def testRegroupOneSpeakerTwoUtterances_obeyingOriginalBoundary(self):
utterances = ["would you like to", "sit down"]
words = [("would", 1, 0.1, 0.2), ("you", 1, 0.2, 0.3),
("like", 1, 0.3, 0.4), ("to", 1, 0.5, 0.6),
("sit", 1, 1.6, 1.7), ("down", 1, 1.7, 1.8)]
regrouped = audio_asr.regroup_utterances(utterances, words)
self.assertEqual(regrouped, [("would you like to", 1, 0.1, 0.6),
("sit down", 1, 1.6, 1.8)])
def testRegroupThreeSpeakersOneUtterances(self):
utterances = ["would you like to sit down no me neither"]
words = [("would", 1, 0.1, 0.2), ("you", 1, 0.2, 0.3),
("like", 1, 0.3, 0.4), ("to", 1, 0.5, 0.6),
("sit", 1, 0.6, 0.7), ("down", 1, 0.7, 0.8),
("no", 2, 0.8, 0.9), ("me", 3, 0.9, 1.0),
("neither", 3, 1.0, 1.1)]
regrouped = audio_asr.regroup_utterances(utterances, words)
self.assertEqual(regrouped,
[("would you like to sit down", 1, 0.1, 0.8),
("no", 2, 0.8, 0.9), ("me neither", 3, 0.9, 1.1)])
def testRegroupTwoSpeakersTwoUtterances(self):
utterances = ["would you like to sit down no", "been sitting all day"]
words = [("would", 1, 0.1, 0.2), ("you", 1, 0.2, 0.3),
("like", 1, 0.3, 0.4), ("to", 1, 0.5, 0.6),
("sit", 1, 0.6, 0.7), ("down", 1, 0.7, 0.8),
("no", 2, 0.8, 0.9), ("been", 2, 0.9, 1.0),
("sitting", 2, 1.0, 1.1), ("all", 2, 1.1, 1.2),
("day", 2, 1.2, 1.3)]
regrouped = audio_asr.regroup_utterances(utterances, words)
self.assertEqual(regrouped,
[("would you like to sit down", 1, 0.1, 0.8),
("no been sitting all day", 2, 0.8, 1.3)])
def testMismatchInWordsRaisesValueError(self):
utterances = ["would you like to sit down no", "been sitting all day "]
words = [("would", 1, 0.1, 0.2), ("thou", 1, 0.2, 0.3)]
with self.assertRaises(ValueError):
audio_asr.regroup_utterances(utterances, words)
def testUnmatchedWordsRaisesValuError(self):
utterances = ["would you like to", "sit down"]
words = [("would", 1, 0.1, 0.2), ("you", 1, 0.2, 0.3)]
with self.assertRaisesRegex(ValueError, r"Some words .* missing"):
audio_asr.regroup_utterances(utterances, words)