pycountry provides the ISO databases for the standards:
- 639-3 Languages
- 3166 Codes for representation of names of countries and their subdivisions
- 3166-1 Countries
- 3166-3 Deleted countries
- 3166-2 Subdivisions of countries
- 4217 Currencies
- 15924 Scripts
The package includes a copy from Debian's pkg-isocodes and makes the data accessible through a Python API.
Translation files for the various strings are included as well.
No changes to the data will be accepted into pycountry. This is a pure
wrapper around the ISO standard using the pkg-isocodes database from
Debian as is. If you need changes to the political situation in the
world, please talk to the ISO or Debian people, not me.
This is a small project that I maintain in my personal time. I am not interested in personal financial gain. However, if you would like to support the project then I would love if you would donate to Feminist Frequency instead. Also, let the world know you did so, so that others can follow your path.
The code lives in a git repository on GitHub, and issues must be reported in there as well.
Countries are accessible through a database object that is already configured upon import of pycountry and works as an iterable:
>>> import pycountry
>>> len(pycountry.countries)
249
>>> list(pycountry.countries)[0]
Country(alpha_2='AF', alpha_3='AFG', name='Afghanistan', numeric='004', official_name='Islamic Republic of Afghanistan')Specific countries can be looked up by their various codes and provide the information included in the standard as attributes:
>>> germany = pycountry.countries.get(alpha_2='DE')
>>> germany
Country(alpha_2='DE', alpha_3='DEU', name='Germany', numeric='276', official_name='Federal Republic of Germany')
>>> germany.alpha_2
'DE'
>>> germany.alpha_3
'DEU'
>>> germany.numeric
'276'
>>> germany.name
'Germany'
>>> germany.official_name
'Federal Republic of Germany'There's also a "fuzzy" search to help people discover "proper" countries for names that might only actually be subdivisions. The fuzziness also includes normalizing unicode accents. There's also a bit of prioritization included to prefer matches on country names before subdivision names and have countries with more matches be listed before ones with fewer matches:
>>> pycountry.countries.search_fuzzy('England')
[Country(alpha_2='GB', alpha_3='GBR', name='United Kingdom', numeric='826', official_name='United Kingdom of Great Britain and Northern Ireland')]
>>> pycountry.countries.search_fuzzy('Cote')
[Country(alpha_2='CI', alpha_3='CIV', name="Côte d'Ivoire", numeric='384', official_name="Republic of Côte d'Ivoire"),
Country(alpha_2='FR', alpha_3='FRA', name='France', numeric='250', official_name='French Republic'),
Country(alpha_2='HN', alpha_3='HND', name='Honduras', numeric='340', official_name='Republic of Honduras')]Attributes for the country class can be accessed using the
__getattr__ method. If the requested attribute is a key for the
country class, it will return the corresponding value. In the special
cases of missing 'common_name' or 'official_name' attributes,
__getattr__ will return 'name'. Here are some examples:
>>> aland = pycountry.countries.get(alpha_2='AX')
>>> print(aland)
Country(alpha_2='AX', alpha_3='ALA', flag='🇦🇽', name='Åland Islands', numeric='248')
>>> aland.common_name
UserWarning: Country's common_name not found. Country name provided instead.
warnings.warn(warning_message, UserWarning)
'Åland Islands'
>>> aland.official_name
Country's official_name not found. Country name provided instead.
warnings.warn(warning_message, UserWarning)
'Åland Islands'
>>> aland.flag
'🇦🇽'
>>> aland.foo # Raises AttributeErrorThe historic_countries database contains former countries that have
been removed from the standard and are now included in ISO 3166-3,
excluding existing ones:
>>> ussr = pycountry.historic_countries.get(alpha_3='SUN')
>>> ussr
Country(alpha_3='SUN', alpha_4='SUHH', withdrawal_date='1992-08-30', name='USSR, Union of Soviet Socialist Republics', numeric='810')
>>> ussr.alpha_4
'SUHH'
>>> ussr.alpha_3
'SUN'
>>> ussr.name
'USSR, Union of Soviet Socialist Republics'
>>> ussr.withdrawal_date
'1992-08-30'The country subdivisions are a little more complex than the countries itself because they provide a nested and typed structure.
All subdivisons can be accessed directly:
>>> len(pycountry.subdivisions)
4847
>>> list(pycountry.subdivisions)[0]
Subdivision(code='AD-07', country_code='AD', name='Andorra la Vella', parent_code=None, type='Parish')Subdivisions can be accessed using their unique code. The resulting object will provide at least their code, name and type:
>>> de_st = pycountry.subdivisions.get(code='DE-ST')
>>> de_st.code
'DE-ST'
>>> de_st.name
'Sachsen-Anhalt'
>>> de_st.type
'State'
>>> de_st.country
Country(alpha_2='DE', alpha_3='DEU', name='Germany', numeric='276', official_name='Federal Republic of Germany')Some subdivisions specify another subdivision as a parent:
>>> al_br = pycountry.subdivisions.get(code='AL-BU')
>>> al_br.code
'AL-BU'
>>> al_br.name
'Bulqiz\xeb'
>>> al_br.type
'District'
>>> al_br.parent_code
'AL-09'
>>> al_br.parent
Subdivision(code='AL-09', country_code='AL', name='Dib\xebr', parent_code=None, type='County')
>>> al_br.parent.name
'Dib\xebr'The divisions of a single country can be queried using the country_code index:
>>> len(pycountry.subdivisions.get(country_code='DE'))
16
>>> len(pycountry.subdivisions.get(country_code='US'))
57Similar to countries, the search_fuzzy method has been implemented
for subdivisions to facilitate finding relevant subdivision entries.
This method includes unicode normalization for accents and prioritizes
matches on subdivision names. The search algorithm is designed to return
more relevant matches first:
This method is especially useful for cases where the exact name or code of the subdivision is not known.
>>> pycountry.subdivisions.search_fuzzy('York')
[Subdivision(code='GB-YOR', country_code='GB', name='York', parent='GB-ENG', parent_code='GB-GB-ENG', type='Unitary authority')
Subdivision(code='GB-ERY', country_code='GB', name='East Riding of Yorkshire', parent='GB-ENG', parent_code='GB-GB-ENG', type='Unitary authority')
Subdivision(code='GB-NYK', country_code='GB', name='North Yorkshire', parent='GB-ENG', parent_code='GB-GB-ENG', type='Two-tier county')
Subdivision(code='US-NY', country_code='US', name='New York', parent_code=None, type='State')]Scripts are available from a database similar to the countries:
>>> len(pycountry.scripts)
169
>>> list(pycountry.scripts)[0]
Script(alpha_4='Afak', name='Afaka', numeric='439')
>>> latin = pycountry.scripts.get(name='Latin')
>>> latin
Script(alpha_4='Latn', name='Latin', numeric='215')
>>> latin.alpha4
'Latn'
>>> latin.name
'Latin'
>>> latin.numeric
'215'The currencies database is, again, similar to the ones before:
>>> len(pycountry.currencies)
182
>>> list(pycountry.currencies)[0]
Currency(alpha_3='AED', name='UAE Dirham', numeric='784')
>>> argentine_peso = pycountry.currencies.get(alpha_3='ARS')
>>> argentine_peso
Currency(alpha_3='ARS', name='Argentine Peso', numeric='032')
>>> argentine_peso.alpha_3
'ARS'
>>> argentine_peso.name
'Argentine Peso'
>>> argentine_peso.numeric
'032'The languages database is similar too:
>>> len(pycountry.languages)
7874
>>> list(pycountry.languages)[0]
Language(alpha_3='aaa', name='Ghotuo', scope='I', type='L')
>>> aragonese = pycountry.languages.get(alpha_2='an')
>>> aragonese.alpha_2
'an'
>>> aragonese.alpha_3
'arg'
>>> aragonese.name
'Aragonese'
>>> bengali = pycountry.languages.get(alpha_2='bn')
>>> bengali.name
'Bengali'
>>> bengali.common_name
'Bangla'Locales are available in the pycountry.LOCALES_DIR subdirectory of
this package. The translation domains are called isoXXX according to
the standard they provide translations for. The directory is structured
in a way compatible to Python's gettext module.
Here is an example translating language names:
>>> import gettext
>>> german = gettext.translation('iso3166-1', pycountry.LOCALES_DIR,
... languages=['de'])
>>> german.install()
>>> _('Germany')
'Deutschland'For each database (countries, languages, scripts, etc.), you can also look up entities case insensitively without knowing which key the value may match. For example:
>>> pycountry.countries.lookup('de')
<pycountry.db.Country object at 0x...>The search ends with the first match, which is returned.
You can cast each object type into a dict:
>>> country = pycountry.countries.lookup('de')
>>> dict(country)
{'alpha_2': 'DE', 'name': 'Germany', ...}While pycountry will not be adding non-ISO values to its standard library, you can add or remove entries at runtime to fit your needs.
Add a non-ISO country:
>>> pycountry.countries.add_entry(alpha_2="XK", alpha_3="XXK", name="Kosovo", numeric="926")Remove a country from a database:
>>> pycountry.countries.remove_entry(alpha_2="XK")Some users have reported issues using PyCountry with PyInstaller guidance on how to handle the issues can be found in the PyInstaller Google Group.
![@pre-commit-ci[bot]](https://avatars.githubusercontent.com/in/68672?s=64&v=4){kind=small}
