|
1
|
|
|
# -*- coding: utf-8 -*- |
|
2
|
|
|
|
|
3
|
|
|
# Copyright 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_phonetic_es. |
|
20
|
|
|
|
|
21
|
|
|
This module contains unit tests for abydos.phonetic.es |
|
22
|
|
|
""" |
|
23
|
|
|
|
|
24
|
|
|
from __future__ import unicode_literals |
|
25
|
|
|
|
|
26
|
|
|
import unittest |
|
27
|
|
|
|
|
28
|
|
|
from abydos.phonetic.es import phonetic_spanish, spanish_metaphone |
|
29
|
|
|
|
|
30
|
|
|
|
|
31
|
|
|
class PhoneticSpanishTestCases(unittest.TestCase): |
|
32
|
|
|
"""Test PhoneticSpanish functions. |
|
33
|
|
|
|
|
34
|
|
|
test cases for abydos.phonetic.phonetic_spanish |
|
35
|
|
|
""" |
|
36
|
|
|
|
|
37
|
|
|
def test_phonetic_spanish(self): |
|
38
|
|
|
"""Test abydos.phonetic.phonetic_spanish.""" |
|
39
|
|
|
# Base case |
|
40
|
|
|
self.assertEqual(phonetic_spanish(''), '') |
|
41
|
|
|
|
|
42
|
|
|
# Examples given in |
|
43
|
|
|
self.assertEqual(phonetic_spanish('Giraldo'), '8953') |
|
44
|
|
|
self.assertEqual(phonetic_spanish('Jiraldo'), '8953') |
|
45
|
|
|
self.assertEqual(phonetic_spanish('Halla'), '25') |
|
46
|
|
|
self.assertEqual(phonetic_spanish('Haya'), '25') |
|
47
|
|
|
self.assertEqual(phonetic_spanish('Cielo'), '45') |
|
48
|
|
|
self.assertEqual(phonetic_spanish('Sielo'), '45') |
|
49
|
|
|
|
|
50
|
|
|
# Test to maximize coverage |
|
51
|
|
|
self.assertEqual(phonetic_spanish('Giraldo', max_length=2), '89') |
|
52
|
|
|
|
|
53
|
|
|
|
|
54
|
|
|
class SpanishMetaphoneTestCases(unittest.TestCase): |
|
55
|
|
|
"""Test Spanish Metaphone functions. |
|
56
|
|
|
|
|
57
|
|
|
test cases for abydos.phonetic.spanish_metaphone |
|
58
|
|
|
""" |
|
59
|
|
|
|
|
60
|
|
|
def test_spanish_metaphone(self): |
|
61
|
|
|
"""Test abydos.phonetic.spanish_metaphone.""" |
|
62
|
|
|
# Base case |
|
63
|
|
|
self.assertEqual(spanish_metaphone(''), '') |
|
64
|
|
|
|
|
65
|
|
|
# Examples given in |
|
66
|
|
|
# https://github.com/amsqr/Spanish-Metaphone/blob/master/phonetic_algorithms_es.py |
|
67
|
|
|
self.assertEqual(spanish_metaphone('X'), 'X') |
|
68
|
|
|
self.assertEqual(spanish_metaphone('xplosion'), 'EXPLSN') |
|
69
|
|
|
self.assertEqual(spanish_metaphone('escalera'), 'ESKLR') |
|
70
|
|
|
self.assertEqual(spanish_metaphone('scalera'), 'ESKLR') |
|
71
|
|
|
self.assertEqual(spanish_metaphone('mi'), 'M') |
|
72
|
|
|
self.assertEqual(spanish_metaphone('tu'), 'T') |
|
73
|
|
|
self.assertEqual(spanish_metaphone('su'), 'S') |
|
74
|
|
|
self.assertEqual(spanish_metaphone('te'), 'T') |
|
75
|
|
|
self.assertEqual(spanish_metaphone('ochooomiiiillllllll'), 'OXMYY') |
|
76
|
|
|
self.assertEqual(spanish_metaphone('complicado'), 'KMPLKD') |
|
77
|
|
|
self.assertEqual(spanish_metaphone('ácaro'), 'AKR') |
|
78
|
|
|
self.assertEqual(spanish_metaphone('ácido'), 'AZD') |
|
79
|
|
|
self.assertEqual(spanish_metaphone('clown'), 'KLUN') |
|
80
|
|
|
self.assertEqual(spanish_metaphone('down'), 'DUN') |
|
81
|
|
|
self.assertEqual(spanish_metaphone('col'), 'KL') |
|
82
|
|
|
self.assertEqual(spanish_metaphone('clon'), 'KLN') |
|
83
|
|
|
self.assertEqual(spanish_metaphone('waterpolo'), 'UTRPL') |
|
84
|
|
|
self.assertEqual(spanish_metaphone('aquino'), 'AKN') |
|
85
|
|
|
self.assertEqual(spanish_metaphone('rebosar'), 'RVSR') |
|
86
|
|
|
self.assertEqual(spanish_metaphone('rebozar'), 'RVZR') |
|
87
|
|
|
self.assertEqual(spanish_metaphone('grajea'), 'GRJ') |
|
88
|
|
|
self.assertEqual(spanish_metaphone('gragea'), 'GRJ') |
|
89
|
|
|
self.assertEqual(spanish_metaphone('encima'), 'ENZM') |
|
90
|
|
|
self.assertEqual(spanish_metaphone('enzima'), 'ENZM') |
|
91
|
|
|
self.assertEqual(spanish_metaphone('alhamar'), 'ALAMR') |
|
92
|
|
|
self.assertEqual(spanish_metaphone('abollar'), 'AVYR') |
|
93
|
|
|
self.assertEqual(spanish_metaphone('aboyar'), 'AVYR') |
|
94
|
|
|
self.assertEqual(spanish_metaphone('huevo'), 'UV') |
|
95
|
|
|
self.assertEqual(spanish_metaphone('webo'), 'UV') |
|
96
|
|
|
self.assertEqual(spanish_metaphone('macho'), 'MX') |
|
97
|
|
|
self.assertEqual(spanish_metaphone('xocolate'), 'XKLT') |
|
98
|
|
|
self.assertEqual(spanish_metaphone('chocolate'), 'XKLT') |
|
99
|
|
|
self.assertEqual(spanish_metaphone('axioma'), 'AXM') |
|
100
|
|
|
self.assertEqual(spanish_metaphone('abedul'), 'AVDL') |
|
101
|
|
|
self.assertEqual(spanish_metaphone('a'), 'A') |
|
102
|
|
|
self.assertEqual(spanish_metaphone('gengibre'), 'JNJVR') |
|
103
|
|
|
self.assertEqual(spanish_metaphone('yema'), 'YM') |
|
104
|
|
|
self.assertEqual(spanish_metaphone('wHISKY'), 'UISKY') |
|
105
|
|
|
self.assertEqual(spanish_metaphone('google'), 'GGL') |
|
106
|
|
|
self.assertEqual(spanish_metaphone('xilófono'), 'XLFN') |
|
107
|
|
|
self.assertEqual(spanish_metaphone('web'), 'UV') |
|
108
|
|
|
self.assertEqual(spanish_metaphone('guerra'), 'GRR') |
|
109
|
|
|
self.assertEqual(spanish_metaphone('pingüino'), 'PNUN') |
|
110
|
|
|
self.assertEqual(spanish_metaphone('si'), 'S') |
|
111
|
|
|
self.assertEqual(spanish_metaphone('ke'), 'K') |
|
112
|
|
|
self.assertEqual(spanish_metaphone('que'), 'K') |
|
113
|
|
|
self.assertEqual(spanish_metaphone('tu'), 'T') |
|
114
|
|
|
self.assertEqual(spanish_metaphone('gato'), 'GT') |
|
115
|
|
|
self.assertEqual(spanish_metaphone('gitano'), 'JTN') |
|
116
|
|
|
self.assertEqual(spanish_metaphone('queso'), 'KS') |
|
117
|
|
|
self.assertEqual(spanish_metaphone('paquete'), 'PKT') |
|
118
|
|
|
self.assertEqual(spanish_metaphone('cuco'), 'KK') |
|
119
|
|
|
self.assertEqual(spanish_metaphone('perro'), 'PRR') |
|
120
|
|
|
self.assertEqual(spanish_metaphone('pero'), 'PR') |
|
121
|
|
|
self.assertEqual(spanish_metaphone('arrebato'), 'ARRVT') |
|
122
|
|
|
self.assertEqual(spanish_metaphone('hola'), 'OL') |
|
123
|
|
|
self.assertEqual(spanish_metaphone('zapato'), 'ZPT') |
|
124
|
|
|
self.assertEqual(spanish_metaphone('españa'), 'ESPNY') |
|
125
|
|
|
self.assertEqual(spanish_metaphone('garrulo'), 'GRRL') |
|
126
|
|
|
self.assertEqual(spanish_metaphone('expansión'), 'EXPNSN') |
|
127
|
|
|
self.assertEqual(spanish_metaphone('membrillo'), 'MMVRY') |
|
128
|
|
|
self.assertEqual(spanish_metaphone('jamón'), 'JMN') |
|
129
|
|
|
self.assertEqual(spanish_metaphone('risa'), 'RS') |
|
130
|
|
|
self.assertEqual(spanish_metaphone('caricia'), 'KRZ') |
|
131
|
|
|
self.assertEqual(spanish_metaphone('llaves'), 'YVS') |
|
132
|
|
|
self.assertEqual(spanish_metaphone('paella'), 'PY') |
|
133
|
|
|
self.assertEqual(spanish_metaphone('cerilla'), 'ZRY') |
|
134
|
|
|
|
|
135
|
|
|
# tests from file:///home/chrislit/Downloads/ICTRS_2016_12.pdf |
|
136
|
|
|
# including of the modified version of the algorithm |
|
137
|
|
|
self.assertEqual(spanish_metaphone('Caricia'), 'KRZ') |
|
138
|
|
|
self.assertEqual(spanish_metaphone('Caricia', modified=True), 'KRZ') |
|
139
|
|
|
self.assertEqual(spanish_metaphone('Llaves'), 'YVS') |
|
140
|
|
|
self.assertEqual(spanish_metaphone('Llaves', modified=True), 'YVZ') |
|
141
|
|
|
self.assertEqual(spanish_metaphone('Paella'), 'PY') |
|
142
|
|
|
self.assertEqual(spanish_metaphone('Paella', modified=True), 'PY') |
|
143
|
|
|
self.assertEqual(spanish_metaphone('Cerilla'), 'ZRY') |
|
144
|
|
|
self.assertEqual(spanish_metaphone('Cerilla', modified=True), 'ZRY') |
|
145
|
|
|
self.assertEqual(spanish_metaphone('Empeorar'), 'EMPRR') |
|
146
|
|
|
self.assertEqual(spanish_metaphone('Empeorar', modified=True), 'ENPRR') |
|
147
|
|
|
self.assertEqual(spanish_metaphone('Embotellar'), 'EMVTYR') |
|
148
|
|
|
self.assertEqual(spanish_metaphone('Embotellar', modified=True), |
|
149
|
|
|
'ENVTYR') |
|
150
|
|
|
self.assertEqual(spanish_metaphone('Hoy'), 'OY') |
|
151
|
|
|
self.assertEqual(spanish_metaphone('Hoy', modified=True), 'OY') |
|
152
|
|
|
self.assertEqual(spanish_metaphone('Xochimilco'), 'XXMLK') |
|
153
|
|
|
self.assertEqual(spanish_metaphone('Xochimilco', modified=True), |
|
154
|
|
|
'XXMLK') |
|
155
|
|
|
self.assertEqual(spanish_metaphone('Psiquiatra'), 'PSKTR') |
|
156
|
|
|
self.assertEqual(spanish_metaphone('Psiquiatra', modified=True), |
|
157
|
|
|
'ZKTR') |
|
158
|
|
|
self.assertEqual(spanish_metaphone('siquiatra'), 'SKTR') |
|
159
|
|
|
self.assertEqual(spanish_metaphone('siquiatra', modified=True), 'ZKTR') |
|
160
|
|
|
self.assertEqual(spanish_metaphone('Obscuro'), 'OVSKR') |
|
161
|
|
|
self.assertEqual(spanish_metaphone('Obscuro', modified=True), 'OZKR') |
|
162
|
|
|
self.assertEqual(spanish_metaphone('Oscuro'), 'OSKR') |
|
163
|
|
|
self.assertEqual(spanish_metaphone('Oscuro', modified=True), 'OZKR') |
|
164
|
|
|
self.assertEqual(spanish_metaphone('Combate'), 'KMVT') |
|
165
|
|
|
self.assertEqual(spanish_metaphone('Combate', modified=True), 'KNVT') |
|
166
|
|
|
self.assertEqual(spanish_metaphone('Convate'), 'KNVT') |
|
167
|
|
|
self.assertEqual(spanish_metaphone('Convate', modified=True), 'KNVT') |
|
168
|
|
|
self.assertEqual(spanish_metaphone('Conbate'), 'KNVT') |
|
169
|
|
|
self.assertEqual(spanish_metaphone('Conbate', modified=True), 'KNVT') |
|
170
|
|
|
self.assertEqual(spanish_metaphone('Comportar'), 'KMPRTR') |
|
171
|
|
|
self.assertEqual(spanish_metaphone('Comportar', modified=True), |
|
172
|
|
|
'KNPRTR') |
|
173
|
|
|
self.assertEqual(spanish_metaphone('Conportar'), 'KNPRTR') |
|
174
|
|
|
self.assertEqual(spanish_metaphone('Conportar', modified=True), |
|
175
|
|
|
'KNPRTR') |
|
176
|
|
|
self.assertEqual(spanish_metaphone('Zapato'), 'ZPT') |
|
177
|
|
|
self.assertEqual(spanish_metaphone('Zapato', modified=True), 'ZPT') |
|
178
|
|
|
self.assertEqual(spanish_metaphone('Sapato'), 'SPT') |
|
179
|
|
|
self.assertEqual(spanish_metaphone('Sapato', modified=True), 'ZPT') |
|
180
|
|
|
self.assertEqual(spanish_metaphone('Escalera'), 'ESKLR') |
|
181
|
|
|
self.assertEqual(spanish_metaphone('Escalera', modified=True), 'EZKLR') |
|
182
|
|
|
self.assertEqual(spanish_metaphone('scalera'), 'ESKLR') |
|
183
|
|
|
self.assertEqual(spanish_metaphone('scalera', modified=True), 'EZKLR') |
|
184
|
|
|
|
|
185
|
|
|
# terms from algorithm/source |
|
186
|
|
|
self.assertEqual(spanish_metaphone('acción'), 'AXN') |
|
187
|
|
|
self.assertEqual(spanish_metaphone('reacción'), 'RXN') |
|
188
|
|
|
self.assertEqual(spanish_metaphone('cesar'), 'ZSR') |
|
189
|
|
|
self.assertEqual(spanish_metaphone('cien'), 'ZN') |
|
190
|
|
|
self.assertEqual(spanish_metaphone('cid'), 'ZD') |
|
191
|
|
|
self.assertEqual(spanish_metaphone('conciencia'), 'KNZNZ') |
|
192
|
|
|
self.assertEqual(spanish_metaphone('gente'), 'JNT') |
|
193
|
|
|
self.assertEqual(spanish_metaphone('ecologia'), 'EKLJ') |
|
194
|
|
|
|
|
195
|
|
|
# completing coverage |
|
196
|
|
|
self.assertEqual(spanish_metaphone('hola'), 'OL') |
|
197
|
|
|
self.assertEqual(spanish_metaphone('aqi'), 'AK') |
|
198
|
|
|
self.assertEqual(spanish_metaphone('hjordis'), 'HJRDS') |
|
199
|
|
|
|
|
200
|
|
|
|
|
201
|
|
|
if __name__ == '__main__': |
|
202
|
|
|
unittest.main() |
|
203
|
|
|
|