Codecov Report

❌ Patch coverage is 87.50000% with 9 lines in your changes missing coverage. Please review.
✅ Project coverage is 66.01%. Comparing base (a9b90a0) to head (730d18b).
⚠️ Report is 1 commits behind head on main.

@@            Coverage Diff             @@
##             main     #170      +/-   ##
==========================================
+ Coverage   65.84%   66.01%   +0.17%     
==========================================
  Files          34       34              
  Lines        4793     4841      +48     
  Branches     1670     1670              
==========================================
+ Hits         3156     3196      +40     
- Misses        683      691       +8     
  Partials      954      954              

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

Thanks for updating the qcschema support. I tested it locally but found that I could not get it to work with Python 3.14 and an older qcelemental version (which seems to install just fine together).

Right in that Python 3.14 can work only with QCSchema v2, and QCSchema v2 is only available from QCElemental >=0.50.0rc3, which one can get from pypi or -c conda-forge/label/qcelemental_dev. So Python 3.14 and an older qcelemental version can't work. I didn't try to constrain the installation versions since qcel is an optional dependency and nothing that was working should break.

This PR adds a lot of code duplication and stray imports, which will make this very hard to maintain. I started to do some cleanup, but I think it will probably need more work.

Yes, there's a lot of code duplication in the tests. I factored out molecule and some other large chunks, and I could rearrange for less duplication — I wasn't sure what balance between dupl. and lines with only whitespace change was best. I'm not quite following "stray imports," as the below was the only import change in the qcschema.py file.

    if (schver := atomic_input.schema_version) == 1:
        from qcelemental.models import AtomicInput, AtomicResult, ComputeError

    elif schver == 2:
        from qcelemental.models.v2 import AtomicInput, AtomicResult, ComputeError, FailedOperation

I rebased this locally after the changes to main, but I see you've also reconciled with main. I'll have to compare my local and this version closer later.

Thanks for the review and comments!

I did a preliminary cleanup of the version, please verify that it works as you intended and then I think it is good to merge.

Excluding pyscf for py314 should help that lane. I'm working on local testing ...

Ok, CI is passing again and py314 is being tested. I've tested the new versions of the schema files against a local QCEngine and Psi4 with py313. So I think this is good to go.

Thanks for the cleanup. I'll follow this pattern for xtb rather than the dftd4 pattern, if that sounds good.

Thanks for the cleanup. I'll follow this pattern for xtb rather than the dftd4 pattern, if that sounds good.

For tblite and xtb that would be preferable. Maybe it would also be good to refactor the dftd4 one to be consistent.