A barnard59 step that validates RDF using the Shapes Constraint Language (SHACL).
npm install barnard59-shacl --save
barnard59-validate-shacl. Please update your dependencies accordingly.
Validates a chunk of RDF against a set of conditions specified in a SHACL graph.
The argument is an RDF stream containing the SHACL shapes.
prefix : <https://pipeline.described.at/>
prefix code: <https://code.described.at/>
prefix shacl: <https://barnard59.zazuko.com/operations/shacl/>
prefix fs: <https://barnard59.zazuko.com/operations/core/fs/>
[
a :Pipeline ;
:steps
[
:stepList
(
[ fs:createReadStream ( "path/to/data" ) ]
[ shacl:validate ( <#CubeShapes> ) ]
)
] ;
] .Note that this operation does not take care of partitioning the data, using this operation requires to prepare the data accordingly.
To pass additional options, initialize the step with named arguments instead. All except shape are optional.
prefix : <https://pipeline.described.at/>
prefix code: <https://code.described.at/>
prefix shacl: <https://barnard59.zazuko.com/operations/shacl/>
[ shacl:validate
[ code:name "shape" ; code:value <#CubeShapes> ] ,
# validation will stop when the given number is reached (default 1)
# set to 0 to report all errors
[ code:name "maxErrors" ; code:value 100 ] ,
# callback function which allows pipeline authors to perform
# additional actions and/or decide to continue the pipeline
# when SHACL violations are encountered (see below)
[
code:name "onViolation" ;
code:value [ a code:EcmaScriptModule ; code:link <file:...> ] ;
]
].The function below could be used to continue the pipeline when SHACL violations are found but none of them are sh:Violation.
import { sh } from '@tpluscode/rdf-ns-builders'
/**
* @param context {Object} Pipeline context
* @param data {DatasetCore} Data graph which failed validation
* @param report {ValidationReport}
*/
export function continueOnWarnings({ context, data, report }) {
const hasViolations = report.results.some(({ severity }) => severity.equals(sh.Violation))
return !hasViolations
}Similar to validate, but instead of returning the data graph, it streams SHACL validation report(s).
prefix : <https://pipeline.described.at/>
prefix code: <https://code.described.at/>
prefix shacl: <https://barnard59.zazuko.com/operations/shacl/>
prefix fs: <https://barnard59.zazuko.com/operations/core/fs/>
[
a :Pipeline, :Readable ;
:steps
[
:stepList
(
[ fs:createReadStream ( "path/to/data" ) ]
# TODO: parse input and group into dataset chunks for validation
[ shacl:report ( <#CubeShapes> ) ]
)
] ;
] .By default, the pipeline will stop when 500 validation errors are encountered across all processed datasets. This can be changed by passing a maxErrors argument. Specifically, to fail on first violation, set it to 0.
[
shacl:report
[ code:name "shape" ; code:value <#CubeShapes> ] ,
[ code:name "maxErrors" ; code:value 0 ]
] .Peer dependencies must be explicitly installed
npm i barnard59-base barnard59-formats barnard59-rdf
Reads standard input and validates the data against the given SHACL shapes.
cat data.ttl | barnard59 shacl validate --shapes https://example.com/shapes.ttlShapes can also be path relative to the working directory.
In cases when a remote address give to --shapes option does not include a correct content-type header (or does not provide a content-type header at all), the pipeline will fail. In such cases, it is possible to use the --shapesFormat option to select the correct RDF parser. Its value must be a media type, such as text/turtle.
Creates a human-readable summary of the machine-readable standard output of the validation pipeline.
cat data.ttl \
| barnard59 shacl validate --shapes https://example.com/shapes.ttl \
| barnard59 shacl report-summary