"""
Retrieve a list of information about countries, pulled from GeoNames.

Example entry:

 {u'Area(in sq km)': u'33843',
  u'Capital': u'Chi\u015fin\u0103u',
  u'Continent': u'EU',
  u'Country': u'Moldova',
  u'CurrencyCode': u'MDL',
  u'CurrencyName': u'Leu',
  u'EquivalentFipsCode': u'',
  u'ISO': u'MD',
  u'ISO-Numeric': u'498',
  u'ISO3': u'MDA',
  u'Languages': u'mo,ro,ru,gag,tr',
  u'Phone': u'373',
  u'Population': u'4324000',
  u'Postal Code Format': u'MD-####',
  u'Postal Code Regex': u'^(?:MD)*(\\d{4})$',
  u'fips': u'MD',
  u'geonameid': u'617790',
  u'neighbours': u'RO,UA',
  u'tld': u'.md'}
"""

import urllib, codecs

COUNTRY_INFO_URL = "http://download.geonames.org/export/dump/countryInfo.txt"

def get_geonames_country_data():
    "Returns a list of dictionaries, each representing a country"
    udata = urllib.urlopen(COUNTRY_INFO_URL).read().decode('utf8')
    # Strip the BOM
    if udata[0] == codecs.BOM_UTF8.decode('utf8'):
        udata = udata[1:]
    # Ignore blank lines
    lines = [l for l in udata.split('\n') if l]
    # Find the line with the headers (starts #ISO)
    header_line = [l for l in lines if l.startswith('#ISO')][0]
    headers = header_line[1:].split('\t')
    # Now get all the countries
    country_lines = [l for l in lines if not l.startswith('#')]
    countries = []
    for line in country_lines:
        countries.append(dict(zip(headers, line.split('\t'))))
    return countries