JarCastingjDruid
Simple Java Druid client
##Build Status
jDruid is configured on Travis CI. The current status of the master branch is given below:
##Setting it up
###From scratch
You can clone this repository and build the client by using
mvn clean package
This will generate jdruid-0.1.0.jar in the target folder which can be used in your project.
###Maven
jDruid is available on Bintray and Maven Central:
<dependency>
<groupId>com.inferlytics</groupId>
<artifactId>jdruid</artifactId>
<version>0.1.1</version>
</dependency>
##Usage
jDruid provides the following methods:
- topN
- groupBy
- timeseries
- query
The first three return results in Lists with native Druid types. The query is a generic "catch all" method, which can run any valid Query and returns a List of Map<String, Object>
It should be fairly simple to use:
Query query = Druids.newTimeseriesQueryBuilder().build();
DruidClient client = new DruidClient("http://localhost:8082/druid/v2");
List<Result> results = client.timeseries(query);
##Extending functionality
The query() method should be generic enough to support any valid Druid Query, which returns a List<Map<String, Object>>, which can be converted to your custom POJO using the either the Jackson ObjectMapper provided by Druid, which can be instantiated by calling new DefaultObjectMapper(), or use Gson or any other valid SerDe provider.
##Help
If you run into any issues while using jDruid, please feel free to drop an email to sriram@raremile.com, or raise an issue.
