Beginning of manifest validations using SHACL

Only validates some fields of RDF tests for now

Author: Tpt

.github/workflows/validate.yml

@@ -0,0 +1,15 @@
+name: validate
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  validate_manifest:
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v5
+      - uses: astral-sh/setup-uv@v7
+      - run: uv run validate_manifest.py

ns/shapes.ttl

@@ -0,0 +1,40 @@
+# SHACL shapes for tests
+
+PREFIX mf: <http://www.w3.org/2001/sw/DataAccess/tests/test-manifest#>
+PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+PREFIX rdft: <http://www.w3.org/ns/rdftest#>
+PREFIX sh: <http://www.w3.org/ns/shacl#>
+
+[
+    a sh:NodeShape ;
+    sh:targetClass mf:Manifest ;
+    sh:property [
+        sh:path mf:include ;
+        sh:maxCount 1
+    ] ;
+    sh:property [
+        sh:path mf:entries ;
+        sh:maxCount 1
+    ]
+] .
+
+[
+    a sh:NodeShape ;
+    sh:targetClass rdft:Test ;
+    sh:property [
+        sh:path mf:action ;
+        sh:nodeKind sh:IRI ;
+        sh:maxCount 1
+    ] ;
+    sh:property [
+        sh:path mf:result ;
+        sh:or ( [ sh:nodeKind sh:IRI ] [ sh:hasValue false ] ) ;
+        sh:maxCount 1
+    ] ;
+    sh:property [
+        sh:path rdft:approval ;
+        sh:nodeKind sh:IRI ;
+        sh:in (rdft:Approved rdft:Proposed rdft:Rejected rdft:NotClassified) ; # TODO: rdft:NotClassified is not in the vocabulary but used
+        sh:maxCount 1
+    ]
+] .

rdf/rdf11/rdf-mt/manifest-az.ttl

@@ -39,7 +39,7 @@
 <#ill-formed-string> a mf:PositiveEntailmentTest;
   mf:name "ill-formed-string";
   rdfs:comment """
-    XSD Strings do not allow the null code point \0000.
+    XSD Strings do not allow the null code point \\0000.
   """;
   mf:entailmentRegime "Simple" ;
   mf:recognizedDatatypes ( xsd:string ) ;
@@ -50,7 +50,7 @@
 <#well-formed-html> a mf:NegativeEntailmentTest;
   mf:name "well-formed-html";
   rdfs:comment """
-    rdf:HTML does allow the null code point \0000.
+    rdf:HTML does allow the null code point \\0000.
   """;
   mf:entailmentRegime "Simple" ;
   mf:recognizedDatatypes ( rdf:HTML ) ;
@@ -87,7 +87,7 @@
   """;
   mf:entailmentRegime "RDFS" ;
   mf:recognizedDatatypes ( xsd:integer ) ;
-  mf:unrecognizedDatatypes ( ex:dt ) ;
+  mf:unrecognizedDatatypes ( <http://example.org/dt> ) ;
   mf:action <az-tests/unrecognized-datatype001.ttl>;
   mf:result false .
 
@@ -100,7 +100,7 @@
   """;
   mf:entailmentRegime "RDFS" ;
   mf:recognizedDatatypes ( ) ;
-  mf:unrecognizedDatatypes ( ex:dt ) ;
+  mf:unrecognizedDatatypes ( <http://example.org/dt> ) ;
   mf:action <az-tests/unrecognized-datatype002.ttl>;
   mf:result <az-tests/unrecognized-datatype003.ttl> .
 

rdf/rdf12/rdf-semantics/manifest.ttl

@@ -244,7 +244,7 @@ trs:literal-type a mf:PositiveEntailmentTest;
   mf:recognizedDatatypes (xsd:integer);
   mf:result <literal-type.ttl>;
   mf:unrecognizedDatatypes ();
-  test:approval test:NotClassified .
+  rdft:approval rdft:NotClassified .
 
 trs:malformed-literal a mf:PositiveEntailmentTest;
   rdfs:comment "Malformed literals are allowed in triple terms, but cause inconsistency.";

validate_manifest.py

@@ -0,0 +1,52 @@
+# /// script
+# dependencies = [
+#   "pyshacl~=0.30.0",
+# ]
+# ///
+"""
+Small Python script to validate manifest files
+"""
+
+import os
+from pathlib import Path
+from pyshacl import validate
+
+repository_root = Path(__file__).parent
+shape_path = repository_root / "ns" / "shapes.ttl"
+vocab_path = repository_root / "ns" / "rdftest.ttl"
+
+
+def log_line(message: str, level: str, file: Path | None = None) -> None:
+    if "GITHUB_ACTIONS" in os.environ:
+        print(
+            f"::{level}{f' file={file.relative_to(repository_root)}' if file else ''}::{message}"
+        )
+    else:
+        print(message)
+
+
+failure_counter = 0
+total_counter = 0
+for manifest_path in repository_root.rglob("manifest*.ttl"):
+    try:
+        (conforms, _, results_text) = validate(
+            str(manifest_path),
+            shacl_graph=str(shape_path),
+            ont_graph=str(vocab_path),
+            inference="rdfs",
+        )
+    except SyntaxError as e:
+        conforms = False
+        results_text = str(e)
+    if not conforms:
+        log_line(results_text, "error", manifest_path)
+        failure_counter += 1
+    total_counter += 1
+
+log_line(
+    f"Found {failure_counter} files with error"
+    if failure_counter > 0
+    else f"Validated {total_counter} files without errors",
+    "notice",
+)
+exit(int(failure_counter > 0))