hcls-validator.shex

PREFIX cito: <http://purl.org/spar/cito/>
PREFIX dcat: <http://www.w3.org/ns/dcat#>
PREFIX dct: <http://purl.org/dc/terms/>
PREFIX dctypes: <http://purl.org/dc/dcmitype/>
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX freq: <http://purl.org/cld/freq/>
PREFIX idot: <http://identifiers.org/idot/>
PREFIX lexvo: <http://lexvo.org/id/iso639-3/>
PREFIX pav: <http://purl.org/pav/>
PREFIX prov: <http://www.w3.org/ns/prov#>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX schemaorg: <http://schema.org/>
PREFIX sd: <http://www.w3.org/ns/sparql-service-description#>
PREFIX sio: <http://semanticscience.org/resource/>
PREFIX void: <http://rdfs.org/ns/void#>
PREFIX void-ext: <http://ldf.fi/void-ext#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>

<SummaryLevelShape> {
#Core
    rdf:type (dctypes:Dataset),
    !rdf:type (void:Dataset),
    !rdf:type (dcat:Distribution),
    dct:title rdf:langString,
    ###dct:alternative rdf:langString, is a MAY property
    dct:description rdf:langString,
    !dct:creator .,
    !pav:createBy .,
    !dct:contributor .,
    !pav:authoredBy .,
    !pav:curatedBy .,
    !dct:created .,
    !pav:createdOn .,
    !pav:authoredOn .,
    !pav:curatedOn .,
    dct:publisher IRI,
    !dct:issued .,
    (foaf:page IRI)?,
    (schemaorg:logo IRI)?,
    (dct:license IRI)* , ### License is now a MAY property
    ###dct:rights rdf:langString, ### Rights is a MAY property
    !dct:language .,
    ###dcat:theme IRI*, ### Theme is a MAY property, should match something that is a skos:Concept
    ###dcat:keyword xsd:string, ### Keyword is a MAY property
    !void:vocabulary .,
    !dct:conformsTo .,
    ###dct:references IRI, ###MAY property
    ###cito:citesAsAuthority IRI, ### citation is a MAY property
    ###rdfs:seeAlso IRI, ###MAY property
    ###dct:hasPart IRI, ###part is a MAY property, what should be the shape of the description it matches?
    !void:subset .,
#Identifiers
    ###idot:preferredPrefix xsd:string, ### Preferred prefix is a MAY property
    ###idot:alternatePrefix xsd:string, ### Alternate prefix is a MAY property
    !idot:identifierPattern .,
    !void:uriRegexPattern .,
    !idot:accessPattern ., ### Access pattern is a MAY property
    !idot:exampleIdentifier .,
    !void:exampleResource .,
#Provenance and Change
    !void:inDataset .,
    !sio:has-data-item .,
    !pav:version .,
    !dct:isVersionOf .,
    !pav:previousVersion .,
    ###pav:hasCurrentVersion IRI, ###  is a MAY property, should this match an @VersionShape?
    !dct:source .,
    !pav:retrievedFrom .,
    !prov:wasDerivedFrom .,
    !pav:createdWith .,
    dct:accrualPeriodicity IRI?,###freq:~?, ###TODO Check freq:~ is the correct syntax
#Availability/Distributions
    !dcat:distribution .,
    !dcat:landingPage .,
    !dct:format .,
    ###dcat:accessURL IRI, ### is a MAY property
    !dcat:downloadURL .,
    !void:dataDump .,
    (void:sparqlEndpoint IRI)?,
#Statistics
    !void:triples .,
    !void:entities .,
    !void:distinctSubjects .,
    !void:properties .,
    !void:distinctObjects .,
    !void:classPartition .,
    !void:propertyPartition .,
    !void:subset .
}

<VersionLevelShape> {
#Core
    rdf:type (dctypes:Dataset),
    !rdf:type (void:Dataset),
    !rdf:type (dcat:Distribution),
    dct:title rdf:langString,
    ###dct:alternative rdf:langString, is a MAY property
    dct:description rdf:langString,
    (dct:creator IRI | pav:createdBy IRI)+,
    ###(dct:contributor IRI)*, ###MAY property
    ###(pav:authoredBy IRI)*, ###MAY property
    ###(pav:curatedBy IRI)*, ###MAY property
    ###TODO add test that one of created or issued is present
    (dct:created xsd:date | dct:created  xsd:dateTime | dct:created xsd:gYearMonth | dct:created xsd:gYear | pav:createdOn xsd:dateTime)?,
    ###(pav:createdOn xsd:date | pav:createdOn  xsd:dateTime | pav:createdOn xsd:gYearMonth | pav:createdOn xsd:gYear), ###MAY property
    ###(pav:authoredOn xsd:date | pav:authoredOn  xsd:dateTime | pav:authoredOn xsd:gYearMonth | pav:authoredOn xsd:gYear), ###MAY property
    ###(pav:curatedOn xsd:date | pav:curatedOn  xsd:dateTime | pav:curatedOn xsd:gYearMonth | pav:curatedOn xsd:gYear), ###MAY property
    dct:publisher IRI,
    (dct:issued xsd:date | dct:issued xsd:dateTime | dct:issued xsd:gYearMonth | dct:issued xsd:gYear)?,
    (foaf:page IRI)?,
    (schemaorg:logo IRI)?,
    (dct:license IRI)+ , ### License is SHOULD property, but multiple allowed
    ###dct:rights rdf:langString, ### Rights is a MAY property
    (dct:language .)*, ###TODO define the object
    ###dcat:theme IRI*, ### Theme is a MAY property, should match something that is a skos:Concept
    ###dcat:keyword xsd:string, ### Keyword is a MAY property
    !void:vocabulary .,
    ###dct:conformsTo IRI, ###MAY property
    ###dct:references IRI, ###MAY property
    ###cito:citesAsAuthority IRI, ### citation is a MAY property
    ###rdfs:seeAlso IRI, ###MAY property
    ###dct:hasPart IRI, ###part is a MAY property, what should be the shape of the description it matches?
    !void:subset .,
#Identifiers
    ###idot:preferredPrefix xsd:string, ### Preferred prefix is a MAY property
    ###idot:alternatePrefix xsd:string, ### Alternate prefix is a MAY property
    !idot:identifierPattern .,
    !void:uriRegexPattern .,
    !idot:accessPattern .,
    !idot:exampleIdentifier .,
    !void:exampleResource .,
#Provenance and Change
    !sio:has-data-item .,
    !void:inDataset .,
    pav:version xsd:string,
    dct:isVersionOf IRI,
    (pav:previousVersion IRI)?,
    !pav:hasCurrentVersion .,
    (dct:source IRI)*,
    (pav:retrievedFrom IRI)?,
    (prov:wasDerivedFrom IRI)*,
    (pav:createdWith IRI)?,
    !dct:accrualPeriodicity . ,
#Availability/Distributions
    (dcat:distribution IRI)*, ###TODO restrict the shape of the IRI but could be hard to test in practice
    ###dcat:landingPage IRI, ###MAY property
    !dct:format .,
    ###dcat:accessURL IRI, ### is a MAY property
    !dcat:downloadURL .,
    !void:dataDump .,
    !void:sparqlEndpoint ., ###SHOULD NOT being strict here
#Statistics
    !void:triples .,
    !void:entities .,
    !void:distinctSubjects .,
    !void:properties .,
    !void:distinctObjects .,
    !void:classPartition .,
    !void:propertyPartition .,
    !void:subset .
}

<DistributionLevelShape> {
#Core
    (rdf:type (dctypes:Dataset))?,
    ###RDF datasets are being validated separately
    !rdf:type (void:Dataset),
    rdf:type (dcat:Distribution),
    dct:title rdf:langString,
    ###dct:alternative rdf:langString, is a MAY property
    dct:description rdf:langString,
    (dct:creator IRI | pav:createdBy IRI)+,
    ###(dct:contributor IRI)*,
    ###(pav:authoredBy IRI)*,
    ###(pav:curatedBy IRI)*, MAY property
    ###TODO add test that one of created or issued is present
    (dct:created xsd:date | dct:created  xsd:dateTime | dct:created xsd:gYearMonth | dct:created xsd:gYear | pav:createdOn xsd:dateTime)?,
    ###pav:createdOn xsd:date | pav:createdOn  xsd:dateTime | pav:createdOn xsd:gYearMonth | pav:createdOn xsd:gYear, ###MAY property
    ###pav:authoredOn xsd:date | pav:authoredOn  xsd:dateTime | pav:authoredOn xsd:gYearMonth | pav:authoredOn xsd:gYear, ###MAY property
    ###pav:curatedOn xsd:date | pav:curatedOn  xsd:dateTime | pav:curatedOn xsd:gYearMonth | pav:curatedOn xsd:gYear, ###MAY property
    dct:publisher IRI,
    (dct:issued xsd:date | dct:issued xsd:dateTime | dct:issued xsd:gYearMonth | dct:issued xsd:gYear)?,
    (foaf:page IRI)?,
    (schemaorg:logo IRI)?,
    (dct:license IRI)+ , ### License is MUST property, but multiple allowed
    ###dct:rights rdf:langString, ### Rights is a MAY property
    (dct:language .)*, ###TODO define the object
    ###dcat:theme IRI*, ### Theme is a MAY property, should match something that is a skos:Concept
    ###dcat:keyword xsd:string*, ### Keyword is a MAY property
    !void:vocabulary .,
    (dct:conformsTo IRI)?,
    ###dct:references IRI, ###MAY property
    ###cito:citesAsAuthority IRI, ### citation is a MAY property
    ###rdfs:seeAlso IRI, ###MAY property
    !dct:hasPart .,
    !void:subset ., ### Not for a non-RDF dataset
#Identifiers
    ###idot:preferredPrefix xsd:string, ### Preferred prefix is a MAY property
    ###idot:alternatePrefix xsd:string, ### Alternate prefix is a MAY property
    ###idot:identifierPattern xsd:string, ###MAY property
    !void:uriRegexPattern ., ### Not for an non-RDF dataset
    ###idot:accessPattern idot:AccessPattern, ### Access pattern is a MAY property
    (idot:exampleIdentifier xsd:string)*,
    !void:exampleResource ., ### Not for an non-RDF dataset
#Provenance and Change
    !sio:has-data-item ., ###Only for RDF dataset
    !void:inDataset ., ### Not for a non-RDF dataset
    (pav:version xsd:string)?,
    !dct:isVersionOf .,
    (pav:previousVersion IRI)?,
    !pav:hasCurrentVersion .,
    (dct:source IRI)*,
    (pav:retrievedFrom IRI)?,
    (prov:wasDerivedFrom IRI)*,
    (pav:createdWith IRI)*,
    !dct:accrualPeriodicity . ,
#Availability/Distributions
    !dcat:distribution .,
    ###dcat:landingPage IRI,###MAY property
    (dct:format IRI | dct:format xsd:string),
    ###dcat:accessURL IRI, ### is a MAY property
    (dcat:downloadURL IRI)?,
    !void:dataDump .,  ### Not for a non-RDF dataset
    !void:sparqlEndpoint ., #Applying strict semantics here
#Statistics
    !void:triples .,  ### Not for a non-RDF dataset
    !void:entities ., ### Not for a non-RDF dataset
    !void:distinctSubjects ., ### Not for a non-RDF dataset
    !void:properties ., ### Not for a non-RDF dataset
    !void:distinctObjects ., ### Not for a non-RDF dataset
    !void:classPartition ., ### Not for a non-RDF dataset
    !void:propertyPartition ., ### Not for a non-RDF dataset
    !void:subset . ### Not for a non-RDF dataset
}

<RDFDistributionLevelShape> {
#Core
    (rdf:type (dctypes:Dataset))?,
    rdf:type (void:Dataset),
    rdf:type (dcat:Distribution),
    dct:title rdf:langString,
    ###dct:alternative rdf:langString, is a MAY property
    dct:description rdf:langString,
    (dct:creator IRI | pav:createdBy IRI)+,
    ###(dct:contributor IRI)*,
    ###(pav:authoredBy IRI)*,
    ###(pav:curatedBy IRI)*, MAY property
    ###TODO add test that one of created or issued is present
    (dct:created xsd:date | dct:created  xsd:dateTime | dct:created xsd:gYearMonth | dct:created xsd:gYear | pav:createdOn xsd:dateTime)?,
    ###pav:createdOn xsd:date | pav:createdOn  xsd:dateTime | pav:createdOn xsd:gYearMonth | pav:createdOn xsd:gYear, ###MAY property
    ###pav:authoredOn xsd:date | pav:authoredOn  xsd:dateTime | pav:authoredOn xsd:gYearMonth | pav:authoredOn xsd:gYear, ###MAY property
    ###pav:curatedOn xsd:date | pav:curatedOn  xsd:dateTime | pav:curatedOn xsd:gYearMonth | pav:curatedOn xsd:gYear, ###MAY property
    dct:publisher IRI,
    (dct:issued xsd:date | dct:issued xsd:dateTime | dct:issued xsd:gYearMonth | dct:issued xsd:gYear)?,
    (foaf:page IRI)?,
    (schemaorg:logo IRI)?,
    (dct:license IRI)+ , ### License is MUST property, but multiple allowed
    ###dct:rights rdf:langString, ### Rights is a MAY property
    (dct:language .)*, ###TODO define the object
    ###dcat:theme IRI*, ### Theme is a MAY property, should match something that is a skos:Concept
    ###dcat:keyword xsd:string*, ### Keyword is a MAY property
    void:vocabulary IRI*,
    (dct:conformsTo IRI)?,
    ###dct:references IRI, ###MAY property
    ###cito:citesAsAuthority IRI, ### citation is a MAY property
    ###rdfs:seeAlso IRI, ###MAY property
    !dct:hasPart .,
    !void:subset ., ### Not for a non-RDF dataset
#Identifiers
    ###idot:preferredPrefix xsd:string, ### Preferred prefix is a MAY property
    ###idot:alternatePrefix xsd:string, ### Alternate prefix is a MAY property
    ###idot:identifierPattern xsd:string, ###MAY property
    ###void:uriRegexPattern xsd:string, ### May property
    ###idot:accessPattern idot:AccessPattern, ### Access pattern is a MAY property
    (idot:exampleIdentifier xsd:string)*,
    (void:exampleResource IRI)*,
#Provenance and Change
    ###(sio:has-data-item IRI)*, ###MAY property
    (pav:version xsd:string)?,
    !dct:isVersionOf .,
    (pav:previousVersion IRI)?,
    !pav:hasCurrentVersion .,
    (dct:source IRI)*,
    (pav:retrievedFrom IRI)?,
    (prov:wasDerivedFrom IRI)*,
    (pav:createdWith IRI)?,
    !dct:accrualPeriodicity . ,
#Availability/Distributions
    !dcat:distribution .,
    ###dcat:landingPage IRI,###MAY property
    (dct:format IRI | dct:format xsd:string), ###Could limit the string to an RDF format
    ###dcat:accessURL IRI, ### is a MAY property
    (dcat:downloadURL IRI)?,
    (void:dataDump IRI)?,
    !void:sparqlEndpoint ., #Applying strict semantics here
#Statistics
    (void:triples xsd:integer)?,
    (void:entities xsd:integer)?,
    (void:distinctSubjects xsd:integer)?,
    (void:properties xsd:integer)?,
    (void:distinctObjects xsd:integer)?,
    ### Number of classes
    (void:classPartition @<NumberOfClassesShape>)?,
    ### Number of graphs
    (void:classPartition @<NumberofGraphsShape>)?
#Enhanced Statistics
    ###Not validating these!!!
    ###MAY properties
}

<NumberOfClassesShape> {
    void:class rdfs:Class ,
    void:distinctSubjects xsd:integer
}

<NumberOfGraphsShape> {
    void:class sd:Graph ,
    void:distinctSubjects xsd:integer
}