# Requests Examples
# Individual Resources
Individual resources can be retrieved by including their ID directly in the URL path, after the URL of the collection they belong to. This is shown in the examples below:
Top level collections:
http://demo.opensensorhub.org/api/systems/2ecd4713http://demo.opensensorhub.org/api/datastreams/56de11a
One or more JSON properties can be included or excluded specifically with the select query parameter. The ! character is used to exclude fields:
http://demo.opensensorhub.org/api/systems/2ecd4713?select=id,namehttp://demo.opensensorhub.org/api/systems/2ecd4713?select=!validTime
# Resource Collections
Resource collections are accessed using plural in the URL path.
For example, to retrieve the entire collection of systems, datastreams, observations and features, you would use respectively:
http://demo.opensensorhub.org/api/systemshttp://demo.opensensorhub.org/api/datastreamshttp://demo.opensensorhub.org/api/observationshttp://demo.opensensorhub.org/api/features
# Nested Collections
Nested collections can be accessed in the same way, provided the parent resource is identified:
http://demo.opensensorhub.org/api/systems/2ecd4713/membershttp://demo.opensensorhub.org/api/systems/2ecd4713/featuresOfInteresthttp://demo.opensensorhub.org/api/datastreams/56de11a/observations
# Filtering Items
Collections can be filtered using query parameters:
http://demo.opensensorhub.org/api/systems?validTime=2019-03-18Z/2020-04-30Zhttp://demo.opensensorhub.org/api/systems?q=temp
# Paging
Paging is done by providing a offset and limit parameters. The offset is ommitted for the first fetch, and then included when continuing paging through results. The next offset is always provided along with the last page fetched:
http://demo.opensensorhub.org/api/systems?limit=100http://demo.opensensorhub.org/api/systems?offset=44def6589c&limit=100
Note that the offset parameter should be considered an opaque identifier, not necessarily the sequential index of the next item. For certain data stores, it could be convenient to implement it as the row number at the start of the page (equivalent to the skip parameter in SQL) but it is not a requirement, and is not the most efficient way to implement paging in large datastores.