fix(dataset): fix datasets shacl shape (#2368)

renku/core/commands/dataset.py

@@ -419,10 +419,10 @@ def _export_dataset(name, provider_name, publish, tag, client_dispatcher: IClien
         selected_tag = _prompt_tag_selection(tags)
 
     if selected_tag:
-        dataset = datasets_provenance.get_by_id(selected_tag.dataset_id, immutable=True)
+        dataset = datasets_provenance.get_by_id(selected_tag.dataset_id.value, immutable=True)
 
         if not dataset:
-            raise DatasetNotFound(message=f"Cannot find dataset with id: '{selected_tag.dataset_id}'")
+            raise DatasetNotFound(message=f"Cannot find dataset with id: '{selected_tag.dataset_id.value}'")
 
     data_dir = get_dataset_data_dir(client, dataset)
     dataset = DynamicProxy(dataset)

renku/core/commands/format/dataset_tags.py

@@ -35,7 +35,7 @@ def tabular(tags):
                 ("name", None),
                 ("description", None),
                 ("dataset", None),
-                ("dataset_id", "dataset id"),
+                ("dataset_id.value", "dataset id"),
             )
         ),
         # workaround for tabulate issue 181

renku/core/management/dataset/datasets_provenance.py

@@ -24,7 +24,7 @@
 from renku.core import errors
 from renku.core.management.command_builder.command import inject
 from renku.core.management.interface.dataset_gateway import IDatasetGateway
-from renku.core.models.dataset import Dataset, DatasetTag
+from renku.core.models.dataset import Dataset, DatasetTag, Url
 from renku.core.models.provenance.agent import Person
 from renku.core.utils import communication
 
@@ -194,6 +194,9 @@ def _process_dataset_tags(self, dataset: Dataset, tags: List[DatasetTag]):
             if tag.name in current_tag_names:
                 continue
             tag = DatasetTag(
-                dataset_id=dataset.id, date_created=tag.date_created, description=tag.description, name=tag.name
+                dataset_id=Url(url_id=dataset.id),
+                date_created=tag.date_created,
+                description=tag.description,
+                name=tag.name,
             )
             self.add_tag(dataset, tag)

renku/core/management/dataset/tag.py

@@ -22,7 +22,7 @@
 
 from renku.core import errors
 from renku.core.management.dataset.datasets_provenance import DatasetsProvenance
-from renku.core.models.dataset import Dataset, DatasetTag
+from renku.core.models.dataset import Dataset, DatasetTag, Url
 
 
 def add_dataset_tag(dataset: Dataset, tag: str, description="", force=False):
@@ -49,7 +49,7 @@ def add_dataset_tag(dataset: Dataset, tag: str, description="", force=False):
             raise errors.ParameterError(f"Tag '{tag}' already exists")
         datasets_provenance.remove_tag(dataset, existing_tag)
 
-    new_tag = DatasetTag(dataset_id=dataset.id, description=description, name=tag)
+    new_tag = DatasetTag(dataset_id=Url(url_id=dataset.id), description=description, name=tag)
 
     datasets_provenance.add_tag(dataset, new_tag)
 

renku/core/management/migrations/utils/conversion.py

@@ -36,7 +36,11 @@ def _convert_dataset_tag(tag: Optional[old_datasets.DatasetTag]) -> Optional[Dat
     """Convert old DatasetTag to new DatasetTag."""
     if not tag:
         return
-    return DatasetTag(dataset_id="dummy-id", date_created=tag.created, description=tag.description, name=tag.name)
+
+    # NOTE: ``dataset_id`` field will be set later when processing the migrated commit.
+    return DatasetTag(
+        dataset_id=Url(url_id="dummy-id"), date_created=tag.created, description=tag.description, name=tag.name
+    )
 
 
 def _convert_language(language: Optional[old_datasets.Language]) -> Optional[Language]:

renku/core/metadata/gateway/dataset_gateway.py

@@ -59,7 +59,7 @@ def add_tag(self, dataset: Dataset, tag: DatasetTag):
             tags = PersistentList()
             self.database_dispatcher.current_database["datasets-tags"].add(tags, key=dataset.name)
 
-        assert tag.dataset_id == dataset.id, f"Tag has wrong dataset id: {tag.dataset_id} != {dataset.id}"
+        assert tag.dataset_id.value == dataset.id, f"Tag has wrong dataset id: {tag.dataset_id.value} != {dataset.id}"
 
         tags.append(tag)
 

renku/core/models/dataset.py

@@ -116,16 +116,16 @@ class DatasetTag(Persistent):
     def __init__(
         self,
         *,
-        dataset_id: str,
+        dataset_id: Url,
         date_created: datetime = None,
         description: str = None,
         id: str = None,
         name: str,
     ):
         if not id:
-            id = DatasetTag.generate_id(dataset_id=dataset_id, name=name)
+            id = DatasetTag.generate_id(dataset_id=dataset_id.value, name=name)
 
-        self.dataset_id: str = dataset_id
+        self.dataset_id: Url = dataset_id
         self.date_created: datetime = parse_date(date_created) or local_now()
         self.description: str = description
         self.id: str = id
@@ -614,7 +614,7 @@ class Meta:
         model = DatasetTag
         unknown = EXCLUDE
 
-    dataset_id = fields.String(schema.location)
+    dataset_id = Nested(schema.about, UrlSchema, missing=None)
     date_created = fields.DateTime(schema.startDate, missing=None, format="iso", extra_formats=("%Y-%m-%d",))
     description = fields.String(schema.description, missing=None)
     id = fields.Id()

renku/data/shacl_shape.json

@@ -321,10 +321,7 @@
                "datatype": {
                   "@id": "xsd:string"
                },
-               "maxCount": 1,
-               "sh:lessThanOrEquals": {
-                  "@id": "schema:datePublished"
-               }
+               "maxCount": 1
             },
             {
                "nodeKind": "sh:Literal",
@@ -553,19 +550,9 @@
                "maxCount": 1
             },
             {
-               "nodeKind": "sh:Literal",
-               "path": "schema:location",
-               "datatype": {
-                  "@id": "xsd:string"
-               },
-               "minCount": 1,
-               "maxCount": 1
-            },
-            {
-               "nodeKind": "sh:Literal",
                "path": "schema:about",
-               "datatype": {
-                  "@id": "xsd:string"
+               "sh:class": {
+                  "@id": "schema:URL"
                },
                "minCount": 1,
                "maxCount": 1

tests/cli/test_datasets.py

@@ -1114,7 +1114,7 @@ def test_dataset_provider_resolution_dataverse(doi_responses, uri):
     assert type(provider) is DataverseProvider
 
 
-def test_dataset_tag(tmpdir, runner, project, client, subdirectory):
+def test_dataset_tag(tmpdir, runner, client, subdirectory, get_datasets_provenance_with_injection):
     result = runner.invoke(cli, ["dataset", "create", "my-dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
     assert "OK" in result.output
@@ -1139,6 +1139,11 @@ def test_dataset_tag(tmpdir, runner, project, client, subdirectory):
     result = runner.invoke(cli, ["dataset", "tag", "my-dataset", "aBc9.34-11_55.t"], catch_exceptions=False)
     assert 0 == result.exit_code, format_result_exception(result)
 
+    with get_datasets_provenance_with_injection(client) as datasets_provenance:
+        dataset = datasets_provenance.get_by_name("my-dataset")
+        all_tags = datasets_provenance.get_all_tags(dataset)
+        assert {dataset.id} == {t.dataset_id.value for t in all_tags}
+
 
 @pytest.mark.parametrize("form", ["tabular", "json-ld"])
 def test_dataset_ls_tags(tmpdir, runner, project, client, form, load_dataset_with_injection):