country-iso-search

Resolve country names and codes to their canonical ISO 3166-1 alpha-3 form.

---

Accepts alpha-3 ("FRA"), alpha-2 ("FR"), UN M49 numeric (250, "250", "04", or "0250" — leading zeros are stripped and the result is zero-padded to 3 digits), or a country name / alias. Name/alias matching is case-insensitive and input is sanitized before comparison: diacritics are stripped (Türkiye → turkiye), apostrophes and . () [] , are dropped, & becomes and, -/–/— become spaces, st expands to saint, the is dropped when it leads the input or follows ,/( (so ISO short-name forms like Korea (the Republic of) and Korea, the Republic of both resolve), and internal whitespace is collapsed. Returns the canonical alpha-3, or undefined when no record matches.

Install

npm install country-iso-search

Usage

import { lookup, lookupAlpha3 } from 'country-iso-search';

lookupAlpha3('FRA');                 // 'FRA'
lookupAlpha3('FR');                  // 'FRA'
lookupAlpha3(250);                   // 'FRA' (M49 numeric)
lookupAlpha3('04');                  // 'AFG' (zero-padded to '004')
lookupAlpha3('0250');                // 'FRA' (leading zeros stripped)
lookupAlpha3('France');              // 'FRA'
lookupAlpha3('Burma');               // 'MMR' (historical name)
lookupAlpha3('Türkiye');             // 'TUR'
lookupAlpha3('cote d ivoire');       // 'CIV' (apostrophes / diacritics ignored)
lookupAlpha3('St. Kitts and Nevis'); // 'KNA'
lookupAlpha3('🇯🇵');                  // 'JPN' (flag emoji)
lookupAlpha3('not a country');       // undefined

// `lookup` returns the full record:
lookup('France')?.iso2;              // 'FR'
lookup('France')?.m49;               // '250'

Custom records

Need codes for sub-national regions, historical countries, disputed areas, or anything else outside ISO 3166-1? Pass your own records to createLookup and use the returned scoped helpers. Spread COUNTRIES to layer custom records on top of the canonical data:

import { createLookup, COUNTRIES, type CountryRecord } from 'country-iso-search';

const CUSTOM: CountryRecord[] = [
    { iso3: 'XAC', iso2: '', m49: '', name: 'Aksai Chin', aliases: [] },
    { iso3: 'XJK', iso2: '', m49: '', name: 'Jammu and Kashmir', aliases: [] },
];

const { lookupAlpha3, lookup, byAlpha3, byAlpha2, byM49 } = createLookup([
    ...COUNTRIES,
    ...CUSTOM,
]);

lookupAlpha3('FRA');               // 'FRA' (canonical)
lookupAlpha3('XAC');               // 'XAC' (custom)
lookupAlpha3('Aksai Chin');        // 'XAC'

Leave iso2 and m49 blank on user-assigned codes — createLookup skips blank values when building the byAlpha2 and byM49 maps. Cross-country alias collisions throw at construction time so they can be fixed before any lookups happen.

API

lookupAlpha3(input)

• input: string | number — country identifier in any supported form.
• Returns string | undefined.

lookup(input)

Same input shape as lookupAlpha3, but returns the full CountryRecord (or undefined). Use this when you need iso2, m49, the canonical name, or the aliases list instead of just the alpha-3.

createLookup(records)

• records: ReadonlyArray<CountryRecord> — every record the returned lookup should resolve.
• Returns { lookup, lookupAlpha3, byAlpha3, byAlpha2, byM49 } scoped to those records.
• Throws on cross-record name/alias collisions (after sanitization).

The top-level lookup / lookupAlpha3 and byAlpha3 / byAlpha2 / byM49 exports are a createLookup(COUNTRIES) instance.

sanitize(input)

• input: string — text to normalize.
• Returns the same string lowercased with diacritics / apostrophes / . () [] , stripped, & mapped to and, hyphen-likes turned into spaces, st expanded to saint, and the dropped when it leads the input or follows ,/(. Exported for advanced use — call it to produce keys consistent with the internal name/alias index.

Exports

• COUNTRIES: ReadonlyArray<CountryRecord> — the bundled ISO 3166-1 / M49 records.
• byAlpha3, byAlpha2, byM49 — ReadonlyMap<string, CountryRecord> lookups over COUNTRIES (use .get(key); e.g. byAlpha3.get('FRA')?.name).
• CountryRecord, CountryLookup — TypeScript types.

CountryRecord

interface CountryRecord {
    iso3: string;               // ISO 3166-1 alpha-3
    iso2: string;               // ISO 3166-1 alpha-2 (may be blank on custom records)
    m49: string;                // UN M49, 3-digit zero-padded (may be blank on custom records)
    name: string;               // English short name from UN M49
    aliases: readonly string[]; // additional lowercased forms — historical names,
                                //   common alternates, native-language variants,
                                //   and the country's flag emoji
}

Data

COUNTRIES mirrors UN M49 / ISO 3166-1 for iso3, iso2, m49, and name. The ~1,700 aliases were assembled from CLDR, Wikidata (filtered to languages tagged official via P37), and GeoNames, then curated by hand to resolve substring collisions (e.g. Niger vs. Nigeria, Guinea vs. Guinea-Bissau vs. Equatorial Guinea vs. Papua New Guinea), normalize transliterations, and drop variants that would produce false matches. Aliases are stored in their sanitized form (diacritics / apostrophes / punctuation stripped, hyphens turned into spaces, st expanded to saint), so don't add accented or punctuated variants by hand — lookupAlpha3 applies the same transform to user input at lookup time. To add an alias or fix an entry, edit src/countries.ts directly and open a pull request.

GeoNames (CC BY 4.0) and Unicode CLDR (Unicode License V3) require attribution. The full third-party notices live in NOTICE in this repository — consult that file before redistributing this package or its data, since it carries the attribution texts those licenses require to travel with downstream redistributions. Wikidata is CC0 and is acknowledged there as a courtesy.

License

MIT — see LICENSE. Third-party data attributions: NOTICE.