LabelSegmenter interface

[shuttle1987]

While looking through the code to figure out handling of label segmentation I saw this bit:

LabelSegmenter = NamedTuple("LabelSegmenter",
                            [("segment_labels", Callable[[Utterance], Utterance]),
                            ("labels", Set[str])])

LabelSegmenter.__doc__ = (
    """ An immutable object that segments the phonemes of an utterance. This
    could probably actually have a __call__ implementation. That won't work
    because namedtuples can't have special methods. Perhaps it could instead
    just be a function which we give a labels attribute. Perhaps that
    obfuscates things a bit, but it could be okay.

What's mentioned in the docstring actually isn't true, a namedtuple is actually just a shortcut for creating a python class. You can subtype or just define the magic methods on it directly.

So something like:

LabelSegmenter.__call__ = lambda self, utterance: self.segment_labels(utterance)

Should be perfectly legal, no esoteric messing around is needed here to get that behavior. This might make the interface cleaner.