@lde/dataset

Classes for working with DCAT Datasets and Distributions.

Installation

npm install @lde/dataset

API

The package exports:

• Dataset – a DCAT dataset with iri, title, description, language, license, creator, publisher and its distributions. getDownloadDistributions() returns the downloadable (non-SPARQL) distributions, preferring compressed ones.
• DatasetArguments – the options object the Dataset constructor takes; only iri and distributions are required.
• Creator and Publisher – an agent’s iri plus its language-mapped name.
• Distribution – a DCAT distribution with accessUrl, mediaType (an IANA media type URI per DCAT-AP 3.0, or a plain content type for convenience), mimeType, compressFormat, byteSize, lastModified and more. The static Distribution.sparql(endpoint, namedGraph?) creates a SPARQL distribution.
  • isSparql() – true when the distribution’s conformsTo is the SPARQL 1.1 Protocol URI (https://www.w3.org/TR/sparql11-protocol/) or its mimeType is application/sparql-query or application/sparql-results+json.
  • compressMimeType – plain content type derived from compressFormat (e.g. application/gzip), or undefined when no compression format is declared.
  • compressedMimeType – the full content type a server is expected to send for the compressed download: mimeType with the compression suffix appended (e.g. application/n-quads+gzip), or undefined when no media type or no recognised compression format is declared.
  • namedGraph and subjectFilter – optional fields narrowing a SPARQL distribution: the named graph to query, and a SPARQL pattern fragment that pipeline queries splice in to restrict the subjects they select.
• IANA_MEDIA_TYPE_PREFIX – the https://www.iana.org/assignments/media-types/ prefix that turns a plain content type into a DCAT-AP 3.0 media type URI.
• RdfFormat and rdfFormatToFileExtension() – RDF serialization formats and their file extensions.
• rdfMediaTypes, sparqlMediaTypes and compressionMediaTypes – media type constants for matching distributions. Note that compressionMediaTypes includes application/octet-stream: servers commonly label compressed bodies as opaque bytes, so consumers matching a declared RDF media type should ignore it like the gzip/zip types.
• skolemIri() and hashSuffix() – mint deterministic IRIs for otherwise-anonymous structural nodes, instead of blank nodes whose labels are not preserved across separately serialised documents.
• assertSafeIri() – throw if an IRI contains characters that could break out of a SPARQL <…> IRI reference.

Most exports carry JSDoc, so your editor shows inline documentation as you use them.