Segment fragment

[Luke-Poeppel]

There are a number of instances (particularly Livre d'Orgue movements and moments of SF) where the rhythm is consecutive X notes (e.g. 16th, 32nd, etc.). In these cases, we don't really care about extracting paths. It would be cool to have a Segment fragment for these regions. So a search/result might look like:

>>> path = path_finder(
...        ....,
...        use_segments=True
>>> )
>>> for x in path:
...         print(x.fragment, x.fragment.num_onsets)
<fragment.GreekFoot Trochee> 2
<fragment.GreekFoot Iamb> 2
<fragment.GreekFoot Iamb> 2
<fragment.Segment> 37
...

This would require some kind of clustering testing (essentially an addition to the neutral source/target pick of Dijkstra). Would also have to set a threshold for segment length. So alternatively maybe:

>>> path = path_finder(
...        ....,
...        min_segment_length=10
>>> )
>>> for x in path:
...         print(x.fragment, x.fragment.num_onsets)
<fragment.GreekFoot Trochee> 2
<fragment.GreekFoot Iamb> 2
<fragment.GreekFoot Iamb> 2
<fragment.Segment> 37
...

The shortest segment would, I think, be of length 5.