problems with postproc.get_trial_targets()

i'm not really sure what the point of get_trial_targets() is, i think i wrote it but i can't remember why. it probably belongs in preproc rather than postproc, because it's bmi3d specific. but basically it is redundant and not necessary because we are using events to keep track of targets anyway.

if we keep it, i think it should be renamed parse_bmi3d_sequence() or something similar, and it should be modified to return the sequence indices as well as the target locations. and the docstring updated to reflect that it is returning the targets in the order they were supplied by the generator rather than the ones that ended up getting used in the task, since sometimes they repeat on failures.