This document collects experiences and findings on using OGC API Features (OAF) in all kinds of clients. A client could be a custom application, a software library
OGC has already collected information on clients. Some of the results we collect here, could be input for that document too.
Many APIs have a default limit on the number of items in a response. Sometimes the serverside limit is higher than this default limit. To get the maximum number of items an API supports in a single request, one should add the limit parameter to a request and set it to -1. For example:
https://demo.pygeoapi.io/master/collections/dutch_castles/items?f=json&limit=-1
Note that a server may still have a limit in a response.
Screencast (YouTube) of adding GeoJSON from an OGC API Features implementation in QGIS
Support is limited. If the API provides GeoJSON, the data can easily be used / opened in QGIS as a Vector Layer. Because QGIS supports loading a remote dataset, like GeoJSON.
Basic screencast how to use QGIS for visualising data from OGC API Features
Properly accessing the API is not possible now, like accessing the OpenAPI doc and using the paging mechanism.
Question: what is the best way to use / consume OAF implementations?
Implemented an early wfs3 client to facilitate data downloads from metadata advertising a wfs3 endpoint. Question: in some xslt cases we’d prefer to have the root and collections response encoded as xml, how does the service advertise that it supports that encoding, only via the api-yaml file? So we’d need to parse that file before connecting to the service (requires a parser for yet another encoding)? Question: what are the conventions to identify a link to external collection metadata between the other links?
There is experimental support for WFS 3 in GDAL / OGR, a well-known and much used conversion/transformation library for many geospatial formats. See the GDAL documentation on WFS3 for implementation details.
No tests done yet
The OAF / WFS3 test client uses jQuery to consume an API and display GeoJSON data on a Leaflet JS map. Source code is available at Github.