Fix elastic conventional structure

[mjwen]

Summary

• Fix the bug in BUG:ElasticMaker breaks when conventional=True #654
• Update test data to use those generated from conventional cells
• Use the new test data to test both conventional and unconventional cells

TODO (if any)

• The elastic workflow cannot finish no matter conventional is set to True or False. There seems to be another bug (see BUG: ElasticMaker deepcopy error #707). Let's wait for that to be corrected first. (It is a pymatgen bug, and fix in Revert back TransformedStructure.__getattr__ pymatgen#3617)
• add a test for conventional (done)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (e5ab301) 76.32% compared to head (da7a22e) 76.21%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #706      +/-   ##
==========================================
- Coverage   76.32%   76.21%   -0.11%     
==========================================
  Files          87       87              
  Lines        7164     7165       +1     
  Branches     1057     1057              
==========================================
- Hits         5468     5461       -7     
- Misses       1373     1383      +10     
+ Partials      323      321       -2     

Files	Coverage Δ
src/atomate2/common/flows/elastic.py	84.00% <100.00%> (+6.44%)	⬆️

... and 1 file with indirect coverage changes

[JaGeo]

Maybe interesting:

atomate2/src/atomate2/common/jobs/utils.py

Line 39 in e5ab301

def structure_to_conventional(

[utf]

Thanks very much @mjwen

[tlsong]

Thanks very much @mjwen