RDF¶

udata has built-in RDF support allowing it to both expose and harvest RDF metadata. It uses the Data Catalog Vocabulary (or DCAT) as base vocabulary.

Endpoints¶

udata exposes instance metadata through different RDF endpoints and tries to follow some best practices.

All relative URLs are relative to the API root of the udata instance, ie. https://www.data.gouv.fr/api/1/

Content Negotiation¶

The following formats are supported (default in bold):

Format	Extension	MIME type
RDF/XML	xml, rdf	application/rdf+xml, application/xml
Turtle	ttl	text/turtle, application/x-turtle
Notation3	n3	text/n3
JSON-LD	jsonld, json	application/ld+json, application/json
N-Triples	nt	application/n-triples
TriG	trig	application/trig

Each endpoint is available through a generic URL which performs content negotiation and redirects to a set of format specific URLs. The default format is RDF/XML.

Organization¶

Organizations are available through the following URL:

/organizations/{id}/catalog

where id is the organization’s identifier or slug on the udata instance.

This URL performs content negotiation and redirects to:

/organizations/{id}/catalog.{format}

It is exposed as a DCAT Catalog and a Hydra Collection This allows pagination through the hydra:PartialCollectionView class.

The organization’s catalog embeds the organization’s datasets.

Dataset¶

Datasets are available through the following URL:

/datasets/{id}/rdf

where id is the dataset’s identifier or slug on the udata instance.

This URL performs content negotiation and redirects to:

/datasets/{id}/rdf.{format}

Dataservices expose the same pair of endpoints under /dataservices/{id}/rdf.

A Dataset is exposed as a DCAT Dataset, a Resource as DCAT Distribution and fields are mapped according to:

Dataset	dcat:Dataset	notes
id	dct:identifier
title	dct:title
description	dct:description
tags	dct:keyword
created_at	dct:issued
last_modified	dct:modified
resources	dcat:distribution
temporal_coverage	dct:temporal	Uses schema.org startDate and endDate
frequency	dct:accrualPeriodicity	Frequencies without Dublin Core equivalent are mapped to the closest one
license	dct:license + dct:right	License URL in dct:license and license label in dct:right

Resource	dcat:Distribution	notes
id	dct:identifier
title	dct:title
description	dct:description
url	dcat:downloadURL	as URI reference
permanent url	dcat:accessURL	as URI reference
created_at	dct:issued
last_modified	dct:modified
format	dct:format
mime	dcat:mediaType
filesize	dcat:byteSize
checksum	spdx:checksum

TemporalCoverage	dct:PeriodOfTime
start	schema:startDate
end	schema:endDate

Checksum	spdx:Checksum
type	spdx:algorithm
value	spdx:checksumValue

Catalog¶

The site catalog is exposed through:

/site/catalog

and performs content negotiation to

/site/catalog.{format}

It is exposed as a DCAT Catalog and a Hydra Collection This allows pagination through the hydra:PartialCollectionView class.

Harvester¶

udata can harvest other RDF/DCAT enabled data portals with the DCAT Harvester.

References¶

The udata rdf implementation and its harvester were created using these references:

The used namespaces are: