Add check for Flux conservation

[dschwoerer]

This adds a check that B*J/sqrt(g_22) stays constant along the traced flux tube.

[ZedThree]

Probably useful to have an input option for the tolerance on this to automatically fail if it's too large.

The plotting should also be user controllable, rather than unconditional

[dschwoerer]

I have removed the plotting, and only print the error. I do not see how we manage to get this reliably small enough without some fudging, which I think is probably better done in BOUT++, to ensure things are consistent. Maybe we should completely remove the check, i.e. not add this PR.

Other things in this PR:

• EMC3 field + FieldTracing
• Add check for swapping inner and outer RZline (causes perfectly valid grids, but then the normal input files are not-matching)
• Add check for finite data for stored data
• store parallel magnetic field - this is needed for parallel derivatives to replace the current, problematic method using parallel metric coefficients or rather fudging them up to match the value based on B.

@ZedThree I can separate them, if you think that makes sense.

[ZedThree]

Yes please, smaller PRs are just easier to look over :)

[dschwoerer]

Splitted up 🎉
There should be only one merge conflict between #49 and #48 - but that should be trivial.