tests.distance.test_distance_synoname.SynonameTestCases.test_synoname() - Code Metrics - Inspection of "87ccc137f1c4bdd6a747877328db37ccf7a1299d" - chrislit/abydos - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Branch — master (87ccc1)

by Chris

created 2018-10-23 01:27 UTC

tests.distance.test_distance_synoname.SynonameTestCases.test_synoname() B

↳ Parent: Project

Complexity

Conditions

Size

Total Lines	120
Code Lines	109

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	1
eloc	109
nop	1
dl	0
loc	120
rs	7
c	0
b	0
f	0

How to fix Long Method

# -*- coding: utf-8 -*-

# Copyright 2018 by Christopher C. Little.
# This file is part of Abydos.
#
# Abydos is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Abydos is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Abydos. If not, see <http://www.gnu.org/licenses/>.

"""abydos.tests.test_distance.synoname.

This module contains unit tests for abydos.distance.synoname
"""

from __future__ import division, unicode_literals

import unittest

from abydos.distance.synoname import _synoname_strip_punct, \
    _synoname_word_approximation, synoname


class SynonameTestCases(unittest.TestCase):
    """Test Synoname functions.

    abydos.distance._synoname_strip_punct, _synoname_word_approximation, &
    synoname
    """

    def test_synoname_strip_punct(self):
        """Test abydos.distance._synoname_strip_punct."""
        # Base cases
        self.assertEqual(_synoname_strip_punct(''), '')
        self.assertEqual(_synoname_strip_punct('abcdefg'), 'abcdefg')
        self.assertEqual(_synoname_strip_punct('a\'b-c,d!e:f%g'), 'abcdefg')

    def test_synoname_word_approximation(self):
        """Test abydos.distance._synoname_word_approximation."""
        # Base cases
        self.assertEqual(_synoname_word_approximation('', ''), 0)

        self.assertEqual(
            _synoname_word_approximation('di Domenico di Bonaventura',
                                         'di Tomme di Nuto',
                                         'Cosimo', 'Luca'), 0.4)
        self.assertEqual(
            _synoname_word_approximation('Antonello da Messina',
                                         'Messina', '', 'Antonello da',
                                         {'gen_conflict': False,
                                          'roman_conflict': False,
                                          'src_specials':
                                              [(35, 'b'), (35, 'c')],
                                          'tar_specials':
                                              [(35, 'b'), (35, 'c')]}), 0)
        self.assertEqual(
            _synoname_word_approximation('louis ii', 'louis ii',
                                         'sr jean', 'sr  pierre',
                                         {'gen_conflict': False,
                                          'roman_conflict': False,
                                          'src_specials':
                                              [(49, 'b'), (68, 'd'),
                                               (121, 'b')],
                                          'tar_specials':
                                              [(49, 'b'), (68, 'd'),
                                               (121, 'b')]}), 0)
        self.assertEqual(
            _synoname_word_approximation('louis ii', 'louis ii',
                                         'il giovane', 'sr cadet',
                                         {'gen_conflict': False,
                                          'roman_conflict': False,
                                          'src_specials':
                                              [(46, 'a'), (49, 'b'),
                                               (52, 'a'), (68, 'd')],
                                          'tar_specials':
                                              [(8, 'a'), (49, 'b'),
                                               (68, 'd'), (121, 'a')]}), 1)
        self.assertAlmostEqual(
            _synoname_word_approximation('louis ii', 'louis ii',
                                         'ste.-geo ste.', 'ste.-jo ste.',
                                         {'gen_conflict': False,
                                          'roman_conflict': False,
                                          'src_specials':
                                              [(49, 'b'), (68, 'd'),
                                               (127, 'b'), (127, 'X')],
                                          'tar_specials':
                                              [(49, 'b'), (68, 'd'),
                                               (127, 'b'), (127, 'X')]}), 2/3)
        self.assertAlmostEqual(
            _synoname_word_approximation('louis ii', 'louis',
                                         'ste.-geo ste.', '',
                                         {'gen_conflict': False,
                                          'roman_conflict': False,
                                          'src_specials':
                                              [(49, 'b'), (68, 'd'),
                                               (127, 'b'), (127, 'X')],
                                          'tar_specials': []}), 0)
        self.assertAlmostEqual(
            _synoname_word_approximation('lou ii', 'louis', 'louis iv', 'ste.',
                                         {}), 0)
        self.assertEqual(
            _synoname_word_approximation('ren', 'loren ste.', '', '',
                                         {'tar_specials': [(68, 'd'),
                                                           (127, 'X')],
                                          'src_specials': [(0, '')]}), 1)

    def test_synoname(self):
        """Test abydos.distance.synoname."""
        # Base cases
        self.assertEqual(synoname('', ''), 1)
        self.assertEqual(synoname('', '', tests=['exact']), 1)
        self.assertEqual(synoname('', '', tests=[]), 13)
        self.assertEqual(synoname('', '', tests=['nonsense-test']), 13)
        self.assertEqual(synoname('', '', ret_name=True), 'exact')

        # Test input formats
        self.assertEqual(synoname(('Brueghel II (the Younger)', 'Pieter',
                                   'Workshop of'),
                                  ('Brueghel II (the Younger)', 'Pieter',
                                   'Workshop of')), 1)
        self.assertEqual(synoname('Brueghel II (the Younger)#Pieter#' +
                                  'Workshop of',
                                  'Brueghel II (the Younger)#Pieter#' +
                                  'Workshop of'), 1)
        self.assertEqual(synoname('22#Brueghel II (the Younger)#Pieter#' +
                                  'Workshop of',
                                  '44#Brueghel II (the Younger)#Pieter#' +
                                  'Workshop of'), 1)

        # approx_c tests
        self.assertEqual(synoname(('Master of Brueghel II (the Younger)',
                                   'Pieter', 'Workshop of'),
                                  ('Brueghel I (the Elder)', 'Pieter',
                                   'Workshop of')), 13)
        self.assertEqual(synoname(('Master of Brueghel II',
                                   'Pieter', 'Workshop of'),
                                  ('Master known as the Brueghel II', 'Pieter',
                                   'Workshop of')), 10)

        # Types 1-12
        self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
                                  ('Brueghel', 'Pieter', ''),
                                  ret_name=True), 'exact')

        self.assertEqual(synoname(('Brueghel II', 'Pieter', ''),
                                  ('Brueghel I', 'Pieter', ''),
                                  ret_name=True), 'no_match')
        self.assertEqual(synoname(('Breghel', 'Pieter', ''),
                                  ('Brueghel', 'Pieter', ''),
                                  ret_name=True), 'omission')
        self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
                                  ('Breghel', 'Pieter', ''),
                                  ret_name=True), 'omission')
        self.assertEqual(synoname(('Brueghel', 'Piter', ''),
                                  ('Brueghel', 'Pieter', ''),
                                  ret_name=True), 'omission')
        self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
                                  ('Brueghel', 'Piter', ''),
                                  ret_name=True), 'omission')
        self.assertEqual(synoname(('Brughel', 'Pieter', ''),
                                  ('Breghel', 'Pieter', ''),
                                  ret_name=True), 'substitution')
        self.assertEqual(synoname(('Breughel', 'Peter', ''),
                                  ('Breughel', 'Piter', ''),
                                  ret_name=True), 'substitution')
        self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
                                  ('Breughel', 'Pieter', ''),
                                  ret_name=True), 'transposition')
        self.assertEqual(synoname(('Brueghel', 'Peiter', ''),
                                  ('Brueghel', 'Pieter', ''),
                                  ret_name=True), 'transposition')

        self.assertEqual(synoname(('Brueghel:', 'Pieter', ''),
                                  ('Brueghel', 'Pi-eter', ''),
                                  ret_name=True), 'punctuation')
        self.assertEqual(synoname(('Brueghel,', 'Pieter', ''),
                                  ('Brueghel', 'Pieter...', ''),
                                  ret_name=True), 'punctuation')
        self.assertEqual(synoname(('Seu rat', 'George Pierre', ''),
                                  ('Seu-rat', 'George-Pierre', ''),
                                  ret_name=True), 'punctuation')
        self.assertEqual(synoname(('Picasso', '', ''),
                                  ('Picasso', 'Pablo', ''),
                                  ret_name=True), 'no_first')
        self.assertEqual(synoname(('Pereira', 'I. R.', ''),
                                  ('Pereira', 'Irene Rice', ''),
                                  ret_name=True), 'initials')
        self.assertEqual(synoname(('Pereira', 'I.', ''),
                                  ('Pereira', 'Irene Rice', ''),
                                  ret_name=True), 'initials')
        self.assertNotEqual(synoname(('Pereira', 'I. R.', ''),
                                     ('Pereira', 'I. Smith', ''),
                                     ret_name=True), 'initials')
        self.assertNotEqual(synoname(('Pereira', 'I. R. S.', ''),
                                     ('Pereira', 'I. S. R.', ''),
                                     ret_name=True), 'initials')
        self.assertEqual(synoname(('de Goya', 'Francisco', ''),
                                  ('de Goya y Lucientes', 'Francisco', ''),
                                  ret_name=True), 'extension')
        self.assertEqual(synoname(('Seurat', 'George', ''),
                                  ('Seurat', 'George-Pierre', ''),
                                  ret_name=True), 'extension')
        self.assertEqual(synoname(('Gericault', 'Theodore', ''),
                                  ('Gericault', 'Jean Louis Andre Theodore',
                                   ''),
                                  ret_name=True), 'inclusion')
        self.assertEqual(synoname(('Dore', 'Gustave', ''),
                                  ('Dore', 'Paul Gustave Louis Christophe',
                                   ''),
                                  ret_name=True), 'inclusion')

        self.assertEqual(synoname(('Rosetti', 'Dante Gabriel', ''),
                                  ('Rosetti', 'Gabriel Charles Dante', ''),
                                  ret_name=True), 'word_approx')
        self.assertEqual(synoname(('di Domenico di Bonaventura', 'Cosimo', ''),
                                  ('di Tomme di Nuto', 'Luca', ''),
                                  ret_name=True), 'no_match')
        self.assertEqual(synoname(('Pereira', 'I. R.', ''),
                                  ('Pereira', 'I. Smith', ''),
                                  ret_name=True), 'word_approx')
        self.assertEqual(synoname(('Antonello da Messina', '', ''),
                                  ('Messina', 'Antonello da', ''),
                                  ret_name=True), 'confusions')
        self.assertEqual(synoname(('Brueghel', 'Pietter', ''),
                                  ('Bruegghel', 'Pieter', ''),
                                  ret_name=True), 'char_approx')


if __name__ == '__main__':
    unittest.main()


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_distance.synoname.
20
21			This module contains unit tests for abydos.distance.synoname
22			"""
23
24			from __future__ import division, unicode_literals
25
26			import unittest
27
28			from abydos.distance.synoname import _synoname_strip_punct, \
29			_synoname_word_approximation, synoname
30
31
32			class SynonameTestCases(unittest.TestCase):
33			"""Test Synoname functions.
34
35			abydos.distance._synoname_strip_punct, _synoname_word_approximation, &
36			synoname
37			"""
38
39			def test_synoname_strip_punct(self):
40			"""Test abydos.distance._synoname_strip_punct."""
41			# Base cases
42			self.assertEqual(_synoname_strip_punct(''), '')
43			self.assertEqual(_synoname_strip_punct('abcdefg'), 'abcdefg')
44			self.assertEqual(_synoname_strip_punct('a\'b-c,d!e:f%g'), 'abcdefg')
45
46			def test_synoname_word_approximation(self):
47			"""Test abydos.distance._synoname_word_approximation."""
48			# Base cases
49			self.assertEqual(_synoname_word_approximation('', ''), 0)
50
51			self.assertEqual(
52			_synoname_word_approximation('di Domenico di Bonaventura',
53			'di Tomme di Nuto',
54			'Cosimo', 'Luca'), 0.4)
55			self.assertEqual(
56			_synoname_word_approximation('Antonello da Messina',
57			'Messina', '', 'Antonello da',
58			{'gen_conflict': False,
59			'roman_conflict': False,
60			'src_specials':
61			[(35, 'b'), (35, 'c')],
62			'tar_specials':
63			[(35, 'b'), (35, 'c')]}), 0)
64			self.assertEqual(
65			_synoname_word_approximation('louis ii', 'louis ii',
66			'sr jean', 'sr pierre',
67			{'gen_conflict': False,
68			'roman_conflict': False,
69			'src_specials':
70			[(49, 'b'), (68, 'd'),
71			(121, 'b')],
72			'tar_specials':
73			[(49, 'b'), (68, 'd'),
74			(121, 'b')]}), 0)
75			self.assertEqual(
76			_synoname_word_approximation('louis ii', 'louis ii',
77			'il giovane', 'sr cadet',
78			{'gen_conflict': False,
79			'roman_conflict': False,
80			'src_specials':
81			[(46, 'a'), (49, 'b'),
82			(52, 'a'), (68, 'd')],
83			'tar_specials':
84			[(8, 'a'), (49, 'b'),
85			(68, 'd'), (121, 'a')]}), 1)
86			self.assertAlmostEqual(
87			_synoname_word_approximation('louis ii', 'louis ii',
88			'ste.-geo ste.', 'ste.-jo ste.',
89			{'gen_conflict': False,
90			'roman_conflict': False,
91			'src_specials':
92			[(49, 'b'), (68, 'd'),
93			(127, 'b'), (127, 'X')],
94			'tar_specials':
95			[(49, 'b'), (68, 'd'),
96			(127, 'b'), (127, 'X')]}), 2/3)
97			self.assertAlmostEqual(
98			_synoname_word_approximation('louis ii', 'louis',
99			'ste.-geo ste.', '',
100			{'gen_conflict': False,
101			'roman_conflict': False,
102			'src_specials':
103			[(49, 'b'), (68, 'd'),
104			(127, 'b'), (127, 'X')],
105			'tar_specials': []}), 0)
106			self.assertAlmostEqual(
107			_synoname_word_approximation('lou ii', 'louis', 'louis iv', 'ste.',
108			{}), 0)
109			self.assertEqual(
110			_synoname_word_approximation('ren', 'loren ste.', '', '',
111			{'tar_specials': [(68, 'd'),
112			(127, 'X')],
113			'src_specials': [(0, '')]}), 1)
114
115			def test_synoname(self):
116			"""Test abydos.distance.synoname."""
117			# Base cases
118			self.assertEqual(synoname('', ''), 1)
119			self.assertEqual(synoname('', '', tests=['exact']), 1)
120			self.assertEqual(synoname('', '', tests=[]), 13)
121			self.assertEqual(synoname('', '', tests=['nonsense-test']), 13)
122			self.assertEqual(synoname('', '', ret_name=True), 'exact')
123
124			# Test input formats
125			self.assertEqual(synoname(('Brueghel II (the Younger)', 'Pieter',
126			'Workshop of'),
127			('Brueghel II (the Younger)', 'Pieter',
128			'Workshop of')), 1)
129			self.assertEqual(synoname('Brueghel II (the Younger)#Pieter#' +
130			'Workshop of',
131			'Brueghel II (the Younger)#Pieter#' +
132			'Workshop of'), 1)
133			self.assertEqual(synoname('22#Brueghel II (the Younger)#Pieter#' +
134			'Workshop of',
135			'44#Brueghel II (the Younger)#Pieter#' +
136			'Workshop of'), 1)
137
138			# approx_c tests
139			self.assertEqual(synoname(('Master of Brueghel II (the Younger)',
140			'Pieter', 'Workshop of'),
141			('Brueghel I (the Elder)', 'Pieter',
142			'Workshop of')), 13)
143			self.assertEqual(synoname(('Master of Brueghel II',
144			'Pieter', 'Workshop of'),
145			('Master known as the Brueghel II', 'Pieter',
146			'Workshop of')), 10)
147
148			# Types 1-12
149			self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
150			('Brueghel', 'Pieter', ''),
151			ret_name=True), 'exact')
152
153			self.assertEqual(synoname(('Brueghel II', 'Pieter', ''),
154			('Brueghel I', 'Pieter', ''),
155			ret_name=True), 'no_match')
156			self.assertEqual(synoname(('Breghel', 'Pieter', ''),
157			('Brueghel', 'Pieter', ''),
158			ret_name=True), 'omission')
159			self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
160			('Breghel', 'Pieter', ''),
161			ret_name=True), 'omission')
162			self.assertEqual(synoname(('Brueghel', 'Piter', ''),
163			('Brueghel', 'Pieter', ''),
164			ret_name=True), 'omission')
165			self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
166			('Brueghel', 'Piter', ''),
167			ret_name=True), 'omission')
168			self.assertEqual(synoname(('Brughel', 'Pieter', ''),
169			('Breghel', 'Pieter', ''),
170			ret_name=True), 'substitution')
171			self.assertEqual(synoname(('Breughel', 'Peter', ''),
172			('Breughel', 'Piter', ''),
173			ret_name=True), 'substitution')
174			self.assertEqual(synoname(('Brueghel', 'Pieter', ''),
175			('Breughel', 'Pieter', ''),
176			ret_name=True), 'transposition')
177			self.assertEqual(synoname(('Brueghel', 'Peiter', ''),
178			('Brueghel', 'Pieter', ''),
179			ret_name=True), 'transposition')
180
181			self.assertEqual(synoname(('Brueghel:', 'Pieter', ''),
182			('Brueghel', 'Pi-eter', ''),
183			ret_name=True), 'punctuation')
184			self.assertEqual(synoname(('Brueghel,', 'Pieter', ''),
185			('Brueghel', 'Pieter...', ''),
186			ret_name=True), 'punctuation')
187			self.assertEqual(synoname(('Seu rat', 'George Pierre', ''),
188			('Seu-rat', 'George-Pierre', ''),
189			ret_name=True), 'punctuation')
190			self.assertEqual(synoname(('Picasso', '', ''),
191			('Picasso', 'Pablo', ''),
192			ret_name=True), 'no_first')
193			self.assertEqual(synoname(('Pereira', 'I. R.', ''),
194			('Pereira', 'Irene Rice', ''),
195			ret_name=True), 'initials')
196			self.assertEqual(synoname(('Pereira', 'I.', ''),
197			('Pereira', 'Irene Rice', ''),
198			ret_name=True), 'initials')
199			self.assertNotEqual(synoname(('Pereira', 'I. R.', ''),
200			('Pereira', 'I. Smith', ''),
201			ret_name=True), 'initials')
202			self.assertNotEqual(synoname(('Pereira', 'I. R. S.', ''),
203			('Pereira', 'I. S. R.', ''),
204			ret_name=True), 'initials')
205			self.assertEqual(synoname(('de Goya', 'Francisco', ''),
206			('de Goya y Lucientes', 'Francisco', ''),
207			ret_name=True), 'extension')
208			self.assertEqual(synoname(('Seurat', 'George', ''),
209			('Seurat', 'George-Pierre', ''),
210			ret_name=True), 'extension')
211			self.assertEqual(synoname(('Gericault', 'Theodore', ''),
212			('Gericault', 'Jean Louis Andre Theodore',
213			''),
214			ret_name=True), 'inclusion')
215			self.assertEqual(synoname(('Dore', 'Gustave', ''),
216			('Dore', 'Paul Gustave Louis Christophe',
217			''),
218			ret_name=True), 'inclusion')
219
220			self.assertEqual(synoname(('Rosetti', 'Dante Gabriel', ''),
221			('Rosetti', 'Gabriel Charles Dante', ''),
222			ret_name=True), 'word_approx')
223			self.assertEqual(synoname(('di Domenico di Bonaventura', 'Cosimo', ''),
224			('di Tomme di Nuto', 'Luca', ''),
225			ret_name=True), 'no_match')
226			self.assertEqual(synoname(('Pereira', 'I. R.', ''),
227			('Pereira', 'I. Smith', ''),
228			ret_name=True), 'word_approx')
229			self.assertEqual(synoname(('Antonello da Messina', '', ''),
230			('Messina', 'Antonello da', ''),
231			ret_name=True), 'confusions')
232			self.assertEqual(synoname(('Brueghel', 'Pietter', ''),
233			('Bruegghel', 'Pieter', ''),
234			ret_name=True), 'char_approx')
235
236
237			if __name__ == '__main__':
238			unittest.main()
239

chrislit / abydos

Branch — master (87ccc1)

tests.distance.test_distance_synoname.SynonameTestCases.test_synoname() B

Complexity

Size

Duplication

Importance

How to fix Long Method

Long Method

Duplication Side-by-Side

Filter issues like