Hotfix/issue128 fix track line

[martinschwinzerl]

Attempts to fix the problems encountered in isseu #128 by changing the behavior of the global aperture check and the track_line method:

• by using state &= ((|x|<LIMIT_X)&&(|y|<LIMIT_Y)) instead of "updating" the state only with the right-hand-side of the expression above, we no longer rely on external logic to ensure that already lost particles are not updated again. I.e. by virtue of the bit-wise &, particles can only be marked as lost by this method

• Adding additional checks at the entry of the NS(Track_particle_line_objs) to repel already lost particles, the behaviour of the outer loop in track_untilis reproduced even though this outer loop is missing in the context of track_line. This should ensure that invoking track_line always skips already lost particles, regardless of the history of the particles and how they have been lost.

Feedback & tests would be appreciated

[aoeftiger]

Thanks so much for the quick fix! It surely solves the https://github.com/SixTrack/sixtracklib/pull/128 MWE already!

$ python reproducing.py 

[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
Indeed, half of the particles have been lost in the aperture. They are all marked `state==0`.

Let's track the first drift again after having finished the previous turn:
[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]

[aoeftiger]

Yes!! I can confirm it works perfectly fine also in the full SIS100 tracking setup for SixTrackLib+PyHEADTAIL with losses in both, this bug was the source for the issue I had discussed with you earlier! Thanks a million times @martinschwinzerl ! :+1:

[martinschwinzerl]

Adds tests ensuring the expected behavior based on the minimal reproducible example for issue #128. Adds also missing API to NS(MultiPole) and sixtrack::MultiPole required for the tests to actually work.