Make sure disulphide search is converted to a SelectorBond.

[lohedges]

This PR closes #223 and closes OpenBioSim/biosimspace_tutorials#20 by making sure any (succesful) search for disulphide bonds is converted to a SelectorBond, thus is iterable. At present, if a single result is found, then it is returned as a Bond, hence the logic that follows the search will fail, i.e. getting an item from will extract an Atom, not a Bond.

I've tested the search logic for other bonds in a molecule, e.g. those between specific atom names, or indices. I don't (yet) have a single disulphide bond test molecule to add.

• I confirm that I have merged the latest version of devel into this branch before issuing this pull request (e.g. by running git pull origin devel): [y]
• I confirm that I have permission to release this code under the GPL3 license: [y]

Suggested reviewers:

@chryswoods

[lohedges]

Not sure if it's related to OpenMM 8.1, but those Exscientia restraint tests appear to be failing again (even with reduced tolerance).

[chryswoods]

All good - I accept this is confusing and a source of bugs. But, it makes for a much nicer interactive API. For scripts, there are the .bond() / .bonds() and associated functions that let us control what type of answer is needed from the query. I almost want a tool that could scan the code to auto-fix these bugs ;-)