Update OptimizationResultCollection.create_basic_dataset to preserve molecule IDs

[ntBre]

Description

The goal of this PR is a long-term fix to #297. In short, the current OptimizationResultCollection.create_basic_dataset implementation creates a new dataset by round-tripping through OpenFF Molecules, which causes issues with the hashing/equivalence checks in QCArchive, leading to "the same molecules" being considered different. In turn, this causes OptimizationResultCollection.to_basic_result_collection to return fewer records than expected.

The fix is simply to pass qcelemental Molecule objects along directly instead of reconstructing OpenFF Molecules. As it turns out, this was essentially already built into qcsubmit and just required a two-line change to update some keyword arguments in the BasicDataset.add_molecule call.

Todos

• Update create_basic_dataset to preserve molecule hashes from the optimization dataset

Status

• Ready to go

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.43%. Comparing base (cc3f23e) to head (bd3bf3b).

Additional details and impacted files

[ntBre]

Well, that turned out to be a pretty easy fix. The hardest part was adding a new mock get_molecules method to the _PortalClient test fixture. The main limitation now is just that the new test is pulling down a full dataset from a live QCArchive instance. I think I can probably pull out a small subset of the real dataset to speed that up at least.