support pydantic 3.13

[atravitz]

resolves #807

Note that this will push us to drop python 3.11, which is out of SPEC0.

Tips

• Comment "pre-commit.ci autofix" to have pre-commit.ci atomically format your PR.
  Since this will create a commit, it is best to make this comment when you are finished with your work.

LLM / AI generated code disclosure

LLMs or other AI-powered tools (beyond simple IDE use cases) were used in this contribution: no

Checklist

• Added a news entry
• Filled in the AI-generated code disclosure.

Developers certificate of origin

• I certify that this contribution is covered by the MIT License here and the Developer Certificate of Origin at https://developercertificate.org/.

[atravitz]

currently, these are ways we could fix things, but the existing model_dump() behavior is a problem IMO.

for example, in the test src/gufe/tests/test_models.py(340)test_frozen_equality_changed()

s1.model_dump()
  > {'forcefield_settings': {}, 'thermo_settings': {'temperature': {'val': 300, 'unit': 'kelvin'}, 'pressure': None, 'ph': None, 'redox_potential': None}}`

(note 'forcefield_settings':{})
vs

s1.forcefield_settings.model_dump()
 > {'constraints': 'allbonds', 'rigid_water': True, 'hydrogen_mass': 3.0, 'forcefields': ['amber/ff14SB.xml', 'amber/tip3p_standard.xml', 'amber/tip3p_HFE_multivalent.xml', 'amber/phosaa10.xml', 'amber/lipid17_merged.xml'], 'small_molecule_forcefield': 'openff-2.2.1', 'nonbonded_method': 'PME', 'nonbonded_cutoff': {'val': 0.9, 'unit': 'nanometer'}}

I think we should be able to resolve this in model settings or something similarly upstream, but I haven't sorted it out yet.

[github-actions]

No API break detected ✅

View workflow run

Griffe output

$ griffe check "gufe" -s src --no-inspection --no-color --verbose -a origin/main

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.64%. Comparing base (b18420e) to head (f56747f).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #808   +/-   ##
=======================================
  Coverage   98.64%   98.64%           
=======================================
  Files          43       43           
  Lines        2807     2807           
=======================================
  Hits         2769     2769           
  Misses         38       38           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[atravitz]

pull in this comment from @mattwthompson :

I'm not sure why the serialized representation changes - whether or not the default value of "nonbonded_cutoff" is serialized might not be important, but I've seen problems¹ in which a serialized non-default value is ignored and set to the default value².

The failures due to dependencies not being installed is just from the switch to the Conda package to the development install - cachetools, python-constraint, and xmltodict are probably all you need

Footnotes

1. https://github.com/openforcefield/openff-interchange/pull/1501#issuecomment-4545442911 ↩

2. Bad! ↩