chrislit /
abydos
| Total Complexity | 476 |
| Total Lines | 1365 |
| Duplicated Lines | 5.35 % |
| Coverage | 100% |
| Changes | 0 | ||
Duplicate code is one of the most pungent code smells. A rule that is often used is to re-structure code once it is duplicated in three or more places.
Common duplication problems, and corresponding solutions are:
Complex classes like abydos.stemmer._snowball often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
| 1 | # -*- coding: utf-8 -*- |
||
|
coding-style
introduced
by
|
|||
| 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 | 1 | """abydos.stemmer._snowball. |
|
| 20 | |||
| 21 | The stemmer._snowball module defines the stemmers: |
||
| 22 | |||
| 23 | - Porter |
||
| 24 | - Porter2 (Snowball English) |
||
| 25 | - Snowball German |
||
| 26 | - Snowball Dutch |
||
| 27 | - Snowball Norwegian |
||
| 28 | - Snowball Swedish |
||
| 29 | - Snowball Danish |
||
| 30 | """ |
||
| 31 | |||
| 32 | 1 | from __future__ import unicode_literals |
|
| 33 | |||
| 34 | 1 | from unicodedata import normalize |
|
| 35 | |||
| 36 | 1 | from six import text_type |
|
| 37 | 1 | from six.moves import range |
|
| 38 | |||
| 39 | 1 | __all__ = [ |
|
| 40 | 'porter', |
||
| 41 | 'porter2', |
||
| 42 | 'sb_danish', |
||
| 43 | 'sb_dutch', |
||
| 44 | 'sb_german', |
||
| 45 | 'sb_norwegian', |
||
| 46 | 'sb_swedish', |
||
| 47 | ] |
||
| 48 | |||
| 49 | |||
| 50 | 1 | def _m_degree(term, vowels): |
|
| 51 | """Return Porter helper function _m_degree value. |
||
| 52 | |||
| 53 | m-degree is equal to the number of V to C transitions |
||
| 54 | |||
| 55 | :param str term: the word for which to calculate the m-degree |
||
| 56 | :param set vowels: the set of vowels in the language |
||
| 57 | :returns: the m-degree as defined in the Porter stemmer definition |
||
| 58 | :rtype: int |
||
| 59 | """ |
||
| 60 | 1 | mdeg = 0 |
|
| 61 | 1 | last_was_vowel = False |
|
| 62 | 1 | for letter in term: |
|
| 63 | 1 | if letter in vowels: |
|
| 64 | 1 | last_was_vowel = True |
|
| 65 | else: |
||
| 66 | 1 | if last_was_vowel: |
|
| 67 | 1 | mdeg += 1 |
|
| 68 | 1 | last_was_vowel = False |
|
| 69 | 1 | return mdeg |
|
| 70 | |||
| 71 | |||
| 72 | 1 | def _sb_has_vowel(term, vowels): |
|
| 73 | """Return Porter helper function _sb_has_vowel value. |
||
| 74 | |||
| 75 | :param str term: the word to scan for vowels |
||
| 76 | :param set vowels: the set of vowels in the language |
||
| 77 | :returns: true iff a vowel exists in the term (as defined in the Porter |
||
| 78 | stemmer definition) |
||
| 79 | :rtype: bool |
||
| 80 | """ |
||
| 81 | 1 | for letter in term: |
|
| 82 | 1 | if letter in vowels: |
|
| 83 | 1 | return True |
|
| 84 | 1 | return False |
|
| 85 | |||
| 86 | |||
| 87 | 1 | def _ends_in_doubled_cons(term, vowels): |
|
| 88 | """Return Porter helper function _ends_in_doubled_cons value. |
||
| 89 | |||
| 90 | :param str term: the word to check for a final doubled consonant |
||
| 91 | :param set vowels: the set of vowels in the language |
||
| 92 | :returns: true iff the stem ends in a doubled consonant (as defined in the |
||
| 93 | Porter stemmer definition) |
||
| 94 | :rtype: bool |
||
| 95 | """ |
||
| 96 | 1 | return len(term) > 1 and term[-1] not in vowels and term[-2] == term[-1] |
|
| 97 | |||
| 98 | |||
| 99 | 1 | def _ends_in_cvc(term, vowels): |
|
| 100 | """Return Porter helper function _ends_in_cvc value. |
||
| 101 | |||
| 102 | :param str term: the word to scan for cvc |
||
| 103 | :param set vowels: the set of vowels in the language |
||
| 104 | :returns: true iff the stem ends in cvc (as defined in the Porter stemmer |
||
| 105 | definition) |
||
| 106 | :rtype: bool |
||
| 107 | """ |
||
| 108 | 1 | return len(term) > 2 and ( |
|
| 109 | term[-1] not in vowels |
||
| 110 | and term[-2] in vowels |
||
| 111 | and term[-3] not in vowels |
||
| 112 | and term[-1] not in tuple('wxY') |
||
| 113 | ) |
||
| 114 | |||
| 115 | |||
| 116 | 1 | def porter(word, early_english=False): |
|
| 117 | """Return Porter stem. |
||
| 118 | |||
| 119 | The Porter stemmer is described in :cite:`Porter:1980`. |
||
| 120 | |||
| 121 | :param str word: the word to calculate the stem of |
||
| 122 | :param bool early_english: set to True in order to remove -eth & -est |
||
| 123 | (2nd & 3rd person singular verbal agreement suffixes) |
||
| 124 | :returns: word stem |
||
| 125 | :rtype: str |
||
| 126 | |||
| 127 | >>> porter('reading') |
||
| 128 | 'read' |
||
| 129 | >>> porter('suspension') |
||
| 130 | 'suspens' |
||
| 131 | >>> porter('elusiveness') |
||
| 132 | 'elus' |
||
| 133 | |||
| 134 | >>> porter('eateth', early_english=True) |
||
| 135 | 'eat' |
||
| 136 | """ |
||
| 137 | # lowercase, normalize, and compose |
||
| 138 | 1 | word = normalize('NFC', text_type(word.lower())) |
|
| 139 | |||
| 140 | # Return word if stem is shorter than 2 |
||
| 141 | 1 | if len(word) < 3: |
|
| 142 | 1 | return word |
|
| 143 | |||
| 144 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y'} |
|
| 145 | # Re-map consonantal y to Y (Y will be C, y will be V) |
||
| 146 | 1 | if word[0] == 'y': |
|
| 147 | 1 | word = 'Y' + word[1:] |
|
| 148 | 1 | for i in range(1, len(word)): |
|
| 149 | 1 | if word[i] == 'y' and word[i - 1] in _vowels: |
|
| 150 | 1 | word = word[:i] + 'Y' + word[i + 1 :] |
|
| 151 | |||
| 152 | # Step 1a |
||
| 153 | 1 | if word[-1] == 's': |
|
| 154 | 1 | if word[-4:] == 'sses': |
|
| 155 | 1 | word = word[:-2] |
|
| 156 | 1 | elif word[-3:] == 'ies': |
|
| 157 | 1 | word = word[:-2] |
|
| 158 | 1 | elif word[-2:] == 'ss': |
|
| 159 | 1 | pass |
|
| 160 | else: |
||
| 161 | 1 | word = word[:-1] |
|
| 162 | |||
| 163 | # Step 1b |
||
| 164 | 1 | step1b_flag = False |
|
| 165 | 1 | if word[-3:] == 'eed': |
|
| 166 | 1 | if _m_degree(word[:-3], _vowels) > 0: |
|
| 167 | 1 | word = word[:-1] |
|
| 168 | 1 | elif word[-2:] == 'ed': |
|
| 169 | 1 | if _sb_has_vowel(word[:-2], _vowels): |
|
| 170 | 1 | word = word[:-2] |
|
| 171 | 1 | step1b_flag = True |
|
| 172 | 1 | elif word[-3:] == 'ing': |
|
| 173 | 1 | if _sb_has_vowel(word[:-3], _vowels): |
|
| 174 | 1 | word = word[:-3] |
|
| 175 | 1 | step1b_flag = True |
|
| 176 | 1 | elif early_english: |
|
| 177 | 1 | if word[-3:] == 'est': |
|
| 178 | 1 | if _sb_has_vowel(word[:-3], _vowels): |
|
| 179 | 1 | word = word[:-3] |
|
| 180 | 1 | step1b_flag = True |
|
| 181 | 1 | elif word[-3:] == 'eth': |
|
| 182 | 1 | if _sb_has_vowel(word[:-3], _vowels): |
|
| 183 | 1 | word = word[:-3] |
|
| 184 | 1 | step1b_flag = True |
|
| 185 | |||
| 186 | 1 | if step1b_flag: |
|
| 187 | 1 | if word[-2:] in {'at', 'bl', 'iz'}: |
|
| 188 | 1 | word += 'e' |
|
| 189 | 1 | elif _ends_in_doubled_cons(word, _vowels) and word[-1] not in { |
|
| 190 | 'l', |
||
| 191 | 's', |
||
| 192 | 'z', |
||
| 193 | }: |
||
| 194 | 1 | word = word[:-1] |
|
| 195 | 1 | elif _m_degree(word, _vowels) == 1 and _ends_in_cvc(word, _vowels): |
|
| 196 | 1 | word += 'e' |
|
| 197 | |||
| 198 | # Step 1c |
||
| 199 | 1 | if word[-1] in {'Y', 'y'} and _sb_has_vowel(word[:-1], _vowels): |
|
| 200 | 1 | word = word[:-1] + 'i' |
|
| 201 | |||
| 202 | # Step 2 |
||
| 203 | 1 | if len(word) > 1: |
|
| 204 | 1 | if word[-2] == 'a': |
|
| 205 | 1 | if word[-7:] == 'ational': |
|
| 206 | 1 | if _m_degree(word[:-7], _vowels) > 0: |
|
| 207 | 1 | word = word[:-5] + 'e' |
|
| 208 | 1 | elif word[-6:] == 'tional': |
|
| 209 | 1 | if _m_degree(word[:-6], _vowels) > 0: |
|
| 210 | 1 | word = word[:-2] |
|
| 211 | 1 | elif word[-2] == 'c': |
|
| 212 | 1 | if word[-4:] in {'enci', 'anci'}: |
|
| 213 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 214 | 1 | word = word[:-1] + 'e' |
|
| 215 | 1 | elif word[-2] == 'e': |
|
| 216 | 1 | if word[-4:] == 'izer': |
|
| 217 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 218 | 1 | word = word[:-1] |
|
| 219 | 1 | elif word[-2] == 'g': |
|
| 220 | 1 | if word[-4:] == 'logi': |
|
| 221 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 222 | 1 | word = word[:-1] |
|
| 223 | 1 | elif word[-2] == 'l': |
|
| 224 | 1 | if word[-3:] == 'bli': |
|
| 225 | 1 | if _m_degree(word[:-3], _vowels) > 0: |
|
| 226 | 1 | word = word[:-1] + 'e' |
|
| 227 | 1 | elif word[-4:] == 'alli': |
|
| 228 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 229 | 1 | word = word[:-2] |
|
| 230 | 1 | elif word[-5:] == 'entli': |
|
| 231 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 232 | 1 | word = word[:-2] |
|
| 233 | 1 | elif word[-3:] == 'eli': |
|
| 234 | 1 | if _m_degree(word[:-3], _vowels) > 0: |
|
| 235 | 1 | word = word[:-2] |
|
| 236 | 1 | elif word[-5:] == 'ousli': |
|
| 237 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 238 | 1 | word = word[:-2] |
|
| 239 | 1 | elif word[-2] == 'o': |
|
| 240 | 1 | if word[-7:] == 'ization': |
|
| 241 | 1 | if _m_degree(word[:-7], _vowels) > 0: |
|
| 242 | 1 | word = word[:-5] + 'e' |
|
| 243 | 1 | elif word[-5:] == 'ation': |
|
| 244 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 245 | 1 | word = word[:-3] + 'e' |
|
| 246 | 1 | elif word[-4:] == 'ator': |
|
| 247 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 248 | 1 | word = word[:-2] + 'e' |
|
| 249 | 1 | elif word[-2] == 's': |
|
| 250 | 1 | if word[-5:] == 'alism': |
|
| 251 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 252 | 1 | word = word[:-3] |
|
| 253 | 1 | elif word[-7:] in {'iveness', 'fulness', 'ousness'}: |
|
| 254 | 1 | if _m_degree(word[:-7], _vowels) > 0: |
|
| 255 | 1 | word = word[:-4] |
|
| 256 | 1 | elif word[-2] == 't': |
|
| 257 | 1 | if word[-5:] == 'aliti': |
|
| 258 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 259 | 1 | word = word[:-3] |
|
| 260 | 1 | elif word[-5:] == 'iviti': |
|
| 261 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 262 | 1 | word = word[:-3] + 'e' |
|
| 263 | 1 | elif word[-6:] == 'biliti': |
|
| 264 | 1 | if _m_degree(word[:-6], _vowels) > 0: |
|
| 265 | 1 | word = word[:-5] + 'le' |
|
| 266 | |||
| 267 | # Step 3 |
||
| 268 | 1 | if word[-5:] == 'icate': |
|
| 269 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 270 | 1 | word = word[:-3] |
|
| 271 | 1 | elif word[-5:] == 'ative': |
|
| 272 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 273 | 1 | word = word[:-5] |
|
| 274 | 1 | elif word[-5:] in {'alize', 'iciti'}: |
|
| 275 | 1 | if _m_degree(word[:-5], _vowels) > 0: |
|
| 276 | 1 | word = word[:-3] |
|
| 277 | 1 | elif word[-4:] == 'ical': |
|
| 278 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 279 | 1 | word = word[:-2] |
|
| 280 | 1 | elif word[-3:] == 'ful': |
|
| 281 | 1 | if _m_degree(word[:-3], _vowels) > 0: |
|
| 282 | 1 | word = word[:-3] |
|
| 283 | 1 | elif word[-4:] == 'ness': |
|
| 284 | 1 | if _m_degree(word[:-4], _vowels) > 0: |
|
| 285 | 1 | word = word[:-4] |
|
| 286 | |||
| 287 | # Step 4 |
||
| 288 | 1 | if word[-2:] == 'al': |
|
| 289 | 1 | if _m_degree(word[:-2], _vowels) > 1: |
|
| 290 | 1 | word = word[:-2] |
|
| 291 | 1 | elif word[-4:] == 'ance': |
|
| 292 | 1 | if _m_degree(word[:-4], _vowels) > 1: |
|
| 293 | 1 | word = word[:-4] |
|
| 294 | 1 | elif word[-4:] == 'ence': |
|
| 295 | 1 | if _m_degree(word[:-4], _vowels) > 1: |
|
| 296 | 1 | word = word[:-4] |
|
| 297 | 1 | elif word[-2:] == 'er': |
|
| 298 | 1 | if _m_degree(word[:-2], _vowels) > 1: |
|
| 299 | 1 | word = word[:-2] |
|
| 300 | 1 | elif word[-2:] == 'ic': |
|
| 301 | 1 | if _m_degree(word[:-2], _vowels) > 1: |
|
| 302 | 1 | word = word[:-2] |
|
| 303 | 1 | elif word[-4:] == 'able': |
|
| 304 | 1 | if _m_degree(word[:-4], _vowels) > 1: |
|
| 305 | 1 | word = word[:-4] |
|
| 306 | 1 | elif word[-4:] == 'ible': |
|
| 307 | 1 | if _m_degree(word[:-4], _vowels) > 1: |
|
| 308 | 1 | word = word[:-4] |
|
| 309 | 1 | elif word[-3:] == 'ant': |
|
| 310 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 311 | 1 | word = word[:-3] |
|
| 312 | 1 | elif word[-5:] == 'ement': |
|
| 313 | 1 | if _m_degree(word[:-5], _vowels) > 1: |
|
| 314 | 1 | word = word[:-5] |
|
| 315 | 1 | elif word[-4:] == 'ment': |
|
| 316 | 1 | if _m_degree(word[:-4], _vowels) > 1: |
|
| 317 | 1 | word = word[:-4] |
|
| 318 | 1 | elif word[-3:] == 'ent': |
|
| 319 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 320 | 1 | word = word[:-3] |
|
| 321 | 1 | elif word[-4:] in {'sion', 'tion'}: |
|
| 322 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 323 | 1 | word = word[:-3] |
|
| 324 | 1 | elif word[-2:] == 'ou': |
|
| 325 | 1 | if _m_degree(word[:-2], _vowels) > 1: |
|
| 326 | 1 | word = word[:-2] |
|
| 327 | 1 | elif word[-3:] == 'ism': |
|
| 328 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 329 | 1 | word = word[:-3] |
|
| 330 | 1 | elif word[-3:] == 'ate': |
|
| 331 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 332 | 1 | word = word[:-3] |
|
| 333 | 1 | elif word[-3:] == 'iti': |
|
| 334 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 335 | 1 | word = word[:-3] |
|
| 336 | 1 | elif word[-3:] == 'ous': |
|
| 337 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 338 | 1 | word = word[:-3] |
|
| 339 | 1 | elif word[-3:] == 'ive': |
|
| 340 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 341 | 1 | word = word[:-3] |
|
| 342 | 1 | elif word[-3:] == 'ize': |
|
| 343 | 1 | if _m_degree(word[:-3], _vowels) > 1: |
|
| 344 | 1 | word = word[:-3] |
|
| 345 | |||
| 346 | # Step 5a |
||
| 347 | 1 | if word[-1] == 'e': |
|
| 348 | 1 | if _m_degree(word[:-1], _vowels) > 1: |
|
| 349 | 1 | word = word[:-1] |
|
| 350 | 1 | elif _m_degree(word[:-1], _vowels) == 1 and not _ends_in_cvc( |
|
| 351 | word[:-1], _vowels |
||
| 352 | ): |
||
| 353 | 1 | word = word[:-1] |
|
| 354 | |||
| 355 | # Step 5b |
||
| 356 | 1 | if word[-2:] == 'll' and _m_degree(word, _vowels) > 1: |
|
| 357 | 1 | word = word[:-1] |
|
| 358 | |||
| 359 | # Change 'Y' back to 'y' if it survived stemming |
||
| 360 | 1 | for i in range(len(word)): |
|
| 361 | 1 | if word[i] == 'Y': |
|
| 362 | 1 | word = word[:i] + 'y' + word[i + 1 :] |
|
| 363 | |||
| 364 | 1 | return word |
|
| 365 | |||
| 366 | |||
| 367 | 1 | def _sb_r1(term, vowels, r1_prefixes=None): |
|
| 368 | """Return the R1 region, as defined in the Porter2 specification.""" |
||
| 369 | 1 | vowel_found = False |
|
| 370 | 1 | if hasattr(r1_prefixes, '__iter__'): |
|
| 371 | 1 | for prefix in r1_prefixes: |
|
| 372 | 1 | if term[: len(prefix)] == prefix: |
|
| 373 | 1 | return len(prefix) |
|
| 374 | |||
| 375 | 1 | for i in range(len(term)): |
|
| 376 | 1 | if not vowel_found and term[i] in vowels: |
|
| 377 | 1 | vowel_found = True |
|
| 378 | 1 | elif vowel_found and term[i] not in vowels: |
|
| 379 | 1 | return i + 1 |
|
| 380 | 1 | return len(term) |
|
| 381 | |||
| 382 | |||
| 383 | 1 | def _sb_r2(term, vowels, r1_prefixes=None): |
|
| 384 | """Return the R2 region, as defined in the Porter2 specification.""" |
||
| 385 | 1 | r1_start = _sb_r1(term, vowels, r1_prefixes) |
|
| 386 | 1 | return r1_start + _sb_r1(term[r1_start:], vowels) |
|
| 387 | |||
| 388 | |||
| 389 | 1 | def _sb_ends_in_short_syllable(term, vowels, codanonvowels): |
|
| 390 | """Return True iff term ends in a short syllable. |
||
| 391 | |||
| 392 | (...according to the Porter2 specification.) |
||
| 393 | |||
| 394 | NB: This is akin to the CVC test from the Porter stemmer. The description |
||
| 395 | is unfortunately poor/ambiguous. |
||
| 396 | """ |
||
| 397 | 1 | if not term: |
|
| 398 | 1 | return False |
|
| 399 | 1 | if len(term) == 2: |
|
| 400 | 1 | if term[-2] in vowels and term[-1] not in vowels: |
|
| 401 | 1 | return True |
|
| 402 | 1 | elif len(term) >= 3: |
|
| 403 | 1 | if ( |
|
| 404 | term[-3] not in vowels |
||
| 405 | and term[-2] in vowels |
||
| 406 | and term[-1] in codanonvowels |
||
| 407 | ): |
||
| 408 | 1 | return True |
|
| 409 | 1 | return False |
|
| 410 | |||
| 411 | |||
| 412 | 1 | def _sb_short_word(term, vowels, codanonvowels, r1_prefixes=None): |
|
| 413 | """Return True iff term is a short word. |
||
| 414 | |||
| 415 | (...according to the Porter2 specification.) |
||
| 416 | """ |
||
| 417 | 1 | if _sb_r1(term, vowels, r1_prefixes) == len( |
|
| 418 | term |
||
| 419 | ) and _sb_ends_in_short_syllable(term, vowels, codanonvowels): |
||
| 420 | 1 | return True |
|
| 421 | 1 | return False |
|
| 422 | |||
| 423 | |||
| 424 | 1 | def porter2(word, early_english=False): |
|
| 425 | """Return the Porter2 (Snowball English) stem. |
||
| 426 | |||
| 427 | The Porter2 (Snowball English) stemmer is defined in :cite:`Porter:2002`. |
||
| 428 | |||
| 429 | :param str word: the word to calculate the stem of |
||
| 430 | :param bool early_english: set to True in order to remove -eth & -est |
||
| 431 | (2nd & 3rd person singular verbal agreement suffixes) |
||
| 432 | :returns: word stem |
||
| 433 | :rtype: str |
||
| 434 | |||
| 435 | >>> porter2('reading') |
||
| 436 | 'read' |
||
| 437 | >>> porter2('suspension') |
||
| 438 | 'suspens' |
||
| 439 | >>> porter2('elusiveness') |
||
| 440 | 'elus' |
||
| 441 | |||
| 442 | >>> porter2('eateth', early_english=True) |
||
| 443 | 'eat' |
||
| 444 | """ |
||
| 445 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y'} |
|
| 446 | 1 | _codanonvowels = { |
|
| 447 | "'", |
||
| 448 | 'b', |
||
| 449 | 'c', |
||
| 450 | 'd', |
||
| 451 | 'f', |
||
| 452 | 'g', |
||
| 453 | 'h', |
||
| 454 | 'j', |
||
| 455 | 'k', |
||
| 456 | 'l', |
||
| 457 | 'm', |
||
| 458 | 'n', |
||
| 459 | 'p', |
||
| 460 | 'q', |
||
| 461 | 'r', |
||
| 462 | 's', |
||
| 463 | 't', |
||
| 464 | 'v', |
||
| 465 | 'z', |
||
| 466 | } |
||
| 467 | 1 | _doubles = {'bb', 'dd', 'ff', 'gg', 'mm', 'nn', 'pp', 'rr', 'tt'} |
|
| 468 | 1 | _li = {'c', 'd', 'e', 'g', 'h', 'k', 'm', 'n', 'r', 't'} |
|
| 469 | |||
| 470 | # R1 prefixes should be in order from longest to shortest to prevent |
||
| 471 | # masking |
||
| 472 | 1 | _r1_prefixes = ('commun', 'gener', 'arsen') |
|
| 473 | 1 | _exception1dict = { # special changes: |
|
| 474 | 'skis': 'ski', |
||
| 475 | 'skies': 'sky', |
||
| 476 | 'dying': 'die', |
||
| 477 | 'lying': 'lie', |
||
| 478 | 'tying': 'tie', |
||
| 479 | # special -LY cases: |
||
| 480 | 'idly': 'idl', |
||
| 481 | 'gently': 'gentl', |
||
| 482 | 'ugly': 'ugli', |
||
| 483 | 'early': 'earli', |
||
| 484 | 'only': 'onli', |
||
| 485 | 'singly': 'singl', |
||
| 486 | } |
||
| 487 | 1 | _exception1set = { |
|
| 488 | 'sky', |
||
| 489 | 'news', |
||
| 490 | 'howe', |
||
| 491 | 'atlas', |
||
| 492 | 'cosmos', |
||
| 493 | 'bias', |
||
| 494 | 'andes', |
||
| 495 | } |
||
| 496 | 1 | _exception2set = { |
|
| 497 | 'inning', |
||
| 498 | 'outing', |
||
| 499 | 'canning', |
||
| 500 | 'herring', |
||
| 501 | 'earring', |
||
| 502 | 'proceed', |
||
| 503 | 'exceed', |
||
| 504 | 'succeed', |
||
| 505 | } |
||
| 506 | |||
| 507 | # lowercase, normalize, and compose |
||
| 508 | 1 | word = normalize('NFC', text_type(word.lower())) |
|
| 509 | # replace apostrophe-like characters with U+0027, per |
||
| 510 | # http://snowball.tartarus.org/texts/apostrophe.html |
||
| 511 | 1 | word = word.replace('’', '\'') |
|
| 512 | 1 | word = word.replace('’', '\'') |
|
| 513 | |||
| 514 | # Exceptions 1 |
||
| 515 | 1 | if word in _exception1dict: |
|
| 516 | 1 | return _exception1dict[word] |
|
| 517 | 1 | elif word in _exception1set: |
|
| 518 | 1 | return word |
|
| 519 | |||
| 520 | # Return word if stem is shorter than 3 |
||
| 521 | 1 | if len(word) < 3: |
|
| 522 | 1 | return word |
|
| 523 | |||
| 524 | # Remove initial ', if present. |
||
| 525 | 1 | while word and word[0] == '\'': |
|
| 526 | 1 | word = word[1:] |
|
| 527 | # Return word if stem is shorter than 2 |
||
| 528 | 1 | if len(word) < 2: |
|
| 529 | 1 | return word |
|
| 530 | |||
| 531 | # Re-map vocalic Y to y (Y will be C, y will be V) |
||
| 532 | 1 | if word[0] == 'y': |
|
| 533 | 1 | word = 'Y' + word[1:] |
|
| 534 | 1 | for i in range(1, len(word)): |
|
| 535 | 1 | if word[i] == 'y' and word[i - 1] in _vowels: |
|
| 536 | 1 | word = word[:i] + 'Y' + word[i + 1 :] |
|
| 537 | |||
| 538 | 1 | r1_start = _sb_r1(word, _vowels, _r1_prefixes) |
|
| 539 | 1 | r2_start = _sb_r2(word, _vowels, _r1_prefixes) |
|
| 540 | |||
| 541 | # Step 0 |
||
| 542 | 1 | if word[-3:] == '\'s\'': |
|
| 543 | 1 | word = word[:-3] |
|
| 544 | 1 | elif word[-2:] == '\'s': |
|
| 545 | 1 | word = word[:-2] |
|
| 546 | 1 | elif word[-1:] == '\'': |
|
| 547 | 1 | word = word[:-1] |
|
| 548 | # Return word if stem is shorter than 2 |
||
| 549 | 1 | if len(word) < 3: |
|
| 550 | 1 | return word |
|
| 551 | |||
| 552 | # Step 1a |
||
| 553 | 1 | if word[-4:] == 'sses': |
|
| 554 | 1 | word = word[:-2] |
|
| 555 | 1 | elif word[-3:] in {'ied', 'ies'}: |
|
| 556 | 1 | if len(word) > 4: |
|
| 557 | 1 | word = word[:-2] |
|
| 558 | else: |
||
| 559 | 1 | word = word[:-1] |
|
| 560 | 1 | elif word[-2:] in {'us', 'ss'}: |
|
| 561 | 1 | pass |
|
| 562 | 1 | elif word[-1] == 's': |
|
| 563 | 1 | if _sb_has_vowel(word[:-2], _vowels): |
|
| 564 | 1 | word = word[:-1] |
|
| 565 | |||
| 566 | # Exceptions 2 |
||
| 567 | 1 | if word in _exception2set: |
|
| 568 | 1 | return word |
|
| 569 | |||
| 570 | # Step 1b |
||
| 571 | 1 | step1b_flag = False |
|
| 572 | 1 | if word[-5:] == 'eedly': |
|
| 573 | 1 | if len(word[r1_start:]) >= 5: |
|
| 574 | 1 | word = word[:-3] |
|
| 575 | 1 | elif word[-5:] == 'ingly': |
|
| 576 | 1 | if _sb_has_vowel(word[:-5], _vowels): |
|
| 577 | 1 | word = word[:-5] |
|
| 578 | 1 | step1b_flag = True |
|
| 579 | 1 | elif word[-4:] == 'edly': |
|
| 580 | 1 | if _sb_has_vowel(word[:-4], _vowels): |
|
| 581 | 1 | word = word[:-4] |
|
| 582 | 1 | step1b_flag = True |
|
| 583 | 1 | elif word[-3:] == 'eed': |
|
| 584 | 1 | if len(word[r1_start:]) >= 3: |
|
| 585 | 1 | word = word[:-1] |
|
| 586 | 1 | elif word[-3:] == 'ing': |
|
| 587 | 1 | if _sb_has_vowel(word[:-3], _vowels): |
|
| 588 | 1 | word = word[:-3] |
|
| 589 | 1 | step1b_flag = True |
|
| 590 | 1 | elif word[-2:] == 'ed': |
|
| 591 | 1 | if _sb_has_vowel(word[:-2], _vowels): |
|
| 592 | 1 | word = word[:-2] |
|
| 593 | 1 | step1b_flag = True |
|
| 594 | 1 | elif early_english: |
|
| 595 | 1 | if word[-3:] == 'est': |
|
| 596 | 1 | if _sb_has_vowel(word[:-3], _vowels): |
|
| 597 | 1 | word = word[:-3] |
|
| 598 | 1 | step1b_flag = True |
|
| 599 | 1 | elif word[-3:] == 'eth': |
|
| 600 | 1 | if _sb_has_vowel(word[:-3], _vowels): |
|
| 601 | 1 | word = word[:-3] |
|
| 602 | 1 | step1b_flag = True |
|
| 603 | |||
| 604 | 1 | if step1b_flag: |
|
| 605 | 1 | if word[-2:] in {'at', 'bl', 'iz'}: |
|
| 606 | 1 | word += 'e' |
|
| 607 | 1 | elif word[-2:] in _doubles: |
|
| 608 | 1 | word = word[:-1] |
|
| 609 | 1 | elif _sb_short_word(word, _vowels, _codanonvowels, _r1_prefixes): |
|
| 610 | 1 | word += 'e' |
|
| 611 | |||
| 612 | # Step 1c |
||
| 613 | 1 | if len(word) > 2 and word[-1] in {'Y', 'y'} and word[-2] not in _vowels: |
|
| 614 | 1 | word = word[:-1] + 'i' |
|
| 615 | |||
| 616 | # Step 2 |
||
| 617 | 1 | if word[-2] == 'a': |
|
| 618 | 1 | if word[-7:] == 'ational': |
|
| 619 | 1 | if len(word[r1_start:]) >= 7: |
|
| 620 | 1 | word = word[:-5] + 'e' |
|
| 621 | 1 | elif word[-6:] == 'tional': |
|
| 622 | 1 | if len(word[r1_start:]) >= 6: |
|
| 623 | 1 | word = word[:-2] |
|
| 624 | 1 | elif word[-2] == 'c': |
|
| 625 | 1 | if word[-4:] in {'enci', 'anci'}: |
|
| 626 | 1 | if len(word[r1_start:]) >= 4: |
|
| 627 | 1 | word = word[:-1] + 'e' |
|
| 628 | 1 | elif word[-2] == 'e': |
|
| 629 | 1 | if word[-4:] == 'izer': |
|
| 630 | 1 | if len(word[r1_start:]) >= 4: |
|
| 631 | 1 | word = word[:-1] |
|
| 632 | 1 | elif word[-2] == 'g': |
|
| 633 | 1 | if word[-3:] == 'ogi': |
|
| 634 | 1 | if r1_start >= 1 and len(word[r1_start:]) >= 3 and word[-4] == 'l': |
|
| 635 | 1 | word = word[:-1] |
|
| 636 | 1 | elif word[-2] == 'l': |
|
| 637 | 1 | if word[-6:] == 'lessli': |
|
| 638 | 1 | if len(word[r1_start:]) >= 6: |
|
| 639 | 1 | word = word[:-2] |
|
| 640 | 1 | elif word[-5:] in {'entli', 'fulli', 'ousli'}: |
|
| 641 | 1 | if len(word[r1_start:]) >= 5: |
|
| 642 | 1 | word = word[:-2] |
|
| 643 | 1 | elif word[-4:] == 'abli': |
|
| 644 | 1 | if len(word[r1_start:]) >= 4: |
|
| 645 | 1 | word = word[:-1] + 'e' |
|
| 646 | 1 | elif word[-4:] == 'alli': |
|
| 647 | 1 | if len(word[r1_start:]) >= 4: |
|
| 648 | 1 | word = word[:-2] |
|
| 649 | 1 | elif word[-3:] == 'bli': |
|
| 650 | 1 | if len(word[r1_start:]) >= 3: |
|
| 651 | 1 | word = word[:-1] + 'e' |
|
| 652 | 1 | elif word[-2:] == 'li': |
|
| 653 | 1 | if r1_start >= 1 and len(word[r1_start:]) >= 2 and word[-3] in _li: |
|
| 654 | 1 | word = word[:-2] |
|
| 655 | 1 | elif word[-2] == 'o': |
|
| 656 | 1 | if word[-7:] == 'ization': |
|
| 657 | 1 | if len(word[r1_start:]) >= 7: |
|
| 658 | 1 | word = word[:-5] + 'e' |
|
| 659 | 1 | elif word[-5:] == 'ation': |
|
| 660 | 1 | if len(word[r1_start:]) >= 5: |
|
| 661 | 1 | word = word[:-3] + 'e' |
|
| 662 | 1 | elif word[-4:] == 'ator': |
|
| 663 | 1 | if len(word[r1_start:]) >= 4: |
|
| 664 | 1 | word = word[:-2] + 'e' |
|
| 665 | 1 | elif word[-2] == 's': |
|
| 666 | 1 | if word[-7:] in {'fulness', 'ousness', 'iveness'}: |
|
| 667 | 1 | if len(word[r1_start:]) >= 7: |
|
| 668 | 1 | word = word[:-4] |
|
| 669 | 1 | elif word[-5:] == 'alism': |
|
| 670 | 1 | if len(word[r1_start:]) >= 5: |
|
| 671 | 1 | word = word[:-3] |
|
| 672 | 1 | elif word[-2] == 't': |
|
| 673 | 1 | if word[-6:] == 'biliti': |
|
| 674 | 1 | if len(word[r1_start:]) >= 6: |
|
| 675 | 1 | word = word[:-5] + 'le' |
|
| 676 | 1 | elif word[-5:] == 'aliti': |
|
| 677 | 1 | if len(word[r1_start:]) >= 5: |
|
| 678 | 1 | word = word[:-3] |
|
| 679 | 1 | elif word[-5:] == 'iviti': |
|
| 680 | 1 | if len(word[r1_start:]) >= 5: |
|
| 681 | 1 | word = word[:-3] + 'e' |
|
| 682 | |||
| 683 | # Step 3 |
||
| 684 | 1 | if word[-7:] == 'ational': |
|
| 685 | 1 | if len(word[r1_start:]) >= 7: |
|
| 686 | 1 | word = word[:-5] + 'e' |
|
| 687 | 1 | elif word[-6:] == 'tional': |
|
| 688 | 1 | if len(word[r1_start:]) >= 6: |
|
| 689 | 1 | word = word[:-2] |
|
| 690 | 1 | elif word[-5:] in {'alize', 'icate', 'iciti'}: |
|
| 691 | 1 | if len(word[r1_start:]) >= 5: |
|
| 692 | 1 | word = word[:-3] |
|
| 693 | 1 | elif word[-5:] == 'ative': |
|
| 694 | 1 | if len(word[r2_start:]) >= 5: |
|
| 695 | 1 | word = word[:-5] |
|
| 696 | 1 | elif word[-4:] == 'ical': |
|
| 697 | 1 | if len(word[r1_start:]) >= 4: |
|
| 698 | 1 | word = word[:-2] |
|
| 699 | 1 | elif word[-4:] == 'ness': |
|
| 700 | 1 | if len(word[r1_start:]) >= 4: |
|
| 701 | 1 | word = word[:-4] |
|
| 702 | 1 | elif word[-3:] == 'ful': |
|
| 703 | 1 | if len(word[r1_start:]) >= 3: |
|
| 704 | 1 | word = word[:-3] |
|
| 705 | |||
| 706 | # Step 4 |
||
| 707 | 1 | for suffix in ( |
|
| 708 | 'ement', |
||
| 709 | 'ance', |
||
| 710 | 'ence', |
||
| 711 | 'able', |
||
| 712 | 'ible', |
||
| 713 | 'ment', |
||
| 714 | 'ant', |
||
| 715 | 'ent', |
||
| 716 | 'ism', |
||
| 717 | 'ate', |
||
| 718 | 'iti', |
||
| 719 | 'ous', |
||
| 720 | 'ive', |
||
| 721 | 'ize', |
||
| 722 | 'al', |
||
| 723 | 'er', |
||
| 724 | 'ic', |
||
| 725 | ): |
||
| 726 | 1 | if word[-len(suffix) :] == suffix: |
|
| 727 | 1 | if len(word[r2_start:]) >= len(suffix): |
|
| 728 | 1 | word = word[: -len(suffix)] |
|
| 729 | 1 | break |
|
| 730 | else: |
||
| 731 | 1 | if word[-3:] == 'ion': |
|
| 732 | 1 | if ( |
|
| 733 | len(word[r2_start:]) >= 3 |
||
| 734 | and len(word) >= 4 |
||
| 735 | and word[-4] in tuple('st') |
||
| 736 | ): |
||
| 737 | 1 | word = word[:-3] |
|
| 738 | |||
| 739 | # Step 5 |
||
| 740 | 1 | if word[-1] == 'e': |
|
| 741 | 1 | if len(word[r2_start:]) >= 1 or ( |
|
| 742 | len(word[r1_start:]) >= 1 |
||
| 743 | and not _sb_ends_in_short_syllable( |
||
| 744 | word[:-1], _vowels, _codanonvowels |
||
| 745 | ) |
||
| 746 | ): |
||
| 747 | 1 | word = word[:-1] |
|
| 748 | 1 | elif word[-1] == 'l': |
|
| 749 | 1 | if len(word[r2_start:]) >= 1 and word[-2] == 'l': |
|
| 750 | 1 | word = word[:-1] |
|
| 751 | |||
| 752 | # Change 'Y' back to 'y' if it survived stemming |
||
| 753 | 1 | for i in range(0, len(word)): |
|
| 754 | 1 | if word[i] == 'Y': |
|
| 755 | 1 | word = word[:i] + 'y' + word[i + 1 :] |
|
| 756 | |||
| 757 | 1 | return word |
|
| 758 | |||
| 759 | |||
| 760 | 1 | def sb_german(word, alternate_vowels=False): |
|
| 761 | """Return Snowball German stem. |
||
| 762 | |||
| 763 | The Snowball German stemmer is defined at: |
||
| 764 | http://snowball.tartarus.org/algorithms/german/stemmer.html |
||
| 765 | |||
| 766 | :param str word: the word to calculate the stem of |
||
| 767 | :param bool alternate_vowels: composes ae as ä, oe as ö, and ue as ü before |
||
| 768 | running the algorithm |
||
| 769 | :returns: word stem |
||
| 770 | :rtype: str |
||
| 771 | |||
| 772 | >>> sb_german('lesen') |
||
| 773 | 'les' |
||
| 774 | >>> sb_german('graues') |
||
| 775 | 'grau' |
||
| 776 | >>> sb_german('buchstabieren') |
||
| 777 | 'buchstabi' |
||
| 778 | """ |
||
| 779 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y', 'ä', 'ö', 'ü'} |
|
| 780 | 1 | _s_endings = {'b', 'd', 'f', 'g', 'h', 'k', 'l', 'm', 'n', 'r', 't'} |
|
| 781 | 1 | _st_endings = {'b', 'd', 'f', 'g', 'h', 'k', 'l', 'm', 'n', 't'} |
|
| 782 | |||
| 783 | # lowercase, normalize, and compose |
||
| 784 | 1 | word = normalize('NFC', word.lower()) |
|
| 785 | 1 | word = word.replace('ß', 'ss') |
|
| 786 | |||
| 787 | 1 | if len(word) > 2: |
|
| 788 | 1 | for i in range(2, len(word)): |
|
| 789 | 1 | if word[i] in _vowels and word[i - 2] in _vowels: |
|
| 790 | 1 | if word[i - 1] == 'u': |
|
| 791 | 1 | word = word[: i - 1] + 'U' + word[i:] |
|
| 792 | 1 | elif word[i - 1] == 'y': |
|
| 793 | 1 | word = word[: i - 1] + 'Y' + word[i:] |
|
| 794 | |||
| 795 | 1 | if alternate_vowels: |
|
| 796 | 1 | word = word.replace('ae', 'ä') |
|
| 797 | 1 | word = word.replace('oe', 'ö') |
|
| 798 | 1 | word = word.replace('que', 'Q') |
|
| 799 | 1 | word = word.replace('ue', 'ü') |
|
| 800 | 1 | word = word.replace('Q', 'que') |
|
| 801 | |||
| 802 | 1 | r1_start = max(3, _sb_r1(word, _vowels)) |
|
| 803 | 1 | r2_start = _sb_r2(word, _vowels) |
|
| 804 | |||
| 805 | # Step 1 |
||
| 806 | 1 | niss_flag = False |
|
| 807 | 1 | if word[-3:] == 'ern': |
|
| 808 | 1 | if len(word[r1_start:]) >= 3: |
|
| 809 | 1 | word = word[:-3] |
|
| 810 | 1 | elif word[-2:] == 'em': |
|
| 811 | 1 | if len(word[r1_start:]) >= 2: |
|
| 812 | 1 | word = word[:-2] |
|
| 813 | 1 | elif word[-2:] == 'er': |
|
| 814 | 1 | if len(word[r1_start:]) >= 2: |
|
| 815 | 1 | word = word[:-2] |
|
| 816 | 1 | elif word[-2:] == 'en': |
|
| 817 | 1 | if len(word[r1_start:]) >= 2: |
|
| 818 | 1 | word = word[:-2] |
|
| 819 | 1 | niss_flag = True |
|
| 820 | 1 | elif word[-2:] == 'es': |
|
| 821 | 1 | if len(word[r1_start:]) >= 2: |
|
| 822 | 1 | word = word[:-2] |
|
| 823 | 1 | niss_flag = True |
|
| 824 | 1 | elif word[-1:] == 'e': |
|
| 825 | 1 | if len(word[r1_start:]) >= 1: |
|
| 826 | 1 | word = word[:-1] |
|
| 827 | 1 | niss_flag = True |
|
| 828 | 1 | elif word[-1:] == 's': |
|
| 829 | 1 | if ( |
|
| 830 | len(word[r1_start:]) >= 1 |
||
| 831 | and len(word) >= 2 |
||
| 832 | and word[-2] in _s_endings |
||
| 833 | ): |
||
| 834 | 1 | word = word[:-1] |
|
| 835 | |||
| 836 | 1 | if niss_flag and word[-4:] == 'niss': |
|
| 837 | 1 | word = word[:-1] |
|
| 838 | |||
| 839 | # Step 2 |
||
| 840 | 1 | if word[-3:] == 'est': |
|
| 841 | 1 | if len(word[r1_start:]) >= 3: |
|
| 842 | 1 | word = word[:-3] |
|
| 843 | 1 | elif word[-2:] == 'en': |
|
| 844 | 1 | if len(word[r1_start:]) >= 2: |
|
| 845 | 1 | word = word[:-2] |
|
| 846 | 1 | elif word[-2:] == 'er': |
|
| 847 | 1 | if len(word[r1_start:]) >= 2: |
|
| 848 | 1 | word = word[:-2] |
|
| 849 | 1 | elif word[-2:] == 'st': |
|
| 850 | 1 | if ( |
|
| 851 | len(word[r1_start:]) >= 2 |
||
| 852 | and len(word) >= 6 |
||
| 853 | and word[-3] in _st_endings |
||
| 854 | ): |
||
| 855 | 1 | word = word[:-2] |
|
| 856 | |||
| 857 | # Step 3 |
||
| 858 | 1 | if word[-4:] == 'isch': |
|
| 859 | 1 | if len(word[r2_start:]) >= 4 and word[-5] != 'e': |
|
| 860 | 1 | word = word[:-4] |
|
| 861 | 1 | elif word[-4:] in {'lich', 'heit'}: |
|
| 862 | 1 | if len(word[r2_start:]) >= 4: |
|
| 863 | 1 | word = word[:-4] |
|
| 864 | 1 | if word[-2:] in {'er', 'en'} and len(word[r1_start:]) >= 2: |
|
| 865 | 1 | word = word[:-2] |
|
| 866 | 1 | elif word[-4:] == 'keit': |
|
| 867 | 1 | if len(word[r2_start:]) >= 4: |
|
| 868 | 1 | word = word[:-4] |
|
| 869 | 1 | if word[-4:] == 'lich' and len(word[r2_start:]) >= 4: |
|
| 870 | 1 | word = word[:-4] |
|
| 871 | 1 | elif word[-2:] == 'ig' and len(word[r2_start:]) >= 2: |
|
| 872 | 1 | word = word[:-2] |
|
| 873 | 1 | elif word[-3:] in {'end', 'ung'}: |
|
| 874 | 1 | if len(word[r2_start:]) >= 3: |
|
| 875 | 1 | word = word[:-3] |
|
| 876 | 1 | if ( |
|
| 877 | word[-2:] == 'ig' |
||
| 878 | and len(word[r2_start:]) >= 2 |
||
| 879 | and word[-3] != 'e' |
||
| 880 | ): |
||
| 881 | 1 | word = word[:-2] |
|
| 882 | 1 | elif word[-2:] in {'ig', 'ik'}: |
|
| 883 | 1 | if len(word[r2_start:]) >= 2 and word[-3] != 'e': |
|
| 884 | 1 | word = word[:-2] |
|
| 885 | |||
| 886 | # Change 'Y' and 'U' back to lowercase if survived stemming |
||
| 887 | 1 | for i in range(0, len(word)): |
|
| 888 | 1 | if word[i] == 'Y': |
|
| 889 | 1 | word = word[:i] + 'y' + word[i + 1 :] |
|
| 890 | 1 | elif word[i] == 'U': |
|
| 891 | 1 | word = word[:i] + 'u' + word[i + 1 :] |
|
| 892 | |||
| 893 | # Remove umlauts |
||
| 894 | 1 | _umlauts = dict(zip((ord(_) for _ in 'äöü'), 'aou')) |
|
| 895 | 1 | word = word.translate(_umlauts) |
|
| 896 | |||
| 897 | 1 | return word |
|
| 898 | |||
| 899 | |||
| 900 | 1 | def sb_dutch(word): |
|
| 901 | """Return Snowball Dutch stem. |
||
| 902 | |||
| 903 | The Snowball Dutch stemmer is defined at: |
||
| 904 | http://snowball.tartarus.org/algorithms/dutch/stemmer.html |
||
| 905 | |||
| 906 | :param str word: the word to calculate the stem of |
||
| 907 | :returns: word stem |
||
| 908 | :rtype: str |
||
| 909 | |||
| 910 | >>> sb_dutch('lezen') |
||
| 911 | 'lez' |
||
| 912 | >>> sb_dutch('opschorting') |
||
| 913 | 'opschort' |
||
| 914 | >>> sb_dutch('ongrijpbaarheid') |
||
| 915 | 'ongrijp' |
||
| 916 | """ |
||
| 917 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y', 'è'} |
|
| 918 | 1 | _not_s_endings = {'a', 'e', 'i', 'j', 'o', 'u', 'y', 'è'} |
|
| 919 | |||
| 920 | 1 | def _undouble(word): |
|
| 921 | """Undouble endings -kk, -dd, and -tt.""" |
||
| 922 | 1 | if ( |
|
| 923 | len(word) > 1 |
||
| 924 | and word[-1] == word[-2] |
||
| 925 | and word[-1] in {'d', 'k', 't'} |
||
| 926 | ): |
||
| 927 | 1 | return word[:-1] |
|
| 928 | 1 | return word |
|
| 929 | |||
| 930 | # lowercase, normalize, decompose, filter umlauts & acutes out, and compose |
||
| 931 | 1 | word = normalize('NFC', text_type(word.lower())) |
|
| 932 | 1 | _accented = dict(zip((ord(_) for _ in 'äëïöüáéíóú'), 'aeiouaeiou')) |
|
| 933 | 1 | word = word.translate(_accented) |
|
| 934 | |||
| 935 | 1 | for i in range(len(word)): |
|
| 936 | 1 | if i == 0 and word[0] == 'y': |
|
| 937 | 1 | word = 'Y' + word[1:] |
|
| 938 | 1 | elif word[i] == 'y' and word[i - 1] in _vowels: |
|
| 939 | 1 | word = word[:i] + 'Y' + word[i + 1 :] |
|
| 940 | 1 | elif ( |
|
| 941 | word[i] == 'i' |
||
| 942 | and word[i - 1] in _vowels |
||
| 943 | and i + 1 < len(word) |
||
| 944 | and word[i + 1] in _vowels |
||
| 945 | ): |
||
| 946 | 1 | word = word[:i] + 'I' + word[i + 1 :] |
|
| 947 | |||
| 948 | 1 | r1_start = max(3, _sb_r1(word, _vowels)) |
|
| 949 | 1 | r2_start = _sb_r2(word, _vowels) |
|
| 950 | |||
| 951 | # Step 1 |
||
| 952 | 1 | if word[-5:] == 'heden': |
|
| 953 | 1 | if len(word[r1_start:]) >= 5: |
|
| 954 | 1 | word = word[:-3] + 'id' |
|
| 955 | 1 | elif word[-3:] == 'ene': |
|
| 956 | 1 | if len(word[r1_start:]) >= 3 and ( |
|
| 957 | word[-4] not in _vowels and word[-6:-3] != 'gem' |
||
| 958 | ): |
||
| 959 | 1 | word = _undouble(word[:-3]) |
|
| 960 | 1 | elif word[-2:] == 'en': |
|
| 961 | 1 | if len(word[r1_start:]) >= 2 and ( |
|
| 962 | word[-3] not in _vowels and word[-5:-2] != 'gem' |
||
| 963 | ): |
||
| 964 | 1 | word = _undouble(word[:-2]) |
|
| 965 | 1 | elif word[-2:] == 'se': |
|
| 966 | 1 | if len(word[r1_start:]) >= 2 and word[-3] not in _not_s_endings: |
|
| 967 | 1 | word = word[:-2] |
|
| 968 | 1 | elif word[-1:] == 's': |
|
| 969 | 1 | if len(word[r1_start:]) >= 1 and word[-2] not in _not_s_endings: |
|
| 970 | 1 | word = word[:-1] |
|
| 971 | |||
| 972 | # Step 2 |
||
| 973 | 1 | e_removed = False |
|
| 974 | 1 | if word[-1:] == 'e': |
|
| 975 | 1 | if len(word[r1_start:]) >= 1 and word[-2] not in _vowels: |
|
| 976 | 1 | word = _undouble(word[:-1]) |
|
| 977 | 1 | e_removed = True |
|
| 978 | |||
| 979 | # Step 3a |
||
| 980 | 1 | if word[-4:] == 'heid': |
|
| 981 | 1 | if len(word[r2_start:]) >= 4 and word[-5] != 'c': |
|
| 982 | 1 | word = word[:-4] |
|
| 983 | 1 | if word[-2:] == 'en': |
|
| 984 | 1 | if len(word[r1_start:]) >= 2 and ( |
|
| 985 | word[-3] not in _vowels and word[-5:-2] != 'gem' |
||
| 986 | ): |
||
| 987 | 1 | word = _undouble(word[:-2]) |
|
| 988 | |||
| 989 | # Step 3b |
||
| 990 | 1 | if word[-4:] == 'lijk': |
|
| 991 | 1 | if len(word[r2_start:]) >= 4: |
|
| 992 | 1 | word = word[:-4] |
|
| 993 | # Repeat step 2 |
||
| 994 | 1 | if word[-1:] == 'e': |
|
| 995 | 1 | if len(word[r1_start:]) >= 1 and word[-2] not in _vowels: |
|
| 996 | 1 | word = _undouble(word[:-1]) |
|
| 997 | 1 | elif word[-4:] == 'baar': |
|
| 998 | 1 | if len(word[r2_start:]) >= 4: |
|
| 999 | 1 | word = word[:-4] |
|
| 1000 | 1 | elif word[-3:] in ('end', 'ing'): |
|
| 1001 | 1 | if len(word[r2_start:]) >= 3: |
|
| 1002 | 1 | word = word[:-3] |
|
| 1003 | 1 | if ( |
|
| 1004 | word[-2:] == 'ig' |
||
| 1005 | and len(word[r2_start:]) >= 2 |
||
| 1006 | and word[-3] != 'e' |
||
| 1007 | ): |
||
| 1008 | 1 | word = word[:-2] |
|
| 1009 | else: |
||
| 1010 | 1 | word = _undouble(word) |
|
| 1011 | 1 | elif word[-3:] == 'bar': |
|
| 1012 | 1 | if len(word[r2_start:]) >= 3 and e_removed: |
|
| 1013 | 1 | word = word[:-3] |
|
| 1014 | 1 | elif word[-2:] == 'ig': |
|
| 1015 | 1 | if len(word[r2_start:]) >= 2 and word[-3] != 'e': |
|
| 1016 | 1 | word = word[:-2] |
|
| 1017 | |||
| 1018 | # Step 4 |
||
| 1019 | 1 | if ( |
|
| 1020 | len(word) >= 4 |
||
| 1021 | and word[-3] == word[-2] |
||
| 1022 | and word[-2] in {'a', 'e', 'o', 'u'} |
||
| 1023 | and word[-4] not in _vowels |
||
| 1024 | and word[-1] not in _vowels |
||
| 1025 | and word[-1] != 'I' |
||
| 1026 | ): |
||
| 1027 | 1 | word = word[:-2] + word[-1] |
|
| 1028 | |||
| 1029 | # Change 'Y' and 'U' back to lowercase if survived stemming |
||
| 1030 | 1 | for i in range(0, len(word)): |
|
| 1031 | 1 | if word[i] == 'Y': |
|
| 1032 | 1 | word = word[:i] + 'y' + word[i + 1 :] |
|
| 1033 | 1 | elif word[i] == 'I': |
|
| 1034 | 1 | word = word[:i] + 'i' + word[i + 1 :] |
|
| 1035 | |||
| 1036 | 1 | return word |
|
| 1037 | |||
| 1038 | |||
| 1039 | 1 | def sb_norwegian(word): |
|
| 1040 | """Return Snowball Norwegian stem. |
||
| 1041 | |||
| 1042 | The Snowball Norwegian stemmer is defined at: |
||
| 1043 | http://snowball.tartarus.org/algorithms/norwegian/stemmer.html |
||
| 1044 | |||
| 1045 | :param str word: the word to calculate the stem of |
||
| 1046 | :returns: word stem |
||
| 1047 | :rtype: str |
||
| 1048 | |||
| 1049 | >>> sb_norwegian('lese') |
||
| 1050 | 'les' |
||
| 1051 | >>> sb_norwegian('suspensjon') |
||
| 1052 | 'suspensjon' |
||
| 1053 | >>> sb_norwegian('sikkerhet') |
||
| 1054 | 'sikker' |
||
| 1055 | """ |
||
| 1056 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y', 'å', 'æ', 'ø'} |
|
| 1057 | 1 | _s_endings = { |
|
| 1058 | 'b', |
||
| 1059 | 'c', |
||
| 1060 | 'd', |
||
| 1061 | 'f', |
||
| 1062 | 'g', |
||
| 1063 | 'h', |
||
| 1064 | 'j', |
||
| 1065 | 'l', |
||
| 1066 | 'm', |
||
| 1067 | 'n', |
||
| 1068 | 'o', |
||
| 1069 | 'p', |
||
| 1070 | 'r', |
||
| 1071 | 't', |
||
| 1072 | 'v', |
||
| 1073 | 'y', |
||
| 1074 | 'z', |
||
| 1075 | } |
||
| 1076 | # lowercase, normalize, and compose |
||
| 1077 | 1 | word = normalize('NFC', text_type(word.lower())) |
|
| 1078 | |||
| 1079 | 1 | r1_start = min(max(3, _sb_r1(word, _vowels)), len(word)) |
|
| 1080 | |||
| 1081 | # Step 1 |
||
| 1082 | 1 | _r1 = word[r1_start:] |
|
| 1083 | 1 | if _r1[-7:] == 'hetenes': |
|
| 1084 | 1 | word = word[:-7] |
|
| 1085 | 1 | elif _r1[-6:] in {'hetene', 'hetens'}: |
|
| 1086 | 1 | word = word[:-6] |
|
| 1087 | 1 | elif _r1[-5:] in {'heten', 'heter', 'endes'}: |
|
| 1088 | 1 | word = word[:-5] |
|
| 1089 | 1 | elif _r1[-4:] in {'ande', 'ende', 'edes', 'enes', 'erte'}: |
|
| 1090 | 1 | if word[-4:] == 'erte': |
|
| 1091 | 1 | word = word[:-2] |
|
| 1092 | else: |
||
| 1093 | 1 | word = word[:-4] |
|
| 1094 | 1 | elif _r1[-3:] in { |
|
| 1095 | 'ede', |
||
| 1096 | 'ane', |
||
| 1097 | 'ene', |
||
| 1098 | 'ens', |
||
| 1099 | 'ers', |
||
| 1100 | 'ets', |
||
| 1101 | 'het', |
||
| 1102 | 'ast', |
||
| 1103 | 'ert', |
||
| 1104 | }: |
||
| 1105 | 1 | if word[-3:] == 'ert': |
|
| 1106 | 1 | word = word[:-1] |
|
| 1107 | else: |
||
| 1108 | 1 | word = word[:-3] |
|
| 1109 | 1 | elif _r1[-2:] in {'en', 'ar', 'er', 'as', 'es', 'et'}: |
|
| 1110 | 1 | word = word[:-2] |
|
| 1111 | 1 | elif _r1[-1:] in {'a', 'e'}: |
|
| 1112 | 1 | word = word[:-1] |
|
| 1113 | 1 | elif _r1[-1:] == 's': |
|
| 1114 | 1 | if (len(word) > 1 and word[-2] in _s_endings) or ( |
|
| 1115 | len(word) > 2 and word[-2] == 'k' and word[-3] not in _vowels |
||
| 1116 | ): |
||
| 1117 | 1 | word = word[:-1] |
|
| 1118 | |||
| 1119 | # Step 2 |
||
| 1120 | 1 | if word[r1_start:][-2:] in {'dt', 'vt'}: |
|
| 1121 | 1 | word = word[:-1] |
|
| 1122 | |||
| 1123 | # Step 3 |
||
| 1124 | 1 | _r1 = word[r1_start:] |
|
| 1125 | 1 | if _r1[-7:] == 'hetslov': |
|
| 1126 | 1 | word = word[:-7] |
|
| 1127 | 1 | elif _r1[-4:] in {'eleg', 'elig', 'elov', 'slov'}: |
|
| 1128 | 1 | word = word[:-4] |
|
| 1129 | 1 | elif _r1[-3:] in {'leg', 'eig', 'lig', 'els', 'lov'}: |
|
| 1130 | 1 | word = word[:-3] |
|
| 1131 | 1 | elif _r1[-2:] == 'ig': |
|
| 1132 | 1 | word = word[:-2] |
|
| 1133 | |||
| 1134 | 1 | return word |
|
| 1135 | |||
| 1136 | |||
| 1137 | 1 | def sb_swedish(word): |
|
| 1138 | """Return Snowball Swedish stem. |
||
| 1139 | |||
| 1140 | The Snowball Swedish stemmer is defined at: |
||
| 1141 | http://snowball.tartarus.org/algorithms/swedish/stemmer.html |
||
| 1142 | |||
| 1143 | :param str word: the word to calculate the stem of |
||
| 1144 | :returns: word stem |
||
| 1145 | :rtype: str |
||
| 1146 | |||
| 1147 | >>> sb_swedish('undervisa') |
||
| 1148 | 'undervis' |
||
| 1149 | >>> sb_swedish('suspension') |
||
| 1150 | 'suspension' |
||
| 1151 | >>> sb_swedish('visshet') |
||
| 1152 | 'viss' |
||
| 1153 | """ |
||
| 1154 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y', 'ä', 'å', 'ö'} |
|
| 1155 | 1 | _s_endings = { |
|
| 1156 | 'b', |
||
| 1157 | 'c', |
||
| 1158 | 'd', |
||
| 1159 | 'f', |
||
| 1160 | 'g', |
||
| 1161 | 'h', |
||
| 1162 | 'j', |
||
| 1163 | 'k', |
||
| 1164 | 'l', |
||
| 1165 | 'm', |
||
| 1166 | 'n', |
||
| 1167 | 'o', |
||
| 1168 | 'p', |
||
| 1169 | 'r', |
||
| 1170 | 't', |
||
| 1171 | 'v', |
||
| 1172 | 'y', |
||
| 1173 | } |
||
| 1174 | |||
| 1175 | # lowercase, normalize, and compose |
||
| 1176 | 1 | word = normalize('NFC', text_type(word.lower())) |
|
| 1177 | |||
| 1178 | 1 | r1_start = min(max(3, _sb_r1(word, _vowels)), len(word)) |
|
| 1179 | |||
| 1180 | # Step 1 |
||
| 1181 | 1 | _r1 = word[r1_start:] |
|
| 1182 | 1 | View Code Duplication | if _r1[-7:] == 'heterna': |
| 1183 | 1 | word = word[:-7] |
|
| 1184 | 1 | elif _r1[-6:] == 'hetens': |
|
| 1185 | 1 | word = word[:-6] |
|
| 1186 | 1 | elif _r1[-5:] in { |
|
| 1187 | 'anden', |
||
| 1188 | 'heten', |
||
| 1189 | 'heter', |
||
| 1190 | 'arnas', |
||
| 1191 | 'ernas', |
||
| 1192 | 'ornas', |
||
| 1193 | 'andes', |
||
| 1194 | 'arens', |
||
| 1195 | 'andet', |
||
| 1196 | }: |
||
| 1197 | 1 | word = word[:-5] |
|
| 1198 | 1 | elif _r1[-4:] in { |
|
| 1199 | 'arna', |
||
| 1200 | 'erna', |
||
| 1201 | 'orna', |
||
| 1202 | 'ande', |
||
| 1203 | 'arne', |
||
| 1204 | 'aste', |
||
| 1205 | 'aren', |
||
| 1206 | 'ades', |
||
| 1207 | 'erns', |
||
| 1208 | }: |
||
| 1209 | 1 | word = word[:-4] |
|
| 1210 | 1 | elif _r1[-3:] in {'ade', 'are', 'ern', 'ens', 'het', 'ast'}: |
|
| 1211 | 1 | word = word[:-3] |
|
| 1212 | 1 | elif _r1[-2:] in {'ad', 'en', 'ar', 'er', 'or', 'as', 'es', 'at'}: |
|
| 1213 | 1 | word = word[:-2] |
|
| 1214 | 1 | elif _r1[-1:] in {'a', 'e'}: |
|
| 1215 | 1 | word = word[:-1] |
|
| 1216 | 1 | elif _r1[-1:] == 's': |
|
| 1217 | 1 | if len(word) > 1 and word[-2] in _s_endings: |
|
| 1218 | 1 | word = word[:-1] |
|
| 1219 | |||
| 1220 | # Step 2 |
||
| 1221 | 1 | if word[r1_start:][-2:] in {'dd', 'gd', 'nn', 'dt', 'gt', 'kt', 'tt'}: |
|
| 1222 | 1 | word = word[:-1] |
|
| 1223 | |||
| 1224 | # Step 3 |
||
| 1225 | 1 | _r1 = word[r1_start:] |
|
| 1226 | 1 | if _r1[-5:] == 'fullt': |
|
| 1227 | 1 | word = word[:-1] |
|
| 1228 | 1 | elif _r1[-4:] == 'löst': |
|
| 1229 | 1 | word = word[:-1] |
|
| 1230 | 1 | elif _r1[-3:] in {'lig', 'els'}: |
|
| 1231 | 1 | word = word[:-3] |
|
| 1232 | 1 | elif _r1[-2:] == 'ig': |
|
| 1233 | 1 | word = word[:-2] |
|
| 1234 | |||
| 1235 | 1 | return word |
|
| 1236 | |||
| 1237 | |||
| 1238 | 1 | def sb_danish(word): |
|
| 1239 | """Return Snowball Danish stem. |
||
| 1240 | |||
| 1241 | The Snowball Danish stemmer is defined at: |
||
| 1242 | http://snowball.tartarus.org/algorithms/danish/stemmer.html |
||
| 1243 | |||
| 1244 | :param str word: the word to calculate the stem of |
||
| 1245 | :returns: word stem |
||
| 1246 | :rtype: str |
||
| 1247 | |||
| 1248 | >>> sb_danish('underviser') |
||
| 1249 | 'undervis' |
||
| 1250 | >>> sb_danish('suspension') |
||
| 1251 | 'suspension' |
||
| 1252 | >>> sb_danish('sikkerhed') |
||
| 1253 | 'sikker' |
||
| 1254 | """ |
||
| 1255 | 1 | _vowels = {'a', 'e', 'i', 'o', 'u', 'y', 'å', 'æ', 'ø'} |
|
| 1256 | 1 | _s_endings = { |
|
| 1257 | 'a', |
||
| 1258 | 'b', |
||
| 1259 | 'c', |
||
| 1260 | 'd', |
||
| 1261 | 'f', |
||
| 1262 | 'g', |
||
| 1263 | 'h', |
||
| 1264 | 'j', |
||
| 1265 | 'k', |
||
| 1266 | 'l', |
||
| 1267 | 'm', |
||
| 1268 | 'n', |
||
| 1269 | 'o', |
||
| 1270 | 'p', |
||
| 1271 | 'r', |
||
| 1272 | 't', |
||
| 1273 | 'v', |
||
| 1274 | 'y', |
||
| 1275 | 'z', |
||
| 1276 | 'å', |
||
| 1277 | } |
||
| 1278 | |||
| 1279 | # lowercase, normalize, and compose |
||
| 1280 | 1 | word = normalize('NFC', text_type(word.lower())) |
|
| 1281 | |||
| 1282 | 1 | r1_start = min(max(3, _sb_r1(word, _vowels)), len(word)) |
|
| 1283 | |||
| 1284 | # Step 1 |
||
| 1285 | 1 | _r1 = word[r1_start:] |
|
| 1286 | 1 | View Code Duplication | if _r1[-7:] == 'erendes': |
| 1287 | 1 | word = word[:-7] |
|
| 1288 | 1 | elif _r1[-6:] in {'erende', 'hedens'}: |
|
| 1289 | 1 | word = word[:-6] |
|
| 1290 | 1 | elif _r1[-5:] in { |
|
| 1291 | 'ethed', |
||
| 1292 | 'erede', |
||
| 1293 | 'heden', |
||
| 1294 | 'heder', |
||
| 1295 | 'endes', |
||
| 1296 | 'ernes', |
||
| 1297 | 'erens', |
||
| 1298 | 'erets', |
||
| 1299 | }: |
||
| 1300 | 1 | word = word[:-5] |
|
| 1301 | 1 | elif _r1[-4:] in { |
|
| 1302 | 'ered', |
||
| 1303 | 'ende', |
||
| 1304 | 'erne', |
||
| 1305 | 'eren', |
||
| 1306 | 'erer', |
||
| 1307 | 'heds', |
||
| 1308 | 'enes', |
||
| 1309 | 'eres', |
||
| 1310 | 'eret', |
||
| 1311 | }: |
||
| 1312 | 1 | word = word[:-4] |
|
| 1313 | 1 | elif _r1[-3:] in {'hed', 'ene', 'ere', 'ens', 'ers', 'ets'}: |
|
| 1314 | 1 | word = word[:-3] |
|
| 1315 | 1 | elif _r1[-2:] in {'en', 'er', 'es', 'et'}: |
|
| 1316 | 1 | word = word[:-2] |
|
| 1317 | 1 | elif _r1[-1:] == 'e': |
|
| 1318 | 1 | word = word[:-1] |
|
| 1319 | 1 | elif _r1[-1:] == 's': |
|
| 1320 | 1 | if len(word) > 1 and word[-2] in _s_endings: |
|
| 1321 | 1 | word = word[:-1] |
|
| 1322 | |||
| 1323 | # Step 2 |
||
| 1324 | 1 | if word[r1_start:][-2:] in {'gd', 'dt', 'gt', 'kt'}: |
|
| 1325 | 1 | word = word[:-1] |
|
| 1326 | |||
| 1327 | # Step 3 |
||
| 1328 | 1 | if word[-4:] == 'igst': |
|
| 1329 | 1 | word = word[:-2] |
|
| 1330 | |||
| 1331 | 1 | _r1 = word[r1_start:] |
|
| 1332 | 1 | repeat_step2 = False |
|
| 1333 | 1 | if _r1[-4:] == 'elig': |
|
| 1334 | 1 | word = word[:-4] |
|
| 1335 | 1 | repeat_step2 = True |
|
| 1336 | 1 | elif _r1[-4:] == 'løst': |
|
| 1337 | 1 | word = word[:-1] |
|
| 1338 | 1 | elif _r1[-3:] in {'lig', 'els'}: |
|
| 1339 | 1 | word = word[:-3] |
|
| 1340 | 1 | repeat_step2 = True |
|
| 1341 | 1 | elif _r1[-2:] == 'ig': |
|
| 1342 | 1 | word = word[:-2] |
|
| 1343 | 1 | repeat_step2 = True |
|
| 1344 | |||
| 1345 | 1 | if repeat_step2: |
|
| 1346 | 1 | if word[r1_start:][-2:] in {'gd', 'dt', 'gt', 'kt'}: |
|
| 1347 | 1 | word = word[:-1] |
|
| 1348 | |||
| 1349 | # Step 4 |
||
| 1350 | 1 | if ( |
|
| 1351 | len(word[r1_start:]) >= 1 |
||
| 1352 | and len(word) >= 2 |
||
| 1353 | and word[-1] == word[-2] |
||
| 1354 | and word[-1] not in _vowels |
||
| 1355 | ): |
||
| 1356 | 1 | word = word[:-1] |
|
| 1357 | |||
| 1358 | 1 | return word |
|
| 1359 | |||
| 1360 | |||
| 1361 | if __name__ == '__main__': |
||
| 1362 | import doctest |
||
| 1363 | |||
| 1364 | doctest.testmod() |
||
| 1365 |
Loading history...