This library provides play-json Reads/Writes/Formats for GeoJSON.

I don't actively maintain this project, but occassionally, if I have time, I do merge pull requests and cut releases. That said, it is a very simple project and many people have found it useful. There has been no need for a bug fix in four years, the code is stable and mature, the only changes have been upgrades of dependencies. If you are planning to use it in a production application, you will most likely have no problems with it, if you do have problems, you may find help here, but you should be prepared to fork the codebase (a little over 500 lines of code including comments) and maintain it yourself in the worst case.

• Supports all GeoJSON types, including Point, MultiPoint, LineString, MultiLineString, Polygon, MultiPolygon, GeometryCollection, Feature, FeatureCollection.
• Supports any CRS, with WGS84 (latitude, longitude) provided out of the box.
• Type safe and polymorphic serialisation and deserialisation.
• Correctly adds the crs parameter according to the spec.
• Supports bounding boxes.
• Supports optional ids and properties of features.

Add the following dependency to build.sbt:

resolvers += Resolver.bintrayRepo("jroper", "maven")
libraryDependencies += "au.id.jazzy" %% "play-geojson" % "1.7.0"

Starting version 1.5.0, play-geojson is now deployed to bintray rather than Maven central, the group id/organization has changed from com.typesafe.play.extras to au.id.jazzy.play, and the package name has changed from play.extras.geojson to au.id.jazzy.play.geojson. So be sure to add the bintray repo above to your build and migrate your code accordingly.

The motivation behind this change is that this has never been (and probably will never be) an official Play library maintained by Lightbend, I have always maintained it in my spare time, and so the package name and organisation have been updated to reflect this, so as not to set incorrect expectations.

Import the types you need:

import au.id.jazzy.play.geojson._

Serialise a GeoJSON object to a JSON:

val sydney = Feature(Point(LatLng(-33.86, 151.2111)), 
    properties = Json.obj("name" -> "Sydney"))
val json = Json.toJson(sydney)

Deserialise some JSON into a GeoJSON object:

val json = Json.obj(
  "type" -> "Feature",
  "geometry" -> Json.obj(
    "type" -> "Point",
    "coordinates" -> Seq(151.2111, -33.86)
  ),
  "properties" -> Json.obj(
    "name" -> "Sydney"
  )
)
val sydney = Json.fromJson[Feature[LatLng]](json)

Implement a custom CRS:

case class SphericalMercator(x: Double, y: Double)

object SphericalMercator {
  implicit val format: Format[SphericalMercator] = SphericalMercatorCrs.format
  implicit val crs: CrsFormat[SphericalMercator] = SphericalMercatorCrs
}

object SphericalMercatorCrs extends CrsFormat[SphericalMercator] {
  val crs = NamedCrs("urn:ogc:def:crs:EPSG::3857")
  val format = Format[SphericalMercator](
    __.read[Seq[Double]].map {
      case Seq(x, y) => SphericalMercator(x, y)
    }, Writes(m => Json.arr(m.x, m.y))
  )
}

This software is licensed under the Apache 2 license.