vincent-kuwornu / pelias Goto Github PK

Pelias is a modular, open-source geocoder built on top of ElasticSearch for fast geocoding.

Geocoding is the process of transforming input text, such as an address, or a name of a place—to a location on the earth's surface.

Reverse geocoding is the opposite, it transforms your current geographic location in to a list of places nearby.

• It's completely open-source and MIT licensed
• It's based on open-data, so you can run it yourself
• You can install it locally and modify to suit your needs
• It has an impressive list of features, such as fast autocomplete
• It's modular, so you don't need to be an expert to make changes
• It's easy to install and requires no external dependencies
• We run a continuous deployment cycle with a new version shipping weekly

• Provide accurate search results
• Give users query suggestions (typeahead in the search box)
• Account for location bias (places nearer to you appear higher in the results)
• Support multiple data sources (OSM, geonames, quattroshapes etc)
• Flexible software architecture
• Easy to contribute software patches and features to
• Easy to set-up and configure your own instance
• No external dependencies (such as postgres)
• Reliable, configurable & fast import process
• Work equally well for a small city and the entire planet

Sure! Our API lives at pelias.mapzen.com, and is usable without an API key and generous rate-limits. The endpoints are documented here; happy geocoding!

$ curl "pelias.mapzen.com/reverse?lat=40.74358294846026&lon=-73.99047374725342" | python -m json.tool
{
    "bbox": [
        -73.99051,
        40.74361,
        -73.99051,
        40.74361
    ],
    "date": 1427901226020,
    "features": [
        {
            "geometry": {
                "coordinates": [
                    -73.99051,
                    40.74361
                ],
                "type": "Point"
            },
            "properties": {
                "admin0": "United States",
                "admin1": "New York",
                "admin2": "New York County",
                "alpha3": "USA",
                "id": "9851011",
                "layer": "geoname",
                "name": "Arlington",
                "text": "Arlington, New York County, New York",
                "type": "geoname"
            },
            "type": "Feature"
        }
    ],
    "type": "FeatureCollection"
}

Check out our vagrant development environment.

Magic! Well, like any geocoder, Pelias essentially just executes search queries against an enormous amount of geographic data that maps longitude/latitude coordinates on the Earth to searchable names (eg Empire State Building or 28 Elm Street). We run entirely on open datasets, like OpenStreetMap, GeoNames, and OpenAddresses.

The underlying architecture has three components:

• import pipelines: the pipelines used to filter, normalize, and ingest geographic datasets into the Pelias database.
• database: the underlying datastore that does all of the query heavy-lifting and powers our search results. We use ElasticSearch.
• API: the thing that users interact with. A thin layer sitting on top of the datastore that implements additional logic and features.

Here's how they interact:

Pelias itself (the import pipelines and API) is written in Node, which makes it highly accessible for other developers and performant under heavy I/O. It aims to be modular and is distributed across a number of Node packages, each with its own repository under the Pelias GitHub organization. ElasticSearch is our unconventional datastore of choice because of its unparalleled text functionality, which makes text search just work right out of the box, and sufficiently robust geospatial support.