@acusti/matchmaking

Intuitive approximate string matching (i.e. fuzzy searches). See the tests to understand its behavior and evaluate if it’s what you are looking for.

Usage

npm install @acusti/matchmaking
# or
yarn add @acusti/matchmaking

matchmaking exports two functions: getBestMatch and sortByBestMatch. Import them by name:

import { getBestMatch, sortByBestMatch } from '@acusti/matchmaking';

Both functions take the same payload:

type Payload = {
    excludeMismatches?: boolean;
    items: Array<string>;
    text: string;
};

However, sortByBestMatch returns an array of items matching the one passed in but sorted by how close they match the passed-in text, while getBestMatch just returns the text of the single closest match found (i.e. sortByBestMatch(payload)[0]).

The excludeMismatches option for sortByBestMatch allows that function to also filter the results to only include (partial) matches, where a match include fuzzy matches that are strong enough to qualify as a partial match. For example, considering a list of all states in alphabetical order, searching for the text "ma" with excludeMismatches: true returns:

[
    'Maine', // exact match
    'Maryland', // exact match
    'Massachusetts', // exact match
    'Kansas', // partial match
    'Hawaii', // partial match
    'Michigan', // partial match (further distance on 2nd letter)
    'Minnesota', // partial match (further distance on 2nd letter)
    'Mississippi', // partial match (further distance on 2nd letter)
    'Missouri', // partial match (further distance on 2nd letter)
    'Alabama', // end-of-text exact match
    'Oklahoma', // end-of-text exact match
    'Nebraska', // short distance from match
    'Nevada', // short distance from match
    'New Hampshire', // short distance from match
    'New Jersey', // short distance from match
    'New Mexico', // short distance from match
    'New York', // short distance from match
]