fixes replicafiller

[SteffenSeckler]

fixes replicafiller also when using mpi.

assumes:

• The ParticleContainerToBasisWrapper does not have to do any cutting of particles
• Removing particles happens later, i.e., by the ReplicaFiller!

what has been done:

• src/particleContainer/ParticleContainer.h: make various functions virtual, s.t. it actually is used by the ParticleContainerToBasisWrapper!
• src/utils/generator/ReplicaFiller.cpp: make the ParticleContainerToBasisWrapper accept all particles, by implementing getBoundingBoxMin, getBoundingBoxMax, isInBoundingBox
• don't pass any object to the ParticleContainerToBasisWrapper, no check needs to be done here yet!
• abort if no particles have been added to the ParticleContainerToBasisWrapper

in addition:

• some clang-tidy fixes

remark:

• some tidy-up work could still be done inside of the ParticleContainerToBasisWrapper

[tchipev]

Looks great, I even tested it briefly vs revision 6000 - the numbers are good.
But please indicate that it's a dirty fix.
Make an Issue with recommendations, observations, etc. on the ReplicaFiller class for eventual clean-up.
It seems to me that the whole ParticleContainerToBasisWrapper approach is a bad idea. Perhaps a more abstract interface should be introduced, from which ParticleContainer and whatever this should be can also derive.

[tchipev]

OK