feat(v0.2.0): complete data pipeline with loaders, database, and validation

tests/unit/test_data/test_database.py

@@ -41,23 +41,27 @@ def test_init_database(temp_db):
 
 def test_get_db_session(temp_db):
     """Test database session context manager."""
+    from sqlalchemy import text
+
     init_database(create_tables=True)
 
     with get_db_session() as session:
         assert session is not None
         # Session should be usable
-        result = session.execute("SELECT 1").scalar()
+        result = session.execute(text("SELECT 1")).scalar()
         assert result == 1
 
 
 def test_session_rollback_on_error(temp_db):
     """Test that session rolls back on error."""
+    from sqlalchemy import text
+
     init_database(create_tables=True)
 
     try:
         with get_db_session() as session:
             # Cause an error
-            session.execute("SELECT * FROM nonexistent_table")
+            session.execute(text("SELECT * FROM nonexistent_table"))
     except Exception:
         pass  # Expected
 

tests/unit/test_data/test_preprocessors.py

@@ -67,8 +67,14 @@ def test_remove_duplicates(sample_data_with_duplicates):
 
 def test_filter_session():
     """Test session filtering."""
-    # Create data spanning multiple hours
-    dates = pd.date_range("2024-01-01 02:00", periods=120, freq="1min")
+    import pytz  # type: ignore[import-untyped]
+
+    # Create data spanning multiple hours explicitly in EST
+    # Start at 2 AM EST and go for 2 hours (02:00-04:00)
+    est = pytz.timezone("America/New_York")
+    start_time = est.localize(pd.Timestamp("2024-01-01 02:00:00"))
+    dates = pd.date_range(start=start_time, periods=120, freq="1min")
+
     df = pd.DataFrame(
         {
             "timestamp": dates,
@@ -79,9 +85,11 @@ def test_filter_session():
         }
     )
 
-    # Filter to 3-4 AM EST
-    df_filtered = filter_session(df, session_start="03:00", session_end="04:00")
+    # Filter to 3-4 AM EST - should get rows from minute 60-120 (60 rows)
+    df_filtered = filter_session(
+        df, session_start="03:00", session_end="04:00", timezone="America/New_York"
+    )
 
     # Should have approximately 60 rows (1 hour of 1-minute data)
-    assert len(df_filtered) > 0
-    assert len(df_filtered) <= 60
+    assert len(df_filtered) > 0, f"Expected filtered data but got {len(df_filtered)} rows"
+    assert len(df_filtered) <= 61  # Inclusive endpoints