# stemming / stemming / porter2.py

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288``` ```"""An implementation of the Porter2 stemming algorithm. See http://snowball.tartarus.org/algorithms/english/stemmer.html Adapted from pyporter2 by Michael Dirolf. This algorithm is more correct but (at least in this implementation) several times slower than the original porter algorithm as implemented in stemming.porter. """ import re r_exp = re.compile(r"[^aeiouy]*[aeiouy]+[^aeiouy](\w*)") ewss_exp1 = re.compile(r"^[aeiouy][^aeiouy]\$") ewss_exp2 = re.compile(r".*[^aeiouy][aeiouy][^aeiouywxY]\$") ccy_exp = re.compile(r"([aeiouy])y") s1a_exp = re.compile(r"[aeiouy].") s1b_exp = re.compile(r"[aeiouy]") def get_r1(word): # exceptional forms if word.startswith('gener') or word.startswith('arsen'): return 5 if word.startswith('commun'): return 6 # normal form match = r_exp.match(word) if match: return match.start(1) return len(word) def get_r2(word): match = r_exp.match(word, get_r1(word)) if match: return match.start(1) return len(word) def ends_with_short_syllable(word): if len(word) == 2: if ewss_exp1.match(word): return True if ewss_exp2.match(word): return True return False def is_short_word(word): if ends_with_short_syllable(word): if get_r1(word) == len(word): return True return False def remove_initial_apostrophe(word): if word.startswith("'"): return word[1:] return word def capitalize_consonant_ys(word): if word.startswith('y'): word = 'Y' + word[1:] return ccy_exp.sub('\g<1>Y', word) def step_0(word): if word.endswith("'s'"): return word[:-3] if word.endswith("'s"): return word[:-2] if word.endswith("'"): return word[:-1] return word def step_1a(word): if word.endswith('sses'): return word[:-4] + 'ss' if word.endswith('ied') or word.endswith('ies'): if len(word) > 4: return word[:-3] + 'i' else: return word[:-3] + 'ie' if word.endswith('us') or word.endswith('ss'): return word if word.endswith('s'): preceding = word[:-1] if s1a_exp.search(preceding): return preceding return word return word doubles = ('bb', 'dd', 'ff', 'gg', 'mm', 'nn', 'pp', 'rr', 'tt') def ends_with_double(word): for double in doubles: if word.endswith(double): return True return False def step_1b_helper(word): if word.endswith('at') or word.endswith('bl') or word.endswith('iz'): return word + 'e' if ends_with_double(word): return word[:-1] if is_short_word(word): return word + 'e' return word s1b_suffixes = ('ed', 'edly', 'ing', 'ingly') def step_1b(word, r1): if word.endswith('eedly'): if len(word) - 5 >= r1: return word[:-3] return word if word.endswith('eed'): if len(word) - 3 >= r1: return word[:-1] return word for suffix in s1b_suffixes: if word.endswith(suffix): preceding = word[:-len(suffix)] if s1b_exp.search(preceding): return step_1b_helper(preceding) return word return word def step_1c(word): if word.endswith('y') or word.endswith('Y') and len(word) > 1: if word[-2] not in 'aeiouy': if len(word) > 2: return word[:-1] + 'i' return word def step_2_helper(word, r1, end, repl, prev): if word.endswith(end): if len(word) - len(end) >= r1: if prev == []: return word[:-len(end)] + repl for p in prev: if word[:-len(end)].endswith(p): return word[:-len(end)] + repl return word return None s2_triples = (('ization', 'ize', []), ('ational', 'ate', []), ('fulness', 'ful', []), ('ousness', 'ous', []), ('iveness', 'ive', []), ('tional', 'tion', []), ('biliti', 'ble', []), ('lessli', 'less', []), ('entli', 'ent', []), ('ation', 'ate', []), ('alism', 'al', []), ('aliti', 'al', []), ('ousli', 'ous', []), ('iviti', 'ive', []), ('fulli', 'ful', []), ('enci', 'ence', []), ('anci', 'ance', []), ('abli', 'able', []), ('izer', 'ize', []), ('ator', 'ate', []), ('alli', 'al', []), ('bli', 'ble', []), ('ogi', 'og', ['l']), ('li', '', ['c', 'd', 'e', 'g', 'h', 'k', 'm', 'n', 'r', 't'])) def step_2(word, r1): for trip in s2_triples: attempt = step_2_helper(word, r1, trip[0], trip[1], trip[2]) if attempt: return attempt return word def step_3_helper(word, r1, r2, end, repl, r2_necessary): if word.endswith(end): if len(word) - len(end) >= r1: if not r2_necessary: return word[:-len(end)] + repl else: if len(word) - len(end) >= r2: return word[:-len(end)] + repl return word return None s3_triples = (('ational', 'ate', False), ('tional', 'tion', False), ('alize', 'al', False), ('icate', 'ic', False), ('iciti', 'ic', False), ('ative', '', True), ('ical', 'ic', False), ('ness', '', False), ('ful', '', False)) def step_3(word, r1, r2): for trip in s3_triples: attempt = step_3_helper(word, r1, r2, trip[0], trip[1], trip[2]) if attempt: return attempt return word s4_delete_list = ('al', 'ance', 'ence', 'er', 'ic', 'able', 'ible', 'ant', 'ement', 'ment', 'ent', 'ism', 'ate', 'iti', 'ous', 'ive', 'ize') def step_4(word, r2): for end in s4_delete_list: if word.endswith(end): if len(word) - len(end) >= r2: return word[:-len(end)] return word if word.endswith('sion') or word.endswith('tion'): if len(word) - 3 >= r2: return word[:-3] return word def step_5(word, r1, r2): if word.endswith('l'): if len(word) - 1 >= r2 and word[-2] == 'l': return word[:-1] return word if word.endswith('e'): if len(word) - 1 >= r2: return word[:-1] if len(word) - 1 >= r1 and not ends_with_short_syllable(word[:-1]): return word[:-1] return word def normalize_ys(word): return word.replace('Y', 'y') exceptional_forms = {'skis': 'ski', 'skies': 'sky', 'dying': 'die', 'lying': 'lie', 'tying': 'tie', 'idly': 'idl', 'gently': 'gentl', 'ugly': 'ugli', 'early': 'earli', 'only': 'onli', 'singly': 'singl', 'sky': 'sky', 'news': 'news', 'howe': 'howe', 'atlas': 'atlas', 'cosmos': 'cosmos', 'bias': 'bias', 'andes': 'andes'} exceptional_early_exit_post_1a = frozenset(['inning', 'outing', 'canning', 'herring', 'earring', 'proceed', 'exceed', 'succeed']) def stem(word): if len(word) <= 2: return word word = remove_initial_apostrophe(word) # handle some exceptional forms if word in exceptional_forms: return exceptional_forms[word] word = capitalize_consonant_ys(word) r1 = get_r1(word) r2 = get_r2(word) word = step_0(word) word = step_1a(word) # handle some more exceptional forms if word in exceptional_early_exit_post_1a: return word word = step_1b(word, r1) word = step_1c(word) word = step_2(word, r1) word = step_3(word, r1, r2) word = step_4(word, r2) word = step_5(word, r1, r2) word = normalize_ys(word) return word if __name__ == "__main__": assert stem("bill's") == "bill" assert stem("y's") == "y" ```