Since a Splitgraph engine is also a PostgreSQL database, tools that use PostgreSQL can work with Splitgraph tables without any changes.

One such tool is PostgREST that generates a REST API for a PostgreSQL schema. Splitgraph runs PostgREST in Splitgraph Cloud, allowing any Splitgraph dataset to be accessed via HTTP. For example, this link runs the following PostgREST query against the splitgraph/domestic_us_flights:latest image:

flights?and=(origin_airport.eq.JFK,destination_airport.eq.LAX)

You can reproduce a similar setup locally, getting PostgREST to work against a Splitgraph image.

This example will:

• Set up a Splitgraph engine with some sample data
• Run a PostgREST instance against the engine
• Use curl to query the PostgREST instance.
• Swap the schema to be a layered checkout, which still looks like a regular schema to PostgREST but has the ability to lazily download and cache required fragments of the dataset on the fly.

../run_example.py example.yaml and press ENTER when prompted to go through the steps.