The FRaunhofer Opensource SensorThings-Client is a Java-based client library for the SensorThingsAPI and aims to simplify development of SensorThings enabled client applications.
- CRUD operations
- Queries on entity sets
- Loading of referenced entities
- MultiDatastreams
- dataArray (for creating observations)
- Batch requests
- dataArray (for requesting observations)
- MQTT
Add the dependency:
<dependency>
<groupId>de.fraunhofer.iosb.ilt</groupId>
<artifactId>FROST-Client</artifactId>
<version>0.22</version>
</dependency>
If you do not yet have the FraunhoferIOSB bintray repository, add:
<repositories>
<repository>
<id>bintray-fraunhoferiosb-Maven</id>
<url>https://dl.bintray.com/fraunhoferiosb/Maven</url>
<releases>
<enabled>true</enabled>
</releases>
</repository>
</repositories>
Add the dependency:
compile 'de.fraunhofer.iosb.ilt:FROST-Client:0.22'
If you do not yet have the FraunhoferIOSB bintray repository, add:
repositories {
maven {
url "https://dl.bintray.com/fraunhoferiosb/Maven"
}
}
The SensorThingsService
class is central to the library. An instance of it represents a SensorThings service and is identified by an URI.
The source code below demonstrates the CRUD operations for Thing objects. Operations for other entities work similarly.
URL serviceEndpoint = new URL("http://example.org/v1.0/");
SensorThingsService service = new SensorThingsService(serviceEndpoint);
Thing thing = ThingBuilder.builder()
.name("Thingything")
.description("I'm a thing!")
.build();
service.create(thing);
// get Thing with numeric id 1234
thing = service.things().find(1234l);
// get Thing with String id ab12cd
thing = service.things().find("ab12cd");
thing.setDescription("Things change...");
service.update(thing);
service.delete(thing);
Entity Sets are represented by instances of EntityList<>
. The query parameters specified by the SensorThingsAPI standard can be applied to queries.
EntityList<Thing> things = service.things()
.query()
.count()
.orderBy("description")
.select("name","id","description")
.filter("")
.skip(5)
.top(10)
.list();
for (Thing thing : things) {
System.out.println("So many things!");
}
Entity sets only load so many entities at a time. If you want to get all entities,
and there are more entities than the $top parameter allows you get in one request, you can
use the EntityList.fullIterator();
Iterator.
EntityList<Observations> observations = service.observations()
.query()
.count()
.top(1000)
.list();
Iterator<Observation> i = observations.fullIterator();
while (i.hasNext()) {
Observation obs = i.next();
System.out.println("Observation " + obs.getId() + " has result " + obs.getResult());
}
Related entity sets can also be queried.
// Get the thing with ID 1
thing = service.things().find(1l);
// Get the Datastreams of this Thing
EntityList<Datastream> dataStreams = thing.datastreams().query().list();
for (Datastream dataStream : dataStreams) {
Sensor sensor = dataStream.getSensor();
System.out.println("dataStream " + dataStream.getId() + " has Sensor " + sensor.getId());
}
However, $expand
does not work on queries yet.
Loading referenced objects in one operation (and therefore in one request) is supported. The $expand option of the SensorThingsAPI standard is used internally.
Thing thing = service.things().find(1l,
Expansion.of(EntityType.THING)
.with(ExpandedEntity.from(EntityType.LOCATIONS)));
EntityList<Location> locations = thing.getLocations();
Using DataArrays for creating Observations is more efficient, since only one http request is done, and the observations are more efficiently encoded in this request, so the request is smaller than the sum of the separate, normal requests.
Set<DataArrayValue.Property> properties = new HashSet<>();
properties.add(DataArrayValue.Property.Result);
properties.add(DataArrayValue.Property.PhenomenonTime);
DataArrayValue dav1 = new DataArrayValue(datastream1, properties);
dav1.addObservation(observation1);
dav1.addObservation(observation2);
dav1.addObservation(observation3);
DataArrayValue dav2 = new DataArrayValue(multiDatastream1, properties);
dav2.addObservation(observation4);
dav2.addObservation(observation5);
dav2.addObservation(observation6);
DataArrayDocument dad = new DataArrayDocument();
dad.addDataArrayValue(dav1);
dad.addDataArrayValue(dav2);
service.create(dad);
This library emerged from a practical work for a lecture at KIT in collaboration with the Fraunhofer IOSB. A server implementation of the SensorThingsAPI, developed by the Fraunhofer IOSB, is available on GitHub as well.
Contributions are welcome!
- Fork this repository
- Commit your changes
- Create a pull request
The code and the documentation of this work is available under the MIT license.