setup_country_names.py

gistfile1.txt

import json
import urllib.request
import re

def main():
    url = 'https://raw.githubusercontent.com/mledoze/countries/refs/heads/master/countries.json'
    response = urllib.request.urlopen(url)
    data = json.loads(response.read().decode())

    mapping = {}

    for country in data:
        common_name = country['name']['common']
        official_name = country['name']['official']
        native_names = country['name']['native']

        # Collect all possible names
        names = set()
        names.add(common_name)
        names.add(official_name)

        # for lang in native_names.values():
        #     names.add(lang['common'])
        #     names.add(lang['official'])

        # Process names: remove "(the)" and trim whitespace
        for name in names:
            processed_name = re.sub(r'\s*\(the\)\s*', '', name, flags=re.IGNORECASE).strip()
            mapping[processed_name] = common_name

    # Save the mapping to a JSON file
    with open('country-names.json', 'w', encoding='utf-8') as f:
        json.dump(mapping, f, ensure_ascii=False, indent=2)

if __name__ == '__main__':
    main()