Compiling an extensive cities list from free geonames.org data

cities.ex

# This module provides a function to compile city data using the free data
# provided on http://download.geonames.org/export/dump.
# It uses following files to compile a list of 4.4 mio. cities world-wide, 
# including their name, state (administrative level 1), country, time zone, 
# latitude, and longitude:
#
# * allCountries.txt (included in allCountries.zip)
# * countryInfo.txt
# * admin1CodesASCII.txt
#

defmodule Cities do
  defmodule Data do
    @moduledoc """
    Provides functions for compiling a raw data file.
    """

    @doc """
    Compiles the needed raw city data from the given files.
    """
    def compile(location_file, countries_file, states_file) do
      new_file = File.stream!("cities.txt", [:write])
      states = compile_states(states_file)
      countries = compile_countries(countries_file)

      location_file
      |> File.stream!
      |> Stream.filter(&city?/1)
      |> Stream.map(&to_attrs/1)
      |> Stream.map(&(replace_state(&1, states)))
      |> Stream.map(&(replace_country(&1, countries)))
      |> Stream.map(&to_line/1)
      |> Enum.into(new_file)
    end

    defp city?(line) do
      type =
        line
        |> String.split("\t")
        |> Enum.at(6)

      type == "P" # Parish -> city or village
    end

    defp to_attrs(line) do
      [_geonameid,         # integer id of record in geonames database
       name,               # name of geographical point (utf8) varchar(200)
       _asciiname,         # name of geographical point in plain ascii characters, varchar(200)
       _alternatenames,    # alternatenames, comma separated, ascii names automatically transliterated, convenience attribute from alternatename table, varchar(10000)
       latitude,           # latitude in decimal degrees (wgs84)
       longitude,          # longitude in decimal degrees (wgs84)
       _feature_class,     # see http://www.geonames.org/export/codes.html, char(1)
       _feature_code,      # see http://www.geonames.org/export/codes.html, varchar(10)
       country_code,       # ISO-3166 2-letter country code, 2 characters
       _cc2,               # alternate country codes, comma separated, ISO-3166 2-letter country code, 200 characters
       state,              # fipscode (subject to change to iso code), see exceptions below, see file admin1Codes.txt for display names of this code; varchar(20)
       _admin2_code,       # code for the second administrative division, a county in the US, see file admin2Codes.txt; varchar(80)
       _admin3_code,       # code for third level administrative division, varchar(20)
       _admin4_code,       # code for fourth level administrative division, varchar(20)
       _population,        # bigint (8 byte int)
       _elevation,         # in meters, integer
       _dem,               # digital elevation model, srtm3 or gtopo30, average elevation of 3''x3'' (ca 90mx90m) or 30''x30'' (ca 900mx900m) area in meters, integer. srtm processed by cgiar/ciat.
       timezone,           # the iana timezone id (see file timeZone.txt) varchar(40)
       _modification_date  # date of last modification in yyyy-MM-dd format
      ] = String.split(line, "\t")

      %{city: name,
        state: state,
        country: country_code,
        timezone: timezone,
        latitude: latitude,
        longitude: longitude}
    end

    defp replace_state(attrs, state_names) do
      %{attrs | state: state_names["#{attrs.country}.#{attrs.state}"]}
    end

    defp compile_states(file) do
      file
      |> File.stream!
      |> Stream.filter(&(!String.starts_with?(&1, "#")))
      |> Enum.into(%{}, fn line ->
        [key, value | _] = String.split(line, "\t")
        {key, value}
      end)
    end

    defp replace_country(attrs, country_names) do
      %{attrs | country: country_names[attrs.country]}
    end

    defp compile_countries(file) do
      file
      |> File.stream!
      |> Stream.filter(&(!String.starts_with?(&1, "#")))
      |> Enum.into(%{}, fn line ->
        [key, _, _, _, value | _] = String.split(line, "\t")
        {key, value}
      end)
    end

    defp to_line(attrs) do
      line = Enum.join([
        "\"#{attrs.city}\"",
        "\"#{attrs.state}\"",
        "\"#{attrs.country}\"",
        "\"#{attrs.timezone}\"",
        attrs.latitude,
        attrs.longitude
      ], ",")

      "#{line}\n"
    end
  end
end


# Example usage:
Cities.Data.compile("allCountries.txt", "countryInfo.txt", "admin1CodesASCII.txt")