexedPolycurve containing an arc which runs "clockwise" with respect to its circle / the Z axis, to ensure consistency in direction of the overall polycurve:
  
• I added a test for issue Arc.Polyline doesn't work if its domain has negative length. #994's original complaint, and visually verified that it worked.
• All tests pass.

FUTURE WORK:

• We may need to introduce a notion of SenseAgreement as STEP does for more complex curves, such that a subcurve could act like it was running in the opposite direction from its basis curve. Or maybe not ¯\_(ツ)_/¯

REQUIRED:

• All changes are up to date in CHANGELOG.md.

---

This change is 

[wynged]

Reviewed 4 of 4 files at r1, all commit messages.
Reviewable status: complete! 1 of 1 approvals obtained