ci(shacl): add shape validation for sample data

.github/scripts/validate_shacl.py

@@ -0,0 +1,61 @@
+import json
+import sys
+from pathlib import Path
+
+from pyshacl import validate
+from rdflib import Graph
+
+OPENAPI_VER_DIR = Path("openapi/ver")
+SHACL_DIR = Path("data-model/shacl")
+
+
+def load_shapes(version):
+    shapes_path = SHACL_DIR / version / "shacl.ttl"
+    if not shapes_path.exists():
+        return None
+    g = Graph()
+    g.parse(str(shapes_path), format="turtle")
+    return g
+
+
+def validate_file(json_path, shapes_graph):
+    with open(json_path, encoding="utf-8") as f:
+        data = json.load(f)
+    data_graph = Graph()
+    data_graph.parse(data=json.dumps(data), format="json-ld")
+    if len(data_graph) == 0:
+        return False, "Parsed RDF graph is empty (context resolution may have failed)"
+    conforms, _, results_text = validate(
+        data_graph=data_graph, shacl_graph=shapes_graph, debug=False
+    )
+    return conforms, results_text
+
+
+def main():
+    failures = []
+    versions_found = 0
+
+    for version_dir in sorted(OPENAPI_VER_DIR.iterdir()):
+        sample_data_dir = version_dir / "sample_data"
+        json_files = sorted(sample_data_dir.glob("**/*.json")) if sample_data_dir.is_dir() else []
+        if not json_files:
+            continue
+
+        shapes_graph = load_shapes(version_dir.name)
+
+        versions_found += 1
+        for json_path in json_files:
+            conforms, results_text = validate_file(json_path, shapes_graph)
+            if conforms:
+                print(f"PASS {json_path}")
+            else:
+                print(f"FAIL {json_path}\n{results_text}")
+                failures.append(json_path)
+
+    print(f"\n{versions_found} versions validated, {len(failures)} failures")
+    if failures:
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()

.github/workflows/validate_specs_and_jsons.yml

@@ -5,10 +5,12 @@ on:
     paths:
       - 'openapi/ver/**/*.json'
       - 'openapi/ver/**/skg-if-openapi.yaml'
+      - 'data-model'
   pull_request:
     paths:
       - 'openapi/ver/**/*.json'
       - 'openapi/ver/**/skg-if-openapi.yaml'
+      - 'data-model'
 
 jobs:
   validate_yaml_files:
@@ -26,6 +28,23 @@ jobs:
       - if: ${{ failure() }}
         run: exit()
 
+  validate_shacl:
+    runs-on: ubuntu-latest
+    name: Validate SHACL Shapes
+    needs: validate_yaml_files
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6
+        with:
+          submodules: true
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@08807647e7069bb48b6ef5acd8ec9567f424441b # v8
+
+      - name: Validate sample data against SHACL shapes
+        run: uv run python .github/scripts/validate_shacl.py
+
   validate_json_files:
     runs-on: ubuntu-latest
     name: Validate JSON Files

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "data-model"]
+	path = data-model
+	url = https://github.com/skg-if/data-model

pyproject.toml

@@ -19,3 +19,8 @@ dependencies = [
 
 [tool.uv.sources]
 plain-text-markdown-extention = { git = "https://github.com/kostyachum/python-markdown-plain-text.git" }
+
+[dependency-groups]
+dev = [
+    "pyshacl>=0.31.0",
+]