|
1
|
|
|
from topik.tokenizers.ngrams import _collect_ngrams, \ |
|
|
|
|
|
|
2
|
|
|
_collocation_document, ngrams |
|
3
|
|
|
from nose.tools import nottest |
|
|
|
|
|
|
4
|
|
|
sample_data = [ |
|
|
|
|
|
|
5
|
|
|
("doc1", str(u"Frank the Swank-Tank walked his sassy unicorn, Brony," |
|
6
|
|
|
u" to prancercise class daily. Prancercise was " |
|
7
|
|
|
u"a tremendously popular pastime of sassy " |
|
8
|
|
|
u"unicorns and retirees alike.")), |
|
9
|
|
|
("doc2", str(u"Frank the Swank-Tank was also known as Big Daddy Workout Queen " |
|
10
|
|
|
u"as he loved to cross-dress while prancercising." |
|
11
|
|
|
u"Dressing up as a sassy unicorn to match Brony was a key " |
|
12
|
|
|
u"source of enjoyment for both him and the onlooking retirees.")), |
|
13
|
|
|
("doc3", str(u"Big Daddy Workout Queen knew that the best way to get " |
|
14
|
|
|
u"more people to embrace prancercise was to " |
|
15
|
|
|
u"wear flashy outfits with tassels and bells.")), |
|
16
|
|
|
("doc3", str(u"Prancercise is a form of both art and fitness, " |
|
17
|
|
|
u"originally invented by sassy unicorns. It has " |
|
18
|
|
|
u"recently been popularized by such retired " |
|
19
|
|
|
u"celebrities as Frank The Swank-Tank (aka Big Daddy Workout Queen)"))] |
|
20
|
|
|
|
|
21
|
|
|
|
|
22
|
|
|
@nottest |
|
|
|
|
|
|
23
|
|
|
def generator(): |
|
|
|
|
|
|
24
|
|
|
for item in sample_data: |
|
25
|
|
|
yield item |
|
26
|
|
|
|
|
27
|
|
|
|
|
28
|
|
|
def test__collect_ngrams(): |
|
|
|
|
|
|
29
|
|
|
result_ngrams = _collect_ngrams(sample_data, min_freqs=[2, 2, 2]) |
|
30
|
|
|
assert(result_ngrams[0].pattern == |
|
31
|
|
|
u'(big daddy|daddy workout|frank swank|swank tank|workout queen|sassy unicorn|sassy unicorns)') |
|
32
|
|
|
assert(result_ngrams[1].pattern == u'(big daddy workout|daddy workout queen|frank swank tank)') |
|
|
|
|
|
|
33
|
|
|
assert(result_ngrams[2].pattern == u'(big daddy workout queen)') |
|
|
|
|
|
|
34
|
|
|
|
|
35
|
|
|
|
|
36
|
|
|
def test__collocation_document(): |
|
|
|
|
|
|
37
|
|
|
these_ngrams = _collect_ngrams(sample_data, min_freqs=[2, 2, 2]) |
|
38
|
|
|
assert(_collocation_document(sample_data[0][1],these_ngrams) == [ |
|
|
|
|
|
|
39
|
|
|
u'frank_swank', u'tank', u'walked', u'sassy_unicorn', u'brony', |
|
40
|
|
|
u'prancercise', u'class', u'daily', u'prancercise', u'tremendously', |
|
41
|
|
|
u'popular', u'pastime', u'sassy_unicorns', u'retirees', u'alike' |
|
42
|
|
|
]) |
|
43
|
|
|
|
|
44
|
|
|
assert(_collocation_document(sample_data[1][1],these_ngrams) == [ |
|
|
|
|
|
|
45
|
|
|
u'frank_swank', u'tank', u'known', u'big_daddy', u'workout_queen', |
|
46
|
|
|
u'loved', u'cross', u'dress', u'prancercising', u'dressing', |
|
47
|
|
|
u'sassy_unicorn', u'match', u'brony', u'key', u'source', u'enjoyment', |
|
48
|
|
|
u'onlooking', u'retirees']) |
|
49
|
|
|
|
|
50
|
|
|
|
|
51
|
|
|
def test_ngrams_list(): |
|
|
|
|
|
|
52
|
|
|
freq_bounds = [(2, 100), (2, 100), (2, 100)] |
|
53
|
|
|
tokenized_corpora = ngrams(sample_data, freq_bounds=freq_bounds) |
|
54
|
|
|
assert(len(freq_bounds) == 3) |
|
|
|
|
|
|
55
|
|
|
assert(next(tokenized_corpora) == ( |
|
56
|
|
|
'doc1', [ |
|
57
|
|
|
u'frank_swank', u'tank', u'walked', u'sassy_unicorn', u'brony', |
|
58
|
|
|
u'prancercise', u'class', u'daily', u'prancercise', u'tremendously', |
|
59
|
|
|
u'popular', u'pastime', u'sassy_unicorns', u'retirees', u'alike' |
|
60
|
|
|
])) |
|
61
|
|
|
assert(next(tokenized_corpora) == ( |
|
62
|
|
|
'doc2', [ |
|
63
|
|
|
u'frank_swank', u'tank', u'known', u'big_daddy', u'workout_queen', |
|
64
|
|
|
u'loved', u'cross', u'dress', u'prancercising', |
|
65
|
|
|
u'dressing', u'sassy_unicorn', u'match', u'brony', u'key', u'source', |
|
66
|
|
|
u'enjoyment', u'onlooking', u'retirees'])) |
|
67
|
|
|
|
|
68
|
|
|
|
|
69
|
|
|
def test_ngrams_generator(): |
|
|
|
|
|
|
70
|
|
|
freq_bounds = [(2, 100), (2, 100), (2, 100)] |
|
71
|
|
|
corpus_gen = generator() |
|
72
|
|
|
tokenized_corpora = ngrams(corpus_gen, freq_bounds=freq_bounds) |
|
73
|
|
|
assert(len(freq_bounds) == 3) |
|
|
|
|
|
|
74
|
|
|
assert(next(tokenized_corpora) == ( |
|
75
|
|
|
'doc1', [ |
|
76
|
|
|
u'frank_swank', u'tank', u'walked', u'sassy_unicorn', u'brony', |
|
77
|
|
|
u'prancercise', u'class', u'daily', u'prancercise', u'tremendously', |
|
78
|
|
|
u'popular', u'pastime', u'sassy_unicorns', u'retirees', u'alike' |
|
79
|
|
|
])) |
|
80
|
|
|
assert(next(tokenized_corpora) == ( |
|
81
|
|
|
'doc2', [ |
|
82
|
|
|
u'frank_swank', u'tank', u'known', u'big_daddy', u'workout_queen', |
|
83
|
|
|
u'loved', u'cross', u'dress', u'prancercising', |
|
84
|
|
|
u'dressing', u'sassy_unicorn', u'match', u'brony', u'key', u'source', |
|
85
|
|
|
u'enjoyment', u'onlooking', u'retirees'])) |
|
|
|
|
|
ContinuumIO /
topik
Loading history...
The coding style of this project requires that you add a docstring to this code element. Below, you find an example for methods:
If you would like to know more about docstrings, we recommend to read PEP-257: Docstring Conventions.