How to Expose Data Via Feature Services

After installing the connector, the following describes the basic steps required to expose your data as features via a feature service:

1. Load the data that you want into MarkLogic – This could be done via a DHF-based application or as a standalone dataset (not a data integration use case)
   1. GeoJSON is the easiest to deal with
   2. MarkLogic supports other formats as well: https://docs.marklogic.com/guide/search-dev/geospatial#id_77391
2. Determine what the “features” are in the documents – I.E. if you have a bunch of flight data about airplane locations, the features are likely the airplanes
3. Create TDE templates to extract the properties from the documents that you want to have in the features – I.E. for flight data, maybe the airplane tail number, airline, etc.
4. Create a service descriptor for the feature service you want to expose
   1. Define the service name
   2. Configure the layers you want the service to serve – Each layer has the following
      1. number (starts at 0)
      2. name
      3. extents
      4. data source configuration (what schema and view to use), and bounding query (an option MarkLogic query that limits what documents to pull features from)
   3. Here’s an example of a service descriptor: https://github.com/koopjs/koop-provider-marklogic/blob/master/config/example/services/GDeltExample.json
5. Load the service descriptor to the database