SHACL (Shapes Constraint Language)

SHACL relies on constraint components—atomic, reusable validation primitives. These are the building blocks of shapes, defining specific conditions that must be met.

• Cardinality constraints: sh:minCount, sh:maxCount to control the number of value nodes.
• Value type constraints: sh:class, sh:datatype to enforce the type of a property's value.
• Value range constraints: sh:minInclusive, sh:maxExclusive for numeric or date comparisons.
• String constraints: sh:pattern, sh:minLength, sh:languageIn for text-based rules.

SHACL cleanly separates the validation logic from the data being validated.

• The Shapes Graph contains the SHACL shape definitions, which are themselves expressed as RDF. It defines the 'what' of validation.
• The Data Graph is the target RDF graph containing the actual instance data to be checked.
• A SHACL engine takes both graphs as input and produces a validation report, detailing any violations found in the data graph against the shapes graph.

SHACL defines two fundamental types of shapes to scope validation:

• Node Shapes (sh:NodeShape): Define constraints that apply directly to a focus node itself, such as requiring it to be an instance of a specific class or to have a certain IRI pattern.
• Property Shapes (sh:PropertyShape): Define constraints on the values associated with a specific property (sh:path) of the focus node. This is the most common way to validate attributes and relationships.

A SHACL engine produces a machine-readable validation report conforming to the SHACL Validation Report vocabulary. This report is itself an RDF graph.

• Conformance: A top-level boolean property (sh:conforms) indicating true or false.
• Validation Results (sh:result): Individual violations, each detailing:
  • sh:focusNode: The specific node that failed.
  • sh:resultPath: The property that caused the violation.
  • sh:resultMessage: A human-readable description of the error.
  • sh:sourceConstraintComponent: The exact constraint that was violated.

SHACL provides multiple ways to specify which nodes in a data graph a shape should apply to, using target declarations.

• Class-based targets (sh:targetClass): Applies the shape to all instances of a specified RDF/OWL class.
• Node-based targets (sh:targetNode): Directly targets a specific IRI.
• Subject-of targets (sh:targetSubjectsOf): Targets all nodes that are the subject of a given predicate.
• SPARQL-based targets (sh:target): Uses a SPARQL SELECT query to dynamically identify target nodes, offering maximum flexibility.

For complex validation logic beyond built-in constraint components, SHACL allows defining custom constraints using SPARQL.

• sh:sparql: A shape component that uses an ASK query. The shape is violated if the query returns true for a focus node.
• sh:select: A shape component that uses a SELECT query. The shape is violated if the query returns any results, which are then used to construct the validation result.
• This extensibility makes SHACL Turing-complete for validation, capable of expressing any computable condition.