Creates the Geo Location XML with data from dbip

generate_xml.py

import csv
from unicodedata import normalize

root = """<mappings>
{elements}</mappings>"""

element = """<mapping>
        <ip-range from="{ip_from}" to="{ip_to}"/>
        <location country="{country}" region="{region}" city="{city}"/>
    </mapping>"""


regions = {'Acre': 'Acre',
           'Alagoas': 'Alagoas',
           'Am': 'Amazonas',
           'Amapa': 'Amapa',
           'Amazonas': 'Amazonas',
           'Ba': 'Bahia',
           'Bahia': 'Bahia',
           'Ce': 'Ceara',
           'Ceara': 'Ceara',
           'Df': 'Federal District',
           'Distrito Federal': 'Federal District',
           'Espirito Santo': 'Espirito Santo',
           'Federal District': 'Federal District',
           'Go': 'Goias',
           'Goias': 'Goias',
           'Maranhao': 'Maranhao',
           'Mato Grosso': 'Mato Grosso',
           'Mato Grosso Do Sul': 'Mato Grosso do Sul',
           'Mg': 'Minas Gerais',
           'Minas Gerais': 'Minas Gerais',
           'Ms': 'Mato Grosso do Sul',
           'Para': 'Para',
           'Paraiba': 'Paraiba',
           'Parana': 'Parana',
           'Pe': 'Pernambuco',
           'Pernambuco': 'Pernambuco',
           'Piaui': 'Piaui',
           'Pr': 'Parana',
           'Rio De Janeiro': 'Rio De Janeiro',
           'Rio Grande Do Norte': 'Rio Gramde do Norte',
           'Rio Grande Do Sul': 'Rio Grande do Sul',
           'Rj': 'Rio de Janeiro',
           'Rn': 'Rio Grande do Norte',
           'Rondonia': 'Rondonia',
           'Roraima': 'Roraima',
           'Rs': 'Rio Grande do Sul',
           'Santa Catarina': 'Santa Catarina',
           'Sao Paulo': 'Sao Paulo',
           'Sc': 'Santa Catarina',
           'Sergipe': 'Sergipe',
           'Sp': 'Sao Paulo',
           'Tocantins': 'Tocantins'
           }


def remove_accents(txt):
    return normalize('NFKD', txt).encode('ASCII', 'ignore').decode('ASCII')


def main():
    mappings = ''
    lines = 1
    with open('dbip-city-2018-03.csv', "r") as csvfile:
        datareader = csv.reader(csvfile)
        for row in datareader:
            ip_from, ip_to, country, region, city = row
            if country == 'BR' and ':' not in ip_from:
                country = 'Brazil'
                region = remove_accents(region).replace('State of ', '')
                region = regions[region]

                lines += 1
                new_location = element.format(ip_from=ip_from,
                                              ip_to=ip_to,
                                              country=country,
                                              region=region,
                                              city=city)
                mappings += new_location
                if lines % 100 == 0:
                    print(lines / 128055 * 100, '%')

    print(lines)
    mappings += """<mapping>
        <ip-range from="172.20.0.0" to="172.20.255.255"/>
        <location country="Brazil" region="Sao Paulo" city="Escritorio Porto"/>
    </mapping>"""

    final_xml = root.format(elements=mappings)
    with open('geo-ip-mappings.xml', 'w') as outfile:
        outfile.write(final_xml)


if __name__ == '__main__':
    main()