Merge pull request #13 from iolanta-tech/html-manifest-te004-failing

html manifest te004 failing

Author: anatoly-scherbakov

ldtest/facets.py

@@ -17,11 +17,18 @@ def show(self) -> Iterable[TestCase]:
         rows = self.stored_query('tests.sparql', test_class=self.iri)
         for row in rows:
             test_url = URL(row['test'])
+
+            try:
+                extract_all_scripts = row['extract_all_scripts'].value
+            except KeyError:
+                extract_all_scripts = False
+
             yield TestCase(
                 test=f'{test_url.name}#{test_url.fragment}',
                 input=URL(row['input']),
                 result=self._process_result(row['result']),
                 req=(req := row.get('req')) and req.value,
+                extract_all_scripts=extract_all_scripts,
             )
 
     def _process_result(

ldtest/models.py

@@ -10,6 +10,7 @@ class TestCase:
     input: Path
     result: Path | str | Exception   # noqa: WPS110
     req: str
+    extract_all_scripts: bool = False
 
     @property
     def raw_document(self) -> bytes:

ldtest/sparql/tests.sparql

@@ -1,4 +1,4 @@
-SELECT ?test ?result ?input ?req WHERE {
+SELECT ?test ?result ?input ?req ?extract_all_scripts WHERE {
     ?test
         a $test_class ;
         mf:result ?result ;
@@ -8,6 +8,10 @@ SELECT ?test ?result ?input ?req WHERE {
         ?test jld:req ?req .
     }
 
+    OPTIONAL {
+        ?test jld:option / jld:extractAllScripts ?extract_all_scripts .
+    }
+
     # FIXME: We mimic how `pyld` tests work, skipping 1.0 tests. This is probably incorrect though.
     FILTER NOT EXISTS {
         ?test jld:option / jld:specVersion "json-ld-1.0"^^xsd:string .

tests/test_specification.py

@@ -39,7 +39,10 @@ def test_to_rdf(test_case: TestCase):
 def test_expand(test_case: TestCase):
     if isinstance(test_case.result, str):
         try:
-            expanded_document = yaml_ld.expand(test_case.input)
+            expanded_document = yaml_ld.expand(
+                test_case.input,
+                extract_all_scripts=test_case.extract_all_scripts,
+            )
         except YAMLLDError as error:
             assert error.code == test_case.result
         else:
@@ -51,7 +54,10 @@ def test_expand(test_case: TestCase):
 
     elif isinstance(test_case.result, Path):
         expected = yaml_ld.parse(test_case.result.read_text())
-        actual = yaml_ld.expand(test_case.input)
+        actual = yaml_ld.expand(
+            test_case.input,
+            extract_all_scripts=test_case.extract_all_scripts,
+        )
         assert actual == expected
     else:
         raise ValueError(f'What to do with this test? {test_case}')

yaml_ld/expand.py

@@ -36,7 +36,7 @@ def expand(   # noqa: C901, WPS211
         if isinstance(document, Path) and base is None:
             base = f'file://{document.parent}/'
 
-        document = parse(document)
+        document = parse(document, extract_all_scripts=extract_all_scripts)
 
     options = ExpandOptions(
         base=base,

yaml_ld/parse.py

@@ -14,7 +14,7 @@
     UndefinedAliasFound, MappingKeyError, InvalidEncoding,
 )
 from yaml_ld.loader import YAMLLDLoader
-from yaml_ld.models import Document, DocumentType
+from yaml_ld.models import Document, DocumentType, ExtractAllScripts
 
 
 def try_extracting_yaml_from_html(
@@ -34,7 +34,11 @@ def try_extracting_yaml_from_html(
         yield script.text
 
 
-def _parse_html(html_string: str, fragment: str | None) -> Document:
+def _parse_html(
+    html_string: str,
+    fragment: str | None,
+    extract_all_scripts: ExtractAllScripts = False,
+) -> Document:
     """Parse all YAML-LD scripts embedded into HTML."""
     html_yaml_scripts = list(try_extracting_yaml_from_html(
         html_string,
@@ -44,12 +48,16 @@ def _parse_html(html_string: str, fragment: str | None) -> Document:
     if not html_yaml_scripts:
         return {}
 
+    if extract_all_scripts:
+        return [parse(script) for script in html_yaml_scripts]
+
     first_script, *_other_scripts = html_yaml_scripts
     return parse(first_script)
 
 
 def parse(   # noqa: WPS238, WPS231, C901
     raw_document: str | bytes | Path | URL,
+    extract_all_scripts: ExtractAllScripts = False,
 ) -> Document:
     """Parse a YAML-LD document."""
     document_type = None
@@ -83,7 +91,11 @@ def parse(   # noqa: WPS238, WPS231, C901
         )
     except ScannerError as err:
         if document_type != DocumentType.YAML:
-            return _parse_html(raw_document, fragment=fragment)
+            return _parse_html(
+                raw_document,
+                fragment=fragment,
+                extract_all_scripts=extract_all_scripts,
+            )
 
         raise LoadingDocumentFailed() from err