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
|
1 |
|
"""abydos.distance.synoname. |
20
|
|
|
|
21
|
|
|
The distance.synoname module implements Synoname. |
22
|
|
|
""" |
23
|
|
|
|
24
|
1 |
|
from __future__ import division, unicode_literals |
25
|
|
|
|
26
|
1 |
|
from collections import Iterable |
27
|
|
|
|
28
|
1 |
|
from ._distance import Distance |
29
|
1 |
|
from ._levenshtein import levenshtein |
30
|
1 |
|
from ._sequence import sim_ratcliff_obershelp |
31
|
|
|
|
32
|
|
|
# noinspection PyProtectedMember |
33
|
1 |
|
from ..fingerprint._synoname import SynonameToolcode |
34
|
|
|
|
35
|
1 |
|
__all__ = ['Synoname', 'synoname'] |
36
|
|
|
|
37
|
|
|
|
38
|
1 |
|
class Synoname(Distance): |
|
|
|
|
39
|
|
|
"""Synoname. |
40
|
|
|
|
41
|
|
|
Cf. :cite:`Getty:1991,Gross:1991` |
42
|
|
|
""" |
43
|
|
|
|
44
|
1 |
|
stc = SynonameToolcode() |
45
|
|
|
|
46
|
1 |
|
test_dict = { |
47
|
|
|
val: 2 ** n |
48
|
|
|
for n, val in enumerate( |
49
|
|
|
( |
50
|
|
|
'exact', |
51
|
|
|
'omission', |
52
|
|
|
'substitution', |
53
|
|
|
'transposition', |
54
|
|
|
'punctuation', |
55
|
|
|
'initials', |
56
|
|
|
'extension', |
57
|
|
|
'inclusion', |
58
|
|
|
'no_first', |
59
|
|
|
'word_approx', |
60
|
|
|
'confusions', |
61
|
|
|
'char_approx', |
62
|
|
|
) |
63
|
|
|
) |
64
|
|
|
} |
65
|
1 |
|
match_name = ( |
66
|
|
|
'', |
67
|
|
|
'exact', |
68
|
|
|
'omission', |
69
|
|
|
'substitution', |
70
|
|
|
'transposition', |
71
|
|
|
'punctuation', |
72
|
|
|
'initials', |
73
|
|
|
'extension', |
74
|
|
|
'inclusion', |
75
|
|
|
'no_first', |
76
|
|
|
'word_approx', |
77
|
|
|
'confusions', |
78
|
|
|
'char_approx', |
79
|
|
|
'no_match', |
80
|
|
|
) |
81
|
1 |
|
match_type_dict = {val: n for n, val in enumerate(match_name)} |
82
|
|
|
|
83
|
1 |
|
def _synoname_strip_punct(self, word): |
|
|
|
|
84
|
|
|
"""Return a word with punctuation stripped out. |
85
|
|
|
|
86
|
|
|
Args: |
87
|
|
|
word (str): a word to strip punctuation from |
88
|
|
|
|
89
|
|
|
Returns: |
90
|
|
|
str: The word stripped of punctuation |
91
|
|
|
|
92
|
|
|
Examples: |
93
|
|
|
>>> pe = Synoname() |
94
|
|
|
>>> pe._synoname_strip_punct('AB;CD EF-GH$IJ') |
95
|
|
|
'ABCD EFGHIJ' |
96
|
|
|
|
97
|
|
|
""" |
98
|
1 |
|
stripped = '' |
99
|
1 |
|
for char in word: |
100
|
1 |
|
if char not in set(',-./:;"&\'()!{|}?$%*+<=>[\\]^_`~'): |
101
|
1 |
|
stripped += char |
102
|
1 |
|
return stripped.strip() |
103
|
|
|
|
104
|
1 |
|
def _synoname_word_approximation( |
|
|
|
|
105
|
|
|
self, src_ln, tar_ln, src_fn='', tar_fn='', features=None |
|
|
|
|
106
|
|
|
): |
107
|
|
|
"""Return the Synoname word approximation score for two names. |
108
|
|
|
|
109
|
|
|
Args: |
110
|
|
|
src_ln (str): Last name of the source |
111
|
|
|
tar_ln (str): Last name of the target |
112
|
|
|
src_fn (str): First name of the source (optional) |
113
|
|
|
tar_fn (str): First name of the target (optional) |
114
|
|
|
features (dict): A dict containing special features calculated |
115
|
|
|
using :py:class:`fingerprint.SynonameToolcode` (optional) |
116
|
|
|
|
117
|
|
|
Returns: |
118
|
|
|
float: The word approximation score |
119
|
|
|
|
120
|
|
|
Examples: |
121
|
|
|
>>> pe = Synoname() |
122
|
|
|
>>> pe._synoname_word_approximation('Smith Waterman', 'Waterman', |
123
|
|
|
... 'Tom Joe Bob', 'Tom Joe') |
124
|
|
|
0.6 |
125
|
|
|
|
126
|
|
|
""" |
127
|
1 |
|
if features is None: |
128
|
1 |
|
features = {} |
129
|
1 |
|
if 'src_specials' not in features: |
130
|
1 |
|
features['src_specials'] = [] |
131
|
1 |
|
if 'tar_specials' not in features: |
132
|
1 |
|
features['tar_specials'] = [] |
133
|
|
|
|
134
|
1 |
|
src_len_specials = len(features['src_specials']) |
135
|
1 |
|
tar_len_specials = len(features['tar_specials']) |
136
|
|
|
|
137
|
|
|
# 1 |
138
|
1 |
|
if ('gen_conflict' in features and features['gen_conflict']) or ( |
139
|
|
|
'roman_conflict' in features and features['roman_conflict'] |
|
|
|
|
140
|
|
|
): |
141
|
1 |
|
return 0 |
142
|
|
|
|
143
|
|
|
# 3 & 7 |
144
|
1 |
|
full_tar1 = ' '.join((tar_ln, tar_fn)).replace('-', ' ').strip() |
145
|
1 |
|
for s_pos, s_type in features['tar_specials']: |
146
|
1 |
|
if s_type == 'a': |
147
|
1 |
|
full_tar1 = full_tar1[ |
148
|
|
|
: -(1 + len(self.stc.synoname_special_table[s_pos][1])) |
149
|
|
|
] |
150
|
1 |
|
elif s_type == 'b': |
151
|
1 |
|
loc = ( |
152
|
|
|
full_tar1.find( |
153
|
|
|
' ' + self.stc.synoname_special_table[s_pos][1] + ' ' |
154
|
|
|
) |
155
|
|
|
+ 1 |
156
|
|
|
) |
157
|
1 |
|
full_tar1 = ( |
158
|
|
|
full_tar1[:loc] |
159
|
|
|
+ full_tar1[ |
160
|
|
|
loc + len(self.stc.synoname_special_table[s_pos][1]) : |
161
|
|
|
] |
162
|
|
|
) |
163
|
1 |
|
elif s_type == 'c': |
164
|
1 |
|
full_tar1 = full_tar1[ |
165
|
|
|
1 + len(self.stc.synoname_special_table[s_pos][1]) : |
166
|
|
|
] |
167
|
|
|
|
168
|
1 |
|
full_src1 = ' '.join((src_ln, src_fn)).replace('-', ' ').strip() |
169
|
1 |
|
for s_pos, s_type in features['src_specials']: |
170
|
1 |
|
if s_type == 'a': |
171
|
1 |
|
full_src1 = full_src1[ |
172
|
|
|
: -(1 + len(self.stc.synoname_special_table[s_pos][1])) |
173
|
|
|
] |
174
|
1 |
|
elif s_type == 'b': |
175
|
1 |
|
loc = ( |
176
|
|
|
full_src1.find( |
177
|
|
|
' ' + self.stc.synoname_special_table[s_pos][1] + ' ' |
178
|
|
|
) |
179
|
|
|
+ 1 |
180
|
|
|
) |
181
|
1 |
|
full_src1 = ( |
182
|
|
|
full_src1[:loc] |
183
|
|
|
+ full_src1[ |
184
|
|
|
loc + len(self.stc.synoname_special_table[s_pos][1]) : |
185
|
|
|
] |
186
|
|
|
) |
187
|
1 |
|
elif s_type == 'c': |
188
|
1 |
|
full_src1 = full_src1[ |
189
|
|
|
1 + len(self.stc.synoname_special_table[s_pos][1]) : |
190
|
|
|
] |
191
|
|
|
|
192
|
1 |
|
full_tar2 = full_tar1 |
193
|
1 |
|
for s_pos, s_type in features['tar_specials']: |
194
|
1 |
|
if s_type == 'd': |
195
|
1 |
|
full_tar2 = full_tar2[ |
196
|
|
|
len(self.stc.synoname_special_table[s_pos][1]) : |
197
|
|
|
] |
198
|
1 |
|
elif ( |
199
|
|
|
s_type == 'X' |
|
|
|
|
200
|
|
|
and self.stc.synoname_special_table[s_pos][1] in full_tar2 |
|
|
|
|
201
|
|
|
): |
202
|
1 |
|
loc = full_tar2.find( |
203
|
|
|
' ' + self.stc.synoname_special_table[s_pos][1] |
204
|
|
|
) |
205
|
1 |
|
full_tar2 = ( |
206
|
|
|
full_tar2[:loc] |
207
|
|
|
+ full_tar2[ |
208
|
|
|
loc + len(self.stc.synoname_special_table[s_pos][1]) : |
209
|
|
|
] |
210
|
|
|
) |
211
|
|
|
|
212
|
1 |
|
full_src2 = full_src1 |
213
|
1 |
|
for s_pos, s_type in features['src_specials']: |
214
|
1 |
|
if s_type == 'd': |
215
|
1 |
|
full_src2 = full_src2[ |
216
|
|
|
len(self.stc.synoname_special_table[s_pos][1]) : |
217
|
|
|
] |
218
|
1 |
|
elif ( |
219
|
|
|
s_type == 'X' |
|
|
|
|
220
|
|
|
and self.stc.synoname_special_table[s_pos][1] in full_src2 |
|
|
|
|
221
|
|
|
): |
222
|
1 |
|
loc = full_src2.find( |
223
|
|
|
' ' + self.stc.synoname_special_table[s_pos][1] |
224
|
|
|
) |
225
|
1 |
|
full_src2 = ( |
226
|
|
|
full_src2[:loc] |
227
|
|
|
+ full_src2[ |
228
|
|
|
loc + len(self.stc.synoname_special_table[s_pos][1]) : |
229
|
|
|
] |
230
|
|
|
) |
231
|
|
|
|
232
|
1 |
|
full_tar1 = self._synoname_strip_punct(full_tar1) |
233
|
1 |
|
tar1_words = full_tar1.split() |
234
|
1 |
|
tar1_num_words = len(tar1_words) |
235
|
|
|
|
236
|
1 |
|
full_src1 = self._synoname_strip_punct(full_src1) |
237
|
1 |
|
src1_words = full_src1.split() |
238
|
1 |
|
src1_num_words = len(src1_words) |
239
|
|
|
|
240
|
1 |
|
full_tar2 = self._synoname_strip_punct(full_tar2) |
241
|
1 |
|
tar2_words = full_tar2.split() |
242
|
1 |
|
tar2_num_words = len(tar2_words) |
243
|
|
|
|
244
|
1 |
|
full_src2 = self._synoname_strip_punct(full_src2) |
245
|
1 |
|
src2_words = full_src2.split() |
246
|
1 |
|
src2_num_words = len(src2_words) |
247
|
|
|
|
248
|
|
|
# 2 |
249
|
1 |
|
if ( |
250
|
|
|
src1_num_words < 2 |
|
|
|
|
251
|
|
|
and src_len_specials == 0 |
|
|
|
|
252
|
|
|
and src2_num_words < 2 |
|
|
|
|
253
|
|
|
and tar_len_specials == 0 |
|
|
|
|
254
|
|
|
): |
255
|
1 |
|
return 0 |
256
|
|
|
|
257
|
|
|
# 4 |
258
|
1 |
|
if ( |
259
|
|
|
tar1_num_words == 1 |
|
|
|
|
260
|
|
|
and src1_num_words == 1 |
|
|
|
|
261
|
|
|
and tar1_words[0] == src1_words[0] |
|
|
|
|
262
|
|
|
): |
263
|
1 |
|
return 1 |
264
|
1 |
|
if tar1_num_words < 2 and tar_len_specials == 0: |
265
|
1 |
|
return 0 |
266
|
|
|
|
267
|
|
|
# 5 |
268
|
1 |
|
last_found = False |
269
|
1 |
|
for word in tar1_words: |
270
|
1 |
|
if src_ln.endswith(word) or word + ' ' in src_ln: |
271
|
1 |
|
last_found = True |
272
|
|
|
|
273
|
1 |
|
if not last_found: |
274
|
1 |
|
for word in src1_words: |
275
|
1 |
|
if tar_ln.endswith(word) or word + ' ' in tar_ln: |
276
|
1 |
|
last_found = True |
277
|
|
|
|
278
|
|
|
# 6 |
279
|
1 |
|
matches = 0 |
280
|
1 |
|
if last_found: |
281
|
1 |
|
for i, s_word in enumerate(src1_words): |
282
|
1 |
|
for j, t_word in enumerate(tar1_words): |
283
|
1 |
|
if s_word == t_word: |
284
|
1 |
|
src1_words[i] = '@' |
285
|
1 |
|
tar1_words[j] = '@' |
286
|
1 |
|
matches += 1 |
287
|
1 |
|
w_ratio = matches / max(tar1_num_words, src1_num_words) |
288
|
1 |
|
if matches > 1 or ( |
|
|
|
|
289
|
|
|
matches == 1 |
|
|
|
|
290
|
|
|
and src1_num_words == 1 |
|
|
|
|
291
|
|
|
and tar1_num_words == 1 |
|
|
|
|
292
|
|
|
and (tar_len_specials > 0 or src_len_specials > 0) |
|
|
|
|
293
|
|
|
): |
294
|
1 |
|
return w_ratio |
295
|
|
|
|
296
|
|
|
# 8 |
297
|
1 |
|
if ( |
298
|
|
|
tar2_num_words == 1 |
|
|
|
|
299
|
|
|
and src2_num_words == 1 |
|
|
|
|
300
|
|
|
and tar2_words[0] == src2_words[0] |
|
|
|
|
301
|
|
|
): |
302
|
1 |
|
return 1 |
303
|
|
|
# I see no way that the following can be True if the equivalent in |
304
|
|
|
# #4 was False. |
305
|
|
|
if tar2_num_words < 2 and tar_len_specials == 0: # pragma: no cover |
306
|
|
|
return 0 |
307
|
|
|
|
308
|
|
|
# 9 |
309
|
1 |
|
last_found = False |
310
|
1 |
|
for word in tar2_words: |
311
|
1 |
|
if src_ln.endswith(word) or word + ' ' in src_ln: |
312
|
1 |
|
last_found = True |
313
|
|
|
|
314
|
1 |
|
if not last_found: |
315
|
1 |
|
for word in src2_words: |
316
|
1 |
|
if tar_ln.endswith(word) or word + ' ' in tar_ln: |
317
|
1 |
|
last_found = True |
318
|
|
|
|
319
|
1 |
|
if not last_found: |
320
|
1 |
|
return 0 |
321
|
|
|
|
322
|
|
|
# 10 |
323
|
1 |
|
matches = 0 |
324
|
1 |
|
if last_found: |
325
|
1 |
|
for i, s_word in enumerate(src2_words): |
326
|
1 |
|
for j, t_word in enumerate(tar2_words): |
327
|
1 |
|
if s_word == t_word: |
328
|
1 |
|
src2_words[i] = '@' |
329
|
1 |
|
tar2_words[j] = '@' |
330
|
1 |
|
matches += 1 |
331
|
1 |
|
w_ratio = matches / max(tar2_num_words, src2_num_words) |
332
|
1 |
|
if matches > 1 or ( |
|
|
|
|
333
|
|
|
matches == 1 |
|
|
|
|
334
|
|
|
and src2_num_words == 1 |
|
|
|
|
335
|
|
|
and tar2_num_words == 1 |
|
|
|
|
336
|
|
|
and (tar_len_specials > 0 or src_len_specials > 0) |
|
|
|
|
337
|
|
|
): |
338
|
|
|
return w_ratio |
339
|
|
|
|
340
|
1 |
|
return 0 |
341
|
|
|
|
342
|
1 |
|
def dist_abs( |
|
|
|
|
343
|
|
|
self, |
|
|
|
|
344
|
|
|
src, |
|
|
|
|
345
|
|
|
tar, |
|
|
|
|
346
|
|
|
word_approx_min=0.3, |
|
|
|
|
347
|
|
|
char_approx_min=0.73, |
|
|
|
|
348
|
|
|
tests=2 ** 12 - 1, |
|
|
|
|
349
|
|
|
ret_name=False, |
|
|
|
|
350
|
|
|
): |
351
|
|
|
"""Return the Synoname similarity type of two words. |
352
|
|
|
|
353
|
|
|
Args: |
354
|
|
|
src (str): Source string for comparison |
355
|
|
|
tar (str): Target string for comparison |
356
|
|
|
word_approx_min (float): the minimum word approximation value to |
357
|
|
|
signal a 'word_approx' match |
358
|
|
|
char_approx_min (float): the minimum character approximation value |
359
|
|
|
to signal a 'char_approx' match |
360
|
|
|
tests (int or Iterable): either an integer indicating tests to |
361
|
|
|
perform or a list of test names to perform (defaults to |
362
|
|
|
performing all tests) |
363
|
|
|
ret_name (bool): If True, returns the match name rather than its |
364
|
|
|
integer equivalent |
365
|
|
|
|
366
|
|
|
Returns: |
367
|
|
|
int (or str if ret_name is True): Synoname value |
368
|
|
|
|
369
|
|
|
Examples: |
370
|
|
|
>>> synoname(('Breghel', 'Pieter', ''), ('Brueghel', 'Pieter', '')) |
371
|
|
|
2 |
372
|
|
|
>>> synoname(('Breghel', 'Pieter', ''), ('Brueghel', 'Pieter', ''), |
373
|
|
|
... ret_name=True) |
374
|
|
|
'omission' |
375
|
|
|
>>> synoname(('Dore', 'Gustave', ''), |
376
|
|
|
... ('Dore', 'Paul Gustave Louis Christophe', ''), |
377
|
|
|
... ret_name=True) |
378
|
|
|
'inclusion' |
379
|
|
|
>>> synoname(('Pereira', 'I. R.', ''), ('Pereira', 'I. Smith', ''), |
380
|
|
|
... ret_name=True) |
381
|
|
|
'word_approx' |
382
|
|
|
|
383
|
|
|
""" |
384
|
1 |
|
if isinstance(tests, Iterable): |
385
|
1 |
|
new_tests = 0 |
386
|
1 |
|
for term in tests: |
387
|
1 |
|
if term in self.test_dict: |
388
|
1 |
|
new_tests += self.test_dict[term] |
389
|
1 |
|
tests = new_tests |
390
|
|
|
|
391
|
1 |
|
if isinstance(src, tuple): |
392
|
1 |
|
src_ln, src_fn, src_qual = src |
393
|
1 |
|
elif '#' in src: |
394
|
1 |
|
src_ln, src_fn, src_qual = src.split('#')[-3:] |
395
|
|
|
else: |
396
|
1 |
|
src_ln, src_fn, src_qual = src, '', '' |
397
|
|
|
|
398
|
1 |
|
if isinstance(tar, tuple): |
399
|
1 |
|
tar_ln, tar_fn, tar_qual = tar |
400
|
1 |
|
elif '#' in tar: |
401
|
1 |
|
tar_ln, tar_fn, tar_qual = tar.split('#')[-3:] |
402
|
|
|
else: |
403
|
1 |
|
tar_ln, tar_fn, tar_qual = tar, '', '' |
404
|
|
|
|
405
|
1 |
|
def _split_special(spec): |
406
|
1 |
|
spec_list = [] |
407
|
1 |
|
while spec: |
408
|
1 |
|
spec_list.append((int(spec[:3]), spec[3:4])) |
409
|
1 |
|
spec = spec[4:] |
410
|
1 |
|
return spec_list |
411
|
|
|
|
412
|
1 |
|
def _fmt_retval(val): |
413
|
1 |
|
if ret_name: |
414
|
1 |
|
return self.match_name[val] |
415
|
1 |
|
return val |
416
|
|
|
|
417
|
|
|
# 1. Preprocessing |
418
|
|
|
|
419
|
|
|
# Lowercasing |
420
|
1 |
|
src_fn = src_fn.strip().lower() |
421
|
1 |
|
src_ln = src_ln.strip().lower() |
422
|
1 |
|
src_qual = src_qual.strip().lower() |
423
|
|
|
|
424
|
1 |
|
tar_fn = tar_fn.strip().lower() |
425
|
1 |
|
tar_ln = tar_ln.strip().lower() |
426
|
1 |
|
tar_qual = tar_qual.strip().lower() |
427
|
|
|
|
428
|
|
|
# Create toolcodes |
429
|
1 |
|
src_ln, src_fn, src_tc = self.stc.fingerprint(src_ln, src_fn, src_qual) |
430
|
1 |
|
tar_ln, tar_fn, tar_tc = self.stc.fingerprint(tar_ln, tar_fn, tar_qual) |
431
|
|
|
|
432
|
1 |
|
src_generation = int(src_tc[2]) |
433
|
1 |
|
src_romancode = int(src_tc[3:6]) |
434
|
1 |
|
src_len_fn = int(src_tc[6:8]) |
435
|
1 |
|
src_tc = src_tc.split('$') |
436
|
1 |
|
src_specials = _split_special(src_tc[1]) |
437
|
|
|
|
438
|
1 |
|
tar_generation = int(tar_tc[2]) |
439
|
1 |
|
tar_romancode = int(tar_tc[3:6]) |
440
|
1 |
|
tar_len_fn = int(tar_tc[6:8]) |
441
|
1 |
|
tar_tc = tar_tc.split('$') |
442
|
1 |
|
tar_specials = _split_special(tar_tc[1]) |
443
|
|
|
|
444
|
1 |
|
gen_conflict = (src_generation != tar_generation) and bool( |
445
|
|
|
src_generation or tar_generation |
446
|
|
|
) |
447
|
1 |
|
roman_conflict = (src_romancode != tar_romancode) and bool( |
448
|
|
|
src_romancode or tar_romancode |
449
|
|
|
) |
450
|
|
|
|
451
|
1 |
|
ln_equal = src_ln == tar_ln |
452
|
1 |
|
fn_equal = src_fn == tar_fn |
453
|
|
|
|
454
|
|
|
# approx_c |
455
|
1 |
|
def _approx_c(): |
456
|
1 |
|
if gen_conflict or roman_conflict: |
457
|
1 |
|
return False, 0 |
458
|
|
|
|
459
|
1 |
|
full_src = ' '.join((src_ln, src_fn)) |
460
|
1 |
|
if full_src.startswith('master '): |
461
|
1 |
|
full_src = full_src[len('master ') :] |
462
|
1 |
|
for intro in [ |
463
|
|
|
'of the ', |
|
|
|
|
464
|
|
|
'of ', |
|
|
|
|
465
|
|
|
'known as the ', |
|
|
|
|
466
|
|
|
'with the ', |
|
|
|
|
467
|
|
|
'with ', |
|
|
|
|
468
|
|
|
]: |
469
|
1 |
|
if full_src.startswith(intro): |
470
|
1 |
|
full_src = full_src[len(intro) :] |
471
|
|
|
|
472
|
1 |
|
full_tar = ' '.join((tar_ln, tar_fn)) |
473
|
1 |
|
if full_tar.startswith('master '): |
474
|
1 |
|
full_tar = full_tar[len('master ') :] |
475
|
1 |
|
for intro in [ |
476
|
|
|
'of the ', |
|
|
|
|
477
|
|
|
'of ', |
|
|
|
|
478
|
|
|
'known as the ', |
|
|
|
|
479
|
|
|
'with the ', |
|
|
|
|
480
|
|
|
'with ', |
|
|
|
|
481
|
|
|
]: |
482
|
1 |
|
if full_tar.startswith(intro): |
483
|
1 |
|
full_tar = full_tar[len(intro) :] |
484
|
|
|
|
485
|
1 |
|
loc_ratio = sim_ratcliff_obershelp(full_src, full_tar) |
486
|
1 |
|
return loc_ratio >= char_approx_min, loc_ratio |
487
|
|
|
|
488
|
1 |
|
approx_c_result, ca_ratio = _approx_c() |
|
|
|
|
489
|
|
|
|
490
|
1 |
|
if tests & self.test_dict['exact'] and fn_equal and ln_equal: |
491
|
1 |
|
return _fmt_retval(self.match_type_dict['exact']) |
492
|
1 |
|
if tests & self.test_dict['omission']: |
493
|
1 |
View Code Duplication |
if ( |
|
|
|
|
494
|
|
|
fn_equal |
|
|
|
|
495
|
|
|
and levenshtein(src_ln, tar_ln, cost=(1, 1, 99, 99)) == 1 |
|
|
|
|
496
|
|
|
): |
497
|
1 |
|
if not roman_conflict: |
498
|
1 |
|
return _fmt_retval(self.match_type_dict['omission']) |
499
|
1 |
|
elif ( |
500
|
|
|
ln_equal |
|
|
|
|
501
|
|
|
and levenshtein(src_fn, tar_fn, cost=(1, 1, 99, 99)) == 1 |
|
|
|
|
502
|
|
|
): |
503
|
1 |
|
return _fmt_retval(self.match_type_dict['omission']) |
504
|
1 |
View Code Duplication |
if tests & self.test_dict['substitution']: |
|
|
|
|
505
|
1 |
|
if ( |
506
|
|
|
fn_equal |
|
|
|
|
507
|
|
|
and levenshtein(src_ln, tar_ln, cost=(99, 99, 1, 99)) == 1 |
|
|
|
|
508
|
|
|
): |
509
|
1 |
|
return _fmt_retval(self.match_type_dict['substitution']) |
510
|
1 |
|
elif ( |
511
|
|
|
ln_equal |
|
|
|
|
512
|
|
|
and levenshtein(src_fn, tar_fn, cost=(99, 99, 1, 99)) == 1 |
|
|
|
|
513
|
|
|
): |
514
|
1 |
|
return _fmt_retval(self.match_type_dict['substitution']) |
515
|
1 |
View Code Duplication |
if tests & self.test_dict['transposition']: |
|
|
|
|
516
|
1 |
|
if fn_equal and ( |
517
|
|
|
levenshtein(src_ln, tar_ln, mode='osa', cost=(99, 99, 99, 1)) |
|
|
|
|
518
|
|
|
== 1 |
|
|
|
|
519
|
|
|
): |
520
|
1 |
|
return _fmt_retval(self.match_type_dict['transposition']) |
521
|
1 |
|
elif ln_equal and ( |
522
|
|
|
levenshtein(src_fn, tar_fn, mode='osa', cost=(99, 99, 99, 1)) |
|
|
|
|
523
|
|
|
== 1 |
|
|
|
|
524
|
|
|
): |
525
|
1 |
|
return _fmt_retval(self.match_type_dict['transposition']) |
526
|
1 |
|
if tests & self.test_dict['punctuation']: |
527
|
1 |
|
np_src_fn = self._synoname_strip_punct(src_fn) |
528
|
1 |
|
np_tar_fn = self._synoname_strip_punct(tar_fn) |
529
|
1 |
|
np_src_ln = self._synoname_strip_punct(src_ln) |
530
|
1 |
|
np_tar_ln = self._synoname_strip_punct(tar_ln) |
531
|
|
|
|
532
|
1 |
|
if (np_src_fn == np_tar_fn) and (np_src_ln == np_tar_ln): |
533
|
1 |
|
return _fmt_retval(self.match_type_dict['punctuation']) |
534
|
|
|
|
535
|
1 |
|
np_src_fn = self._synoname_strip_punct(src_fn.replace('-', ' ')) |
536
|
1 |
|
np_tar_fn = self._synoname_strip_punct(tar_fn.replace('-', ' ')) |
537
|
1 |
|
np_src_ln = self._synoname_strip_punct(src_ln.replace('-', ' ')) |
538
|
1 |
|
np_tar_ln = self._synoname_strip_punct(tar_ln.replace('-', ' ')) |
539
|
|
|
|
540
|
1 |
|
if (np_src_fn == np_tar_fn) and (np_src_ln == np_tar_ln): |
541
|
1 |
|
return _fmt_retval(self.match_type_dict['punctuation']) |
542
|
|
|
|
543
|
1 |
|
if tests & self.test_dict['initials'] and ln_equal: |
544
|
1 |
|
if src_fn and tar_fn: |
545
|
1 |
|
src_initials = self._synoname_strip_punct(src_fn).split() |
546
|
1 |
|
tar_initials = self._synoname_strip_punct(tar_fn).split() |
547
|
1 |
|
initials = bool( |
548
|
|
|
(len(src_initials) == len(''.join(src_initials))) |
549
|
|
|
or (len(tar_initials) == len(''.join(tar_initials))) |
550
|
|
|
) |
551
|
1 |
|
if initials: |
552
|
1 |
|
src_initials = ''.join(_[0] for _ in src_initials) |
553
|
1 |
|
tar_initials = ''.join(_[0] for _ in tar_initials) |
554
|
1 |
|
if src_initials == tar_initials: |
555
|
1 |
|
return _fmt_retval(self.match_type_dict['initials']) |
556
|
1 |
|
initial_diff = abs(len(src_initials) - len(tar_initials)) |
557
|
1 |
|
if initial_diff and ( |
558
|
|
|
( |
|
|
|
|
559
|
|
|
initial_diff |
560
|
|
|
== levenshtein( |
561
|
|
|
src_initials, |
562
|
|
|
tar_initials, |
563
|
|
|
cost=(1, 99, 99, 99), |
564
|
|
|
) |
565
|
|
|
) |
566
|
|
|
or ( |
|
|
|
|
567
|
|
|
initial_diff |
568
|
|
|
== levenshtein( |
569
|
|
|
tar_initials, |
570
|
|
|
src_initials, |
571
|
|
|
cost=(1, 99, 99, 99), |
572
|
|
|
) |
573
|
|
|
) |
574
|
|
|
): |
575
|
1 |
|
return _fmt_retval(self.match_type_dict['initials']) |
576
|
1 |
|
if tests & self.test_dict['extension']: |
577
|
1 |
|
if src_ln[1] == tar_ln[1] and ( |
578
|
|
|
src_ln.startswith(tar_ln) or tar_ln.startswith(src_ln) |
|
|
|
|
579
|
|
|
): |
580
|
1 |
|
if ( |
|
|
|
|
581
|
|
|
(not src_len_fn and not tar_len_fn) |
|
|
|
|
582
|
|
|
or (tar_fn and src_fn.startswith(tar_fn)) |
|
|
|
|
583
|
|
|
or (src_fn and tar_fn.startswith(src_fn)) |
|
|
|
|
584
|
|
|
) and not roman_conflict: |
585
|
1 |
|
return _fmt_retval(self.match_type_dict['extension']) |
586
|
1 |
|
if tests & self.test_dict['inclusion'] and ln_equal: |
587
|
1 |
|
if (src_fn and src_fn in tar_fn) or (tar_fn and tar_fn in src_ln): |
588
|
1 |
|
return _fmt_retval(self.match_type_dict['inclusion']) |
589
|
1 |
|
if tests & self.test_dict['no_first'] and ln_equal: |
590
|
1 |
|
if src_fn == '' or tar_fn == '': |
591
|
1 |
|
return _fmt_retval(self.match_type_dict['no_first']) |
592
|
1 |
|
if tests & self.test_dict['word_approx']: |
593
|
1 |
|
ratio = self._synoname_word_approximation( |
594
|
|
|
src_ln, |
595
|
|
|
tar_ln, |
596
|
|
|
src_fn, |
597
|
|
|
tar_fn, |
598
|
|
|
{ |
599
|
|
|
'gen_conflict': gen_conflict, |
600
|
|
|
'roman_conflict': roman_conflict, |
601
|
|
|
'src_specials': src_specials, |
602
|
|
|
'tar_specials': tar_specials, |
603
|
|
|
}, |
604
|
|
|
) |
605
|
1 |
|
if ratio == 1 and tests & self.test_dict['confusions']: |
606
|
1 |
|
if ( |
607
|
|
|
' '.join((src_fn, src_ln)).strip() |
|
|
|
|
608
|
|
|
== ' '.join((tar_fn, tar_ln)).strip() |
|
|
|
|
609
|
|
|
): |
610
|
1 |
|
return _fmt_retval(self.match_type_dict['confusions']) |
611
|
1 |
|
if ratio >= word_approx_min: |
612
|
1 |
|
return _fmt_retval(self.match_type_dict['word_approx']) |
613
|
1 |
|
if tests & self.test_dict['char_approx']: |
614
|
1 |
|
if ca_ratio >= char_approx_min: |
615
|
1 |
|
return _fmt_retval(self.match_type_dict['char_approx']) |
616
|
1 |
|
return _fmt_retval(self.match_type_dict['no_match']) |
617
|
|
|
|
618
|
1 |
|
def dist( |
|
|
|
|
619
|
|
|
self, |
|
|
|
|
620
|
|
|
src, |
|
|
|
|
621
|
|
|
tar, |
|
|
|
|
622
|
|
|
word_approx_min=0.3, |
|
|
|
|
623
|
|
|
char_approx_min=0.73, |
|
|
|
|
624
|
|
|
tests=2 ** 12 - 1, |
|
|
|
|
625
|
|
|
): |
626
|
|
|
"""Return the normalized Synoname distance between two words. |
627
|
|
|
|
628
|
|
|
Args: |
629
|
|
|
src (str): Source string for comparison |
630
|
|
|
tar (str): Target string for comparison |
631
|
|
|
word_approx_min (float): the minimum word approximation value to |
632
|
|
|
signal a 'word_approx' match |
633
|
|
|
char_approx_min (float): the minimum character approximation value |
634
|
|
|
to signal a 'char_approx' match |
635
|
|
|
tests (int or Iterable): either an integer indicating tests to |
636
|
|
|
perform or a list of test names to perform (defaults to |
637
|
|
|
performing all tests) |
638
|
|
|
|
639
|
|
|
Returns: |
640
|
|
|
float: Normalized Synoname distance |
641
|
|
|
|
642
|
|
|
""" |
643
|
|
|
return ( |
644
|
|
|
synoname(src, tar, word_approx_min, char_approx_min, tests, False) |
645
|
|
|
/ 14 |
646
|
|
|
) |
647
|
|
|
|
648
|
|
|
|
649
|
1 |
|
def synoname( |
|
|
|
|
650
|
|
|
src, |
|
|
|
|
651
|
|
|
tar, |
|
|
|
|
652
|
|
|
word_approx_min=0.3, |
|
|
|
|
653
|
|
|
char_approx_min=0.73, |
|
|
|
|
654
|
|
|
tests=2 ** 12 - 1, |
|
|
|
|
655
|
|
|
ret_name=False, |
|
|
|
|
656
|
|
|
): |
657
|
|
|
"""Return the Synoname similarity type of two words. |
658
|
|
|
|
659
|
|
|
This is a wrapper for :py:meth:`Synoname.synoname`. |
660
|
|
|
|
661
|
|
|
Args: |
662
|
|
|
src (str): Source string for comparison |
663
|
|
|
tar (str): Target string for comparison |
664
|
|
|
word_approx_min (float): the minimum word approximation value to signal |
665
|
|
|
a 'word_approx' match |
666
|
|
|
char_approx_min (float): the minimum character approximation value to |
667
|
|
|
signal a 'char_approx' match |
668
|
|
|
tests (int or Iterable): either an integer indicating tests to perform |
669
|
|
|
or a list of test names to perform (defaults to performing all |
670
|
|
|
tests) |
671
|
|
|
ret_name (bool): If True, returns the match name rather than its |
672
|
|
|
integer equivalent |
673
|
|
|
|
674
|
|
|
Returns: |
675
|
|
|
int (or str if ret_name is True): Synoname value |
676
|
|
|
|
677
|
|
|
Examples: |
678
|
|
|
>>> synoname(('Breghel', 'Pieter', ''), ('Brueghel', 'Pieter', '')) |
679
|
|
|
2 |
680
|
|
|
>>> synoname(('Breghel', 'Pieter', ''), ('Brueghel', 'Pieter', ''), |
681
|
|
|
... ret_name=True) |
682
|
|
|
'omission' |
683
|
|
|
>>> synoname(('Dore', 'Gustave', ''), |
684
|
|
|
... ('Dore', 'Paul Gustave Louis Christophe', ''), |
685
|
|
|
... ret_name=True) |
686
|
|
|
'inclusion' |
687
|
|
|
>>> synoname(('Pereira', 'I. R.', ''), ('Pereira', 'I. Smith', ''), |
688
|
|
|
... ret_name=True) |
689
|
|
|
'word_approx' |
690
|
|
|
|
691
|
|
|
""" |
692
|
1 |
|
return Synoname().dist_abs( |
693
|
|
|
src, tar, word_approx_min, char_approx_min, tests, ret_name |
694
|
|
|
) |
695
|
|
|
|
696
|
|
|
|
697
|
|
|
if __name__ == '__main__': |
698
|
|
|
import doctest |
699
|
|
|
|
700
|
|
|
doctest.testmod() |
701
|
|
|
|