good sql unit testing strategy

Author: bernatsort

.gitignore

@@ -2,9 +2,5 @@ __pycache__/
 .pytest_cache/
 archive/
 output.json
-data/validated_reference_dataset.xlsx
+#ata/validated_reference_dataset.xlsx
 
-# YAML validation rules instead of Python
-# rules_config.yaml
-# validation_rules_yaml.py
-# tests/test_sql_logic_yaml.py

data/validated_reference_dataset.xlsx

@@ -20,5 +20,10 @@
         "expected": [
             {"week": "2017-31", "active_patients": 30}
         ]
+    },
+    "active_patients_study_1368_0004": {
+        "expected": [
+            {"week": "2017-34", "active_patients": 60}
+        ]
     }
 }

data/validated_reference_dataset_errors.xlsx

@@ -1,6 +1,4 @@
 SELECT week, active_patients
 FROM clinical_trials
-WHERE week = "2017-31"
+WHERE week = "2017-31" AND study = "1368-0004"
 
--- Como hacer para que el test podamos generalizarlo para que sea active_patients en una fecha concreta. 
--- De esta forma podemos pasarle distintas fechas y mirar si en todas el test passa. 

expected_results/expected_results.json

@@ -0,0 +1,4 @@
+SELECT week, active_patients
+FROM clinical_trials
+WHERE week = "2017-XX" AND study = "1368-0004" 
+-- error introduced week = "2017-XX" should be week = "2017-34"

expected_results/expected_results_negative.json

@@ -14,15 +14,3 @@ def db_connection():
     connection = engine.connect()
     yield connection
     connection.close()
-
-@pytest.fixture(scope='module')
-def db_connection_negative():
-    engine = create_engine('sqlite:///:memory:')
-    file_path = os.path.join(os.path.dirname(__file__), '..', 'data', 'validated_reference_dataset_errors.xlsx')
-    df = pd.read_excel(file_path)
-    df.to_sql('clinical_trials', engine, if_exists='replace', index=False)
-
-    connection = engine.connect()
-    yield connection
-    connection.close()
-