The quality of RDF based solutions requires the possibility to describe and validate RDF graphs. In the recent years, two technologies have been proposed for RDF validation: ShEx and SHACL. RDFShape is a web service that can be used as an RDF playground to validate RDF with those technologies, but can also be used as a general RDF playground allowing to parse, convert, visualize and query RDF data. The validation engine supports both ShEx and SHACL, allows the conversion between di erent schema formats, and is able visualize schemas using UML-like class diagrams. The system o ers the possibility to play with di erent RDF technologies in a uni ed setting and has already been used to teach RDF technologies in several courses and tutorials. It has also been employed in industrial RDF work ows that needed a web service for RDF data validation and management.
The adoption of any technology depends on the available tools. Although RDF
is not new, there is a lack of integrated tools that can be used for the di erent
aspects of its management, like RDF parsing, conversion, visualization, validation,
querying and inferencing. RDFShape was initially created as an online demo to
play with the recently created Shape Expressions language in 2014[
The RDFShape landing page contains several menu options, the most important of which are: { RDF Data. This pane o ers information about RDF data. It has an option to parse and show information about RDF. For educational purposes, we added the possibility of visualizing small RDF graphs converting the graphs to SVG using Graphviz. This url: http://rdfshape.weso.es/links/i1 shows the result of visualizing a simple RDF graph inspired by the one presented in the RDF 1.1 primer (https://www.w3.org/TR/rdf11-primer/). RDFShape also allows to convert between di erent RDF formats. It handles the common RDF formats supported by the underlying RDF library: Turtle, N-Triples, RDF/XML, JSON-LD, etc. As an example, the result of converting the previous RDF le to JSON-LD is shown in http://rdfshape.weso. es/links/i2. The service has a permalink option which allows to generate permanent URLs that can be copy-pasted and shared. { Schema. This option can be used to obtain information about ShEx and SHACL schema shapes. It allows to convert between the di erent schema formats available and to visualize them. Both ShEx and SHACL support the traditional RDF formats Turtle, N-Triples, JSON-LD, etc., while ShEx also supports a compact syntax. Link http://rdfshape.weso.es/links/i3 shows a ShEx schema converted from ShEx compact syntax to its turtle serialization. It is also possible to visualize ShEx shapes as UML-like class diagrams. The system transforms the schema shapes to UML classes using PlantUML. Those UML class diagrams are later converted to SVG for visualization. Fig. 2 shows an example of a ShEx Schema and its visualization (http://rdfshape.weso.es/links/i4) { Validation. RDFShape supports both ShEx and SHACL. As an example, http://rdfshape.weso.es/links/i5 shows the result of validating an RDF graph using ShEx, while http://rdfshape.weso.es/links/i6 shows the result of SHACL validation report. In the case of SHACL, it shows information about the errors and evidences of the di erent results. This information is useful to distinguish between nodes that are not signalled as problematic because they have been validated to conform to a shape and nodes that are not signalled just because the validator does not reach them. It is also possible to validate against an SPARQL endpoint. For example, the result of validating a simple Schema against the Wikidata SPARQL endpoint can be seen at http://rdfshape.weso.es/links/i7. It is also possible to enable inference over the RDF graphs before querying or validating them. As an example, the following link shows the results of validating a ShEx schema over an RDF graph after applying RDFS inference on it: http://rdfshape.weso.es/links/i8. { SPARQL. RDFShape can also be used to run SPARQL queries over RDF data. Although there are several online SPARQL tools like YASGUI 1, we consider this feature to be very handy to query the same RDF graphs that are being validated, specially in the case of SPARQL-based constraints in SHACL. Link http://rdfshape.weso.es/links/i9 shows the result of a SPARQL query over the previous RDF graph. { API. RDFShape provides a REST API through which it is possible to invoke the previous actions in a programmatic way. 3
RDFShape is available as open source software using the MIT licence at a github repository 2. It has been designed with the goal of being exible enough to work with di erent RDF libraries like Apache Jena or RDf4j. The main components are depicted in gure 2. 1 http://yasgui.org/ 2 https://github.com/labra/rdfshape { APIService and WebService provide the external interfaces which o er both a REST API and an HTML frontend. These two services call di erent methods from that APIHelper module which is based on http4s3, a simple http framework using a purely functional programming approach that o ers support for streaming. For validation, it relies on the Shaclex library, while for RDF parsing and querying, it relies on SRDF. { The Shaclex library has two modules that contain the ShEx and SHACL implementation. Both implementations use a purely functional approach based on monad transformers provided by the Scala Cats library4. { In order to work with di erent underlying implementations, all RDF processing is done through a SRDF (Simple RDF, a generic RDF interface which de ned common methods to manipulate RDF data. It has 3 sub-interfaces: RDFReader for querying RDF without modifying it, RDFBuilder to update and modify RDF graphs, and RDFReasoner to add inference capabilities to RDF graphs. { SRDF4Jena is an implementation of SRDF for Apache Jena 5 while SRDF4J is an implementation of SRDF for RDF4J 6.
RDFShape has already been used in several semantic web courses and
tutorials7, and to test the examples from the Validating RDF data book [
Acknowledgements: This work is partially funded by the Spanish Ministry of Economy and Competitiveness (Society challenges: TIN2017-88877-R) 3 https://http4s.org/ 4 https://typelevel.org/cats/ 5 https://jena.apache.org/ 6 http://rdf4j.org/ 7 http://www.validatingrdf.com/tutorial/iswc2018/