Update dependency gemmi to >=0.5.5, <=0.5.5

[renovate]

This PR contains the following updates:

Package	Change	Age	Adoption	Passing	Confidence
gemmi	>=0.4.2, <=0.5.4 -> >=0.5.5, <=0.5.5

---

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Renovate will not automatically rebase this PR, because other commits have been found.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, click this checkbox. ⚠ Warning: custom changes will be lost.

---

This PR has been generated by Mend Renovate. View repository job log here.

[JBGreisman]

Per project-gemmi/gemmi#215, Op.negated() negated the real-space translation of the symmetry operations (not just the rotational component). This had a direct impact on phase shifts, requiring them to be flipped for Friedel symmetry operations, as was done here with phic * phis.

By using op.combine(friedel), we can now directly compare the phase shifts.

[JBGreisman]

Summary: this PR updates reciprocalspaceship to use gemmi v0.5.5, and sets this to the new minimal version requirement due to the following changes to gemmi namespace:

• gemmi.GroupOps.is_centric() was changed to gemmi.GroupOps.is_centrosymmetric(). This gets called in the multiplicity calculation in rs.utils
• gemmi.Op.negated() was removed, and we instead use gemmi.Op.combine() with the Friedel operation -x,-y,-z. As commented above, Op.negated() had also negated the real-space translation of the symmetry operations, which impacted how phase shifts were compared in one test

[kmdalton]

this looks fine to me. i wanted to double check why compute_structurefactor_mutliplicity needed modifying. was go.is_centric moved to sg.is_centrosymmetric? do they do equivalent things?

[JBGreisman]

the is_centric() method was renamed to is_centrosymmetric(). Before, only the GroupOps object had this method, but now it is in the namespace of both SpaceGroup and GroupOps objects. Here, I opted to call it directly from the SpaceGroup, but either way could have worked -- the change in method name is certainly required though.