GOAF

GOAF is a OGC API - Features implementation in golang.

For who is it?

If you are looking for a very fast OGC API - Features application and are not afraid of some data tuning, this could be the application for you!

Datasources

GOAF supports the following datasources:

• OGC GeoPackage
• PostGIS (*Postgresql 9.4+)

PDOK server implementation of OGCAPI-FEATURES

A a GeoJSON implementation with a Geopackage as a data provider.

The specification is a preliminary one, with go generate the routing based on api spec, provider interfaces en types structs and convenient parameter extractions are generated to stay easily up to date.

• FeatureCollection is overridden in the GeoPackage provider to use the GeoJSON equivalent for decoding blobs
• https://github.com/opengeospatial/ogcapi-features/blob/master/core/openapi/ogcapi-features-1.yaml

Build

docker build -t pdok/goaf:latest .

GeoPackage

The GeoPackage provider is a minimal config for GeoPackages that tend to be relative small e.g. < 3 GB.

docker run --rm -v `pwd`/example:/example -e CONFIG=/example/config-addresses-gpkg-minimal.yaml -p 8080:8080 pdok/goaf:latest

PostGIS

More elaborate config optimised performance for huge db (10M+ features/collection)

docker run -v `pwd`/example:/example -e CONFIG=/example/config-addresses-postgis-localhost.yaml' -p 8080:8080 pdok/goaf:latest

Example table

CREATE TABLE addresses.addresses
(
    fid text COLLATE pg_catalog."default" NOT NULL,
    offsetid bigint NOT NULL,
    properties jsonb,
    geom geometry,
    bbox geometry,
      
    CONSTRAINT addresses_addresses_pk PRIMARY KEY (fid)
)
WITH (
    OIDS = FALSE
)

CREATE INDEX addresses_geom_sidx ON addresses.addresses USING GIST (geom);
CREATE INDEX addresses_offsetid_idx ON addresses.addresses(offsetid);

Generate

Some of the code is generated based on the given oas.yaml:

codegen/provider.go
codegen/types.go
server/routing.gen.go

go generate generate/gen.go

Test

go test ./... -covermode=atomic

How to Contribute

Make a pull request...

License

Distributed under MIT License, please see license file within the code for more details.

Thanks

Inspiration and code copied from:

• https://github.com/go-spatial/jivan
• https://github.com/go-spatial/tegola

The main differences with regards to jivan is the data provider setup, some geopackage query speedups for larger Geopackages and some tweaks for scanning the SQL features