XAS: Move Startup Logic to _update_structure()

[PNOGillespie]

Addresses requirements for lazy-loading (PR #802) by moving logic for checking core-hole pseudopotentials to _update_structure().

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 67.10%. Comparing base (2787397) to head (f4fd984).
Report is 113 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #806      +/-   ##
==========================================
+ Coverage   67.06%   67.10%   +0.03%     
==========================================
  Files          51       51              
  Lines        4713     4712       -1     
==========================================
+ Hits         3161     3162       +1     
+ Misses       1552     1550       -2     

Flag	Coverage Δ
python-3.11	67.10% <100.00%> (+0.03%)	⬆️
python-3.9	67.13% <100.00%> (+0.03%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[PNOGillespie]

Hi @edan-bainglass, apologies for the slow speed on this one, but this should be ready to go. Let me know if you need anything else.

One other thing I did notice (which @superstar54 should know about) when running test workflows is that the changes which add the tot_charge feature (#402) are actually interfering with the input parameters for XAS calculations. I think this is simply because the default (if no tot_charge is set by the user) explicitly sets tot_charge = 0 in the overrides, rather than removing the parameter from the overrides entirely, which ultimately changes the inputs for XAS and causes calculations to fail on the QE side of things.

I'll check this myself today, open an issue, and get some sort of fix ready.

[edan-bainglass]

@PNOGillespie on the contrary, thank you so much for quickly addressing it 🙏 I'll give it a look later today and test on my end. Thanks again 🙂

[edan-bainglass]

Okay @PNOGillespie, LGTM! Thanks for addressing this.