This library provides a local reverse geocoder for Node.js that is based on GeoNames data. It is local in the sense that there are no calls to a remote service like the Google Maps API, and in consequence the gecoder is suitable for batch reverse geocoding. It is reverse in the sense that you give it a (list of) point(s), i.e., a latitude/longitude pair, and it returns the closest city to that point.
$ npm install local-reverse-geocoderOr, with Yarn:
$ yarn add local-reverse-geocoderFor usage with Docker, a Dockerfile is available in this project. It caches all the required files from GeoNames.
To build and run it:
$ docker build -t local-reverse-geocoder .
$ docker run -it -e PORT=3000 --rm local-reverse-geocoderOr pull and run it:
$ docker run -it -e PORT=3000 --rm ghcr.io/tomayac/local-reverse-geocoderYou must initialize the geocoder prior to the first call to lookUp(). This
ensures that all files are loaded into the cache prior to making the first call.
var geocoder = require('local-reverse-geocoder');
geocoder.init({}, function () {
// geocoder is loaded and ready to run
});Optionally init() also allows you to specify which files to load data from.
This reduces initialization time and the runtime memory footprint of the Node.js
process. By default, all files are loaded.
var geocoder = require('local-reverse-geocoder');
geocoder.init(
{
citiesFileOverride: 'cities500', // one of 'cities500', 'cities1000', 'cities5000', 'cities15000' or null to keep the default city database (cities1000)
load: {
admin1: true,
admin2: false,
admin3And4: false,
alternateNames: false,
},
},
function () {
// Ready to call lookUp
}
);Optionally init() allows you to specify the directory that GeoNames files are
downloaded and cached in, and a specific cities database to be used.
var geocoder = require('local-reverse-geocoder');
geocoder.init({ dumpDirectory: '/tmp/geonames' }, function () {
// Ready to call lookUp and all files will be downloaded to /tmp/geonames
});var geocoder = require('local-reverse-geocoder');
// With just one point
var point = { latitude: 42.083333, longitude: 3.1 };
geocoder.lookUp(point, function (err, res) {
console.log(JSON.stringify(res, null, 2));
});
// In batch mode with many points
var points = [
{ latitude: 42.083333, longitude: 3.1 },
{ latitude: 48.466667, longitude: 9.133333 },
];
geocoder.lookUp(points, function (err, res) {
console.log(JSON.stringify(res, null, 2));
});
// How many results to display at max
var maxResults = 5;
// With just one point
var point = { latitude: 42.083333, longitude: 3.1 };
geocoder.lookUp(point, maxResults, function (err, res) {
console.log(JSON.stringify(res, null, 2));
});
// In batch mode with many points
var points = [
{ latitude: 42.083333, longitude: 3.1 },
{ latitude: 48.466667, longitude: 9.133333 },
];
geocoder.lookUp(points, maxResults, function (err, res) {
console.log(JSON.stringify(res, null, 2));
});You can use the built-in Web service by running node app.js as follows.
$ curl "http://localhost:3000/geocode?latitude=48.466667&longitude=9.133333&latitude=42.083333&longitude=3.1&maxResults=2"An output array that maps each point in the input array (or input object
converted to a single-element array) to the maxResults closest addresses.
The measurement units are used
as defined by GeoNames, for
example, elevation is measured in meters. The distance value is dynamically
calculated based on the
haversine distance for the
input point(s) to each of the particular results points and is measured in
kilometers.
[
[
{
geoNameId: '2919146',
name: 'Gomaringen',
asciiName: 'Gomaringen',
alternateNames: null,
latitude: '48.45349',
longitude: '9.09582',
featureClass: 'P',
featureCode: 'PPLA4',
countryCode: 'DE',
cc2: null,
admin1Code: {
name: 'Baden-Württemberg',
asciiName: 'Baden-Wuerttemberg',
geoNameId: '2953481',
},
admin2Code: {
name: 'Tübingen Region',
asciiName: 'Tuebingen Region',
geoNameId: '3214106',
},
admin3Code: {
name: 'Landkreis Tübingen',
asciiName: 'Landkreis Tubingen',
geoNameId: '2820859',
},
admin4Code: {
name: 'Gomaringen',
asciiName: 'Gomaringen',
geoNameId: '6555939',
},
population: '8400',
elevation: null,
dem: '430',
timezone: 'Europe/Berlin',
modificationDate: '2011-04-25',
alternateName: {
de: {
altName: 'Gomaringen',
isPreferredName: true,
isShortName: false,
isColloquial: false,
isHistoric: false,
},
},
distance: 3.1302317076079285,
},
{
geoNameId: '2814195',
name: 'Wannweil',
asciiName: 'Wannweil',
alternateNames: null,
latitude: '48.51667',
longitude: '9.15',
featureClass: 'P',
featureCode: 'PPLA4',
countryCode: 'DE',
cc2: null,
admin1Code: {
name: 'Baden-Württemberg',
asciiName: 'Baden-Wuerttemberg',
geoNameId: '2953481',
},
admin2Code: {
name: 'Tübingen Region',
asciiName: 'Tuebingen Region',
geoNameId: '3214106',
},
admin3Code: {
name: 'Landkreis Reutlingen',
asciiName: 'Landkreis Reutlingen',
geoNameId: '3220792',
},
admin4Code: {
name: 'Wannweil',
asciiName: 'Wannweil',
geoNameId: '6555933',
},
population: '5092',
elevation: null,
dem: '320',
timezone: 'Europe/Berlin',
modificationDate: '2011-04-25',
distance: 5.694122211376861,
},
],
[
{
geoNameId: '3130634',
name: 'Albons',
asciiName: 'Albons',
alternateNames: null,
latitude: '42.10389',
longitude: '3.08433',
featureClass: 'P',
featureCode: 'PPLA3',
countryCode: 'ES',
cc2: null,
admin1Code: {
name: 'Catalonia',
asciiName: 'Catalonia',
geoNameId: '3336901',
},
admin2Code: {
name: 'Província de Girona',
asciiName: 'Provincia de Girona',
geoNameId: '6355230',
},
admin3Code: {
name: 'Albons',
asciiName: 'Albons',
geoNameId: '6534005',
},
admin4Code: null,
population: '558',
elevation: '13',
dem: '18',
timezone: 'Europe/Madrid',
modificationDate: '2012-03-04',
distance: 2.626176210836868,
},
{
geoNameId: '3118799',
name: "la Tallada d'Empordà",
asciiName: "la Tallada d'Emporda",
alternateNames:
"La Tallada,la Tallada,la Tallada d'Emporda,la Tallada d'Empordà",
latitude: '42.0802',
longitude: '3.05583',
featureClass: 'P',
featureCode: 'PPLA3',
countryCode: 'ES',
cc2: null,
admin1Code: {
name: 'Catalonia',
asciiName: 'Catalonia',
geoNameId: '3336901',
},
admin2Code: {
name: 'Província de Girona',
asciiName: 'Provincia de Girona',
geoNameId: '6355230',
},
admin3Code: {
name: "la Tallada d'Empordà",
asciiName: "la Tallada d'Emporda",
geoNameId: '6534150',
},
admin4Code: null,
population: '0',
elevation: null,
dem: '16',
timezone: 'Europe/Madrid',
modificationDate: '2012-03-04',
distance: 3.6618561653699846,
},
],
];By design, i.e., due to the granularity of the available GeoNames data, this reverse geocoder is limited to city-level, so no streets or house numbers. In many cases this is already sufficient, but obviously your actual mileage may vary. If you need street-level granularity, you are better off with a service like Google's reverse geocoding API. (Full disclosure: the author is currently employed by Google.)
The initial lookup takes quite a while, as the geocoder has to download roughly 2GB(!) of data that it then caches locally (unzipped, this occupies about 1.3GB of disk space). All follow-up requests are lightning fast.
To reduce the time taken to initialize the data, you can manually configure it to only download a specific set of countries from GeoNames. Do note that when you add a country code into the array, it will disable the geocoder from downloading all ~2.29GB(!) worth of data and only load the specified countries' data. If you want to re-enable the geocoder to download all data, the countries array needs to be empty.
const geocoder = require('local-reverse-geocoder');
geocoder.init(
{
load: {
admin1: true,
admin2: true,
admin3And4: true,
alternateNames: true,
},
// Comma-separated list of country codes. An empty array means all countries.
countries: ['SG', 'AU'],
},
function () {
// Ready to call lookUp
}
);There's also the option of downloading the GeoNames files via a post-install
script.
The script is invoked automatically after installation, but won't download any
files without getting at least one of the init options in an env variable.
The options should be specified with a GEOCODER_POSTINSTALL_ prefix.
export GEOCODER_POSTINSTALL_DUMP_DIRECTORY=/usr/src/app
export GEOCODER_POSTINSTALL_ADMIN1=true
export GEOCODER_POSTINSTALL_ADMIN2=true
export GEOCODER_POSTINSTALL_COUNTRIES=SG,AU
npm install local-reverse-geocoder
// As long as the app is started within the same day
// and uses the same options, the init call won't download any files.By default, the local GeoNames dump
data gets refreshed each day, creating files such as
admin1CodesASCII_YYYY-MM-DD.txt in the cache directory. If you wish to reuse
the existing downloaded files, you can rename them to remove the date, such as
admin1CodesASCII.txt, which will suppress the download. If you don't
need admin1, admin2, admin3, admin4 or alternate names, you can turn them off in
a manual init call and decrease load time.
If you run into a
FATAL ERROR: CALL_AND_RETRY_LAST Allocation failed - JavaScript heap out of memory
issue, try running node with the
V8 option
--max-old-space-size=2000.
To turn on debug logging add a DEBUG=local-reverse-geocoder environment variable on the command line.
Copyright 2017 Thomas Steiner (tomac@google.com)
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
This project was inspired by Richard Penman's Python reverse geocoder. It uses Ubilabs' k-d-tree implementation that was ported to Node.js by Luke Arduini.
