feat: add support for fr-BE

[victortaleb]

Summary

In Belgium, we don't say "soixante-dix" or "quatre-vingt-dix" but "septante" and "nonante".

Potential improvements

• I don't know if an option was the right call or if it should have been a separate language.

• CH could also be useful for some people but I don't feel comfortable implementing it myself as I am not Swiss and it may be more complicated than I thought.

  Potential CH implementation

  --- a/lib/i18n/fr.js
  +++ b/lib/i18n/fr.js
  @@ -20,9 +20,9 @@ export class N2WordsFR extends BaseLanguage {
         [1000000n, 'million'],
         [1000n, 'mille'],
         [100n, 'cent'],
  -      ...(['BE'].includes(options.region) ? [[90n, 'nonante']] : []),
  -      [80n, 'quatre-vingts'],
  -      ...(['BE'].includes(options.region) ? [[70n, 'septante']] : []),
  +      ...(['BE', 'CH'].includes(options.region) ? [[90n, 'nonante']] : []),
  +      [80n, ['CH'].includes(options.region) ? 'huitante' : 'quatre-vingts'],
  +      ...(['BE', 'CH'].includes(options.region) ? [[70n, 'septante']] : []),
         [60n, 'soixante'],
         [50n, 'cinquante'],
         [40n, 'quarante'],
  @@ -77,7 +77,7 @@ export class N2WordsFR extends BaseLanguage {
         }
       }
       if (nNum < cNum && cNum < 100) {
  -      if (nNum % 10n == 1 && cNum != 80) return { [`${cText} et ${nText}`]: cNum + nNum };
  +      if (nNum % 10n == 1 && cText != 'quatre-vingt') return { [`${cText} et ${nText}`]: cNum + nNum };
         return { [`${cText}-${nText}`]: cNum + nNum };
       }
       if (nNum > cNum) return { [`${cText} ${nText}`]: cNum * nNum };

[forzagreen]

Hi @victortaleb , thanks for the contribution.
Indeed this is a very interesting topic.

👍 OK for the country code as 2 capital letters, which is compliant with ISO 3166-1 alpha-2.

I'm still thinking on how to represent the feature. 2 options are possible:

1. Your suggestion with {region: 'BE'} (or {country: 'BE'})
2. Appending the region name after the language code, as fr-BE or fr_BE. This is the approach used by our cousin package num2words (python equivalent):
   You can supply values like fr_FR; if the country doesn't exist but the language does, the code will fall back to the base language (i.e. fr).

[victortaleb]

Option 2 would probably be better if there are languages with significantly different dialects.

Here is an implementation.
To keep things DRY, I kept the region option but prefixed with an _ to show that it should only be used internally.
I went with fr-BE as I never saw fr_BE before. Feel free to change it if you want to.

[forzagreen]

That's great @victortaleb . Thank you for your contribution !

[forzagreen]

Released in v1.18.0 🎆