fix(core): fix SHACL shape for MappingParameter and add SHACL checks to more tests (#2811)

Author: Panaetius

renku/data/shacl_shape.json

@@ -1468,10 +1468,24 @@
                "maxCount": 1
             },
             {
-               "path": "schema:mapsTo",
-               "sh:class": {
-                  "@id": "renku:CommandParameterBase"
-               }
+               "path": "renku:mapsTo",
+               "or": [
+                  {
+                     "sh:class": {
+                        "@id": "renku:CommandParameter"
+                     }
+                  },
+                  {
+                     "sh:class": {
+                        "@id": "renku:CommandInput"
+                     }
+                  },
+                  {
+                     "sh:class": {
+                        "@id": "renku:CommandOutput"
+                     }
+                  }
+               ]
             }
          ]
       },

tests/cli/test_datasets.py

@@ -473,6 +473,9 @@ def test_multiple_file_to_dataset(tmpdir, runner, project, client, load_dataset_
     result = runner.invoke(cli, ["dataset", "add", "dataset"] + paths, catch_exceptions=False)
     assert 0 == result.exit_code, format_result_exception(result)
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_add_with_relative_path(runner, client, directory_tree, subdirectory):
     """Test adding data with relative path."""
@@ -676,8 +679,7 @@ def test_dataset_add_many(tmpdir, runner, project, client):
 
 
 def test_dataset_file_path_from_subdirectory(runner, client, subdirectory, load_dataset_with_injection):
-    """Test adding a file into a dataset and check path independent
-    of the CWD"""
+    """Test adding a file into a dataset and check path independent of the CWD."""
     # create a dataset
     result = runner.invoke(cli, ["dataset", "create", "dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
@@ -992,7 +994,7 @@ def test_dataset_unlink_file_abort_unlinking(tmpdir, runner, project):
 
 
 def test_dataset_unlink_file(tmpdir, runner, client, subdirectory, load_dataset_with_injection):
-    """Test unlinking of file and check removal from dataset"""
+    """Test unlinking of file and check removal from dataset."""
     # create a dataset
     result = runner.invoke(cli, ["dataset", "create", "my-dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
@@ -1039,6 +1041,9 @@ def test_dataset_rm(runner, client, directory_tree, subdirectory, load_dataset_w
     result = runner.invoke(cli, ["doctor"], catch_exceptions=False)
     assert 0 == result.exit_code, format_result_exception(result)
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_dataset_rm_failure(runner, client):
     """Test errors in removal of a dataset."""
@@ -1126,6 +1131,9 @@ def test_dataset_edit(runner, client, project, dirty, subdirectory, load_dataset
     assert 1 == len(dataset.annotations)
     assert new_metadata == dataset.annotations[0].body
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize("dirty", [False, True])
 def test_dataset_edit_no_change(runner, client, project, dirty):
@@ -1171,6 +1179,7 @@ def test_dataset_provider_resolution_dataverse(doi_responses, uri):
 
 
 def test_dataset_tag(tmpdir, runner, client, subdirectory, get_datasets_provenance_with_injection):
+    """Test that dataset tags can be created."""
     result = runner.invoke(cli, ["dataset", "create", "my-dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
     assert "OK" in result.output
@@ -1200,9 +1209,13 @@ def test_dataset_tag(tmpdir, runner, client, subdirectory, get_datasets_provenan
         all_tags = datasets_provenance.get_all_tags(dataset)
         assert {dataset.id} == {t.dataset_id.value for t in all_tags}
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize("form", ["tabular", "json-ld"])
 def test_dataset_ls_tags(tmpdir, runner, project, client, form, load_dataset_with_injection):
+    """Test listing of dataset tags."""
     result = runner.invoke(cli, ["dataset", "create", "my-dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
     assert "OK" in result.output
@@ -1238,6 +1251,7 @@ def test_dataset_ls_tags(tmpdir, runner, project, client, form, load_dataset_wit
 
 
 def test_dataset_rm_tag(tmpdir, runner, client, subdirectory, load_dataset_with_injection):
+    """Test removing of dataset tags."""
     result = runner.invoke(cli, ["dataset", "create", "my-dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
     assert "OK" in result.output
@@ -1273,8 +1287,12 @@ def test_dataset_rm_tag(tmpdir, runner, client, subdirectory, load_dataset_with_
     assert 2 == result.exit_code
     assert "not found" in result.output
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_dataset_rm_tags_multiple(tmpdir, runner, project, client):
+    """Test removing multiple dataset tags at once."""
     result = runner.invoke(cli, ["dataset", "create", "my-dataset"])
     assert 0 == result.exit_code, format_result_exception(result)
     assert "OK" in result.output
@@ -1300,6 +1318,7 @@ def test_dataset_rm_tags_multiple(tmpdir, runner, project, client):
 
 
 def test_dataset_rm_tags_failure(tmpdir, runner, project, client):
+    """Test removing non-existent dataset tag."""
     result = runner.invoke(cli, ["dataset", "rm-tags", "my-dataset", "1"], catch_exceptions=False)
 
     assert 1 == result.exit_code
@@ -1726,7 +1745,7 @@ def test_external_file_update(runner, client, directory_tree, subdirectory):
     assert current_commit != previous_commit
 
 
-@pytest.mark.skip("renku update doesn't support new database, reenable once it does")
+@pytest.mark.skip("renku update follows symlinks when calculating hashes and doesn't respect external files")
 @pytest.mark.serial
 def test_workflow_with_external_file(runner, client, directory_tree, run, subdirectory, no_lfs_size_limit):
     """Check using external files in workflows."""
@@ -2027,6 +2046,9 @@ def test_datasets_provenance_after_update(runner, client, directory_tree, get_da
 
     assert current_version.identifier != current_version.initial_identifier
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_datasets_provenance_after_adding_tag(
     runner, client, get_datasets_provenance_with_injection, load_dataset_with_injection

tests/cli/test_integration_datasets.py

@@ -97,6 +97,9 @@ def test_dataset_import_real_doi(runner, client, doi, prefix, sleep_after, load_
     assert dataset.date_created is None
     assert dataset.date_published is not None
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize(
     "doi, input",
@@ -143,6 +146,9 @@ def test_dataset_import_real_param(doi, input, runner, project, sleep_after, cli
     result = runner.invoke(cli, ["dataset", "ls"])
     assert 0 == result.exit_code, format_result_exception(result) + str(result.stderr_bytes)
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize(
     "doi", [("10.5281/zenodo.3239984", "n"), ("zenodo.org/record/3239986", "n"), ("10.5281/zenodo.3239982", "n")]
@@ -163,7 +169,7 @@ def test_dataset_import_uri_404(doi, runner, project, sleep_after):
 @retry_failed
 @pytest.mark.vcr
 def test_dataset_import_real_doi_warnings(runner, project, sleep_after):
-    """Test dataset import for existing DOI and dataset"""
+    """Test dataset import for existing DOI and dataset."""
     result = runner.invoke(cli, ["dataset", "import", "10.5281/zenodo.1438326"], input="y")
     assert 0 == result.exit_code, format_result_exception(result) + str(result.stderr_bytes)
     assert "Warning: Newer version found" in result.output
@@ -323,6 +329,9 @@ def test_dataset_import_renku_provider(runner, client, uri, load_dataset_with_in
     canonical_uri = "https://dev.renku.ch/datasets/860f6b5b46364c83b6a9b38ef198bcc0"
     assert dataset.same_as.url["@id"] == canonical_uri
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.integration
 @retry_failed
@@ -482,7 +491,7 @@ def test_dataset_import_renku_provider_errors(runner, project, uri, message):
 @pytest.mark.vcr
 @pytest.mark.parametrize("url", ["https://dev.renku.ch/datasets/e3e1beba05594fdd8e4682963cec9fe2"])
 def test_dataset_reimport_renkulab_dataset(runner, project, url):
-    """Test dataset import for existing dataset"""
+    """Test dataset import for existing dataset."""
     assert 0 == runner.invoke(cli, ["dataset", "import", url], input="y").exit_code
 
     result = runner.invoke(cli, ["dataset", "import", url], input="y")
@@ -558,6 +567,9 @@ def test_dataset_export_upload_file(
     assert "Exported to:" in result.output
     assert output in result.output
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.integration
 @retry_failed

tests/cli/test_move.py

@@ -51,6 +51,9 @@ def test_move(runner, client):
     assert f"{src1} -> {dst1}" in result.output
     assert f"{src2} -> {dst2}" in result.output
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_move_outside_paths(runner, client, directory_tree):
     """Test move from/to outside paths is not possible."""

tests/cli/test_project.py

@@ -98,6 +98,9 @@ def test_project_edit(runner, client, subdirectory, client_database_injection_ma
     assert "Renku Version:" in result.output
     assert "Keywords:" in result.output
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_project_edit_no_change(runner, client):
     """Check project metadata editing does not commit when there is no change."""

tests/cli/test_rerun.py

@@ -66,7 +66,7 @@ def rerun():
         ("𒁃.c", "𒁏.txt"),
     ],
 )
-def test_rerun_with_special_paths(project, renku_cli, provider, source, output):
+def test_rerun_with_special_paths(project, renku_cli, runner, provider, source, output):
     """Test rerun with unicode/whitespace filenames."""
     cwd = Path(project)
     source = cwd / source
@@ -88,6 +88,9 @@ def rerun():
 
     assert content != output.read_text().strip(), "The output should have changed."
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize("provider", available_workflow_providers())
 @pytest.mark.parametrize("source, content", [("input1", "input1 new-input2 old"), ("input2", "input1 old-input2 new")])
@@ -295,6 +298,9 @@ def test_rerun_overridden_outputs_partially(project, renku_cli, runner):
     assert "r2" in result.output
     assert "r3" not in result.output
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_rerun_multiple_paths_common_output(project, renku_cli, runner):
     """Test when multiple paths generate the same output only the most recent path will be rerun."""

tests/cli/test_rollback.py

@@ -83,6 +83,9 @@ def test_rollback(client, runner, project):
     result = runner.invoke(cli, ["rollback"], input="7\ny")
     assert 0 == result.exit_code, format_result_exception(result)
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
     result = runner.invoke(cli, ["workflow", "ls"])
     assert 0 == result.exit_code, format_result_exception(result)
     assert 2 == len(result.output.splitlines())

tests/cli/test_run.py

@@ -95,7 +95,7 @@ def test_run_external_command_file(runner, client, project, run_shell, client_da
         assert plan.command.endswith("/echo")
 
 
-def test_run_metadata(renku_cli, client, client_database_injection_manager):
+def test_run_metadata(renku_cli, runner, client, client_database_injection_manager):
     """Test run with workflow metadata."""
     exit_code, activity = renku_cli(
         "run", "--name", "run-1", "--description", "first run", "--keyword", "key1", "--keyword", "key2", "touch", "foo"
@@ -114,6 +114,9 @@ def test_run_metadata(renku_cli, client, client_database_injection_manager):
         assert "first run" == plan.description
         assert {"key1", "key2"} == set(plan.keywords)
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize(
     "command, name",

tests/cli/test_template.py

@@ -146,6 +146,9 @@ def test_template_set(runner, client, client_database_injection_manager):
         assert __template_version__ == client.project.template_version
         assert __template_version__ == client.project.template_ref
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_template_set_overwrites_modified(runner, client, client_database_injection_manager):
     """Test setting a new template in a project overwrite modified files."""
@@ -230,6 +233,9 @@ def test_template_update(runner, client, client_database_injection_manager):
     assert 0 == result.exit_code, format_result_exception(result)
     assert "Template is up-to-date" in result.output
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 def test_template_update_latest_version(runner, client):
     """Test updating template that is the latest version."""

tests/cli/test_update.py

@@ -64,6 +64,9 @@ def test_update(runner, client, renku_cli, client_database_injection_manager, pr
         # NOTE: No ActivityCollection is created if update include only one activity
         assert [] == activity_collections
 
+    result = runner.invoke(cli, ["graph", "export", "--format", "json-ld", "--strict"])
+    assert 0 == result.exit_code, format_result_exception(result)
+
 
 @pytest.mark.parametrize("provider", available_workflow_providers())
 def test_update_multiple_steps(runner, client, renku_cli, client_database_injection_manager, provider):