adding tests from tsv->rdf and back again

kgx/cli/cli_utils.py

@@ -219,7 +219,7 @@ def validate(
         },  # streaming processing throws the graph data away
         # ... Second, we inject the Inspector into the transform() call,
         # for the underlying Transformer.process() to use...
-        inspector=validator
+        inspector=validator,
     )
 
     if output:

tests/resources/test-transform-rdf-tsv.yaml

@@ -0,0 +1,25 @@
+configuration:
+  output_directory: ../target
+  checkpoint: false
+transform:
+  source:
+    test_graph:
+      name: "Test Graph"
+      input:
+        format: nt
+        filename:
+          - test-transform-rdf.nt
+      output:
+        format: tsv
+        filename:
+          - test-tranform-rdf
+    valid_graph:
+      name: "Valid JSON Graph"
+      input:
+        format: json
+        filename:
+          - valid.json
+      output:
+        format: tsv
+        filename:
+          - valid_graph

tests/resources/test-transform-rdf.nt

@@ -0,0 +1,15 @@
+<http://identifiers.org/ensembl/ENSG0000000000001> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://purl.obolibrary.org/obo/SO_0000704> .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/category> <https://w3id.org/biolink/vocab/Gene> .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/category> <https://w3id.org/biolink/vocab/GenomicEntity> .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/category> <https://w3id.org/biolink/vocab/MolecularEntity> .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/name> "Test Gene 123" .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/description> "This is a Test Gene 123" .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/provided_by> "Test Dataset" .
+<http://identifiers.org/ensembl/ENSG0000000000002> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://purl.obolibrary.org/obo/SO_0000704> .
+<http://identifiers.org/ensembl/ENSG0000000000002> <https://w3id.org/biolink/vocab/category> <https://w3id.org/biolink/vocab/Gene> .
+<http://identifiers.org/ensembl/ENSG0000000000002> <https://w3id.org/biolink/vocab/category> <https://w3id.org/biolink/vocab/GenomicEntity> .
+<http://identifiers.org/ensembl/ENSG0000000000002> <https://w3id.org/biolink/vocab/category> <https://w3id.org/biolink/vocab/MolecularEntity> .
+<http://identifiers.org/ensembl/ENSG0000000000002> <https://w3id.org/biolink/vocab/name> "Test Gene 456" .
+<http://identifiers.org/ensembl/ENSG0000000000002> <https://w3id.org/biolink/vocab/description> "This is a Test Gene 456" .
+<http://identifiers.org/ensembl/ENSG0000000000002> <https://w3id.org/biolink/vocab/provided_by> "Test Dataset" .
+<http://identifiers.org/ensembl/ENSG0000000000001> <https://w3id.org/biolink/vocab/interacts_with> <http://identifiers.org/ensembl/ENSG0000000000002> .

tests/resources/test-transform-tsv-rdf.yaml

@@ -0,0 +1,26 @@
+configuration:
+  output_directory: ../target
+  checkpoint: false
+transform:
+  source:
+    test_graph:
+      name: "Test Graph"
+      input:
+        format: tsv
+        filename:
+          - test2_nodes.tsv
+          - test2_edges.tsv
+      output:
+        format: nt
+        filename:
+          - test-tranform-tsv-rdf.nt
+    valid_graph:
+      name: "Valid JSON Graph"
+      input:
+        format: json
+        filename:
+          - valid.json
+      output:
+        format: tsv
+        filename:
+          - valid_graph

tests/unit/test_cli_utils.py

@@ -866,6 +866,25 @@ def test_transform2():
     assert os.path.exists(os.path.join(RESOURCE_DIR, "graph_edges.tsv"))
 
 
+def test_transform_rdf_to_tsv():
+    """
+    Transform from a test transform YAML.
+    """
+    transform_config = os.path.join(RESOURCE_DIR, "test-transform-tsv-rdf.yaml")
+    transform(inputs=None, transform_config=transform_config)
+    assert os.path.exists(os.path.join(TARGET_DIR, "test-transform-rdf.tsv"))
+    assert os.path.exists(os.path.join(TARGET_DIR, "test-transform-rdf.tsv"))
+
+
+def test_transform_tsv_to_rdf():
+    """
+    Transform from a test transform YAML.
+    """
+    transform_config = os.path.join(RESOURCE_DIR, "test-transform-tsv-rdf.yaml")
+    transform(inputs=None, transform_config=transform_config)
+    assert os.path.exists(os.path.join(TARGET_DIR, "test-tranform-tsv-rdf.nt"))
+
+
 def test_merge1():
     """
     Transform from test merge YAML.