def test_normalise_corpus():
"""Test we can normalise a corpus."""
corpus = [
'Hello, how\'re you?',
'I\'m good, thanks.',
'Good.',
'beer',
'Gone fishing?',
]
normalised, norm_map = brains.normalise_corpus(corpus)
assert normalised == [
'hello howre you',
'im good thanks',
'good',
'beer',
'gone fishing',
]
assert norm_map == {
'beer': ['beer'],
'fishing': ['fishing?'],
'gone': ['Gone'],
'good': ['good,', 'Good.'],
'hello': ['Hello,'],
'howre': ['how\'re'],
'im': ['I\'m'],
'thanks': ['thanks.'],
'you': ['you?'],
}
def test_bisect():
"""We create the splits using the middle of sentences."""
corpus = [
'Every roadtrip needs someone to tell them to slow down in the corners :)',
'looking forward to F1 tomorrow night - should be super tight at the top!',
]
normalised_sentences = brains.normalise_corpus(corpus)[0]
starts, ends = brains.bisect(normalised_sentences, min_occurances=1)
assert starts == {
'be': [['looking', 'forward', 'to', 'f1', 'tomorrow', 'night', 'should']],
'night': [['looking', 'forward', 'to', 'f1', 'tomorrow']],
'should': [['looking', 'forward', 'to', 'f1', 'tomorrow', 'night']],
'tell': [['every', 'roadtrip', 'needs', 'someone', 'to']],
'them': [['every', 'roadtrip', 'needs', 'someone', 'to', 'tell']],
'to': [['every', 'roadtrip', 'needs', 'someone', 'to', 'tell', 'them']],
}
assert ends == {
'be': [['super', 'tight', 'at', 'the', 'top']],
'night': [['should', 'be', 'super', 'tight', 'at', 'the', 'top']],
'should': [['be', 'super', 'tight', 'at', 'the', 'top']],
'tell': [['them', 'to', 'slow', 'down', 'in', 'the', 'corners']],
'them': [['to', 'slow', 'down', 'in', 'the', 'corners']],
'to': [['slow', 'down', 'in', 'the', 'corners']],
}