node-nlp
Version:
Library for NLU (Natural Language Understanding) done in Node.js
378 lines (333 loc) • 14.6 kB
JavaScript
;
/*
Copyright (c) 2014, Ismaël Héry
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.
*/
/*
* Spec for the French Porter Stemmer can be found at:
* http://snowball.tartarus.org/algorithms/french/stemmer.html
*/
var Stemmer = require('./stemmer-fr');
var PorterStemmer = new Stemmer();
module.exports = PorterStemmer;
// Export
PorterStemmer.stem = stem;
// Exports for test purpose
PorterStemmer.prelude = prelude;
PorterStemmer.regions = regions;
PorterStemmer.endsinArr = endsinArr;
/**
* Stem a word thanks to Porter Stemmer rules
* @param {String} token Word to be stemmed
* @return {String} Stemmed word
*/
function stem(token) {
token = prelude(token.toLowerCase());
if (token.length == 1)
return token;
var regs = regions(token);
var r1_txt, r2_txt, rv_txt;
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
// Step 1
var beforeStep1 = token;
var suf, pref2, pref3, letterBefore, letter2Before, i;
var doStep2a = false;
if ((suf = endsinArr(r2_txt, ['ance', 'iqUe', 'isme', 'able', 'iste', 'eux', 'ances', 'iqUes', 'ismes', 'ables', 'istes'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(token, ['icatrice', 'icateur', 'ication', 'icatrices', 'icateurs', 'ications'])) != '') {
if (endsinArr(r2_txt, ['icatrice', 'icateur', 'ication', 'icatrices', 'icateurs', 'ications']) != '') {
token = token.slice(0, -suf.length); // delete
} else {
token = token.slice(0, -suf.length) + 'iqU'; // replace by iqU
}
} else if ((suf = endsinArr(r2_txt, ['atrice', 'ateur', 'ation', 'atrices', 'ateurs', 'ations'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(r2_txt, ['logie', 'logies'])) != '') {
token = token.slice(0, -suf.length) + 'log'; // replace with log
} else if ((suf = endsinArr(r2_txt, ['usion', 'ution', 'usions', 'utions'])) != '') {
token = token.slice(0, -suf.length) + 'u'; // replace with u
} else if ((suf = endsinArr(r2_txt, ['ence', 'ences'])) != '') {
token = token.slice(0, -suf.length) + 'ent'; // replace with ent
}
// ement(s)
else if ((suf = endsinArr(r1_txt, ['issement', 'issements'])) != '') {
if (!isVowel(token[token.length - suf.length - 1])) {
token = token.slice(0, -suf.length); // delete
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
} else if ((suf = endsinArr(r2_txt, ['ativement', 'ativements'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(r2_txt, ['ivement', 'ivements'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(token, ['eusement', 'eusements'])) != '') {
if ((suf = endsinArr(r2_txt, ['eusement', 'eusements'])) != '')
token = token.slice(0, -suf.length); // delete
else if ((suf = endsinArr(r1_txt, ['eusement', 'eusements'])) != '')
token = token.slice(0, -suf.length) + 'eux'; // replace by eux
else if ((suf = endsinArr(rv_txt, ['ement', 'ements'])) != '')
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(r2_txt, ['ablement', 'ablements', 'iqUement', 'iqUements'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(rv_txt, ['ièrement', 'ièrements', 'Ièrement', 'Ièrements'])) != '') {
token = token.slice(0, -suf.length) + 'i'; // replace by i
} else if ((suf = endsinArr(rv_txt, ['ement', 'ements'])) != '') {
token = token.slice(0, -suf.length); // delete
}
// ité(s)
else if ((suf = endsinArr(token, ['icité', 'icités'])) != '') {
if (endsinArr(r2_txt, ['icité', 'icités']) != '')
token = token.slice(0, -suf.length); // delete
else
token = token.slice(0, -suf.length) + 'iqU'; // replace by iqU
} else if ((suf = endsinArr(token, ['abilité', 'abilités'])) != '') {
if (endsinArr(r2_txt, ['abilité', 'abilités']) != '')
token = token.slice(0, -suf.length); // delete
else
token = token.slice(0, -suf.length) + 'abl'; // replace by abl
} else if ((suf = endsinArr(r2_txt, ['ité', 'ités'])) != '') {
token = token.slice(0, -suf.length); // delete if in R2
} else if ((suf = endsinArr(token, ['icatif', 'icative', 'icatifs', 'icatives'])) != '') {
if ((suf = endsinArr(r2_txt, ['icatif', 'icative', 'icatifs', 'icatives'])) != '') {
token = token.slice(0, -suf.length); // delete
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
if ((suf = endsinArr(r2_txt, ['atif', 'ative', 'atifs', 'atives'])) != '') {
token = token.slice(0, -suf.length - 2) + 'iqU'; // replace with iqU
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
} else if ((suf = endsinArr(r2_txt, ['atif', 'ative', 'atifs', 'atives'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(r2_txt, ['if', 'ive', 'ifs', 'ives'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(token, ['eaux'])) != '') {
token = token.slice(0, -suf.length) + 'eau'; // replace by eau
} else if ((suf = endsinArr(r1_txt, ['aux'])) != '') {
token = token.slice(0, -suf.length) + 'al'; // replace by al
} else if ((suf = endsinArr(r2_txt, ['euse', 'euses'])) != '') {
token = token.slice(0, -suf.length); // delete
} else if ((suf = endsinArr(r1_txt, ['euse', 'euses'])) != '') {
token = token.slice(0, -suf.length) + 'eux'; // replace by eux
} else if ((suf = endsinArr(rv_txt, ['amment'])) != '') {
token = token.slice(0, -suf.length) + 'ant'; // replace by ant
doStep2a = true;
} else if ((suf = endsinArr(rv_txt, ['emment'])) != '') {
token = token.slice(0, -suf.length) + 'ent'; // replace by ent
doStep2a = true;
} else if ((suf = endsinArr(rv_txt, ['ment', 'ments'])) != '') {
// letter before must be a vowel in RV
letterBefore = token[token.length - suf.length - 1];
if (isVowel(letterBefore) && endsin(rv_txt, letterBefore + suf)) {
token = token.slice(0, -suf.length); // delete
doStep2a = true;
}
}
// re compute regions
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
// Step 2a
var beforeStep2a = token;
var step2aDone = false;
if (beforeStep1 === token || doStep2a) {
step2aDone = true;
if ((suf = endsinArr(rv_txt, ['îmes', 'ît', 'îtes', 'i', 'ie', 'Ie', 'ies', 'ir', 'ira', 'irai', 'iraIent', 'irais', 'irait', 'iras', 'irent', 'irez', 'iriez', 'irions', 'irons', 'iront', 'is', 'issaIent', 'issais', 'issait', 'issant', 'issante', 'issantes', 'issants', 'isse', 'issent', 'isses', 'issez', 'issiez', 'issions', 'issons', 'it'])) != '') {
letterBefore = token[token.length - suf.length - 1];
if (!isVowel(letterBefore) && endsin(rv_txt, letterBefore + suf))
token = token.slice(0, -suf.length); // delete
}
}
// Step 2b
if (step2aDone && token === beforeStep2a) {
if ((suf = endsinArr(rv_txt, ['é', 'ée', 'ées', 'és', 'èrent', 'er', 'era', 'erai', 'eraIent', 'erais', 'erait', 'eras', 'erez', 'eriez', 'erions', 'erons', 'eront', 'ez', 'iez', 'Iez'])) != '') {
token = token.slice(0, -suf.length); // delete
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
} else if ((suf = endsinArr(rv_txt, ['ions'])) != '' && endsinArr(r2_txt, ['ions'])) {
token = token.slice(0, -suf.length); // delete
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
// add 'Ie' suffix to pass test for 'évanouie'
else if ((suf = endsinArr(rv_txt, ['âmes', 'ât', 'âtes', 'a', 'ai', 'aIent', 'ais', 'ait', 'ant', 'ante', 'antes', 'ants', 'as', 'asse', 'assent', 'asses', 'assiez', 'assions'])) != '') {
token = token.slice(0, -suf.length); // delete
letterBefore = token[token.length - 1];
if (letterBefore === 'e' && endsin(rv_txt, 'e' + suf))
token = token.slice(0, -1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
}
// Step 3
if (!(token === beforeStep1)) {
if (token[token.length - 1] === 'Y')
token = token.slice(0, -1) + 'i';
if (token[token.length - 1] === 'ç')
token = token.slice(0, -1) + 'c';
} // Step 4
else {
letterBefore = token[token.length - 1];
letter2Before = token[token.length - 2];
if (letterBefore === 's' && ['a', 'i', 'o', 'u', 'è', 's'].indexOf(letter2Before) == -1) {
token = token.slice(0, -1);
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
if ((suf = endsinArr(r2_txt, ['ion'])) != '') {
letterBefore = token[token.length - suf.length - 1];
if (letterBefore === 's' || letterBefore === 't') {
token = token.slice(0, -suf.length); // delete
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
}
if ((suf = endsinArr(rv_txt, ['ier', 'ière', 'Ier', 'Ière'])) != '') {
token = token.slice(0, -suf.length) + 'i'; // replace by i
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
if ((suf = endsinArr(rv_txt, 'e')) != '') {
token = token.slice(0, -suf.length); // delete
r1_txt = token.substring(regs.r1);
r2_txt = token.substring(regs.r2);
rv_txt = token.substring(regs.rv);
}
if ((suf = endsinArr(rv_txt, 'ë')) != '') {
if (token.slice(token.length - 3, -1) === 'gu')
token = token.slice(0, -suf.length); // delete
}
}
// Step 5
if ((suf = endsinArr(token, ['enn', 'onn', 'ett', 'ell', 'eill'])) != '') {
token = token.slice(0, -1); // delete last letter
}
// Step 6
i = token.length - 1;
while (i > 0) {
if (!isVowel(token[i])) {
i--;
} else if (i !== token.length - 1 && (token[i] === 'é' || token[i] === 'è')) {
token = token.substring(0, i) + 'e' + token.substring(i + 1, token.length);
break;
} else {
break;
}
}
return token.toLowerCase();
};
/**
* Compute r1, r2, rv regions as required by french porter stemmer algorithm
* @param {String} token Word to compute regions on
* @return {Object} Regions r1, r2, rv as offsets from the begining of the word
*/
function regions(token) {
var r1, r2, rv, len;
var i;
r1 = r2 = rv = len = token.length;
// R1 is the region after the first non-vowel following a vowel,
for (var i = 0; i < len - 1 && r1 == len; i++) {
if (isVowel(token[i]) && !isVowel(token[i + 1])) {
r1 = i + 2;
}
}
// Or is the null region at the end of the word if there is no such non-vowel.
// R2 is the region after the first non-vowel following a vowel in R1
for (i = r1; i < len - 1 && r2 == len; i++) {
if (isVowel(token[i]) && !isVowel(token[i + 1])) {
r2 = i + 2;
}
}
// Or is the null region at the end of the word if there is no such non-vowel.
// RV region
var three = token.slice(0, 3);
if (isVowel(token[0]) && isVowel(token[1])) {
rv = 3;
}
if (three === 'par' || three == 'col' || three === 'tap')
rv = 3;
// the region after the first vowel not at the beginning of the word or null
else {
for (i = 1; i < len - 1 && rv == len; i++) {
if (isVowel(token[i])) {
rv = i + 1;
}
}
}
return {
r1: r1,
r2: r2,
rv: rv
};
};
/**
* Pre-process/prepare words as required by french porter stemmer algorithm
* @param {String} token Word to be prepared
* @return {String} Prepared word
*/
function prelude(token) {
token = token.toLowerCase();
var result = '';
var i = 0;
// special case for i = 0 to avoid '-1' index
if (token[i] === 'y' && isVowel(token[i + 1])) {
result += token[i].toUpperCase();
} else {
result += token[i];
}
for (i = 1; i < token.length; i++) {
if ((token[i] === 'u' || token[i] === 'i') && isVowel(token[i - 1]) && isVowel(token[i + 1])) {
result += token[i].toUpperCase();
} else if (token[i] === 'y' && (isVowel(token[i - 1]) || isVowel(token[i + 1]))) {
result += token[i].toUpperCase();
} else if (token[i] === 'u' && token[i - 1] === 'q') {
result += token[i].toUpperCase();
} else {
result += token[i];
}
}
return result;
};
/**
* Return longest matching suffixes for a token or '' if no suffix match
* @param {String} token Word to find matching suffix
* @param {Array} suffixes Array of suffixes to test matching
* @return {String} Longest found matching suffix or ''
*/
function endsinArr(token, suffixes) {
var i, longest = '';
for (i = 0; i < suffixes.length; i++) {
if (endsin(token, suffixes[i]) && suffixes[i].length > longest.length)
longest = suffixes[i];
}
return longest;
};
function isVowel(letter) {
return (letter == 'a' || letter == 'e' || letter == 'i' || letter == 'o' || letter == 'u' || letter == 'y' || letter == 'â' || letter == 'à' || letter == 'ë' ||
letter == 'é' || letter == 'ê' || letter == 'è' || letter == 'ï' || letter == 'î' || letter == 'ô' || letter == 'û' || letter == 'ù');
};
function endsin(token, suffix) {
if (token.length < suffix.length) return false;
return (token.slice(-suffix.length) == suffix);
};