Parse screw joints from SDFormat

[scpeters]

Part of #17765.

Depends on gazebosim/sdformat#1168 to provide the //joint/screw_thread_pitch parameter in SDFormat 1.10 and corresponding Joint::ScrewThreadPitch() that uses the same units as drake: meters / revolution. The existing thread_pitch parameter (exposed via Joint::ThreadPitch()) has been used in gazebo classic with inverse units and the opposite sign convention.

---

This change is 

[EricCousineau-TRI]

Will require minor release of sdformat, which should happen w/in next 2 weeks or so.

[scpeters]

Will require minor release of sdformat, which should happen w/in next 2 weeks or so.

the 13.2.0 release has been made, and this is ready for review

[EricCousineau-TRI]

+@EricCousineau-TRI , thanks!
+@ggould-tri for platform review, please!

Reviewed 4 of 4 files at r1, all commit messages.
Reviewable status: LGTM missing from assignee ggould-tri(platform), commits need curation (https://drake.mit.edu/reviewable.html#curated-commits), missing label for release notes (waiting on @ggould-tri)

[ggould-tri]

; one obscure comment nit.

Reviewed 4 of 4 files at r1, all commit messages.
Reviewable status: 1 unresolved discussion, commits need curation (https://drake.mit.edu/reviewable.html#curated-commits), missing label for release notes (waiting on @scpeters)

---

multibody/parsing/detail_sdf_parser.cc line 771 at r1 (raw file):

    case sdf::JointType::SCREW: {
      const double damping = ParseJointDamping(diagnostic, joint_spec);
      // The ScrewThreadPitch() API uses the same units as drake: meters / rev

minor: This comment is a bit misleading from the point of view of Drake's internals, since drake is actually violating its own unit conventions in ScrewJoint (drake by convention uses radians, not revolutions, to express rotation). I've suggested a slight rewording here (and also added the missing end-of-sentence punctuation) but you can decide if or how much of this suggestion to take.

Suggestion:

      // The ScrewThreadPitch() API uses the same representation as
      // Drake's ScrewJoint class (meters / revolution, right-handed).

[scpeters]

Reviewable status: commits need curation (https://drake.mit.edu/reviewable.html#curated-commits), missing label for release notes (waiting on @EricCousineau-TRI and @ggould-tri)

---

multibody/parsing/detail_sdf_parser.cc line 771 at r1 (raw file):

Previously, ggould-tri wrote…

minor: This comment is a bit misleading from the point of view of Drake's internals, since drake is actually violating its own unit conventions in ScrewJoint (drake by convention uses radians, not revolutions, to express rotation). I've suggested a slight rewording here (and also added the missing end-of-sentence punctuation) but you can decide if or how much of this suggestion to take.

Done

[ggould-tri]

Reviewed 2 of 2 files at r2, all commit messages.
Reviewable status: commits need curation (https://drake.mit.edu/reviewable.html#curated-commits), missing label for release notes (waiting on @scpeters)

[ggould-tri]

+(release notes: feature)

Reviewable status: commits need curation (https://drake.mit.edu/reviewable.html#curated-commits) (waiting on @scpeters)

[ggould-tri]

+(status: commits are properly curated)

Reviewable status: when planning a "properly curated" merge commit the PR must always be rebased onto latest master (waiting on @scpeters)

[jwnimmer-tri]

FWIW -- to my first glance, the commits are not really independently revertable so we should squash them in any case.

[ggould-tri]

The commits seemed to work when I one-at-a-timed them, but I'm happy to take your caution and get this merged :-) .
-(status: commits are properly curated) +(status: squashing now)

Reviewable status: when planning a "properly curated" merge commit the PR must always be rebased onto latest master (waiting on @scpeters)

[jwnimmer-tri]

The failure mode would be if macOS couldn't build the libsdformat upgrade, then we'd need to revert not only the upgrade but also the new parsing feature.