WIP add cutoff by type

[christophlohrmann]

Fixes #4819

Description of changes:

• Added cutoffs by type
• no guarantees with regards to execution speed

[christophlohrmann]

@RudolfWeeber does this look like what you had in mind?

[RudolfWeeber]

LGTM. Just please rename the functions working on type lists to make it clear that they only concern non_bonded interactions

• Dipoles are also a non-central interation.

[RudolfWeeber]

Once this is merged, the hzbrid decompositoin can be adapted to determine the cutoff automtically. However, then, electrostatics and dipolar and collision cutoffs need to be taken into account.

[christophlohrmann]

LGTM. Just please rename the functions working on type lists to make it clear that they only concern non_bonded interactions

• Dipoles are also a non-central interaction.

1. Is the rename really necessary if the functions are members of InteractionsNonBonded
2. which of the nonbonded interactions is "dipoles"?