Registration of interior algs needs to use a partvec instead of a part

[psakievich]

The current infrastructure is looping over parts where a partvec could be assembled instead. It will require a lot of code changes but it will improve things.

nalu-wind/src/EquationSystems.C

Lines 342 to 359 in 472317f

	for (size_t itarget = 0; itarget < targetNames.size(); ++itarget) {
	stk::mesh::Part* targetPart = meta_data.get_part(targetNames[itarget]);
	if (NULL == targetPart) {
	throw std::runtime_error(
	"Sorry, no part name found by the name " + targetNames[itarget]);
	} else {
	// found the part; no need to subset
	if (stk::topology::ELEMENT_RANK != targetPart->primary_entity_rank()) {
	throw std::runtime_error(
	"Sorry, parts need to be elements.. " + targetNames[itarget]);
	}
	realm_.register_interior_algorithm(targetPart);
	EquationSystemVector::iterator ii;
	for (ii = equationSystemVector_.begin();
	ii != equationSystemVector_.end(); ++ii)
	(*ii)->register_interior_algorithm(targetPart);
	}

[marchdf]

I think this is fixed, looking at the history. Closing. Please reopen if not.

[psakievich]

It's partially complete but not needed unless someone plans to pursue the SmartFields concept further. I don't see myself having time to devote to that in the near future.