Add types for core.(molecular_orbitals|operations|sites|spectrum|tensor|xcfunc)

[DanielYang59]

Summary

• Add/improve types for core
• Replace hard-coded class call with type(self)
• Replace all [0:x] with [:x]

Add types for remaining part of core, follow up of #3814.

• core.molecular_orbitals
• core.operations
• core.sites
• core.spectrum
• core.tensor
• core.xcfunc

PR too large, following separated to another PR

• core.structure
• core.units
• core.trajectory
• Use type(self) to avoid hard-coded class name in __eq__ and similar methods

[DanielYang59]

For some reason, replacing the magic __hash__ method of SymmOp in bf2a953 breaks a unit test of core.structure (test_from_prototype):

pymatgen/pymatgen/core/operations.py

Lines 92 to 93 in 51180fc

	def __hash__(self) -> int:
	return 7

There must be something in it that depends on the hash that I'm not aware of, and I totally have no idea why.

[janosh]

thank a lot @DanielYang59! 👍

[DanielYang59]

Thanks for reviewing. Do you have any comments on #3829 (comment)?

[janosh]

not sure how the SymmOP.__hash__ comes into play in that test. you could modify this assert to print actual and expected string to see how they differ

- assert str(struct) == expected_struct_str
+ assert str(struct) == expected_struct_str, f"{struct=}, {expected_struct_str=}"

[DanielYang59]

Thanks for the suggestion, this is what I got:

> print(struct)
Full Formula (Li4 Cl4)
Reduced Formula: LiCl
abc   :   2.560000   2.560000   2.560000
angles:  90.000000  90.000000  90.000000
pbc   :       True       True       True
Sites (8)
  #  SP      a    b    c
---  ----  ---  ---  ---
  0  Li    0    0    0
  1  Li    0    0.5  0.5
  2  Li    0.5  0    0.5
  3  Li    0.5  0.5  0
  4  Cl    0    0.5  0.5
  5  Cl    0.5  0    0.5
  6  Cl    0    0    0
  7  Cl    0.5  0.5  0

> print(expected_struct_str)
Full Formula (Li4 Cl4)
Reduced Formula: LiCl
abc   :   2.560000   2.560000   2.560000
angles:  90.000000  90.000000  90.000000
pbc   :       True       True       True
Sites (8)
  #  SP      a    b    c
---  ----  ---  ---  ---
  0  Li    0    0    0
  1  Li    0.5  0.5  0
  2  Li    0.5  0    0.5
  3  Li    0    0.5  0.5
  4  Cl    0.5  0    0.5
  5  Cl    0    0.5  0.5
  6  Cl    0.5  0.5  0
  7  Cl    0    0    0

Looks like the ordering of sites changed, and I don't quite know how hash of SymmOp is related to generation of Structure. I would look closer later :)