similar

finds similar words/abbreviations in the abbreviations-in-code project lists.

meant to avoid duplicate or near-duplicate entries.

Uses the Damerau-Levenshtein distance to calculate
similarities between entries of the machine-readable JSON abbreviations.

• words with score < 3 are included in the results.
• a score: 0 means it's an exact duplicate.
• results are sorted by lowest score, ascending.

The lower the score, the more similar the words are.

The score is the number of additions/deletions/edits required to change a word into another.

e.g: "cat" and "car" distance is 1.

install

npm i https://github.com/nicholaswmin/similar.git

theres no deps, neither prod nor dev/test.

usage

Example usage:

import { compare } from '@nicholaswmin/similar'

const json = await fetch('https://raw.githubusercontent.com/abbrcode/abbreviations-in-code/refs/heads/main/data/abbrs/.json')
  .then(res => res.json())

const similarWords = compare.words(json, { cutoff: 3 })
const similarAbbrs = compare.abbrs(json, { cutoff: 3 })

console.dir(similarWords, { depth: 3 })
// console.dir(similarAbbrs, { depth: 3 })

returns:

{
  name: 'predication',
  i: 202,
  matches: [
    { name: 'prediction', i: 203, matches: [], score: 1 },
    { name: 'production', i: 208, matches: [], score: 3 }
  ]
},
{
   name: 'time',
   i: 260,
   matches: [
     { name: 'timer', i: 261, matches: [], score: 1 },
     { name: 'sine', i: 239, matches: [], score: 2 },
     { name: 'type', i: 266, matches: [], score: 2 },
     { name: 'active', i: 4, matches: [], score: 3 },
     { name: 'image', i: 131, matches: [], score: 3 },
     { name: 'link', i: 156, matches: [], score: 3 },
     { name: 'node', i: 177, matches: [], score: 3 },
     { name: 'pixel', i: 197, matches: [], score: 3 },
     { name: 'text', i: 259, matches: [], score: 3 },
     { name: 'to', i: 263, matches: [], score: 3 }
   ]
 },
}

// and so on...

The above example can be run via node example.js

test

no need for npm i, its a native test runner

node --run test

Authors

Nik. Kyriakides, @nicholaswmin

MIT "No attribution" License