|
1
|
|
|
# -*- coding: utf-8 -*- |
|
2
|
|
|
|
|
3
|
|
|
# Copyright 2014-2018 by Christopher C. Little. |
|
4
|
|
|
# This file is part of Abydos. |
|
5
|
|
|
# |
|
6
|
|
|
# Abydos is free software: you can redistribute it and/or modify |
|
7
|
|
|
# it under the terms of the GNU General Public License as published by |
|
8
|
|
|
# the Free Software Foundation, either version 3 of the License, or |
|
9
|
|
|
# (at your option) any later version. |
|
10
|
|
|
# |
|
11
|
|
|
# Abydos is distributed in the hope that it will be useful, |
|
12
|
|
|
# but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
13
|
|
|
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
14
|
|
|
# GNU General Public License for more details. |
|
15
|
|
|
# |
|
16
|
|
|
# You should have received a copy of the GNU General Public License |
|
17
|
|
|
# along with Abydos. If not, see <http://www.gnu.org/licenses/>. |
|
18
|
|
|
|
|
19
|
|
|
"""abydos.tests.test_stemmer_clef. |
|
20
|
|
|
|
|
21
|
|
|
This module contains unit tests for abydos.stemmer.clef |
|
22
|
|
|
""" |
|
23
|
|
|
|
|
24
|
|
|
from __future__ import unicode_literals |
|
25
|
|
|
|
|
26
|
|
|
import unittest |
|
27
|
|
|
|
|
28
|
|
|
from abydos.stemmer.clef import clef_german, clef_german_plus, clef_swedish |
|
29
|
|
|
|
|
30
|
|
|
|
|
31
|
|
|
class CLEFTestCases(unittest.TestCase): |
|
32
|
|
|
"""Test CLEF functions. |
|
33
|
|
|
|
|
34
|
|
|
abydos.stemmer.clef_german, abydos.stemmer.clef_german_plus, & |
|
35
|
|
|
abydos.stemmer.clef_swedish |
|
36
|
|
|
""" |
|
37
|
|
|
|
|
38
|
|
|
def test_clef_german(self): |
|
39
|
|
|
"""Test abydos.stemmer.clef_german.""" |
|
40
|
|
|
# base case |
|
41
|
|
|
self.assertEqual(clef_german(''), '') |
|
42
|
|
|
|
|
43
|
|
|
# len <= 2 |
|
44
|
|
|
self.assertEqual(clef_german('ä'), 'a') |
|
45
|
|
|
self.assertEqual(clef_german('er'), 'er') |
|
46
|
|
|
self.assertEqual(clef_german('es'), 'es') |
|
47
|
|
|
self.assertEqual(clef_german('äh'), 'ah') |
|
48
|
|
|
|
|
49
|
|
|
# len > 2 |
|
50
|
|
|
self.assertEqual(clef_german('deinen'), 'dein') |
|
51
|
|
|
self.assertEqual(clef_german('können'), 'konn') |
|
52
|
|
|
self.assertEqual(clef_german('Damen'), 'dame') |
|
53
|
|
|
self.assertEqual(clef_german('kleines'), 'klein') |
|
54
|
|
|
self.assertEqual(clef_german('Namen'), 'name') |
|
55
|
|
|
self.assertEqual(clef_german('Äpfel'), 'apfel') |
|
56
|
|
|
self.assertEqual(clef_german('Jahre'), 'jahr') |
|
57
|
|
|
self.assertEqual(clef_german('Mannes'), 'mann') |
|
58
|
|
|
self.assertEqual(clef_german('Häuser'), 'haus') |
|
59
|
|
|
self.assertEqual(clef_german('Motoren'), 'motor') |
|
60
|
|
|
self.assertEqual(clef_german('kleine'), 'klein') |
|
61
|
|
|
self.assertEqual(clef_german('Pfingsten'), 'pfingst') |
|
62
|
|
|
self.assertEqual(clef_german('lautest'), 'lautest') |
|
63
|
|
|
self.assertEqual(clef_german('lauteste'), 'lautest') |
|
64
|
|
|
self.assertEqual(clef_german('lautere'), 'lauter') |
|
65
|
|
|
self.assertEqual(clef_german('lautste'), 'lautst') |
|
66
|
|
|
self.assertEqual(clef_german('kleinen'), 'klei') |
|
67
|
|
|
|
|
68
|
|
|
def test_clef_german_plus(self): |
|
69
|
|
|
"""Test abydos.stemmer.clef_german_plus.""" |
|
70
|
|
|
# base case |
|
71
|
|
|
self.assertEqual(clef_german_plus(''), '') |
|
72
|
|
|
|
|
73
|
|
|
# len <= 2 |
|
74
|
|
|
self.assertEqual(clef_german_plus('ä'), 'a') |
|
75
|
|
|
self.assertEqual(clef_german_plus('er'), 'er') |
|
76
|
|
|
self.assertEqual(clef_german_plus('es'), 'es') |
|
77
|
|
|
self.assertEqual(clef_german_plus('äh'), 'ah') |
|
78
|
|
|
|
|
79
|
|
|
# len > 2 |
|
80
|
|
|
self.assertEqual(clef_german_plus('deinen'), 'dein') |
|
81
|
|
|
self.assertEqual(clef_german_plus('können'), 'konn') |
|
82
|
|
|
self.assertEqual(clef_german_plus('Damen'), 'dam') |
|
83
|
|
|
self.assertEqual(clef_german_plus('kleines'), 'klein') |
|
84
|
|
|
self.assertEqual(clef_german_plus('Namen'), 'nam') |
|
85
|
|
|
self.assertEqual(clef_german_plus('Äpfel'), 'apfel') |
|
86
|
|
|
self.assertEqual(clef_german_plus('Jahre'), 'jahr') |
|
87
|
|
|
self.assertEqual(clef_german_plus('Mannes'), 'mann') |
|
88
|
|
|
self.assertEqual(clef_german_plus('Häuser'), 'haus') |
|
89
|
|
|
self.assertEqual(clef_german_plus('Motoren'), 'motor') |
|
90
|
|
|
self.assertEqual(clef_german_plus('kleine'), 'klein') |
|
91
|
|
|
self.assertEqual(clef_german_plus('Pfingsten'), 'pfing') |
|
92
|
|
|
self.assertEqual(clef_german_plus('lautest'), 'laut') |
|
93
|
|
|
self.assertEqual(clef_german_plus('lauteste'), 'laut') |
|
94
|
|
|
self.assertEqual(clef_german_plus('lautere'), 'laut') |
|
95
|
|
|
self.assertEqual(clef_german_plus('lautste'), 'laut') |
|
96
|
|
|
self.assertEqual(clef_german_plus('kleinen'), 'klein') |
|
97
|
|
|
self.assertEqual(clef_german_plus('Pfarrern'), 'pfarr') |
|
98
|
|
|
|
|
99
|
|
|
def test_clef_swedish(self): |
|
100
|
|
|
"""Test abydos.stemmer.clef_swedish.""" |
|
101
|
|
|
# base case |
|
102
|
|
|
self.assertEqual(clef_swedish(''), '') |
|
103
|
|
|
|
|
104
|
|
|
# unstemmed |
|
105
|
|
|
self.assertEqual(clef_swedish('konung'), 'konung') |
|
106
|
|
|
|
|
107
|
|
|
# len <= 3 |
|
108
|
|
|
self.assertEqual(clef_swedish('km'), 'km') |
|
109
|
|
|
self.assertEqual(clef_swedish('ja'), 'ja') |
|
110
|
|
|
self.assertEqual(clef_swedish('de'), 'de') |
|
111
|
|
|
self.assertEqual(clef_swedish('in'), 'in') |
|
112
|
|
|
self.assertEqual(clef_swedish('a'), 'a') |
|
113
|
|
|
self.assertEqual(clef_swedish('mer'), 'mer') |
|
114
|
|
|
self.assertEqual(clef_swedish('s'), 's') |
|
115
|
|
|
self.assertEqual(clef_swedish('e'), 'e') |
|
116
|
|
|
self.assertEqual(clef_swedish('oss'), 'oss') |
|
117
|
|
|
self.assertEqual(clef_swedish('hos'), 'hos') |
|
118
|
|
|
|
|
119
|
|
|
# genitive |
|
120
|
|
|
self.assertEqual(clef_swedish('svenskars'), 'svensk') |
|
121
|
|
|
self.assertEqual(clef_swedish('stadens'), 'stad') |
|
122
|
|
|
self.assertEqual(clef_swedish('kommuns'), 'kommu') |
|
123
|
|
|
self.assertEqual(clef_swedish('aftonbladets'), 'aftonblad') |
|
124
|
|
|
|
|
125
|
|
|
# len > 7 |
|
126
|
|
|
self.assertEqual(clef_swedish('fängelser'), 'fäng') |
|
127
|
|
|
self.assertEqual(clef_swedish('möjligheten'), 'möjlig') |
|
128
|
|
|
|
|
129
|
|
|
# len > 6 |
|
130
|
|
|
self.assertEqual(clef_swedish('svenskar'), 'svensk') |
|
131
|
|
|
self.assertEqual(clef_swedish('myndigheterna'), 'myndighet') |
|
132
|
|
|
self.assertEqual(clef_swedish('avgörande'), 'avgör') |
|
133
|
|
|
self.assertEqual(clef_swedish('fängelse'), 'fäng') |
|
134
|
|
|
self.assertEqual(clef_swedish('viktigaste'), 'viktig') |
|
135
|
|
|
self.assertEqual(clef_swedish('kvinnorna'), 'kvinn') |
|
136
|
|
|
self.assertEqual(clef_swedish('åklagaren'), 'åklag') |
|
137
|
|
|
|
|
138
|
|
|
# len > 5 |
|
139
|
|
|
self.assertEqual(clef_swedish('tidigare'), 'tidig') |
|
140
|
|
|
self.assertEqual(clef_swedish('senast'), 'sen') |
|
141
|
|
|
self.assertEqual(clef_swedish('möjlighet'), 'möjlig') |
|
142
|
|
|
|
|
143
|
|
|
# len > 4 |
|
144
|
|
|
self.assertEqual(clef_swedish('svenskar'), 'svensk') |
|
145
|
|
|
self.assertEqual(clef_swedish('skriver'), 'skriv') |
|
146
|
|
|
self.assertEqual(clef_swedish('människor'), 'människ') |
|
147
|
|
|
self.assertEqual(clef_swedish('staden'), 'stad') |
|
148
|
|
|
self.assertEqual(clef_swedish('kunnat'), 'kunn') |
|
149
|
|
|
self.assertEqual(clef_swedish('samarbete'), 'samarbe') |
|
150
|
|
|
self.assertEqual(clef_swedish('aftonbladet'), 'aftonblad') |
|
151
|
|
|
|
|
152
|
|
|
# len > 3 |
|
153
|
|
|
self.assertEqual(clef_swedish('allt'), 'all') |
|
154
|
|
|
self.assertEqual(clef_swedish('vilka'), 'vilk') |
|
155
|
|
|
self.assertEqual(clef_swedish('länge'), 'läng') |
|
156
|
|
|
self.assertEqual(clef_swedish('kommun'), 'kommu') |
|
157
|
|
|
|
|
158
|
|
|
|
|
159
|
|
|
if __name__ == '__main__': |
|
160
|
|
|
unittest.main() |
|
161
|
|
|
|