Collect common solver functionality in a base class

[oparry-ukaea]

Description

Creates a base class for solver equation systems from which H3LAPD (DriftReducedSystem) and SimpleSOL (SOLSystem) can inherit. Base class is templated on a nektar equation system (for time-evolving systems, it must derive from UnsteadySystem). Time evolving systems are also templated on a particle system class, which must derive from a new particle system base class.

Type of change

• [x] Refactoring
• [x] New feature (non-breaking change which adds functionality)
• [x] Requires documentation updates

Testing

None added

Checklist:

• [x] New and existing unit tests pass locally with my changes
• [x] Any new dependencies are automatically built for users via cmake
• [x] I have used understandable variable names
• [x] I have run clang-format against my *.hpp and *.cpp changes
• [x] I have run cmake-format against my changes to CMakeLists.txt
• [x] I have performed a self-review of my own code
• [x] I have commented my code, particularly in hard-to-understand areas
• [x] My changes generate no new warnings

[jwscook]

Once https://github.com/ExCALIBUR-NEPTUNE/NESO-hw-impurity-transport/pull/5 is merged then this can follow quite quickly

[jwscook]

Should run this past Stefan and Sanket before merging because this assumes that there is a single ParticleGroup per ParticleSystem, which may not fit their usecase.

[jwscook]

Need to create some instructions in the README after NESO 0.1.0.

[oparry-ukaea]

Having had another look, I don't think it'll be a problem to adapt the particle system base class to work with multiple ParticleGroups in a separate PR if needs be. Marking this as ready-to-merge.

[oparry-ukaea]

Instructions for using the base classes in a new solver need to be added (#239), but that may depend on other changes (rearranging submodules etc.) to some extent - I propose we do that in a separate PR.