Created
July 8, 2021 22:54
-
-
Save manciuszz/790aafd79de174de9b912cb33e21f4a2 to your computer and use it in GitHub Desktop.
This is just a browser friendly version of https://github.com/StephanGeorg/trigram-similarity
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
const nGram = function(n) { | |
if ( | |
typeof n !== 'number' || | |
Number.isNaN(n) || | |
n < 1 || | |
n === Number.POSITIVE_INFINITY | |
) { | |
throw new Error('`' + n + '` is not a valid argument for `n-gram`') | |
} | |
return grams | |
/** | |
* Create n-grams from a given value. | |
* | |
* @template {string|string[]} T | |
* @param {T} [value] | |
* @returns {T[]} | |
*/ | |
function grams(value) { | |
/** @type {T[]} */ | |
var nGrams = [] | |
/** @type {number} */ | |
var index | |
/** @type {string|string[]} */ | |
var source | |
if (value === null || value === undefined) { | |
return nGrams | |
} | |
source = value.slice ? value : String(value) | |
index = source.length - n + 1 | |
if (index < 1) { | |
return nGrams | |
} | |
while (index--) { | |
// @ts-ignore | |
nGrams[index] = source.slice(index, index + n) | |
} | |
return nGrams | |
} | |
}; | |
const bigram = nGram(2); | |
const trigram = nGram(3); | |
/** | |
* Two blank spaces are added at the beginning, and one at the end, | |
* and single spaces are replaced by double ones. | |
* @param {string} input | |
* @returns {string} | |
*/ | |
const convertString = (input = '') => { | |
if (!input.trim()) return ''; | |
return ` ${input | |
.trim() | |
.replace(/\s+/g, ' ') // replace multiple spaces w/ single spaces | |
.replace(/\s/g, ' ')} ` // replace single spaces w/ double spaces | |
.toLowerCase(); | |
}; | |
/** | |
* Sorting them, and taking out repetitions (via Set) | |
* @param {string} input | |
* @returns {string} | |
*/ | |
const generateTrigram = (input = '') => [...new Set(trigram(convertString(input)) | |
.filter((trigramItem) => !/^[\p{Letter}\p{Mark}0-9]\s\s$/giu.test(trigramItem)))]; | |
/** | |
* Calculate trigram similarity between 2 strings | |
* @param {string} input1 | |
* @param {string} input2 | |
* @returns {number} | |
*/ | |
const trigramSimilarity = (input1 = '', input2 = '') => { | |
const trigrams1 = generateTrigram(input1); | |
const trigrams2 = generateTrigram(input2); | |
// Total trigrams | |
const total = [...new Set([...trigrams1, ...trigrams2])]; | |
// Trigrams both have in common | |
const common = []; | |
trigrams1.forEach((trigramItem) => { | |
if (trigrams2.includes(trigramItem)) common.push(trigramItem); | |
}); | |
return (total.length === 0) | |
? 0 | |
: common.length / total.length; | |
}; | |
console.log(trigramSimilarity('Chateau blanc', 'chateau cheval blanc')); // 0.7368421052631579 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment