Remove need for spin projection in spin_magnitude_conservation rule

[redeboer]

The spin_magnitude_conservation() rule requires SpinEdgeInput, which has a spin_projection attribute, even though the rule implementation does not need it. This is a problem for #158. That issue either needs to work with spin projections for now (which is what we eventually want to remove, see #219) or we need to remove the need for providing a spin projection altogether by solving this issue first.

qrules/src/qrules/conservation_rules.py

Lines 694 to 728 in e60ff89

	def spin_magnitude_conservation(
	ingoing_spins: List[SpinEdgeInput],
	outgoing_spins: List[SpinEdgeInput],
	interaction_qns: SpinMagnitudeNodeInput,
	) -> bool:
	r"""Check for spin conservation.
	Implements
	.. math::
	|S_1 - S_2| \leq S \leq |S_1 + S_2|
	and
	.. math::
	|L - S| \leq J \leq |L + S|
	"""
	# L and S can only be used if one side is a single state
	# and the other side contains of two states (isobar)
	# So do a full check if this is the case
	if (len(ingoing_spins) == 1 and len(outgoing_spins) == 2) or (
	len(ingoing_spins) == 2 and len(outgoing_spins) == 1
	):
	return _check_magnitude(
	[x.spin_magnitude for x in ingoing_spins],
	[x.spin_magnitude for x in outgoing_spins],
	interaction_qns,
	)
	# otherwise don't use S and L and just check magnitude
	# are integral or non integral on both sides
	return (
	sum(float(x.spin_magnitude) for x in ingoing_spins).is_integer() # type: ignore[union-attr]
	== sum(float(x.spin_magnitude) for x in outgoing_spins).is_integer() # type: ignore[union-attr]
	)

[grayson-helmholz]

Has been implemented in the remove_spin-projections_in_cons-rules-branch. Issue should be closeable after merging.