Make GSD optional

[IAlibay]

Fixes #3819

Changes made in this Pull Request:

• Make GSD an optional requirement

PR Checklist

• Tests?
• Docs?
• CHANGELOG updated?
• Issue raised/referenced?

---

📚 Documentation preview 📚: https://mdanalysis--4174.org.readthedocs.build/en/4174/

[IAlibay]

Opening this early so I can get some feedback from @MDAnalysis/coredevs - this is probably quite an abrupt change, do we need to have some kind of deprecation window?

[github-actions]

Linter Bot Results:

Hi @IAlibay! Thanks for making this PR. We linted your code and found the following:

Some issues were found with the formatting of your code.

Code Location	Outcome
main package	⚠️ Possible failure
testsuite	⚠️ Possible failure

Please have a look at the darker-main-code and darker-test-code steps here for more details: https://github.com/MDAnalysis/mdanalysis/actions/runs/5304029149/jobs/9600062116

---

Please note: The black linter is purely informational, you can safely ignore these outcomes if there are no flake8 failures!

[codecov]

Codecov Report

Patch coverage: 100.00% and no project coverage change.

Comparison is base (efc51af) 93.60% compared to head (ecabb78) 93.61%.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #4174   +/-   ##
========================================
  Coverage    93.60%   93.61%           
========================================
  Files          193      193           
  Lines        25147    25168   +21     
  Branches      4056     4058    +2     
========================================
+ Hits         23539    23560   +21     
  Misses        1092     1092           
  Partials       516      516           

Impacted Files	Coverage Δ
package/MDAnalysis/coordinates/GSD.py	91.89% <100.00%> (+1.56%)	⬆️
package/MDAnalysis/topology/GSDParser.py	92.85% <100.00%> (+1.36%)	⬆️

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[RMeli]

I don't think so IMO. It might surprise some users, but the error message is very clear and the solution a pip install away.

[IAlibay]

Ok this should be good to go I think (unless I forgot to fix some test somewhere). Since it's a core dependency change, I'm going to leave this here for a week for other @MDAnalysis/coredevs to chime in, and if there are no objections it'll get merged.

Specifically pinging @orbeckst here, since we've spent a lot of time pondering in the past on how best to do these types of changes.

[hmacdope]

LGTM thanks @IAlibay

[orbeckst]

For reference: conda-forge feedstock was already updated conda-forge/mdanalysis-feedstock#57

[orbeckst]

I agree.

Eventually it would be good to also allow gsd > 3.0 but we have to be able to test it. In the short term, the restriction is sensible.

[orbeckst]

@IAlibay I'll let you do the merging (unless you want someone else to do it, then change the assignee). Thanks for working out the problem and the quick-fix.

[joaander]

I agree.

Eventually it would be good to also allow gsd > 3.0 but we have to be able to test it. In the short term, the restriction is sensible.

Please test GSD 3.0.1.

[orbeckst]

@joaander thanks, we'll check that #4172 is solved with gsd 3.0.1. Will just take a few days (people on holidays etc).

If it works we could blacklist 3.0.0 and otherwise allow gsd!=3.0.0. We'll probably still make it optional as we have been trying to slim down dependencies.

[joaander]

@orbeckst understood. If 3.0.1 passes your tests, I expect 3.0.0 will work for your users. I find that most users follow PEP8 guidelines and issue imports at the top of the module, not inside child threads.