Skip to content

Native NodeJS implementation of MaxMind's GeoIP API -- works in node 0.5.4 and above, ask me about other versions

License

Notifications You must be signed in to change notification settings

cpmaynard/node-geoip

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

79 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

GeoIP-lite

A native NodeJS API for the GeoLite data from MaxMind.

This product includes GeoLite data created by MaxMind, available from http://maxmind.com/

introduction

MaxMind provides a set of data files for IP to Geo mapping along with opensource libraries to parse and lookup these data files. One would typically write a wrapper around their C API to get access to this data in other languages (like JavaScript).

GeoIP-lite instead attempts to be a fully native JavaScript library. A converter script converts the CSV files from MaxMind into an internal binary format (note that this is different from the binary data format provided by MaxMind). The geoip module uses this binary file to lookup IP addresses and return the country, region and city that it maps to.

Both IPv4 and IPv6 addresses are supported, however since the GeoLite IPv6 database does not currently contain any city or region information, city and region lookups are only supported for IPv4.

philosophy

I was really aiming for a fast JavaScript native implementation for geomapping of IPs. My prime motivator was the fact that it was really hard to get libgeoip built for Mac OSX without using the library from MacPorts.

why geoip-lite

So why are we called geoip-lite? npm already has a geoip package which provides a JavaScript binding around libgeoip from MaxMind. The geoip package is fully featured and supports everything that the MaxMind APIs support, however, it requires libgeoip to be installed on your system.

geoip-lite on the other hand is a fully JavaScript implementation. It is not as fully featured as geoip however, by reducing its scope, it is about 40% faster at doing lookups. On average, an IP to Location lookup should take 20 microseconds on a Macbook Pro. IPv4 addresses take about 6 microseconds, while IPv6 addresses take about 30 microseconds.

synopsis

var geoip = require('geoip-lite');

var ip = "207.97.227.239";
var geo = geoip.lookup(ip);

console.log(geo);
{ range: [ 3479299040, 3479299071 ],
  country: 'US',
  region: 'CA',
  city: 'San Francisco',
  ll: [37.7484, -122.4156] }

installation

1. get the library

$ npm install geoip-lite

2. get the datafiles

Then download the city data files from https://github.com/bluesmoon/node-geoip/tree/master/data You need to get geoip-city.dat and geoip-city-names.dat and put them into the data/ directory of this package.

API

geoip-lite is completely synchronous. There are no callbacks involved. All blocking file IO is done at startup time, so all runtime calls are executed in-memory and are fast. Startup may take up to 200ms while it reads into memory and indexes data files.

Looking up an IP address

If you have an IP address in dotted quad notation, IPv6 colon notation, or a 32 bit unsigned integer (treated as an IPv4 address), pass it to the lookup method. Note that you should remove any [ and ] around an IPv6 address before passing it to this method.

var geo = geoip.lookup(ip);

If the IP address was found, the lookup method returns an object with the following structure:

{
   range: [ <low bound of IP block>, <high bound of IP block> ],
   country: 'XX',                 // 2 letter ISO-3166-1 country code
   region: 'RR',                  // 2 character region code.  For US states this is the 2 letter
                                  // ISO-3166-2 subcountry code for other countries, this is the
                                  // FIPS 10-4 subcountry code
   city: "City Name",             // This is the full city name
   ll: [<latitude>, <longitude>]  // The latitude and longitude of the city
}

The actual values for the range array depend on whether the IP is IPv4 or IPv6 and should be considered internal to geoip-lite. To get a human readable format, pass them to geoip.pretty()

If the IP address was not found, the lookup returns null

Pretty printing an IP address

If you have a 32 bit unsigned integer, or a number returned as part of the range array from the lookup method, the pretty method can be used to turn it into a human readable string.

    console.log("The IP is %s", geoip.pretty(ip));

This method returns a string if the input was in a format that geoip-lite can recognise, else it returns the input itself.

Built-in Updater

This package contains an update script that can pull the files from MaxMind and handle the conversion from CSV. A npm script alias has been setup to make this process easy. Please keep in mind this requires internet and MaxMind rate limits that amount of downloads on thier servers.

npm run-script updatedb

Caveats

This package includes the GeoLite database from MaxMind. This database is not the most accurate database available, however it is the best available for free. You can use the commercial GeoIP database from MaxMind with better accuracy by buying a license from MaxMind, and then using the conversion utility to convert it to a format that geoip-lite understands. You will need to use the .csv files from MaxMind for conversion.

Also note that on occassion, the library may take up to 5 seconds to load into memory. This is largely dependent on how busy your disk is at that time. It can take as little as 200ms on a lightly loaded disk. This is a one time cost though, and you make it up at run time with very fast lookups.

References

Copyright

geoip-lite is Copyright 2011-2012 Philip Tellis [email protected] and the latest version of the code is available at https://github.com/bluesmoon/node-geoip

License

There are two licenses for the code and data. See the LICENSE file for details.

About

Native NodeJS implementation of MaxMind's GeoIP API -- works in node 0.5.4 and above, ask me about other versions

Resources

License

Stars

Watchers

Forks

Packages

No packages published