Vielzeug

Appearance

similarity ​

Problem ​

You need a 0–1 score for how similar two strings are — for example ranking search results or detecting near-duplicates.

Solution ​

Use similarity(str1, str2) to get the Levenshtein-based similarity score. 1.0 is an exact match; 0.0 means completely different.

ts
import { similarity } from '@vielzeug/arsenal';

similarity('hello', 'hello'); // 1
similarity('hello', 'helo'); // 0.8
similarity('hello', 'world'); // 0.2

Use as a search threshold ​

ts
import { similarity } from '@vielzeug/arsenal';

const candidates = ['TypeScript', 'JavaScript', 'Python'];
candidates.filter((c) => similarity(c.toLowerCase(), 'typescript') >= 0.5);
// ['TypeScript']

Pitfalls ​

  • Throws RangeError if either input exceeds 10 000 characters.
  • Case-sensitive — lowercase both strings before comparing if case-insensitive matching is needed.