fix: make status values consistent across validation types (GoogleClo…

…udPlatform#377)
ajwelch4 · Feb 26, 2022 · a56165f · a56165f
1 parent 3d78ee5
commit a56165f
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 9 deletions.
diff --git a/data_validation/schema_validation.py b/data_validation/schema_validation.py
@@ -100,7 +100,7 @@ def schema_validation_matching(source_fields, target_fields):
                         source_field_name,
                         "1",
                         "1",
-                        "Pass",
+                        "success",
                         "Source_type:{} Target_type:{}".format(
                             source_field_type, target_fields[source_field_name]
                         ),
@@ -114,7 +114,7 @@ def schema_validation_matching(source_fields, target_fields):
                         source_field_name,
                         "1",
                         "1",
-                        "Fail",
+                        "fail",
                         "Data type mismatch between source and target. Source_type:{} Target_type:{}".format(
                             source_field_type, target_fields[source_field_name]
                         ),
@@ -128,7 +128,7 @@ def schema_validation_matching(source_fields, target_fields):
                     "N/A",
                     "1",
                     "0",
-                    "Fail",
+                    "fail",
                     "Target doesn't have a matching field name",
                 ]
             )
@@ -142,7 +142,7 @@ def schema_validation_matching(source_fields, target_fields):
                     target_field_name,
                     "0",
                     "1",
-                    "Fail",
+                    "fail",
                     "Source doesn't have a matching field name",
                 ]
             )

diff --git a/tests/unit/test_schema_validation.py b/tests/unit/test_schema_validation.py
@@ -143,13 +143,13 @@ def test_schema_validation_matching(module_under_test):
     target_fields = {"field1": "string", "field2": "timestamp", "field_3": "string"}
 
     expected_results = [
-        ["field1", "field1", "1", "1", "Pass", "Source_type:string Target_type:string"],
+        ["field1", "field1", "1", "1", "success", "Source_type:string Target_type:string"],
         [
             "field2",
             "field2",
             "1",
             "1",
-            "Fail",
+            "fail",
             "Data type mismatch between source and target. "
             "Source_type:datetime Target_type:timestamp",
         ],
@@ -158,15 +158,15 @@ def test_schema_validation_matching(module_under_test):
             "N/A",
             "1",
             "0",
-            "Fail",
+            "fail",
             "Target doesn't have a matching field name",
         ],
         [
             "N/A",
             "field_3",
             "0",
             "1",
-            "Fail",
+            "fail",
             "Source doesn't have a matching field name",
         ],
     ]
@@ -188,7 +188,7 @@ def test_execute(module_under_test, fs):
 
     dv_client = data_validation.DataValidation(SAMPLE_SCHEMA_CONFIG, verbose=True)
     result_df = dv_client.schema_validator.execute()
-    failures = result_df[result_df["status"].str.contains("Fail")]
+    failures = result_df[result_df["status"].str.contains("fail")]
 
     assert len(result_df) == len(source_data[0]) + 1
     assert result_df["source_agg_value"].astype(float).sum() == 7