Awesome

OverpassFrontend

A JavaScript (NodeJS/Browser) library to easily access data from OpenStreetMap via Overpass API or from an OSM File. The objects can directly be used with LeafletJS or exported to GeoJSON. Data will be cached in the browser memory (persistent caching in LocalStorage or so may be added in the future).

INSTALLATION

npm install --save overpass-frontend

Demo

git clone https://github.com/plepe/overpass-frontend
cd overpass-frontend
npm install
npm start

Browse to http://localhost:8000/demo/

EXAMPLES

BBOX Query

You can execute this example as: node example-bbox.js

const OverpassFrontend = require('overpass-frontend')

// you may specify an OSM file as url, e.g. 'test/data.osm.bz2'
const overpassFrontend = new OverpassFrontend('//overpass-api.de/api/interpreter')

// request restaurants in the specified bounding box
overpassFrontend.BBoxQuery(
  'nwr[amenity=restaurant]',
  { minlat: 48.19, maxlat: 48.20, minlon: 16.33, maxlon: 16.34 },
  {
    properties: OverpassFrontend.ALL
  },
  function (err, result) {
    console.log('* ' + result.tags.name + ' (' + result.id + ')')
    // console.log(result.GeoJSON()) // convert to GeoJSON
  },
  function (err) {
    if (err) { console.log(err) }
  }
)

By ID

You can execute this example as: node example-by-id.js

const OverpassFrontend = require('overpass-frontend')

// you may specify an OSM file as url, e.g. 'test/data.osm.bz2'
const overpassFrontend = new OverpassFrontend('//overpass-api.de/api/interpreter')

// request restaurants in the specified bounding box
overpassFrontend.get(
  ['n27365030', 'w5013364'],
  {
    // only return tags of the items. See below under "'Properties' option" for an explanation. If impatient, use ALL.
    properties: OverpassFrontend.TAGS
  },
  function (err, result) {
    if (result) {
      console.log('* ' + result.tags.name + ' (' + result.id + ')')
    } else {
      console.log('* empty result')
    }
  },
  function (err) {
    if (err) { console.log(err) }
  }
)

DOCUMENTATION

Find documentation in doc. You can re-generate the documentation with npm run doc.

The following file types are supported:

Usually, an Overpass API server is used as backend. Alternatively, a file can be used, e.g. exported from the [https://openstreetmap.org](OpenStreetMap homepage) or [https://overpass-turbo.eu/](Overpass Turbo).

const overpassFrontend = new OverpassFrontend(fileUrl, {
  filename: 'file.osm', // optional, override filename to enable auto-detection
  fileFormat: 'OSMXML', // optional, if detection from url fails
  fileFormatOptions: { ... } // optional, depending on the file type
})

All files can be used raw or compressed with bzip2 (detected from the additional extension .bz2).

data: URLs are supported as well (e.g. generated from a file upload field). Passing the filename and/or fileFormat as options is recommended, as file type detection might not work.

Supported file formats:

Read the section "Additional file formats" to learn how to enable additional file formats.

The following queries are supported:

Type

Filters

Every query can have any amount of filters, e.g. nwr[amenity=restraunt][name] (all restaurants with a name).

Conditional query filters

node(if: <Evaluator>)

Values can either by numbers (3, 3.14159) or strings ("foo" or 'foo'). 0 or empty string counts as false.

The following operators are available (ordered weak to string binding):

The following functions are available:

Examples:

node(if:count_tags() > 5 || t["name"] == "foo")
node(if:debug(id()))

'Properties' option

In .get() or .BBoxQuery() requests, you can set which properties will be requested for the items (with the options.properties parameter). Further properties might be loaded as well, if necessary for caching or further checks. The ID (type, id) will always be loaded.

The following properties are defined:

DEVELOPMENT

To run unit tests, you need to have a local Overpass API server installed. For that, you can either set it up manually or use a pre-made Docker image.

Docker

First, you need to build Docker images for osm3s, the Overpass API server. Please refer to the official repository for building instructions.

Once you have built the images, you can build and run the image containing test data for overpass-frontend:

cd test/

# Use the default test configuration for Docker
cp ./conf.json-docker ./conf.js

# Build the test server image
docker build -t overpass-frontend-test .

# Run the test server (it will be exposed on the port number 8080)
docker run -p 8080:80 -it --rm overpass-frontend-test

Now you can run unit tests:

npm install
npm run test

Manual setup

Please follow the installation instructions for osm3s.

Before running unit tests, you should copy test/conf.json-dist to test/conf.json and change the configuration parameters. They should point to your local server address.

To run the tests, execute the following commands:

git clone https://github.com/plepe/overpass-frontend
cd overpass-frontend
npm install

# Initialize database for unit tests (ignore messages because of missing nodes/ways)
# replace /exec/path by the location where you installed osm3s
init_osm3s.sh test/data.osm.bz2 test/data/ /exec/path --meta

# Run unit tests
npm run test
# Check code style (Standard JS)
npm run lint

Additional file formats

Currently, the following file formats are supported: OSMXML, OSMJSON and GeoJSON. To add support to an additional file format, do this:

import OverpassFronted from 'overpass-frontend'
OverpassFrontend.registerFileFormat({
  // unique id
  id: 'myFileFormat',

  // return true for autodetecting a file of this format
  willLoad (url, content, options) {
    return !!url.match(/\.myff$/i)
  },

  // convert the file to OSMJSON format (see below for an example)
  load (content, options, callback) {
    const data = convertToMyFileFormat(content, options)
    callback(null, data)
  }
})

const database = new OverpassFrontend('path/to/file.myff', {
  fileFormat: 'myFileFormat', // optionally override auto-detection
  fileFormatOptions: { // will be passed as 'options' to willLoad() and load()
    whatever: 'value'
  }
})

Example OSMJSON result:

{
  "version": 0.6,
  "elements": [
    {
      "type": "node",
      "id": 1234,
      "lat": 12.34567,
      "lon": -12.34567
      "tags": {
        "key": "value"
      }
    }
  ]
}

The supported OSMJSON format is quite flexible, it supports (almost?) all combinations of 'out' in the Overpass QL, e.g.: if meta data is present, it will be loaded, but it is not required. Geometry of ways and relations can be loaded inline, member ids may be present.