SHACL Workbench

A desktop application for running SHACL inferencing and validation against RDF/Turtle files, built on Apache Jena 5.x.

Download

The easiest way to use SHACL Workbench is to download the macOS installer from the Releases page. The .dmg bundles its own Java runtime — no JDK or Maven installation required.

First launch: macOS may show a security prompt because the app is not signed. Right-click the app in Finder and choose Open, then confirm. You only need to do this once.

Features

• SHACL-AF inference — applies sh:TripleRule and sh:SPARQLRule rules to a fixed point before validating
• SHACL validation — produces a standard sh:ValidationReport using the Jena SHACL engine
• Workspace root folder — loads all .ttl files recursively; individual sub-paths can be excluded
• Drag-and-drop — drop folders or files onto any shapes list or input field
• Severity filter — show Violations, Warnings, Info, or any combination in the report table
• Prefixed URIs — focus nodes and constraints are displayed using declared prefix mappings, with a toggle to full URI
• Inferred triples viewer — inspect each triple added during the inference pass, with the originating shape and rule type (TripleRule / SPARQLRule)
• Named configurations — save and restore named workspace configurations; multiple configs can coexist for different projects
• Session persistence — last-used configuration is saved automatically on exit and restored on next launch

Prerequisites

Requirement	Version
Java (JDK)	21 or later
Maven	3.8 or later

No other installation is needed. All Jena dependencies are bundled into the fat JAR at build time.

To check your versions:

java -version
mvn -version

Build

Fat JAR (for development / command-line use)

git clone https://github.com/steveraysteveray/shacl-workbench.git
cd shacl-workbench
mvn clean package -DskipTests
java -jar target/shacl-workbench-1.0-SNAPSHOT.jar &

macOS .dmg installer

Requires JDK 21+ and Maven. Produces a self-contained .dmg in dist/ that bundles its own Java runtime.

./package-mac.sh

To add a custom Dock icon, place a icon.icns file at src/main/resources/icon.icns before running the script. See Apple's Human Interface Guidelines for icon dimensions.

Run

java -jar target/shacl-workbench-1.0-SNAPSHOT.jar &

macOS may show a security prompt the first time; if so, open System Settings → Privacy & Security and allow the app to run.

Usage

1. Set your inputs

Field	Purpose
Root folder	Folder whose .ttl files are loaded recursively as background context (ontologies, schemas). Optional if a data file is provided.
Workspace exclusions	Sub-folders or files within the root folder to skip (e.g. a directory of rules files you don't want loaded as data).
Data file	A single .ttl file to validate. Optional if a root folder is provided.

All fields accept drag-and-drop. Use the Browse… and Clear buttons as alternatives.

2. Add shapes files

Zone	Purpose
Inference shapes	SHACL-AF shapes containing sh:TripleRule or sh:SPARQLRule rules. Only used when Infer + Validate is selected.
Validation shapes	Standard SHACL shapes used for constraint checking. At least one file is required.

Drag files or folders onto each zone, or use the Add file… / Add folder… buttons. Dropping a folder adds its immediate .ttl children.

3. Choose a mode and run

• Infer + Validate — inference rules are applied to a fixed point, then validation runs against the expanded graph
• Validate only — inference is skipped

Click Run. Results appear in the tabs below.

4. Review results

Validation Report tab

• The table shows focus node, path, constraint component, severity, and message for each result entry
• Use the Violation / Warning / Info checkboxes to filter by severity; the status line shows how many results are visible
• Toggle Full URI to switch between prefixed and expanded URIs
• Copy Turtle copies the raw sh:ValidationReport graph; Save report… writes it to a file

Inferred Triples tab

• The table shows each added triple (subject, predicate, object), the originating shape, and the rule type (TripleRule or SPARQLRule)
• Copy Turtle copies the inferred triples as a Turtle document; Save inferred… writes it to a file

5. Named configurations

Use File → Save Config As… to save the current workspace under a name. Use File → Load Config to restore it. Configs are stored as individual property files:

~/.shacl-workbench/configs/<name>.properties

Multiple named configs can coexist, making it easy to switch between projects.

6. Session

The last-used configuration is saved automatically when the window is closed and restored on the next launch:

~/.shacl-workbench/session.properties

Running the tests

mvn test

The test suite validates that the Jena pipeline correctly detects constraint violations and that the SHACL-AF inference engine adds the expected triples.

Compatibility notes

The Workbench uses the Apache Jena SHACL engine, which implements the W3C SHACL and SHACL-AF standards. Shapes files that rely on capabilities outside those standards may produce silent failures or incorrect results. The main categories to be aware of:

• Custom SPARQL function libraries (e.g. QUDT's qfn:) — unknown functions are silently skipped
• SHACL-JS (JavaScript constraints) — not supported by Jena; silently ignored
• OWL entailment — validation runs against asserted triples only; OWL-inferred facts are not present
• owl:imports — imported ontologies are not fetched automatically
• Non-Turtle formats — only .ttl files are loaded; RDF/XML, JSON-LD, etc. are skipped
• Named graphs — all files are merged into a single default graph
• Federated SPARQL (SERVICE in constraints) — remote endpoints are not queried

See COMPATIBILITY.md for details and workarounds, including a per-file compatibility table for QUDT shapes.